Add docs category file (#12359)

### What problem does this PR solve?

As title.

### Type of change

- [x] Documentation Update

Signed-off-by: Jin Hai <haijin.chn@gmail.com>

.github/copilot-instructions.md

@@ -0,0 +1 @@
+Refer to [AGENTS.MD](../AGENTS.md) for all repo instructions.

.github/workflows/release.yml

@@ -3,11 +3,18 @@ name: release
 on:
   schedule:
     - cron: '0 13 * * *'  # This schedule runs every 13:00:00Z(21:00:00+08:00)
+  # https://github.com/orgs/community/discussions/26286?utm_source=chatgpt.com#discussioncomment-3251208
+  # "The create event does not support branch filter and tag filter."
   # The "create tags" trigger is specifically focused on the creation of new tags, while the "push tags" trigger is activated when tags are pushed, including both new tag creations and updates to existing tags.
-  create:
+  push:
     tags:
       - "v*.*.*"                  # normal release
-      - "nightly"                 # the only one mutable tag
+
+permissions:
+  contents: write
+  actions: read
+  checks: read
+  statuses: read
 
 # https://docs.github.com/en/actions/using-jobs/using-concurrency
 concurrency:
@@ -19,11 +26,11 @@ jobs:
     runs-on: [ "self-hosted", "ragflow-test" ]
     steps:
       - name: Ensure workspace ownership
-        run: echo "chown -R $USER $GITHUB_WORKSPACE" && sudo chown -R $USER $GITHUB_WORKSPACE
+        run: echo "chown -R ${USER} ${GITHUB_WORKSPACE}" && sudo chown -R ${USER} ${GITHUB_WORKSPACE}
 
-      # https://github.com/actions/checkout/blob/v3/README.md
+      # https://github.com/actions/checkout/blob/v6/README.md
       - name: Check out code
-        uses: actions/checkout@v4
+        uses: actions/checkout@v6
         with:
           token: ${{ secrets.GITHUB_TOKEN }}  # Use the secret as an environment variable
           fetch-depth: 0
@@ -31,37 +38,37 @@ jobs:
 
       - name: Prepare release body
         run: |
-          if [[ $GITHUB_EVENT_NAME == 'create' ]]; then
+          if [[ ${GITHUB_EVENT_NAME} != "schedule" ]]; then
             RELEASE_TAG=${GITHUB_REF#refs/tags/}
-            if [[ $RELEASE_TAG == 'nightly' ]]; then
-              PRERELEASE=true
-            else
+            if [[ ${RELEASE_TAG} == v* ]]; then
               PRERELEASE=false
+            else
+              PRERELEASE=true
             fi
-            echo "Workflow triggered by create tag: $RELEASE_TAG"
+            echo "Workflow triggered by create tag: ${RELEASE_TAG}"
           else
             RELEASE_TAG=nightly
             PRERELEASE=true
             echo "Workflow triggered by schedule"
           fi
-          echo "RELEASE_TAG=$RELEASE_TAG" >> $GITHUB_ENV
-          echo "PRERELEASE=$PRERELEASE" >> $GITHUB_ENV
+          echo "RELEASE_TAG=${RELEASE_TAG}" >> ${GITHUB_ENV}
+          echo "PRERELEASE=${PRERELEASE}" >> ${GITHUB_ENV}
           RELEASE_DATETIME=$(date --rfc-3339=seconds)
-          echo Release $RELEASE_TAG created from $GITHUB_SHA at $RELEASE_DATETIME > release_body.md
+          echo Release ${RELEASE_TAG} created from ${GITHUB_SHA} at ${RELEASE_DATETIME} > release_body.md
 
       - name: Move the existing mutable tag
         # https://github.com/softprops/action-gh-release/issues/171
         run: |
           git fetch --tags
-          if [[ $GITHUB_EVENT_NAME == 'schedule' ]]; then
+          if [[ ${GITHUB_EVENT_NAME} == "schedule" ]]; then
             # Determine if a given tag exists and matches a specific Git commit.
-            # actions/checkout@v4 fetch-tags doesn't work when triggered by schedule
-            if [ "$(git rev-parse -q --verify "refs/tags/$RELEASE_TAG")" = "$GITHUB_SHA" ]; then
-              echo "mutable tag $RELEASE_TAG exists and matches $GITHUB_SHA"
+            # actions/checkout@v6 fetch-tags doesn't work when triggered by schedule
+            if [ "$(git rev-parse -q --verify "refs/tags/${RELEASE_TAG}")" = "${GITHUB_SHA}" ]; then
+              echo "mutable tag ${RELEASE_TAG} exists and matches ${GITHUB_SHA}"
             else
-              git tag -f $RELEASE_TAG $GITHUB_SHA
-              git push -f origin $RELEASE_TAG:refs/tags/$RELEASE_TAG
-              echo "created/moved mutable tag $RELEASE_TAG to $GITHUB_SHA"
+              git tag -f ${RELEASE_TAG} ${GITHUB_SHA}
+              git push -f origin ${RELEASE_TAG}:refs/tags/${RELEASE_TAG}
+              echo "created/moved mutable tag ${RELEASE_TAG} to ${GITHUB_SHA}"
             fi
           fi
 
@@ -75,6 +82,14 @@ jobs:
           # The body field does not support environment variable substitution directly.
           body_path: release_body.md
 
+      - name: Build and push image
+        run: |
+          sudo docker login --username infiniflow --password-stdin <<< ${{ secrets.DOCKERHUB_TOKEN }}
+          sudo docker build --build-arg NEED_MIRROR=1 --build-arg HTTPS_PROXY=${HTTPS_PROXY} --build-arg HTTP_PROXY=${HTTP_PROXY} -t infiniflow/ragflow:${RELEASE_TAG} -f Dockerfile .
+          sudo docker tag infiniflow/ragflow:${RELEASE_TAG} infiniflow/ragflow:latest
+          sudo docker push infiniflow/ragflow:${RELEASE_TAG}
+          sudo docker push infiniflow/ragflow:latest
+
       - name: Build and push ragflow-sdk
         if: startsWith(github.ref, 'refs/tags/v')
         run: |
@@ -84,11 +99,3 @@ jobs:
         if: startsWith(github.ref, 'refs/tags/v')
         run: |
           cd admin/client && uv build && uv publish --token ${{ secrets.PYPI_API_TOKEN }}
-
-      - name: Build and push image
-        run: |
-          echo ${{ secrets.DOCKERHUB_TOKEN }} | sudo docker login --username infiniflow --password-stdin
-          sudo docker build --build-arg NEED_MIRROR=1 -t infiniflow/ragflow:${RELEASE_TAG} -f Dockerfile .
-          sudo docker tag infiniflow/ragflow:${RELEASE_TAG} infiniflow/ragflow:latest
-          sudo docker push infiniflow/ragflow:${RELEASE_TAG}
-          sudo docker push infiniflow/ragflow:latest

.github/workflows/tests.yml

@@ -1,4 +1,6 @@
 name: tests
+permissions:
+  contents: read
 
 on:
   push:
@@ -9,8 +11,11 @@ on:
       - 'docs/**'
       - '*.md'
       - '*.mdx'
+  # The only difference between pull_request and pull_request_target is the context in which the workflow runs:
+  # — pull_request_target workflows use the workflow files from the default branch, and secrets are available.
+  # — pull_request workflows use the workflow files from the pull request branch, and secrets are unavailable.
   pull_request:
-    types: [ labeled, synchronize, reopened ]
+    types: [ synchronize, ready_for_review ]
     paths-ignore:
       - 'docs/**'
       - '*.md'
@@ -28,28 +33,26 @@ jobs:
     name: ragflow_tests
     # https://docs.github.com/en/actions/using-jobs/using-conditions-to-control-job-execution
     # https://github.com/orgs/community/discussions/26261
-    if: ${{ github.event_name != 'pull_request' || contains(github.event.pull_request.labels.*.name, 'ci') }}
+    if: ${{ github.event_name != 'pull_request' || (github.event.pull_request.draft == false && contains(github.event.pull_request.labels.*.name, 'ci')) }}
     runs-on: [ "self-hosted", "ragflow-test" ]
     steps:
-      # https://github.com/hmarr/debug-action
-      #- uses: hmarr/debug-action@v2
-
       - name: Ensure workspace ownership
         run: |
           echo "Workflow triggered by ${{ github.event_name }}"
-          echo "chown -R $USER $GITHUB_WORKSPACE" && sudo chown -R $USER $GITHUB_WORKSPACE
+          echo "chown -R ${USER} ${GITHUB_WORKSPACE}" && sudo chown -R ${USER} ${GITHUB_WORKSPACE}
 
       # https://github.com/actions/checkout/issues/1781
       - name: Check out code
-        uses: actions/checkout@v4
+        uses: actions/checkout@v6
         with:
+          ref: ${{ (github.event_name == 'pull_request' || github.event_name == 'pull_request_target') && format('refs/pull/{0}/merge', github.event.pull_request.number) || github.sha }}
           fetch-depth: 0
           fetch-tags: true
 
       - name: Check workflow duplication
-        if: ${{ !cancelled() && !failure() && (github.event_name != 'pull_request' || contains(github.event.pull_request.labels.*.name, 'ci')) }}
+        if: ${{ !cancelled() && !failure() }}
         run: |
-          if [[ "$GITHUB_EVENT_NAME" != "pull_request" && "$GITHUB_EVENT_NAME" != "schedule" ]]; then
+          if [[ ${GITHUB_EVENT_NAME} != "pull_request" && ${GITHUB_EVENT_NAME} != "schedule" ]]; then
             HEAD=$(git rev-parse HEAD)
             # Find a PR that introduced a given commit
             gh auth login --with-token <<< "${{ secrets.GITHUB_TOKEN }}"
@@ -67,14 +70,14 @@ jobs:
                   gh run cancel ${GITHUB_RUN_ID}
                   while true; do
                     status=$(gh run view ${GITHUB_RUN_ID} --json status -q .status)
-                    [ "$status" = "completed" ] && break
+                    [ "${status}" = "completed" ] && break
                     sleep 5
                   done
                   exit 1
                 fi
               fi
             fi
-          else
+          elif [[ ${GITHUB_EVENT_NAME} == "pull_request" ]]; then
             PR_NUMBER=${{ github.event.pull_request.number }}
             PR_SHA_FP=${RUNNER_WORKSPACE_PREFIX}/artifacts/${GITHUB_REPOSITORY}/PR_${PR_NUMBER}
             # Calculate the hash of the current workspace content
@@ -91,20 +94,60 @@ jobs:
           version: ">=0.11.x"
           args: "check"
 
+      - name: Check comments of changed Python files
+        if: ${{ false }}
+        run: |
+          if [[ ${{ github.event_name }} == 'pull_request' || ${{ github.event_name }} == 'pull_request_target' ]]; then
+            CHANGED_FILES=$(git diff --name-only ${{ github.event.pull_request.base.sha }}...${{ github.event.pull_request.head.sha }} \
+              | grep -E '\.(py)$' || true)
+
+            if [ -n "$CHANGED_FILES" ]; then
+              echo "Check comments of changed Python files with check_comment_ascii.py"
+
+              readarray -t files <<< "$CHANGED_FILES"
+              HAS_ERROR=0
+
+              for file in "${files[@]}"; do
+                if [ -f "$file" ]; then
+                  if python3 check_comment_ascii.py "$file"; then
+                    echo "✅ $file"
+                  else
+                    echo "❌ $file"
+                    HAS_ERROR=1
+                  fi
+                fi
+              done
+
+              if [ $HAS_ERROR -ne 0 ]; then
+                exit 1
+              fi
+            else
+              echo "No Python files changed"
+            fi
+          fi
+
+      - name: Run unit test
+        run: |
+          uv sync --python 3.12 --group test --frozen
+          source .venv/bin/activate
+          which pytest || echo "pytest not in PATH"
+          echo "Start to run unit test"
+          python3 run_tests.py
+
       - name: Build ragflow:nightly
         run: |
-          RUNNER_WORKSPACE_PREFIX=${RUNNER_WORKSPACE_PREFIX:-$HOME}
+          RUNNER_WORKSPACE_PREFIX=${RUNNER_WORKSPACE_PREFIX:-${HOME}}
           RAGFLOW_IMAGE=infiniflow/ragflow:${GITHUB_RUN_ID}
-          echo "RAGFLOW_IMAGE=${RAGFLOW_IMAGE}" >> $GITHUB_ENV
+          echo "RAGFLOW_IMAGE=${RAGFLOW_IMAGE}" >> ${GITHUB_ENV}
           sudo docker pull ubuntu:22.04
-          sudo DOCKER_BUILDKIT=1 docker build --build-arg NEED_MIRROR=1 -f Dockerfile -t ${RAGFLOW_IMAGE} .
-          if [[ "$GITHUB_EVENT_NAME" == "schedule" ]]; then
+          sudo DOCKER_BUILDKIT=1 docker build --build-arg NEED_MIRROR=1 --build-arg HTTPS_PROXY=${HTTPS_PROXY} --build-arg HTTP_PROXY=${HTTP_PROXY} -f Dockerfile -t ${RAGFLOW_IMAGE} .
+          if [[ ${GITHUB_EVENT_NAME} == "schedule" ]]; then
             export HTTP_API_TEST_LEVEL=p3
           else
             export HTTP_API_TEST_LEVEL=p2
           fi
-          echo "HTTP_API_TEST_LEVEL=${HTTP_API_TEST_LEVEL}" >> $GITHUB_ENV
-          echo "RAGFLOW_CONTAINER=${GITHUB_RUN_ID}-ragflow-cpu-1" >> $GITHUB_ENV
+          echo "HTTP_API_TEST_LEVEL=${HTTP_API_TEST_LEVEL}" >> ${GITHUB_ENV}
+          echo "RAGFLOW_CONTAINER=${GITHUB_RUN_ID}-ragflow-cpu-1" >> ${GITHUB_ENV}
 
       - name: Start ragflow:nightly
         run: |
@@ -154,42 +197,43 @@ jobs:
           echo -e "COMPOSE_PROFILES=\${COMPOSE_PROFILES},tei-cpu" >> docker/.env
           echo -e "TEI_MODEL=BAAI/bge-small-en-v1.5" >> docker/.env
           echo -e "RAGFLOW_IMAGE=${RAGFLOW_IMAGE}" >> docker/.env
-          echo "HOST_ADDRESS=http://host.docker.internal:${SVR_HTTP_PORT}" >> $GITHUB_ENV
+          echo "HOST_ADDRESS=http://host.docker.internal:${SVR_HTTP_PORT}" >> ${GITHUB_ENV}
 
           sudo docker compose -f docker/docker-compose.yml -p ${GITHUB_RUN_ID} up -d
-          uv sync --python 3.10 --only-group test --no-default-groups --frozen && uv pip install sdk/python
+          uv sync --python 3.12 --only-group test --no-default-groups --frozen && uv pip install sdk/python --group test
 
       - name: Run sdk tests against Elasticsearch
         run: |
           export http_proxy=""; export https_proxy=""; export no_proxy=""; export HTTP_PROXY=""; export HTTPS_PROXY=""; export NO_PROXY=""
-          until sudo docker exec ${RAGFLOW_CONTAINER} curl -s --connect-timeout 5 ${HOST_ADDRESS} > /dev/null; do
+          until sudo docker exec ${RAGFLOW_CONTAINER} curl -s --connect-timeout 5 ${HOST_ADDRESS}/v1/system/ping > /dev/null; do
             echo "Waiting for service to be available..."
             sleep 5
           done
-          source .venv/bin/activate && pytest -s --tb=short --level=${HTTP_API_TEST_LEVEL} test/testcases/test_sdk_api
+          source .venv/bin/activate && set -o pipefail; pytest -s --tb=short --level=${HTTP_API_TEST_LEVEL} test/testcases/test_sdk_api 2>&1 | tee es_sdk_test.log
 
       - name: Run frontend api tests against Elasticsearch
         run: |
           export http_proxy=""; export https_proxy=""; export no_proxy=""; export HTTP_PROXY=""; export HTTPS_PROXY=""; export NO_PROXY=""
-          until sudo docker exec ${RAGFLOW_CONTAINER} curl -s --connect-timeout 5 ${HOST_ADDRESS} > /dev/null; do
+          until sudo docker exec ${RAGFLOW_CONTAINER} curl -s --connect-timeout 5 ${HOST_ADDRESS}/v1/system/ping > /dev/null; do
             echo "Waiting for service to be available..."
             sleep 5
           done
-          source .venv/bin/activate && pytest -s --tb=short sdk/python/test/test_frontend_api/get_email.py sdk/python/test/test_frontend_api/test_dataset.py
+          source .venv/bin/activate && set -o pipefail; pytest -s --tb=short sdk/python/test/test_frontend_api/get_email.py sdk/python/test/test_frontend_api/test_dataset.py 2>&1 | tee es_api_test.log
           
       - name: Run http api tests against Elasticsearch
         run: |
           export http_proxy=""; export https_proxy=""; export no_proxy=""; export HTTP_PROXY=""; export HTTPS_PROXY=""; export NO_PROXY=""
-          until sudo docker exec ${RAGFLOW_CONTAINER} curl -s --connect-timeout 5 ${HOST_ADDRESS} > /dev/null; do
+          until sudo docker exec ${RAGFLOW_CONTAINER} curl -s --connect-timeout 5 ${HOST_ADDRESS}/v1/system/ping > /dev/null; do
             echo "Waiting for service to be available..."
             sleep 5
           done
-          source .venv/bin/activate && pytest -s --tb=short --level=${HTTP_API_TEST_LEVEL} test/testcases/test_http_api
+          source .venv/bin/activate && set -o pipefail; pytest -s --tb=short --level=${HTTP_API_TEST_LEVEL} test/testcases/test_http_api 2>&1 | tee es_http_api_test.log
 
       - name: Stop ragflow:nightly
         if: always()  # always run this step even if previous steps failed
         run: |
-          sudo docker compose -f docker/docker-compose.yml -p ${GITHUB_RUN_ID} down -v
+          sudo docker compose -f docker/docker-compose.yml -p ${GITHUB_RUN_ID} down -v || true
+          sudo docker ps -a --filter "label=com.docker.compose.project=${GITHUB_RUN_ID}" -q | xargs -r sudo docker rm -f
 
       - name: Start ragflow:nightly
         run: |
@@ -199,32 +243,36 @@ jobs:
       - name: Run sdk tests against Infinity
         run: |
           export http_proxy=""; export https_proxy=""; export no_proxy=""; export HTTP_PROXY=""; export HTTPS_PROXY=""; export NO_PROXY=""
-          until sudo docker exec ${RAGFLOW_CONTAINER} curl -s --connect-timeout 5 ${HOST_ADDRESS} > /dev/null; do
+          until sudo docker exec ${RAGFLOW_CONTAINER} curl -s --connect-timeout 5 ${HOST_ADDRESS}/v1/system/ping > /dev/null; do
             echo "Waiting for service to be available..."
             sleep 5
           done
-          source .venv/bin/activate && DOC_ENGINE=infinity pytest -s --tb=short --level=${HTTP_API_TEST_LEVEL} test/testcases/test_sdk_api
+          source .venv/bin/activate && set -o pipefail; DOC_ENGINE=infinity pytest -s --tb=short --level=${HTTP_API_TEST_LEVEL} test/testcases/test_sdk_api 2>&1 | tee infinity_sdk_test.log
 
       - name: Run frontend api tests against Infinity
         run: |
           export http_proxy=""; export https_proxy=""; export no_proxy=""; export HTTP_PROXY=""; export HTTPS_PROXY=""; export NO_PROXY=""
-          until sudo docker exec ${RAGFLOW_CONTAINER} curl -s --connect-timeout 5 ${HOST_ADDRESS} > /dev/null; do
+          until sudo docker exec ${RAGFLOW_CONTAINER} curl -s --connect-timeout 5 ${HOST_ADDRESS}/v1/system/ping > /dev/null; do
             echo "Waiting for service to be available..."
             sleep 5
           done
-          source .venv/bin/activate && DOC_ENGINE=infinity pytest -s --tb=short sdk/python/test/test_frontend_api/get_email.py sdk/python/test/test_frontend_api/test_dataset.py
+          source .venv/bin/activate && set -o pipefail; DOC_ENGINE=infinity pytest -s --tb=short sdk/python/test/test_frontend_api/get_email.py sdk/python/test/test_frontend_api/test_dataset.py 2>&1 | tee infinity_api_test.log
 
       - name: Run http api tests against Infinity
         run: |
           export http_proxy=""; export https_proxy=""; export no_proxy=""; export HTTP_PROXY=""; export HTTPS_PROXY=""; export NO_PROXY=""
-          until sudo docker exec ${RAGFLOW_CONTAINER} curl -s --connect-timeout 5 ${HOST_ADDRESS} > /dev/null; do
+          until sudo docker exec ${RAGFLOW_CONTAINER} curl -s --connect-timeout 5 ${HOST_ADDRESS}/v1/system/ping > /dev/null; do
             echo "Waiting for service to be available..."
             sleep 5
           done
-          source .venv/bin/activate && DOC_ENGINE=infinity pytest -s --tb=short --level=${HTTP_API_TEST_LEVEL} test/testcases/test_http_api
+          source .venv/bin/activate && set -o pipefail; DOC_ENGINE=infinity pytest -s --tb=short --level=${HTTP_API_TEST_LEVEL} test/testcases/test_http_api 2>&1 | tee infinity_http_api_test.log
 
       - name: Stop ragflow:nightly
         if: always()  # always run this step even if previous steps failed
         run: |
-          sudo docker compose -f docker/docker-compose.yml -p ${GITHUB_RUN_ID} down -v
-          sudo docker rmi -f ${RAGFLOW_IMAGE:-NO_IMAGE} || true
+          # Sometimes `docker compose down` fail due to hang container, heavy load etc. Need to remove such containers to release resources(for example, listen ports).
+          sudo docker compose -f docker/docker-compose.yml -p ${GITHUB_RUN_ID} down -v || true
+          sudo docker ps -a --filter "label=com.docker.compose.project=${GITHUB_RUN_ID}" -q | xargs -r sudo docker rm -f
+          if [[ -n ${RAGFLOW_IMAGE} ]]; then
+            sudo docker rmi -f ${RAGFLOW_IMAGE}
+          fi

.gitignore

@@ -195,3 +195,6 @@ ragflow_cli.egg-info
 
 # Default backup dir
 backup
+
+
+.hypothesis

AGENTS.md

@@ -0,0 +1,110 @@
+# RAGFlow Project Instructions for GitHub Copilot
+
+This file provides context, build instructions, and coding standards for the RAGFlow project.
+It is structured to follow GitHub Copilot's [customization guidelines](https://docs.github.com/en/copilot/concepts/prompting/response-customization).
+
+## 1. Project Overview
+RAGFlow is an open-source RAG (Retrieval-Augmented Generation) engine based on deep document understanding. It is a full-stack application with a Python backend and a React/TypeScript frontend.
+
+- **Backend**: Python 3.10+ (Flask/Quart)
+- **Frontend**: TypeScript, React, UmiJS
+- **Architecture**: Microservices based on Docker.
+  - `api/`: Backend API server.
+  - `rag/`: Core RAG logic (indexing, retrieval).
+  - `deepdoc/`: Document parsing and OCR.
+  - `web/`: Frontend application.
+
+## 2. Directory Structure
+- `api/`: Backend API server (Flask/Quart).
+  - `apps/`: API Blueprints (Knowledge Base, Chat, etc.).
+  - `db/`: Database models and services.
+- `rag/`: Core RAG logic.
+  - `llm/`: LLM, Embedding, and Rerank model abstractions.
+- `deepdoc/`: Document parsing and OCR modules.
+- `agent/`: Agentic reasoning components.
+- `web/`: Frontend application (React + UmiJS).
+- `docker/`: Docker deployment configurations.
+- `sdk/`: Python SDK.
+- `test/`: Backend tests.
+
+## 3. Build Instructions
+
+### Backend (Python)
+The project uses **uv** for dependency management.
+
+1. **Setup Environment**:
+   ```bash
+   uv sync --python 3.12 --all-extras
+   uv run download_deps.py
+   ```
+
+2. **Run Server**:
+   - **Pre-requisite**: Start dependent services (MySQL, ES/Infinity, Redis, MinIO).
+     ```bash
+     docker compose -f docker/docker-compose-base.yml up -d
+     ```
+   - **Launch**:
+     ```bash
+     source .venv/bin/activate
+     export PYTHONPATH=$(pwd)
+     bash docker/launch_backend_service.sh
+     ```
+
+### Frontend (TypeScript/React)
+Located in `web/`.
+
+1. **Install Dependencies**:
+   ```bash
+   cd web
+   npm install
+   ```
+
+2. **Run Dev Server**:
+   ```bash
+   npm run dev
+   ```
+   Runs on port 8000 by default.
+
+### Docker Deployment
+To run the full stack using Docker:
+```bash
+cd docker
+docker compose -f docker-compose.yml up -d
+```
+
+## 4. Testing Instructions
+
+### Backend Tests
+- **Run All Tests**:
+  ```bash
+  uv run pytest
+  ```
+- **Run Specific Test**:
+  ```bash
+  uv run pytest test/test_api.py
+  ```
+
+### Frontend Tests
+- **Run Tests**:
+  ```bash
+  cd web
+  npm run test
+  ```
+
+## 5. Coding Standards & Guidelines
+- **Python Formatting**: Use `ruff` for linting and formatting.
+  ```bash
+  ruff check
+  ruff format
+  ```
+- **Frontend Linting**:
+  ```bash
+  cd web
+  npm run lint
+  ```
+- **Pre-commit**: Ensure pre-commit hooks are installed.
+  ```bash
+  pre-commit install
+  pre-commit run --all-files
+  ```
+

CLAUDE.md

@@ -45,7 +45,7 @@ RAGFlow is an open-source RAG (Retrieval-Augmented Generation) engine based on d
 ### Backend Development
 ```bash
 # Install Python dependencies
-uv sync --python 3.10 --all-extras
+uv sync --python 3.12 --all-extras
 uv run download_deps.py
 pre-commit install
 

Dockerfile

@@ -1,5 +1,5 @@
 # base stage
-FROM ubuntu:22.04 AS base
+FROM ubuntu:24.04 AS base
 USER root
 SHELL ["/bin/bash", "-c"]
 
@@ -10,11 +10,10 @@ WORKDIR /ragflow
 # Copy models downloaded via download_deps.py
 RUN mkdir -p /ragflow/rag/res/deepdoc /root/.ragflow
 RUN --mount=type=bind,from=infiniflow/ragflow_deps:latest,source=/huggingface.co,target=/huggingface.co \
-    cp /huggingface.co/InfiniFlow/huqie/huqie.txt.trie /ragflow/rag/res/ && \
     tar --exclude='.*' -cf - \
         /huggingface.co/InfiniFlow/text_concat_xgb_v1.0 \
         /huggingface.co/InfiniFlow/deepdoc \
-        | tar -xf - --strip-components=3 -C /ragflow/rag/res/deepdoc 
+        | tar -xf - --strip-components=3 -C /ragflow/rag/res/deepdoc
 
 # https://github.com/chrismattmann/tika-python
 # This is the only way to run python-tika without internet access. Without this set, the default is to check the tika version and pull latest every time from Apache.
@@ -34,34 +33,41 @@ ENV DEBIAN_FRONTEND=noninteractive
 # selenium:      libatk-bridge2.0-0                       chrome-linux64-121-0-6167-85
 # Building C extensions: libpython3-dev libgtk-4-1 libnss3 xdg-utils libgbm-dev
 RUN --mount=type=cache,id=ragflow_apt,target=/var/cache/apt,sharing=locked \
+    apt update && \
+    apt --no-install-recommends install -y ca-certificates; \
     if [ "$NEED_MIRROR" == "1" ]; then \
-        sed -i 's|http://ports.ubuntu.com|http://mirrors.tuna.tsinghua.edu.cn|g' /etc/apt/sources.list; \
-        sed -i 's|http://archive.ubuntu.com|http://mirrors.tuna.tsinghua.edu.cn|g' /etc/apt/sources.list; \
+        sed -i 's|http://archive.ubuntu.com/ubuntu|https://mirrors.tuna.tsinghua.edu.cn/ubuntu|g' /etc/apt/sources.list.d/ubuntu.sources; \
+        sed -i 's|http://security.ubuntu.com/ubuntu|https://mirrors.tuna.tsinghua.edu.cn/ubuntu|g' /etc/apt/sources.list.d/ubuntu.sources; \
     fi; \
     rm -f /etc/apt/apt.conf.d/docker-clean && \
     echo 'Binary::apt::APT::Keep-Downloaded-Packages "true";' > /etc/apt/apt.conf.d/keep-cache && \
     chmod 1777 /tmp && \
     apt update && \
-    apt --no-install-recommends install -y ca-certificates && \
-    apt update && \
     apt install -y libglib2.0-0 libglx-mesa0 libgl1 && \
     apt install -y pkg-config libicu-dev libgdiplus && \
     apt install -y default-jdk && \
     apt install -y libatk-bridge2.0-0 && \
     apt install -y libpython3-dev libgtk-4-1 libnss3 xdg-utils libgbm-dev && \
     apt install -y libjemalloc-dev && \
-    apt install -y python3-pip pipx nginx unzip curl wget git vim less && \
-    apt install -y ghostscript
+    apt install -y nginx unzip curl wget git vim less && \
+    apt install -y ghostscript && \
+    apt install -y pandoc && \
+    apt install -y texlive && \
+    apt install -y fonts-freefont-ttf fonts-noto-cjk
 
-RUN if [ "$NEED_MIRROR" == "1" ]; then \
-        pip3 config set global.index-url https://pypi.tuna.tsinghua.edu.cn/simple && \
-        pip3 config set global.trusted-host pypi.tuna.tsinghua.edu.cn; \
+# Install uv
+RUN --mount=type=bind,from=infiniflow/ragflow_deps:latest,source=/,target=/deps \
+    if [ "$NEED_MIRROR" == "1" ]; then \
         mkdir -p /etc/uv && \
-        echo "[[index]]" > /etc/uv/uv.toml && \
+        echo 'python-install-mirror = "https://registry.npmmirror.com/-/binary/python-build-standalone/"' > /etc/uv/uv.toml && \
+        echo '[[index]]' >> /etc/uv/uv.toml && \
         echo 'url = "https://pypi.tuna.tsinghua.edu.cn/simple"' >> /etc/uv/uv.toml && \
-        echo "default = true" >> /etc/uv/uv.toml; \
+        echo 'default = true' >> /etc/uv/uv.toml; \
     fi; \
-    pipx install uv
+    tar xzf /deps/uv-x86_64-unknown-linux-gnu.tar.gz \
+    && cp uv-x86_64-unknown-linux-gnu/* /usr/local/bin/ \
+    && rm -rf uv-x86_64-unknown-linux-gnu \
+    && uv python install 3.11
 
 ENV PYTHONDONTWRITEBYTECODE=1 DOTNET_SYSTEM_GLOBALIZATION_INVARIANT=1
 ENV PATH=/root/.local/bin:$PATH
@@ -77,12 +83,12 @@ RUN --mount=type=cache,id=ragflow_apt,target=/var/cache/apt,sharing=locked \
 # A modern version of cargo is needed for the latest version of the Rust compiler.
 RUN apt update && apt install -y curl build-essential \
     && if [ "$NEED_MIRROR" == "1" ]; then \
-         # Use TUNA mirrors for rustup/rust dist files
+         # Use TUNA mirrors for rustup/rust dist files \
          export RUSTUP_DIST_SERVER="https://mirrors.tuna.tsinghua.edu.cn/rustup"; \
          export RUSTUP_UPDATE_ROOT="https://mirrors.tuna.tsinghua.edu.cn/rustup/rustup"; \
          echo "Using TUNA mirrors for Rustup."; \
        fi; \
-    # Force curl to use HTTP/1.1
+    # Force curl to use HTTP/1.1 \
     curl --proto '=https' --tlsv1.2 --http1.1 -sSf https://sh.rustup.rs | bash -s -- -y --profile minimal \
     && echo 'export PATH="/root/.cargo/bin:${PATH}"' >> /root/.bashrc
 
@@ -99,10 +105,10 @@ RUN --mount=type=cache,id=ragflow_apt,target=/var/cache/apt,sharing=locked \
     apt update && \
     arch="$(uname -m)"; \
     if [ "$arch" = "arm64" ] || [ "$arch" = "aarch64" ]; then \
-        # ARM64 (macOS/Apple Silicon or Linux aarch64)
+        # ARM64 (macOS/Apple Silicon or Linux aarch64) \
         ACCEPT_EULA=Y apt install -y unixodbc-dev msodbcsql18; \
     else \
-        # x86_64 or others
+        # x86_64 or others \
         ACCEPT_EULA=Y apt install -y unixodbc-dev msodbcsql17; \
     fi || \
     { echo "Failed to install ODBC driver"; exit 1; }
@@ -146,7 +152,7 @@ RUN --mount=type=cache,id=ragflow_uv,target=/root/.cache/uv,sharing=locked \
     else \
         sed -i 's|pypi.tuna.tsinghua.edu.cn|pypi.org|g' uv.lock; \
     fi; \
-    uv sync --python 3.10 --frozen
+    uv sync --python 3.12 --frozen
 
 COPY web web
 COPY docs docs
@@ -186,6 +192,7 @@ COPY pyproject.toml uv.lock ./
 COPY mcp mcp
 COPY plugin plugin
 COPY common common
+COPY memory memory
 
 COPY docker/service_conf.yaml.template ./conf/service_conf.yaml.template
 COPY docker/entrypoint.sh ./

Dockerfile.deps

@@ -3,7 +3,7 @@
 FROM scratch
 
 # Copy resources downloaded via download_deps.py
-COPY chromedriver-linux64-121-0-6167-85 chrome-linux64-121-0-6167-85 cl100k_base.tiktoken libssl1.1_1.1.1f-1ubuntu2_amd64.deb libssl1.1_1.1.1f-1ubuntu2_arm64.deb tika-server-standard-3.0.0.jar tika-server-standard-3.0.0.jar.md5 libssl*.deb /
+COPY chromedriver-linux64-121-0-6167-85 chrome-linux64-121-0-6167-85 cl100k_base.tiktoken libssl1.1_1.1.1f-1ubuntu2_amd64.deb libssl1.1_1.1.1f-1ubuntu2_arm64.deb tika-server-standard-3.0.0.jar tika-server-standard-3.0.0.jar.md5 libssl*.deb uv-x86_64-unknown-linux-gnu.tar.gz /
 
 COPY nltk_data /nltk_data
 

README.md

@@ -22,7 +22,7 @@
         <img alt="Static Badge" src="https://img.shields.io/badge/Online-Demo-4e6b99">
     </a>
     <a href="https://hub.docker.com/r/infiniflow/ragflow" target="_blank">
-        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.21.1">
+        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.23.1">
     </a>
     <a href="https://github.com/infiniflow/ragflow/releases/latest">
         <img src="https://img.shields.io/github/v/release/infiniflow/ragflow?color=blue&label=Latest%20Release" alt="Latest Release">
@@ -37,7 +37,7 @@
 
 <h4 align="center">
   <a href="https://ragflow.io/docs/dev/">Document</a> |
-  <a href="https://github.com/infiniflow/ragflow/issues/4214">Roadmap</a> |
+  <a href="https://github.com/infiniflow/ragflow/issues/12241">Roadmap</a> |
   <a href="https://twitter.com/infiniflowai">Twitter</a> |
   <a href="https://discord.gg/NjYzJD3GM3">Discord</a> |
   <a href="https://demo.ragflow.io">Demo</a>
@@ -61,8 +61,7 @@
 - 🔎 [System Architecture](#-system-architecture)
 - 🎬 [Get Started](#-get-started)
 - 🔧 [Configurations](#-configurations)
-- 🔧 [Build a docker image without embedding models](#-build-a-docker-image-without-embedding-models)
-- 🔧 [Build a docker image including embedding models](#-build-a-docker-image-including-embedding-models)
+- 🔧 [Build a Docker image](#-build-a-docker-image)
 - 🔨 [Launch service from source for development](#-launch-service-from-source-for-development)
 - 📚 [Documentation](#-documentation)
 - 📜 [Roadmap](#-roadmap)
@@ -86,6 +85,9 @@ Try our demo at [https://demo.ragflow.io](https://demo.ragflow.io).
 
 ## 🔥 Latest Updates
 
+- 2025-12-26 Supports 'Memory' for AI agent.
+- 2025-11-19 Supports Gemini 3 Pro.
+- 2025-11-12 Supports data synchronization from Confluence, S3, Notion, Discord, Google Drive.
 - 2025-10-23 Supports MinerU & Docling as document parsing methods.
 - 2025-10-15 Supports orchestrable ingestion pipeline.
 - 2025-08-08 Supports OpenAI's latest GPT-5 series models.
@@ -93,9 +95,6 @@ Try our demo at [https://demo.ragflow.io](https://demo.ragflow.io).
 - 2025-05-23 Adds a Python/JavaScript code executor component to Agent.
 - 2025-05-05 Supports cross-language query.
 - 2025-03-19 Supports using a multi-modal model to make sense of images within PDF or DOCX files.
-- 2025-02-28 Combined with Internet search (Tavily), supports reasoning like Deep Research for any LLMs.
-- 2024-12-18 Upgrades Document Layout Analysis model in DeepDoc.
-- 2024-08-22 Support text to SQL statements through RAG.
 
 ## 🎉 Stay Tuned
 
@@ -189,25 +188,31 @@ releases! 🌟
 > All Docker images are built for x86 platforms. We don't currently offer Docker images for ARM64.
 > If you are on an ARM64 platform, follow [this guide](https://ragflow.io/docs/dev/build_docker_image) to build a Docker image compatible with your system.
 
-> The command below downloads the `v0.21.1-slim` edition of the RAGFlow Docker image. See the following table for descriptions of different RAGFlow editions. To download a RAGFlow edition different from `v0.21.1-slim`, update the `RAGFLOW_IMAGE` variable accordingly in **docker/.env** before using `docker compose` to start the server.
+> The command below downloads the `v0.23.1` edition of the RAGFlow Docker image. See the following table for descriptions of different RAGFlow editions. To download a RAGFlow edition different from `v0.23.1`, update the `RAGFLOW_IMAGE` variable accordingly in **docker/.env** before using `docker compose` to start the server.
 
 ```bash
    $ cd ragflow/docker
-   # Use CPU for embedding and DeepDoc tasks:
+  
+   # git checkout v0.23.1
+   # Optional: use a stable tag (see releases: https://github.com/infiniflow/ragflow/releases)
+   # This step ensures the **entrypoint.sh** file in the code matches the Docker image version.
+   
+   # Use CPU for DeepDoc tasks:
    $ docker compose -f docker-compose.yml up -d
 
-   # To use GPU to accelerate embedding and DeepDoc tasks:
+   # To use GPU to accelerate DeepDoc tasks:
    # sed -i '1i DEVICE=gpu' .env
    # docker compose -f docker-compose.yml up -d
 ```
 
-| RAGFlow image tag | Image size (GB) | Has embedding models? | Stable?                    |
-| ----------------- | --------------- | --------------------- | -------------------------- |
-| v0.21.1           | &approx;9       | ✔️                  | Stable release             |
-| v0.21.1-slim      | &approx;2       | ❌                    | Stable release             |
-| nightly           | &approx;2       | ❌                    | _Unstable_ nightly build |
+> Note: Prior to `v0.22.0`, we provided both images with embedding models and slim images without embedding models. Details as follows:
 
-> Note: Starting with `v0.22.0`, we ship only the slim edition and no longer append the **-slim** suffix to the image tag.
+| RAGFlow image tag | Image size (GB) | Has embedding models? | Stable?        |
+|-------------------|-----------------|-----------------------|----------------|
+| v0.21.1           | &approx;9       | ✔️                    | Stable release |
+| v0.21.1-slim      | &approx;2       | ❌                     | Stable release |
+
+> Starting with `v0.22.0`, we ship only the slim edition and no longer append the **-slim** suffix to the image tag.
 
 4. Check the server status after having the server up and running:
 
@@ -228,7 +233,7 @@ releases! 🌟
     * Running on all addresses (0.0.0.0)
    ```
 
-   > If you skip this confirmation step and directly log in to RAGFlow, your browser may prompt a `network anormal`
+   > If you skip this confirmation step and directly log in to RAGFlow, your browser may prompt a `network abnormal`
    > error because, at that moment, your RAGFlow may not be fully initialized.
    >
 5. In your web browser, enter the IP address of your server and log in to RAGFlow.
@@ -288,7 +293,7 @@ RAGFlow uses Elasticsearch by default for storing full text and vectors. To swit
 > [!WARNING]
 > Switching to Infinity on a Linux/arm64 machine is not yet officially supported.
 
-## 🔧 Build a Docker image without embedding models
+## 🔧 Build a Docker image
 
 This image is approximately 2 GB in size and relies on external LLM and embedding services.
 
@@ -298,6 +303,15 @@ cd ragflow/
 docker build --platform linux/amd64 -f Dockerfile -t infiniflow/ragflow:nightly .
 ```
 
+Or if you are behind a proxy, you can pass proxy arguments:
+
+```bash
+docker build --platform linux/amd64 \
+  --build-arg http_proxy=http://YOUR_PROXY:PORT \
+  --build-arg https_proxy=http://YOUR_PROXY:PORT \
+  -f Dockerfile -t infiniflow/ragflow:nightly .
+```
+
 ## 🔨 Launch service from source for development
 
 1. Install `uv` and `pre-commit`, or skip this step if they are already installed:
@@ -310,7 +324,7 @@ docker build --platform linux/amd64 -f Dockerfile -t infiniflow/ragflow:nightly
    ```bash
    git clone https://github.com/infiniflow/ragflow.git
    cd ragflow/
-   uv sync --python 3.10 # install RAGFlow dependent python modules
+   uv sync --python 3.12 # install RAGFlow dependent python modules
    uv run download_deps.py
    pre-commit install
    ```
@@ -382,7 +396,7 @@ docker build --platform linux/amd64 -f Dockerfile -t infiniflow/ragflow:nightly
 
 ## 📜 Roadmap
 
-See the [RAGFlow Roadmap 2025](https://github.com/infiniflow/ragflow/issues/4214)
+See the [RAGFlow Roadmap 2026](https://github.com/infiniflow/ragflow/issues/12241)
 
 ## 🏄 Community
 

README_id.md

@@ -22,7 +22,7 @@
         <img alt="Lencana Daring" src="https://img.shields.io/badge/Online-Demo-4e6b99">
     </a>
     <a href="https://hub.docker.com/r/infiniflow/ragflow" target="_blank">
-        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.21.1">
+        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.23.1">
     </a>
     <a href="https://github.com/infiniflow/ragflow/releases/latest">
         <img src="https://img.shields.io/github/v/release/infiniflow/ragflow?color=blue&label=Rilis%20Terbaru" alt="Rilis Terbaru">
@@ -37,7 +37,7 @@
 
 <h4 align="center">
   <a href="https://ragflow.io/docs/dev/">Dokumentasi</a> |
-  <a href="https://github.com/infiniflow/ragflow/issues/4214">Peta Jalan</a> |
+  <a href="https://github.com/infiniflow/ragflow/issues/12241">Peta Jalan</a> |
   <a href="https://twitter.com/infiniflowai">Twitter</a> |
   <a href="https://discord.gg/NjYzJD3GM3">Discord</a> |
   <a href="https://demo.ragflow.io">Demo</a>
@@ -61,8 +61,7 @@
 - 🔎 [Arsitektur Sistem](#-arsitektur-sistem)
 - 🎬 [Mulai](#-mulai)
 - 🔧 [Konfigurasi](#-konfigurasi)
-- 🔧 [Membangun Image Docker tanpa Model Embedding](#-membangun-image-docker-tanpa-model-embedding)
-- 🔧 [Membangun Image Docker dengan Model Embedding](#-membangun-image-docker-dengan-model-embedding)
+- 🔧 [Membangun Image Docker](#-membangun-docker-image)
 - 🔨 [Meluncurkan aplikasi dari Sumber untuk Pengembangan](#-meluncurkan-aplikasi-dari-sumber-untuk-pengembangan)
 - 📚 [Dokumentasi](#-dokumentasi)
 - 📜 [Peta Jalan](#-peta-jalan)
@@ -86,6 +85,9 @@ Coba demo kami di [https://demo.ragflow.io](https://demo.ragflow.io).
 
 ## 🔥 Pembaruan Terbaru
 
+- 2025-12-26 Mendukung 'Memori' untuk agen AI.
+- 2025-11-19 Mendukung Gemini 3 Pro.
+- 2025-11-12 Mendukung sinkronisasi data dari Confluence, S3, Notion, Discord, Google Drive.
 - 2025-10-23 Mendukung MinerU & Docling sebagai metode penguraian dokumen.
 - 2025-10-15 Dukungan untuk jalur data yang terorkestrasi.
 - 2025-08-08 Mendukung model seri GPT-5 terbaru dari OpenAI.
@@ -93,7 +95,6 @@ Coba demo kami di [https://demo.ragflow.io](https://demo.ragflow.io).
 - 2025-05-23 Menambahkan komponen pelaksana kode Python/JS ke Agen.
 - 2025-05-05 Mendukung kueri lintas bahasa.
 - 2025-03-19 Mendukung penggunaan model multi-modal untuk memahami gambar di dalam file PDF atau DOCX.
-- 2025-02-28 dikombinasikan dengan pencarian Internet (TAVILY), mendukung penelitian mendalam untuk LLM apa pun.
 - 2024-12-18 Meningkatkan model Analisis Tata Letak Dokumen di DeepDoc.
 - 2024-08-22 Dukungan untuk teks ke pernyataan SQL melalui RAG.
 
@@ -187,25 +188,31 @@ Coba demo kami di [https://demo.ragflow.io](https://demo.ragflow.io).
 > Semua gambar Docker dibangun untuk platform x86. Saat ini, kami tidak menawarkan gambar Docker untuk ARM64.
 > Jika Anda menggunakan platform ARM64, [silakan gunakan panduan ini untuk membangun gambar Docker yang kompatibel dengan sistem Anda](https://ragflow.io/docs/dev/build_docker_image).
 
-> Perintah di bawah ini mengunduh edisi v0.21.1 dari gambar Docker RAGFlow. Silakan merujuk ke tabel berikut untuk deskripsi berbagai edisi RAGFlow. Untuk mengunduh edisi RAGFlow yang berbeda dari v0.21.1, perbarui variabel RAGFLOW_IMAGE di docker/.env sebelum menggunakan docker compose untuk memulai server.
+> Perintah di bawah ini mengunduh edisi v0.23.1 dari gambar Docker RAGFlow. Silakan merujuk ke tabel berikut untuk deskripsi berbagai edisi RAGFlow. Untuk mengunduh edisi RAGFlow yang berbeda dari v0.23.1, perbarui variabel RAGFLOW_IMAGE di docker/.env sebelum menggunakan docker compose untuk memulai server.
 
 ```bash
    $ cd ragflow/docker
-   # Use CPU for embedding and DeepDoc tasks:
+   
+   # git checkout v0.23.1
+   # Opsional: gunakan tag stabil (lihat releases: https://github.com/infiniflow/ragflow/releases)
+   # This steps ensures the **entrypoint.sh** file in the code matches the Docker image version.
+
+   # Use CPU for DeepDoc tasks:
    $ docker compose -f docker-compose.yml up -d
 
-   # To use GPU to accelerate embedding and DeepDoc tasks:
+   # To use GPU to accelerate DeepDoc tasks:
    # sed -i '1i DEVICE=gpu' .env
    # docker compose -f docker-compose.yml up -d
 ```
 
-| RAGFlow image tag | Image size (GB) | Has embedding models? | Stable?                    |
-| ----------------- | --------------- | --------------------- | -------------------------- |
-| v0.21.1           | &approx;9       | ✔️                  | Stable release             |
-| v0.21.1-slim      | &approx;2       | ❌                    | Stable release             |
-| nightly           | &approx;2       | ❌                    | _Unstable_ nightly build |
+> Catatan: Sebelum `v0.22.0`, kami menyediakan image dengan model embedding dan image slim tanpa model embedding. Detailnya sebagai berikut:
 
-> Catatan: Mulai dari `v0.22.0`, kami hanya menyediakan edisi slim dan tidak lagi menambahkan akhiran **-slim** pada tag image.
+| RAGFlow image tag | Image size (GB) | Has embedding models? | Stable?        |
+|-------------------|-----------------|-----------------------|----------------|
+| v0.21.1           | &approx;9       | ✔️                    | Stable release |
+| v0.21.1-slim      | &approx;2       | ❌                     | Stable release |
+
+> Mulai dari `v0.22.0`, kami hanya menyediakan edisi slim dan tidak lagi menambahkan akhiran **-slim** pada tag image.
 
 1. Periksa status server setelah server aktif dan berjalan:
 
@@ -226,7 +233,7 @@ Coba demo kami di [https://demo.ragflow.io](https://demo.ragflow.io).
     * Running on all addresses (0.0.0.0)
    ```
 
-   > Jika Anda melewatkan langkah ini dan langsung login ke RAGFlow, browser Anda mungkin menampilkan error `network anormal`
+   > Jika Anda melewatkan langkah ini dan langsung login ke RAGFlow, browser Anda mungkin menampilkan error `network abnormal`
    > karena RAGFlow mungkin belum sepenuhnya siap.
    >
 2. Buka browser web Anda, masukkan alamat IP server Anda, dan login ke RAGFlow.
@@ -260,7 +267,7 @@ Pembaruan konfigurasi ini memerlukan reboot semua kontainer agar efektif:
 > $ docker compose -f docker-compose.yml up -d
 > ```
 
-## 🔧 Membangun Docker Image tanpa Model Embedding
+## 🔧 Membangun Docker Image
 
 Image ini berukuran sekitar 2 GB dan bergantung pada aplikasi LLM eksternal dan embedding.
 
@@ -270,6 +277,15 @@ cd ragflow/
 docker build --platform linux/amd64 -f Dockerfile -t infiniflow/ragflow:nightly .
 ```
 
+Jika berada di belakang proxy, Anda dapat melewatkan argumen proxy:
+
+```bash
+docker build --platform linux/amd64 \
+  --build-arg http_proxy=http://YOUR_PROXY:PORT \
+  --build-arg https_proxy=http://YOUR_PROXY:PORT \
+  -f Dockerfile -t infiniflow/ragflow:nightly .
+```
+
 ## 🔨 Menjalankan Aplikasi dari untuk Pengembangan
 
 1. Instal `uv` dan `pre-commit`, atau lewati langkah ini jika sudah terinstal:
@@ -282,7 +298,7 @@ docker build --platform linux/amd64 -f Dockerfile -t infiniflow/ragflow:nightly
    ```bash
    git clone https://github.com/infiniflow/ragflow.git
    cd ragflow/
-   uv sync --python 3.10 # install RAGFlow dependent python modules
+   uv sync --python 3.12 # install RAGFlow dependent python modules
    uv run download_deps.py
    pre-commit install
    ```
@@ -352,7 +368,7 @@ docker build --platform linux/amd64 -f Dockerfile -t infiniflow/ragflow:nightly
 
 ## 📜 Roadmap
 
-Lihat [Roadmap RAGFlow 2025](https://github.com/infiniflow/ragflow/issues/4214)
+Lihat [Roadmap RAGFlow 2026](https://github.com/infiniflow/ragflow/issues/12241)
 
 ## 🏄 Komunitas
 

README_ja.md

@@ -22,7 +22,7 @@
         <img alt="Static Badge" src="https://img.shields.io/badge/Online-Demo-4e6b99">
     </a>
     <a href="https://hub.docker.com/r/infiniflow/ragflow" target="_blank">
-        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.21.1">
+        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.23.1">
     </a>
     <a href="https://github.com/infiniflow/ragflow/releases/latest">
         <img src="https://img.shields.io/github/v/release/infiniflow/ragflow?color=blue&label=Latest%20Release" alt="Latest Release">
@@ -37,7 +37,7 @@
 
 <h4 align="center">
   <a href="https://ragflow.io/docs/dev/">Document</a> |
-  <a href="https://github.com/infiniflow/ragflow/issues/4214">Roadmap</a> |
+  <a href="https://github.com/infiniflow/ragflow/issues/12241">Roadmap</a> |
   <a href="https://twitter.com/infiniflowai">Twitter</a> |
   <a href="https://discord.gg/NjYzJD3GM3">Discord</a> |
   <a href="https://demo.ragflow.io">Demo</a>
@@ -66,6 +66,9 @@
 
 ## 🔥 最新情報
 
+- 2025-12-26 AIエージェントの「メモリ」機能をサポート。
+- 2025-11-19 Gemini 3 Proをサポートしています。
+- 2025-11-12 Confluence、S3、Notion、Discord、Google Drive からのデータ同期をサポートします。
 - 2025-10-23 ドキュメント解析方法として MinerU と Docling をサポートします。
 - 2025-10-15 オーケストレーションされたデータパイプラインのサポート。
 - 2025-08-08 OpenAI の最新 GPT-5 シリーズモデルをサポートします。
@@ -73,7 +76,6 @@
 - 2025-05-23 エージェントに Python/JS コードエグゼキュータコンポーネントを追加しました。
 - 2025-05-05 言語間クエリをサポートしました。
 - 2025-03-19 PDFまたはDOCXファイル内の画像を理解するために、多モーダルモデルを使用することをサポートします。
-- 2025-02-28 インターネット検索 (TAVILY) と組み合わせて、あらゆる LLM の詳細な調査をサポートします。
 - 2024-12-18 DeepDoc のドキュメント レイアウト分析モデルをアップグレードします。
 - 2024-08-22 RAG を介して SQL ステートメントへのテキストをサポートします。
 
@@ -166,28 +168,34 @@
 > 現在、公式に提供されているすべての Docker イメージは x86 アーキテクチャ向けにビルドされており、ARM64 用の Docker イメージは提供されていません。
 > ARM64 アーキテクチャのオペレーティングシステムを使用している場合は、[このドキュメント](https://ragflow.io/docs/dev/build_docker_image)を参照して Docker イメージを自分でビルドしてください。
 
-> 以下のコマンドは、RAGFlow Docker イメージの v0.21.1 エディションをダウンロードします。異なる RAGFlow エディションの説明については、以下の表を参照してください。v0.21.1 とは異なるエディションをダウンロードするには、docker/.env ファイルの RAGFLOW_IMAGE 変数を適宜更新し、docker compose を使用してサーバーを起動してください。
+> 以下のコマンドは、RAGFlow Docker イメージの v0.23.1 エディションをダウンロードします。異なる RAGFlow エディションの説明については、以下の表を参照してください。v0.23.1 とは異なるエディションをダウンロードするには、docker/.env ファイルの RAGFLOW_IMAGE 変数を適宜更新し、docker compose を使用してサーバーを起動してください。
 
 ```bash
    $ cd ragflow/docker
-   # Use CPU for embedding and DeepDoc tasks:
+
+   # git checkout v0.23.1
+   # 任意: 安定版タグを利用 (一覧: https://github.com/infiniflow/ragflow/releases)
+   # この手順は、コード内の entrypoint.sh ファイルが Docker イメージのバージョンと一致していることを確認します。
+
+   # Use CPU for DeepDoc tasks:
    $ docker compose -f docker-compose.yml up -d
 
-   # To use GPU to accelerate embedding and DeepDoc tasks:
+   # To use GPU to accelerate DeepDoc tasks:
    # sed -i '1i DEVICE=gpu' .env
    # docker compose -f docker-compose.yml up -d
 ```
 
-| RAGFlow image tag | Image size (GB) | Has embedding models? | Stable?                    |
-| ----------------- | --------------- | --------------------- | -------------------------- |
-| v0.21.1           | &approx;9       | ✔️                  | Stable release             |
-| v0.21.1-slim      | &approx;2       | ❌                    | Stable release             |
-| nightly           | &approx;2       | ❌                    | _Unstable_ nightly build |
+> 注意：`v0.22.0` より前のバージョンでは、embedding モデルを含むイメージと、embedding モデルを含まない slim イメージの両方を提供していました。詳細は以下の通りです：
 
-> 注意：`v0.22.0` 以降、当プロジェクトでは slim エディションのみを提供し、イメージタグに **-slim** サフィックスを付けなくなりました。
+| RAGFlow image tag | Image size (GB) | Has embedding models? | Stable?        |
+|-------------------|-----------------|-----------------------|----------------|
+| v0.21.1           | &approx;9       | ✔️                    | Stable release |
+| v0.21.1-slim      | &approx;2       | ❌                     | Stable release |
 
-1. サーバーを立ち上げた後、サーバーの状態を確認する:
+> `v0.22.0` 以降、当プロジェクトでは slim エディションのみを提供し、イメージタグに **-slim** サフィックスを付けなくなりました。
 
+   1. サーバーを立ち上げた後、サーバーの状態を確認する:   
+   
    ```bash
    $ docker logs -f docker-ragflow-cpu-1
    ```
@@ -259,7 +267,7 @@ RAGFlow はデフォルトで Elasticsearch を使用して全文とベクトル
    > Linux/arm64 マシンでの Infinity への切り替えは正式にサポートされていません。
    >
 
-## 🔧 ソースコードで Docker イメージを作成（埋め込みモデルなし）
+## 🔧 ソースコードで Docker イメージを作成
 
 この Docker イメージのサイズは約 1GB で、外部の大モデルと埋め込みサービスに依存しています。
 
@@ -269,6 +277,15 @@ cd ragflow/
 docker build --platform linux/amd64 -f Dockerfile -t infiniflow/ragflow:nightly .
 ```
 
+プロキシ環境下にいる場合は、プロキシ引数を指定できます：
+
+```bash
+docker build --platform linux/amd64 \
+  --build-arg http_proxy=http://YOUR_PROXY:PORT \
+  --build-arg https_proxy=http://YOUR_PROXY:PORT \
+  -f Dockerfile -t infiniflow/ragflow:nightly .
+```
+
 ## 🔨 ソースコードからサービスを起動する方法
 
 1. `uv` と `pre-commit` をインストールする。すでにインストールされている場合は、このステップをスキップしてください:
@@ -281,7 +298,7 @@ docker build --platform linux/amd64 -f Dockerfile -t infiniflow/ragflow:nightly
    ```bash
    git clone https://github.com/infiniflow/ragflow.git
    cd ragflow/
-   uv sync --python 3.10 # install RAGFlow dependent python modules
+   uv sync --python 3.12 # install RAGFlow dependent python modules
    uv run download_deps.py
    pre-commit install
    ```
@@ -351,7 +368,7 @@ docker build --platform linux/amd64 -f Dockerfile -t infiniflow/ragflow:nightly
 
 ## 📜 ロードマップ
 
-[RAGFlow ロードマップ 2025](https://github.com/infiniflow/ragflow/issues/4214) を参照
+[RAGFlow ロードマップ 2026](https://github.com/infiniflow/ragflow/issues/12241) を参照
 
 ## 🏄 コミュニティ
 

README_ko.md

@@ -22,7 +22,7 @@
         <img alt="Static Badge" src="https://img.shields.io/badge/Online-Demo-4e6b99">
     </a>
     <a href="https://hub.docker.com/r/infiniflow/ragflow" target="_blank">
-        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.21.1">
+        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.23.1">
     </a>
     <a href="https://github.com/infiniflow/ragflow/releases/latest">
         <img src="https://img.shields.io/github/v/release/infiniflow/ragflow?color=blue&label=Latest%20Release" alt="Latest Release">
@@ -37,7 +37,7 @@
 
 <h4 align="center">
   <a href="https://ragflow.io/docs/dev/">Document</a> |
-  <a href="https://github.com/infiniflow/ragflow/issues/4214">Roadmap</a> |
+  <a href="https://github.com/infiniflow/ragflow/issues/12241">Roadmap</a> |
   <a href="https://twitter.com/infiniflowai">Twitter</a> |
   <a href="https://discord.gg/NjYzJD3GM3">Discord</a> |
   <a href="https://demo.ragflow.io">Demo</a>
@@ -67,6 +67,9 @@
 
 ## 🔥 업데이트
 
+- 2025-12-26 AI 에이전트의 '메모리' 기능 지원.
+- 2025-11-19 Gemini 3 Pro를 지원합니다.
+- 2025-11-12 Confluence, S3, Notion, Discord, Google Drive에서 데이터 동기화를 지원합니다.
 - 2025-10-23 문서 파싱 방법으로 MinerU 및 Docling을 지원합니다.
 - 2025-10-15 조정된 데이터 파이프라인 지원.
 - 2025-08-08 OpenAI의 최신 GPT-5 시리즈 모델을 지원합니다.
@@ -74,7 +77,6 @@
 - 2025-05-23 Agent에 Python/JS 코드 실행기 구성 요소를 추가합니다.
 - 2025-05-05 언어 간 쿼리를 지원합니다.
 - 2025-03-19 PDF 또는 DOCX 파일 내의 이미지를 이해하기 위해 다중 모드 모델을 사용하는 것을 지원합니다.
-- 2025-02-28 인터넷 검색(TAVILY)과 결합되어 모든 LLM에 대한 심층 연구를 지원합니다.
 - 2024-12-18 DeepDoc의 문서 레이아웃 분석 모델 업그레이드.
 - 2024-08-22 RAG를 통해 SQL 문에 텍스트를 지원합니다.
 
@@ -168,25 +170,31 @@
 > 모든 Docker 이미지는 x86 플랫폼을 위해 빌드되었습니다. 우리는 현재 ARM64 플랫폼을 위한 Docker 이미지를 제공하지 않습니다.
 > ARM64 플랫폼을 사용 중이라면, [시스템과 호환되는 Docker 이미지를 빌드하려면 이 가이드를 사용해 주세요](https://ragflow.io/docs/dev/build_docker_image).
 
-   > 아래 명령어는 RAGFlow Docker 이미지의 v0.21.1 버전을 다운로드합니다. 다양한 RAGFlow 버전에 대한 설명은 다음 표를 참조하십시오. v0.21.1과 다른 RAGFlow 버전을 다운로드하려면, docker/.env 파일에서 RAGFLOW_IMAGE 변수를 적절히 업데이트한 후 docker compose를 사용하여 서버를 시작하십시오.
+   > 아래 명령어는 RAGFlow Docker 이미지의 v0.23.1 버전을 다운로드합니다. 다양한 RAGFlow 버전에 대한 설명은 다음 표를 참조하십시오. v0.23.1과 다른 RAGFlow 버전을 다운로드하려면, docker/.env 파일에서 RAGFLOW_IMAGE 변수를 적절히 업데이트한 후 docker compose를 사용하여 서버를 시작하십시오.
 
    ```bash
    $ cd ragflow/docker
-   # Use CPU for embedding and DeepDoc tasks:
+   
+   # git checkout v0.23.1
+   # Optional: use a stable tag (see releases: https://github.com/infiniflow/ragflow/releases)
+   # 이 단계는 코드의 entrypoint.sh 파일이 Docker 이미지 버전과 일치하도록 보장합니다.
+
+   # Use CPU for DeepDoc tasks:
    $ docker compose -f docker-compose.yml up -d
 
-   # To use GPU to accelerate embedding and DeepDoc tasks:
+   # To use GPU to accelerate DeepDoc tasks:
    # sed -i '1i DEVICE=gpu' .env
    # docker compose -f docker-compose.yml up -d
-   ```
+```
 
-   | RAGFlow image tag | Image size (GB) | Has embedding models? | Stable?                  |
-   | ----------------- | --------------- | --------------------- | ------------------------ |
-   | v0.21.1           | &approx;9       | ✔️                    | Stable release           |
-   | v0.21.1-slim      | &approx;2       | ❌                    | Stable release           |
-   | nightly           | &approx;2       | ❌                    | _Unstable_ nightly build |
+> 참고: `v0.22.0` 이전 버전에서는 embedding 모델이 포함된 이미지와 embedding 모델이 포함되지 않은 slim 이미지를 모두 제공했습니다. 자세한 내용은 다음과 같습니다:
 
-   > 참고: `v0.22.0`부터는 slim 에디션만 배포하며 이미지 태그에 **-slim** 접미사를 더 이상 붙이지 않습니다.
+| RAGFlow image tag | Image size (GB) | Has embedding models? | Stable?        |
+|-------------------|-----------------|-----------------------|----------------|
+| v0.21.1           | &approx;9       | ✔️                    | Stable release |
+| v0.21.1-slim      | &approx;2       | ❌                     | Stable release |
+
+> `v0.22.0`부터는 slim 에디션만 배포하며 이미지 태그에 **-slim** 접미사를 더 이상 붙이지 않습니다.
 
 1. 서버가 시작된 후 서버 상태를 확인하세요:
 
@@ -206,7 +214,7 @@
     * Running on all addresses (0.0.0.0)
    ```
 
-   > 만약 확인 단계를 건너뛰고 바로 RAGFlow에 로그인하면, RAGFlow가 완전히 초기화되지 않았기 때문에 브라우저에서 `network anormal` 오류가 발생할 수 있습니다.
+   > 만약 확인 단계를 건너뛰고 바로 RAGFlow에 로그인하면, RAGFlow가 완전히 초기화되지 않았기 때문에 브라우저에서 `network abnormal` 오류가 발생할 수 있습니다.
 
 2. 웹 브라우저에 서버의 IP 주소를 입력하고 RAGFlow에 로그인하세요.
    > 기본 설정을 사용할 경우, `http://IP_OF_YOUR_MACHINE`만 입력하면 됩니다 (포트 번호는 제외). 기본 HTTP 서비스 포트 `80`은 기본 구성으로 사용할 때 생략할 수 있습니다.
@@ -253,7 +261,7 @@ RAGFlow 는 기본적으로 Elasticsearch 를 사용하여 전체 텍스트 및
    > [!WARNING]
    > Linux/arm64 시스템에서 Infinity로 전환하는 것은 공식적으로 지원되지 않습니다.
 
-## 🔧 소스 코드로 Docker 이미지를 컴파일합니다(임베딩 모델 포함하지 않음)
+## 🔧 소스 코드로 Docker 이미지를 컴파일합니다
 
 이 Docker 이미지의 크기는 약 1GB이며, 외부 대형 모델과 임베딩 서비스에 의존합니다.
 
@@ -263,6 +271,15 @@ cd ragflow/
 docker build --platform linux/amd64 -f Dockerfile -t infiniflow/ragflow:nightly .
 ```
 
+프록시 환경인 경우, 프록시 인수를 전달할 수 있습니다：
+
+```bash
+docker build --platform linux/amd64 \
+  --build-arg http_proxy=http://YOUR_PROXY:PORT \
+  --build-arg https_proxy=http://YOUR_PROXY:PORT \
+  -f Dockerfile -t infiniflow/ragflow:nightly .
+```
+
 ## 🔨 소스 코드로 서비스를 시작합니다.
 
 1. `uv` 와 `pre-commit` 을 설치하거나, 이미 설치된 경우 이 단계를 건너뜁니다:
@@ -276,7 +293,7 @@ docker build --platform linux/amd64 -f Dockerfile -t infiniflow/ragflow:nightly
    ```bash
    git clone https://github.com/infiniflow/ragflow.git
    cd ragflow/
-   uv sync --python 3.10 # install RAGFlow dependent python modules
+   uv sync --python 3.12 # install RAGFlow dependent python modules
    uv run download_deps.py
    pre-commit install
    ```
@@ -355,7 +372,7 @@ docker build --platform linux/amd64 -f Dockerfile -t infiniflow/ragflow:nightly
 
 ## 📜 로드맵
 
-[RAGFlow 로드맵 2025](https://github.com/infiniflow/ragflow/issues/4214)을 확인하세요.
+[RAGFlow 로드맵 2026](https://github.com/infiniflow/ragflow/issues/12241)을 확인하세요.
 
 ## 🏄 커뮤니티
 

README_pt_br.md

@@ -22,7 +22,7 @@
         <img alt="Badge Estático" src="https://img.shields.io/badge/Online-Demo-4e6b99">
     </a>
     <a href="https://hub.docker.com/r/infiniflow/ragflow" target="_blank">
-        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.21.1">
+        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.23.1">
     </a>
     <a href="https://github.com/infiniflow/ragflow/releases/latest">
         <img src="https://img.shields.io/github/v/release/infiniflow/ragflow?color=blue&label=Última%20Relese" alt="Última Versão">
@@ -37,7 +37,7 @@
 
 <h4 align="center">
   <a href="https://ragflow.io/docs/dev/">Documentação</a> |
-  <a href="https://github.com/infiniflow/ragflow/issues/4214">Roadmap</a> |
+  <a href="https://github.com/infiniflow/ragflow/issues/12241">Roadmap</a> |
   <a href="https://twitter.com/infiniflowai">Twitter</a> |
   <a href="https://discord.gg/NjYzJD3GM3">Discord</a> |
   <a href="https://demo.ragflow.io">Demo</a>
@@ -86,6 +86,9 @@ Experimente nossa demo em [https://demo.ragflow.io](https://demo.ragflow.io).
 
 ## 🔥 Últimas Atualizações
 
+- 26-12-2025 Suporte à função 'Memória' para agentes de IA.
+- 19-11-2025 Suporta Gemini 3 Pro.
+- 12-11-2025 Suporta a sincronização de dados do Confluence, S3, Notion, Discord e Google Drive.
 - 23-10-2025 Suporta MinerU e Docling como métodos de análise de documentos.
 - 15-10-2025 Suporte para pipelines de dados orquestrados.
 - 08-08-2025 Suporta a mais recente série GPT-5 da OpenAI.
@@ -93,7 +96,6 @@ Experimente nossa demo em [https://demo.ragflow.io](https://demo.ragflow.io).
 - 23-05-2025 Adicione o componente executor de código Python/JS ao Agente.
 - 05-05-2025 Suporte a consultas entre idiomas.
 - 19-03-2025 Suporta o uso de um modelo multi-modal para entender imagens dentro de arquivos PDF ou DOCX.
-- 28-02-2025 combinado com a pesquisa na Internet (T AVI LY), suporta pesquisas profundas para qualquer LLM.
 - 18-12-2024 Atualiza o modelo de Análise de Layout de Documentos no DeepDoc.
 - 22-08-2024 Suporta conversão de texto para comandos SQL via RAG.
 
@@ -186,25 +188,31 @@ Experimente nossa demo em [https://demo.ragflow.io](https://demo.ragflow.io).
 > Todas as imagens Docker são construídas para plataformas x86. Atualmente, não oferecemos imagens Docker para ARM64.
 > Se você estiver usando uma plataforma ARM64, por favor, utilize [este guia](https://ragflow.io/docs/dev/build_docker_image) para construir uma imagem Docker compatível com o seu sistema.
 
-    > O comando abaixo baixa a edição`v0.21.1` da imagem Docker do RAGFlow. Consulte a tabela a seguir para descrições de diferentes edições do RAGFlow. Para baixar uma edição do RAGFlow diferente da `v0.21.1`, atualize a variável `RAGFLOW_IMAGE` conforme necessário no **docker/.env** antes de usar `docker compose` para iniciar o servidor.
+    > O comando abaixo baixa a edição`v0.23.1` da imagem Docker do RAGFlow. Consulte a tabela a seguir para descrições de diferentes edições do RAGFlow. Para baixar uma edição do RAGFlow diferente da `v0.23.1`, atualize a variável `RAGFLOW_IMAGE` conforme necessário no **docker/.env** antes de usar `docker compose` para iniciar o servidor.
 
 ```bash
    $ cd ragflow/docker
-   # Use CPU for embedding and DeepDoc tasks:
+   
+   # git checkout v0.23.1
+   # Opcional: use uma tag estável (veja releases: https://github.com/infiniflow/ragflow/releases)
+   # Esta etapa garante que o arquivo entrypoint.sh no código corresponda à versão da imagem do Docker.
+
+   # Use CPU for DeepDoc tasks:
    $ docker compose -f docker-compose.yml up -d
 
-   # To use GPU to accelerate embedding and DeepDoc tasks:
+   # To use GPU to accelerate DeepDoc tasks:
    # sed -i '1i DEVICE=gpu' .env
    # docker compose -f docker-compose.yml up -d
 ```
 
-| Tag da imagem RAGFlow | Tamanho da imagem (GB) | Possui modelos de incorporação? | Estável?                      |
-| --------------------- | ---------------------- | --------------------------------- | ------------------------------ |
-| v0.21.1               | &approx;9              | ✔️                              | Lançamento estável           |
-| v0.21.1-slim          | &approx;2              | ❌                                | Lançamento estável           |
-| nightly               | &approx;2              | ❌                                | Construção noturna instável |
+> Nota: Antes da `v0.22.0`, fornecíamos imagens com modelos de embedding e imagens slim sem modelos de embedding. Detalhes a seguir:
 
-    > Observação: A partir da`v0.22.0`, distribuímos apenas a edição slim e não adicionamos mais o sufixo **-slim** às tags das imagens.
+| RAGFlow image tag | Image size (GB) | Has embedding models? | Stable?        |
+|-------------------|-----------------|-----------------------|----------------|
+| v0.21.1           | &approx;9       | ✔️                    | Stable release |
+| v0.21.1-slim      | &approx;2       | ❌                     | Stable release |
+
+> A partir da `v0.22.0`, distribuímos apenas a edição slim e não adicionamos mais o sufixo **-slim** às tags das imagens.
 
 4. Verifique o status do servidor após tê-lo iniciado:
 
@@ -224,7 +232,7 @@ Experimente nossa demo em [https://demo.ragflow.io](https://demo.ragflow.io).
     * Rodando em todos os endereços (0.0.0.0)
    ```
 
-   > Se você pular essa etapa de confirmação e acessar diretamente o RAGFlow, seu navegador pode exibir um erro `network anormal`, pois, nesse momento, seu RAGFlow pode não estar totalmente inicializado.
+   > Se você pular essa etapa de confirmação e acessar diretamente o RAGFlow, seu navegador pode exibir um erro `network abnormal`, pois, nesse momento, seu RAGFlow pode não estar totalmente inicializado.
    >
 5. No seu navegador, insira o endereço IP do seu servidor e faça login no RAGFlow.
 
@@ -274,9 +282,9 @@ O RAGFlow usa o Elasticsearch por padrão para armazenar texto completo e vetore
    ```
 
 > [!ATENÇÃO]
-> A mudança para o Infinity em uma máquina Linux/arm64 ainda não é oficialmente suportada.
+ > A mudança para o Infinity em uma máquina Linux/arm64 ainda não é oficialmente suportada.
 
-## 🔧 Criar uma imagem Docker sem modelos de incorporação
+## 🔧 Criar uma imagem Docker
 
 Esta imagem tem cerca de 2 GB de tamanho e depende de serviços externos de LLM e incorporação.
 
@@ -286,6 +294,15 @@ cd ragflow/
 docker build --platform linux/amd64 -f Dockerfile -t infiniflow/ragflow:nightly .
 ```
 
+Se você estiver atrás de um proxy, pode passar argumentos de proxy:
+
+```bash
+docker build --platform linux/amd64 \
+  --build-arg http_proxy=http://YOUR_PROXY:PORT \
+  --build-arg https_proxy=http://YOUR_PROXY:PORT \
+  -f Dockerfile -t infiniflow/ragflow:nightly .
+```
+
 ## 🔨 Lançar o serviço a partir do código-fonte para desenvolvimento
 
 1. Instale o `uv` e o `pre-commit`, ou pule esta etapa se eles já estiverem instalados:
@@ -298,7 +315,7 @@ docker build --platform linux/amd64 -f Dockerfile -t infiniflow/ragflow:nightly
    ```bash
    git clone https://github.com/infiniflow/ragflow.git
    cd ragflow/
-   uv sync --python 3.10 # instala os módulos Python dependentes do RAGFlow
+   uv sync --python 3.12 # instala os módulos Python dependentes do RAGFlow
    uv run download_deps.py
    pre-commit install
    ```
@@ -368,7 +385,7 @@ docker build --platform linux/amd64 -f Dockerfile -t infiniflow/ragflow:nightly
 
 ## 📜 Roadmap
 
-Veja o [RAGFlow Roadmap 2025](https://github.com/infiniflow/ragflow/issues/4214)
+Veja o [RAGFlow Roadmap 2026](https://github.com/infiniflow/ragflow/issues/12241)
 
 ## 🏄 Comunidade
 

README_tzh.md

@@ -22,7 +22,7 @@
         <img alt="Static Badge" src="https://img.shields.io/badge/Online-Demo-4e6b99">
     </a>
     <a href="https://hub.docker.com/r/infiniflow/ragflow" target="_blank">
-        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.21.1">
+        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.23.1">
     </a>
     <a href="https://github.com/infiniflow/ragflow/releases/latest">
         <img src="https://img.shields.io/github/v/release/infiniflow/ragflow?color=blue&label=Latest%20Release" alt="Latest Release">
@@ -37,7 +37,7 @@
 
 <h4 align="center">
   <a href="https://ragflow.io/docs/dev/">Document</a> |
-  <a href="https://github.com/infiniflow/ragflow/issues/4214">Roadmap</a> |
+  <a href="https://github.com/infiniflow/ragflow/issues/12241">Roadmap</a> |
   <a href="https://twitter.com/infiniflowai">Twitter</a> |
   <a href="https://discord.gg/NjYzJD3GM3">Discord</a> |
   <a href="https://demo.ragflow.io">Demo</a>
@@ -85,14 +85,16 @@
 
 ## 🔥 近期更新
 
+- 2025-12-26 支援AI代理的「記憶」功能。
+- 2025-11-19 支援 Gemini 3 Pro。
+- 2025-11-12 支援從 Confluence、S3、Notion、Discord、Google Drive 進行資料同步。
 - 2025-10-23 支援 MinerU 和 Docling 作為文件解析方法。
 - 2025-10-15 支援可編排的資料管道。
 - 2025-08-08 支援 OpenAI 最新的 GPT-5 系列模型。
-- 2025-08-01 支援 agentic workflow 和 MCP
+- 2025-08-01 支援 agentic workflow 和 MCP。
 - 2025-05-23 為 Agent 新增 Python/JS 程式碼執行器元件。
 - 2025-05-05 支援跨語言查詢。
-- 2025-03-19 PDF和DOCX中的圖支持用多模態大模型去解析得到描述.
-- 2025-02-28 結合網路搜尋（Tavily），對於任意大模型實現類似 Deep Research 的推理功能.
+- 2025-03-19 PDF和DOCX中的圖支持用多模態大模型去解析得到描述。
 - 2024-12-18 升級了 DeepDoc 的文檔佈局分析模型。
 - 2024-08-22 支援用 RAG 技術實現從自然語言到 SQL 語句的轉換。
 
@@ -123,7 +125,7 @@
 
 ### 🍔 **相容各類異質資料來源**
 
-- 支援豐富的文件類型，包括 Word 文件、PPT、excel 表格、txt 檔案、圖片、PDF、影印件、影印件、結構化資料、網頁等。
+- 支援豐富的文件類型，包括 Word 文件、PPT、excel 表格、txt 檔案、圖片、PDF、影印件、複印件、結構化資料、網頁等。
 
 ### 🛀 **全程無憂、自動化的 RAG 工作流程**
 
@@ -185,25 +187,31 @@
 > 所有 Docker 映像檔都是為 x86 平台建置的。目前，我們不提供 ARM64 平台的 Docker 映像檔。
 > 如果您使用的是 ARM64 平台，請使用 [這份指南](https://ragflow.io/docs/dev/build_docker_image) 來建置適合您系統的 Docker 映像檔。
 
-> 執行以下指令會自動下載 RAGFlow slim Docker 映像 `v0.21.1`。請參考下表查看不同 Docker 發行版的說明。如需下載不同於 `v0.21.1` 的 Docker 映像，請在執行 `docker compose` 啟動服務之前先更新 **docker/.env** 檔案內的 `RAGFLOW_IMAGE` 變數。
+> 執行以下指令會自動下載 RAGFlow Docker 映像 `v0.23.1`。請參考下表查看不同 Docker 發行版的說明。如需下載不同於 `v0.23.1` 的 Docker 映像，請在執行 `docker compose` 啟動服務之前先更新 **docker/.env** 檔案內的 `RAGFLOW_IMAGE` 變數。
 
 ```bash
    $ cd ragflow/docker
-   # Use CPU for embedding and DeepDoc tasks:
+   
+   # git checkout v0.23.1
+   # 可選：使用穩定版標籤（查看發佈：https://github.com/infiniflow/ragflow/releases）
+   # 此步驟確保程式碼中的 entrypoint.sh 檔案與 Docker 映像版本一致。
+
+   # Use CPU for DeepDoc tasks:
    $ docker compose -f docker-compose.yml up -d
 
-   # To use GPU to accelerate embedding and DeepDoc tasks:
+   # To use GPU to accelerate DeepDoc tasks:
    # sed -i '1i DEVICE=gpu' .env
    # docker compose -f docker-compose.yml up -d
 ```
 
-| RAGFlow image tag | Image size (GB) | Has embedding models? | Stable?                    |
-| ----------------- | --------------- | --------------------- | -------------------------- |
-| v0.21.1           | &approx;9       | ✔️                  | Stable release             |
-| v0.21.1-slim      | &approx;2       | ❌                    | Stable release             |
-| nightly           | &approx;2       | ❌                    | _Unstable_ nightly build |
+> 注意：在 `v0.22.0` 之前的版本，我們會同時提供包含 embedding 模型的映像和不含 embedding 模型的 slim 映像。具體如下：
 
-> 注意：自 `v0.22.0` 起，我們僅發佈 slim 版本，並且不再在映像標籤後附加 **-slim** 後綴。
+| RAGFlow image tag | Image size (GB) | Has embedding models? | Stable?        |
+|-------------------|-----------------|-----------------------|----------------|
+| v0.21.1           | &approx;9       | ✔️                    | Stable release |
+| v0.21.1-slim      | &approx;2       | ❌                     | Stable release |
+
+> 從 `v0.22.0` 開始，我們只發佈 slim 版本，並且不再在映像標籤後附加 **-slim** 後綴。
 
 > [!TIP]
 > 如果你遇到 Docker 映像檔拉不下來的問題，可以在 **docker/.env** 檔案內根據變數 `RAGFLOW_IMAGE` 的註解提示選擇華為雲或阿里雲的對應映像。
@@ -229,7 +237,7 @@
     * Running on all addresses (0.0.0.0)
    ```
 
-   > 如果您跳過這一步驟系統確認步驟就登入 RAGFlow，你的瀏覽器有可能會提示 `network anormal` 或 `網路異常`，因為 RAGFlow 可能並未完全啟動成功。
+   > 如果您跳過這一步驟系統確認步驟就登入 RAGFlow，你的瀏覽器有可能會提示 `network abnormal` 或 `網路異常`，因為 RAGFlow 可能並未完全啟動成功。
    >
 5. 在你的瀏覽器中輸入你的伺服器對應的 IP 位址並登入 RAGFlow。
 
@@ -285,7 +293,7 @@ RAGFlow 預設使用 Elasticsearch 儲存文字和向量資料. 如果要切換
 > [!WARNING]
 > Infinity 目前官方並未正式支援在 Linux/arm64 架構下的機器上運行.
 
-## 🔧 原始碼編譯 Docker 映像（不含 embedding 模型）
+## 🔧 原始碼編譯 Docker 映像
 
 本 Docker 映像大小約 2 GB 左右並且依賴外部的大模型和 embedding 服務。
 
@@ -295,6 +303,15 @@ cd ragflow/
 docker build --platform linux/amd64 -f Dockerfile -t infiniflow/ragflow:nightly .
 ```
 
+若您位於代理環境，可傳遞代理參數：
+
+```bash
+docker build --platform linux/amd64 \
+  --build-arg http_proxy=http://YOUR_PROXY:PORT \
+  --build-arg https_proxy=http://YOUR_PROXY:PORT \
+  -f Dockerfile -t infiniflow/ragflow:nightly .
+```
+
 ## 🔨 以原始碼啟動服務
 
 1. 安裝 `uv` 和 `pre-commit`。如已安裝，可跳過此步驟：
@@ -308,7 +325,7 @@ docker build --platform linux/amd64 -f Dockerfile -t infiniflow/ragflow:nightly
    ```bash
    git clone https://github.com/infiniflow/ragflow.git
    cd ragflow/
-   uv sync --python 3.10 # install RAGFlow dependent python modules
+   uv sync --python 3.12 # install RAGFlow dependent python modules
    uv run download_deps.py
    pre-commit install
    ```
@@ -382,7 +399,7 @@ docker build --platform linux/amd64 -f Dockerfile -t infiniflow/ragflow:nightly
 
 ## 📜 路線圖
 
-詳見 [RAGFlow Roadmap 2025](https://github.com/infiniflow/ragflow/issues/4214) 。
+詳見 [RAGFlow Roadmap 2026](https://github.com/infiniflow/ragflow/issues/12241) 。
 
 ## 🏄 開源社群
 

README_zh.md

@@ -22,7 +22,7 @@
         <img alt="Static Badge" src="https://img.shields.io/badge/Online-Demo-4e6b99">
     </a>
     <a href="https://hub.docker.com/r/infiniflow/ragflow" target="_blank">
-        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.21.1">
+        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.23.1">
     </a>
     <a href="https://github.com/infiniflow/ragflow/releases/latest">
         <img src="https://img.shields.io/github/v/release/infiniflow/ragflow?color=blue&label=Latest%20Release" alt="Latest Release">
@@ -37,7 +37,7 @@
 
 <h4 align="center">
   <a href="https://ragflow.io/docs/dev/">Document</a> |
-  <a href="https://github.com/infiniflow/ragflow/issues/4214">Roadmap</a> |
+  <a href="https://github.com/infiniflow/ragflow/issues/12241">Roadmap</a> |
   <a href="https://twitter.com/infiniflowai">Twitter</a> |
   <a href="https://discord.gg/NjYzJD3GM3">Discord</a> |
   <a href="https://demo.ragflow.io">Demo</a>
@@ -85,14 +85,16 @@
 
 ## 🔥 近期更新
 
+- 2025-12-26 支持AI代理的“记忆”功能。
+- 2025-11-19 支持 Gemini 3 Pro。
+- 2025-11-12 支持从 Confluence、S3、Notion、Discord、Google Drive 进行数据同步。
 - 2025-10-23 支持 MinerU 和 Docling 作为文档解析方法。
 - 2025-10-15 支持可编排的数据管道。
 - 2025-08-08 支持 OpenAI 最新的 GPT-5 系列模型。
 - 2025-08-01 支持 agentic workflow 和 MCP。
 - 2025-05-23 Agent 新增 Python/JS 代码执行器组件。
 - 2025-05-05 支持跨语言查询。
-- 2025-03-19 PDF 和 DOCX 中的图支持用多模态大模型去解析得到描述.
-- 2025-02-28 结合互联网搜索（Tavily），对于任意大模型实现类似 Deep Research 的推理功能.
+- 2025-03-19 PDF 和 DOCX 中的图支持用多模态大模型去解析得到描述。
 - 2024-12-18 升级了 DeepDoc 的文档布局分析模型。
 - 2024-08-22 支持用 RAG 技术实现从自然语言到 SQL 语句的转换。
 
@@ -186,25 +188,31 @@
 > 请注意，目前官方提供的所有 Docker 镜像均基于 x86 架构构建，并不提供基于 ARM64 的 Docker 镜像。
 > 如果你的操作系统是 ARM64 架构，请参考[这篇文档](https://ragflow.io/docs/dev/build_docker_image)自行构建 Docker 镜像。
 
-   > 运行以下命令会自动下载 RAGFlow slim Docker 镜像 `v0.21.1`。请参考下表查看不同 Docker 发行版的描述。如需下载不同于 `v0.21.1` 的 Docker 镜像，请在运行 `docker compose` 启动服务之前先更新 **docker/.env** 文件内的 `RAGFLOW_IMAGE` 变量。
+   > 运行以下命令会自动下载 RAGFlow Docker 镜像 `v0.23.1`。请参考下表查看不同 Docker 发行版的描述。如需下载不同于 `v0.23.1` 的 Docker 镜像，请在运行 `docker compose` 启动服务之前先更新 **docker/.env** 文件内的 `RAGFLOW_IMAGE` 变量。
 
    ```bash
    $ cd ragflow/docker
-   # Use CPU for embedding and DeepDoc tasks:
+   
+   # git checkout v0.23.1
+   # 可选：使用稳定版本标签（查看发布：https://github.com/infiniflow/ragflow/releases）
+   # 这一步确保代码中的 entrypoint.sh 文件与 Docker 镜像的版本保持一致。
+
+   # Use CPU for DeepDoc tasks:
    $ docker compose -f docker-compose.yml up -d
 
-   # To use GPU to accelerate embedding and DeepDoc tasks:
+   # To use GPU to accelerate DeepDoc tasks:
    # sed -i '1i DEVICE=gpu' .env
    # docker compose -f docker-compose.yml up -d
    ```
+   
+   > 注意：在 `v0.22.0` 之前的版本，我们会同时提供包含 embedding 模型的镜像和不含 embedding 模型的 slim 镜像。具体如下：
 
-   | RAGFlow image tag | Image size (GB) | Has embedding models? | Stable?                  |
-   | ----------------- | --------------- | --------------------- | ------------------------ |
-   | v0.21.1           | &approx;9       | ✔️                    | Stable release           |
-   | v0.21.1-slim      | &approx;2       | ❌                    | Stable release           |
-   | nightly           | &approx;2       | ❌                    | _Unstable_ nightly build |
+   | RAGFlow image tag | Image size (GB) | Has embedding models? | Stable?        |
+   |-------------------|-----------------|-----------------------|----------------|
+   | v0.21.1           | &approx;9       | ✔️                    | Stable release |
+   | v0.21.1-slim      | &approx;2       | ❌                     | Stable release |
 
-   > 注意：从 `v0.22.0` 开始，我们只发布 slim 版本，并且不再在镜像标签后附加 **-slim** 后缀。
+   > 从 `v0.22.0` 开始，我们只发布 slim 版本，并且不再在镜像标签后附加 **-slim** 后缀。
 
    > [!TIP]
    > 如果你遇到 Docker 镜像拉不下来的问题，可以在 **docker/.env** 文件内根据变量 `RAGFLOW_IMAGE` 的注释提示选择华为云或者阿里云的相应镜像。
@@ -230,7 +238,7 @@
     * Running on all addresses (0.0.0.0)
    ```
 
-   > 如果您在没有看到上面的提示信息出来之前，就尝试登录 RAGFlow，你的浏览器有可能会提示 `network anormal` 或 `网络异常`。
+   > 如果您在没有看到上面的提示信息出来之前，就尝试登录 RAGFlow，你的浏览器有可能会提示 `network abnormal` 或 `网络异常`。
 
 5. 在你的浏览器中输入你的服务器对应的 IP 地址并登录 RAGFlow。
    > 上面这个例子中，您只需输入 http://IP_OF_YOUR_MACHINE 即可：未改动过配置则无需输入端口（默认的 HTTP 服务端口 80）。
@@ -284,7 +292,7 @@ RAGFlow 默认使用 Elasticsearch 存储文本和向量数据. 如果要切换
 > [!WARNING]
 > Infinity 目前官方并未正式支持在 Linux/arm64 架构下的机器上运行.
 
-## 🔧 源码编译 Docker 镜像（不含 embedding 模型）
+## 🔧 源码编译 Docker 镜像
 
 本 Docker 镜像大小约 2 GB 左右并且依赖外部的大模型和 embedding 服务。
 
@@ -294,6 +302,15 @@ cd ragflow/
 docker build --platform linux/amd64 -f Dockerfile -t infiniflow/ragflow:nightly .
 ```
 
+如果您处在代理环境下，可以传递代理参数：
+
+```bash
+docker build --platform linux/amd64 \
+  --build-arg http_proxy=http://YOUR_PROXY:PORT \
+  --build-arg https_proxy=http://YOUR_PROXY:PORT \
+  -f Dockerfile -t infiniflow/ragflow:nightly .
+```
+
 ## 🔨 以源代码启动服务
 
 1. 安装 `uv` 和 `pre-commit`。如已经安装，可跳过本步骤：
@@ -308,7 +325,7 @@ docker build --platform linux/amd64 -f Dockerfile -t infiniflow/ragflow:nightly
    ```bash
    git clone https://github.com/infiniflow/ragflow.git
    cd ragflow/
-   uv sync --python 3.10 # install RAGFlow dependent python modules
+   uv sync --python 3.12 # install RAGFlow dependent python modules
    uv run download_deps.py
    pre-commit install
    ```
@@ -385,7 +402,7 @@ docker build --platform linux/amd64 -f Dockerfile -t infiniflow/ragflow:nightly
 
 ## 📜 路线图
 
-详见 [RAGFlow Roadmap 2025](https://github.com/infiniflow/ragflow/issues/4214) 。
+详见 [RAGFlow Roadmap 2026](https://github.com/infiniflow/ragflow/issues/12241) 。
 
 ## 🏄 开源社区
 

SECURITY.md

@@ -6,8 +6,8 @@ Use this section to tell people about which versions of your project are
 currently being supported with security updates.
 
 | Version | Supported          |
-| ------- | ------------------ |
-| <=0.7.0   | :white_check_mark: |
+|---------|--------------------|
+| <=0.7.0 | :white_check_mark: |
 
 ## Reporting a Vulnerability
 

admin/client/README.md

@@ -4,7 +4,7 @@
 
 Admin Service is a dedicated management component designed to monitor, maintain, and administrate the RAGFlow system. It provides comprehensive tools for ensuring system stability, performing operational tasks, and managing users and permissions efficiently.
 
-The service offers real-time monitoring of critical components, including the RAGFlow server, Task Executor processes, and dependent services such as MySQL, Elasticsearch, Redis, and MinIO. It automatically checks their health status, resource usage, and uptime, and performs restarts in case of failures to minimize downtime.
+The service offers real-time monitoring of critical components, including the RAGFlow server, Task Executor processes, and dependent services such as MySQL, Infinity, Elasticsearch, Redis, and MinIO. It automatically checks their health status, resource usage, and uptime, and performs restarts in case of failures to minimize downtime.
 
 For user and system management, it supports listing, creating, modifying, and deleting users and their associated resources like knowledge bases and Agents.
 
@@ -48,7 +48,7 @@ It consists of a server-side Service and a command-line client (CLI), both imple
 1.  Ensure the Admin Service is running.
 2.  Install ragflow-cli.
     ```bash
-    pip install ragflow-cli==0.21.1
+    pip install ragflow-cli==0.23.1
     ```
 3.  Launch the CLI client:
     ```bash

admin/client/admin_client.py

@@ -16,13 +16,14 @@
 
 import argparse
 import base64
+import getpass
 from cmd import Cmd
+from typing import Any, Dict, List
 
-from Cryptodome.PublicKey import RSA
-from Cryptodome.Cipher import PKCS1_v1_5 as Cipher_pkcs1_v1_5
-from typing import Dict, List, Any
-from lark import Lark, Transformer, Tree
 import requests
+from Cryptodome.Cipher import PKCS1_v1_5 as Cipher_pkcs1_v1_5
+from Cryptodome.PublicKey import RSA
+from lark import Lark, Transformer, Tree
 
 GRAMMAR = r"""
 start: command
@@ -51,6 +52,7 @@ sql_command: list_services
            | revoke_permission
            | alter_user_role
            | show_user_permission
+           | show_version
 
 // meta command definition
 meta_command: "\\" meta_command_name [meta_args]
@@ -92,6 +94,7 @@ FOR: "FOR"i
 RESOURCES: "RESOURCES"i
 ON: "ON"i
 SET: "SET"i
+VERSION: "VERSION"i
 
 list_services: LIST SERVICES ";"
 show_service: SHOW SERVICE NUMBER ";"
@@ -120,6 +123,8 @@ revoke_permission: REVOKE action_list ON identifier FROM ROLE identifier ";"
 alter_user_role: ALTER USER quoted_string SET ROLE identifier ";"
 show_user_permission: SHOW USER PERMISSION quoted_string ";"
 
+show_version: SHOW VERSION ";"
+
 action_list: identifier ("," identifier)*
 
 identifier: WORD
@@ -136,7 +141,6 @@ NUMBER: /[0-9]+/
 
 
 class AdminTransformer(Transformer):
-
     def start(self, items):
         return items[0]
 
@@ -144,7 +148,7 @@ class AdminTransformer(Transformer):
         return items[0]
 
     def list_services(self, items):
-        result = {'type': 'list_services'}
+        result = {"type": "list_services"}
         return result
 
     def show_service(self, items):
@@ -231,11 +235,7 @@ class AdminTransformer(Transformer):
         action_list = items[1]
         resource = items[3]
         role_name = items[6]
-        return {
-            "type": "revoke_permission",
-            "role_name": role_name,
-            "resource": resource, "actions": action_list
-        }
+        return {"type": "revoke_permission", "role_name": role_name, "resource": resource, "actions": action_list}
 
     def alter_user_role(self, items):
         user_name = items[2]
@@ -246,6 +246,9 @@ class AdminTransformer(Transformer):
         user_name = items[3]
         return {"type": "show_user_permission", "user_name": user_name}
 
+    def show_version(self, items):
+        return {"type": "show_version"}
+
     def action_list(self, items):
         return items
 
@@ -256,12 +259,12 @@ class AdminTransformer(Transformer):
         # handle quoted parameter
         parsed_args = []
         for arg in args:
-            if hasattr(arg, 'value'):
+            if hasattr(arg, "value"):
                 parsed_args.append(arg.value)
             else:
                 parsed_args.append(str(arg))
 
-        return {'type': 'meta', 'command': command_name, 'args': parsed_args}
+        return {"type": "meta", "command": command_name, "args": parsed_args}
 
     def meta_command_name(self, items):
         return items[0]
@@ -271,22 +274,22 @@ class AdminTransformer(Transformer):
 
 
 def encrypt(input_string):
-    pub = '-----BEGIN PUBLIC KEY-----\nMIIBIjANBgkqhkiG9w0BAQEFAAOCAQ8AMIIBCgKCAQEArq9XTUSeYr2+N1h3Afl/z8Dse/2yD0ZGrKwx+EEEcdsBLca9Ynmx3nIB5obmLlSfmskLpBo0UACBmB5rEjBp2Q2f3AG3Hjd4B+gNCG6BDaawuDlgANIhGnaTLrIqWrrcm4EMzJOnAOI1fgzJRsOOUEfaS318Eq9OVO3apEyCCt0lOQK6PuksduOjVxtltDav+guVAA068NrPYmRNabVKRNLJpL8w4D44sfth5RvZ3q9t+6RTArpEtc5sh5ChzvqPOzKGMXW83C95TxmXqpbK6olN4RevSfVjEAgCydH6HN6OhtOQEcnrU97r9H0iZOWwbw3pVrZiUkuRD1R56Wzs2wIDAQAB\n-----END PUBLIC KEY-----'
+    pub = "-----BEGIN PUBLIC KEY-----\nMIIBIjANBgkqhkiG9w0BAQEFAAOCAQ8AMIIBCgKCAQEArq9XTUSeYr2+N1h3Afl/z8Dse/2yD0ZGrKwx+EEEcdsBLca9Ynmx3nIB5obmLlSfmskLpBo0UACBmB5rEjBp2Q2f3AG3Hjd4B+gNCG6BDaawuDlgANIhGnaTLrIqWrrcm4EMzJOnAOI1fgzJRsOOUEfaS318Eq9OVO3apEyCCt0lOQK6PuksduOjVxtltDav+guVAA068NrPYmRNabVKRNLJpL8w4D44sfth5RvZ3q9t+6RTArpEtc5sh5ChzvqPOzKGMXW83C95TxmXqpbK6olN4RevSfVjEAgCydH6HN6OhtOQEcnrU97r9H0iZOWwbw3pVrZiUkuRD1R56Wzs2wIDAQAB\n-----END PUBLIC KEY-----"
     pub_key = RSA.importKey(pub)
     cipher = Cipher_pkcs1_v1_5.new(pub_key)
-    cipher_text = cipher.encrypt(base64.b64encode(input_string.encode('utf-8')))
+    cipher_text = cipher.encrypt(base64.b64encode(input_string.encode("utf-8")))
     return base64.b64encode(cipher_text).decode("utf-8")
 
 
 def encode_to_base64(input_string):
-    base64_encoded = base64.b64encode(input_string.encode('utf-8'))
-    return base64_encoded.decode('utf-8')
+    base64_encoded = base64.b64encode(input_string.encode("utf-8"))
+    return base64_encoded.decode("utf-8")
 
 
 class AdminCLI(Cmd):
     def __init__(self):
         super().__init__()
-        self.parser = Lark(GRAMMAR, start='start', parser='lalr', transformer=AdminTransformer())
+        self.parser = Lark(GRAMMAR, start="start", parser="lalr", transformer=AdminTransformer())
         self.command_history = []
         self.is_interactive = False
         self.admin_account = "admin@ragflow.io"
@@ -304,7 +307,7 @@ class AdminCLI(Cmd):
             result = self.parse_command(command)
 
             if isinstance(result, dict):
-                if 'type' in result and result.get('type') == 'empty':
+                if "type" in result and result.get("type") == "empty":
                     return False
 
             self.execute_command(result)
@@ -312,7 +315,7 @@ class AdminCLI(Cmd):
             if isinstance(result, Tree):
                 return False
 
-            if result.get('type') == 'meta' and result.get('command') in ['q', 'quit', 'exit']:
+            if result.get("type") == "meta" and result.get("command") in ["q", "quit", "exit"]:
                 return True
 
         except KeyboardInterrupt:
@@ -330,7 +333,7 @@ class AdminCLI(Cmd):
 
     def parse_command(self, command_str: str) -> dict[str, str]:
         if not command_str.strip():
-            return {'type': 'empty'}
+            return {"type": "empty"}
 
         self.command_history.append(command_str)
 
@@ -338,12 +341,12 @@ class AdminCLI(Cmd):
             result = self.parser.parse(command_str)
             return result
         except Exception as e:
-            return {'type': 'error', 'message': f'Parse error: {str(e)}'}
+            return {"type": "error", "message": f"Parse error: {str(e)}"}
 
     def verify_admin(self, arguments: dict, single_command: bool):
-        self.host = arguments['host']
-        self.port = arguments['port']
-        print(f"Attempt to access ip: {self.host}, port: {self.port}")
+        self.host = arguments["host"]
+        self.port = arguments["port"]
+        print("Attempt to access server for admin login")
         url = f"http://{self.host}:{self.port}/api/v1/admin/login"
 
         attempt_count = 3
@@ -357,32 +360,49 @@ class AdminCLI(Cmd):
                 return False
 
             if single_command:
-                admin_passwd = arguments['password']
+                admin_passwd = arguments["password"]
             else:
-                admin_passwd = input(f"password for {self.admin_account}: ").strip()
+                admin_passwd = getpass.getpass(f"password for {self.admin_account}: ").strip()
             try:
                 self.admin_password = encrypt(admin_passwd)
-                response = self.session.post(url, json={'email': self.admin_account, 'password': self.admin_password})
+                response = self.session.post(url, json={"email": self.admin_account, "password": self.admin_password})
                 if response.status_code == 200:
                     res_json = response.json()
-                    error_code = res_json.get('code', -1)
+                    error_code = res_json.get("code", -1)
                     if error_code == 0:
-                        self.session.headers.update({
-                            'Content-Type': 'application/json',
-                            'Authorization': response.headers['Authorization'],
-                            'User-Agent': 'RAGFlow-CLI/0.21.1'
-                        })
+                        self.session.headers.update({"Content-Type": "application/json", "Authorization": response.headers["Authorization"], "User-Agent": "RAGFlow-CLI/0.23.1"})
                         print("Authentication successful.")
                         return True
                     else:
-                        error_message = res_json.get('message', 'Unknown error')
+                        error_message = res_json.get("message", "Unknown error")
                         print(f"Authentication failed: {error_message}, try again")
                         continue
                 else:
                     print(f"Bad response，status: {response.status_code}, password is wrong")
             except Exception as e:
                 print(str(e))
-                print(f"Can't access {self.host}, port: {self.port}")
+                print("Can't access server for admin login (connection failed)")
+
+    def _format_service_detail_table(self, data):
+        if isinstance(data, list):
+            return data
+        if not all([isinstance(v, list) for v in data.values()]):
+            # normal table
+            return data
+        # handle task_executor heartbeats map, for example {'name': [{'done': 2, 'now': timestamp1}, {'done': 3, 'now': timestamp2}]
+        task_executor_list = []
+        for k, v in data.items():
+            # display latest status
+            heartbeats = sorted(v, key=lambda x: x["now"], reverse=True)
+            task_executor_list.append(
+                {
+                    "task_executor_name": k,
+                    **heartbeats[0],
+                }
+                if heartbeats
+                else {"task_executor_name": k}
+            )
+        return task_executor_list
 
     def _print_table_simple(self, data):
         if not data:
@@ -392,16 +412,12 @@ class AdminCLI(Cmd):
             # handle single row data
             data = [data]
 
-        columns = list(data[0].keys())
+        columns = list(set().union(*(d.keys() for d in data)))
+        columns.sort()
         col_widths = {}
 
         def get_string_width(text):
-            half_width_chars = (
-                " !\"#$%&'()*+,-./0123456789:;<=>?@"
-                "ABCDEFGHIJKLMNOPQRSTUVWXYZ[\\]^_`"
-                "abcdefghijklmnopqrstuvwxyz{|}~"
-                "\t\n\r"
-            )
+            half_width_chars = " !\"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\\]^_`abcdefghijklmnopqrstuvwxyz{|}~\t\n\r"
             width = 0
             for char in text:
                 if char in half_width_chars:
@@ -413,7 +429,7 @@ class AdminCLI(Cmd):
         for col in columns:
             max_width = get_string_width(str(col))
             for item in data:
-                value_len = get_string_width(str(item.get(col, '')))
+                value_len = get_string_width(str(item.get(col, "")))
                 if value_len > max_width:
                     max_width = value_len
             col_widths[col] = max(2, max_width)
@@ -431,16 +447,15 @@ class AdminCLI(Cmd):
         for item in data:
             row = "|"
             for col in columns:
-                value = str(item.get(col, ''))
+                value = str(item.get(col, ""))
                 if get_string_width(value) > col_widths[col]:
-                    value = value[:col_widths[col] - 3] + "..."
+                    value = value[: col_widths[col] - 3] + "..."
                 row += f" {value:<{col_widths[col] - (get_string_width(value) - len(value))}} |"
             print(row)
 
         print(separator)
 
     def run_interactive(self):
-
         self.is_interactive = True
         print("RAGFlow Admin command line interface - Type '\\?' for help, '\\q' to quit")
 
@@ -457,7 +472,7 @@ class AdminCLI(Cmd):
                 if isinstance(result, Tree):
                     continue
 
-                if result.get('type') == 'meta' and result.get('command') in ['q', 'quit', 'exit']:
+                if result.get("type") == "meta" and result.get("command") in ["q", "quit", "exit"]:
                     break
 
             except KeyboardInterrupt:
@@ -471,36 +486,30 @@ class AdminCLI(Cmd):
         self.execute_command(result)
 
     def parse_connection_args(self, args: List[str]) -> Dict[str, Any]:
-        parser = argparse.ArgumentParser(description='Admin CLI Client', add_help=False)
-        parser.add_argument('-h', '--host', default='localhost', help='Admin service host')
-        parser.add_argument('-p', '--port', type=int, default=9381, help='Admin service port')
-        parser.add_argument('-w', '--password', default='admin', type=str, help='Superuser password')
-        parser.add_argument('command', nargs='?', help='Single command')
+        parser = argparse.ArgumentParser(description="Admin CLI Client", add_help=False)
+        parser.add_argument("-h", "--host", default="localhost", help="Admin service host")
+        parser.add_argument("-p", "--port", type=int, default=9381, help="Admin service port")
+        parser.add_argument("-w", "--password", default="admin", type=str, help="Superuser password")
+        parser.add_argument("command", nargs="?", help="Single command")
         try:
             parsed_args, remaining_args = parser.parse_known_args(args)
             if remaining_args:
                 command = remaining_args[0]
-                return {
-                    'host': parsed_args.host,
-                    'port': parsed_args.port,
-                    'password': parsed_args.password,
-                    'command': command
-                }
+                return {"host": parsed_args.host, "port": parsed_args.port, "password": parsed_args.password, "command": command}
             else:
                 return {
-                    'host': parsed_args.host,
-                    'port': parsed_args.port,
+                    "host": parsed_args.host,
+                    "port": parsed_args.port,
                 }
         except SystemExit:
-            return {'error': 'Invalid connection arguments'}
+            return {"error": "Invalid connection arguments"}
 
     def execute_command(self, parsed_command: Dict[str, Any]):
-
         command_dict: dict
         if isinstance(parsed_command, Tree):
             command_dict = parsed_command.children[0]
         else:
-            if parsed_command['type'] == 'error':
+            if parsed_command["type"] == "error":
                 print(f"Error: {parsed_command['message']}")
                 return
             else:
@@ -508,54 +517,56 @@ class AdminCLI(Cmd):
 
         # print(f"Parsed command: {command_dict}")
 
-        command_type = command_dict['type']
+        command_type = command_dict["type"]
 
         match command_type:
-            case 'list_services':
+            case "list_services":
                 self._handle_list_services(command_dict)
-            case 'show_service':
+            case "show_service":
                 self._handle_show_service(command_dict)
-            case 'restart_service':
+            case "restart_service":
                 self._handle_restart_service(command_dict)
-            case 'shutdown_service':
+            case "shutdown_service":
                 self._handle_shutdown_service(command_dict)
-            case 'startup_service':
+            case "startup_service":
                 self._handle_startup_service(command_dict)
-            case 'list_users':
+            case "list_users":
                 self._handle_list_users(command_dict)
-            case 'show_user':
+            case "show_user":
                 self._handle_show_user(command_dict)
-            case 'drop_user':
+            case "drop_user":
                 self._handle_drop_user(command_dict)
-            case 'alter_user':
+            case "alter_user":
                 self._handle_alter_user(command_dict)
-            case 'create_user':
+            case "create_user":
                 self._handle_create_user(command_dict)
-            case 'activate_user':
+            case "activate_user":
                 self._handle_activate_user(command_dict)
-            case 'list_datasets':
+            case "list_datasets":
                 self._handle_list_datasets(command_dict)
-            case 'list_agents':
+            case "list_agents":
                 self._handle_list_agents(command_dict)
-            case 'create_role':
+            case "create_role":
                 self._create_role(command_dict)
-            case 'drop_role':
+            case "drop_role":
                 self._drop_role(command_dict)
-            case 'alter_role':
+            case "alter_role":
                 self._alter_role(command_dict)
-            case 'list_roles':
+            case "list_roles":
                 self._list_roles(command_dict)
-            case 'show_role':
+            case "show_role":
                 self._show_role(command_dict)
-            case 'grant_permission':
+            case "grant_permission":
                 self._grant_permission(command_dict)
-            case 'revoke_permission':
+            case "revoke_permission":
                 self._revoke_permission(command_dict)
-            case 'alter_user_role':
+            case "alter_user_role":
                 self._alter_user_role(command_dict)
-            case 'show_user_permission':
+            case "show_user_permission":
                 self._show_user_permission(command_dict)
-            case 'meta':
+            case "show_version":
+                self._show_version(command_dict)
+            case "meta":
                 self._handle_meta_command(command_dict)
             case _:
                 print(f"Command '{command_type}' would be executed with API")
@@ -563,74 +574,77 @@ class AdminCLI(Cmd):
     def _handle_list_services(self, command):
         print("Listing all services")
 
-        url = f'http://{self.host}:{self.port}/api/v1/admin/services'
+        url = f"http://{self.host}:{self.port}/api/v1/admin/services"
         response = self.session.get(url)
         res_json = response.json()
         if response.status_code == 200:
-            self._print_table_simple(res_json['data'])
+            self._print_table_simple(res_json["data"])
         else:
             print(f"Fail to get all services, code: {res_json['code']}, message: {res_json['message']}")
 
     def _handle_show_service(self, command):
-        service_id: int = command['number']
+        service_id: int = command["number"]
         print(f"Showing service: {service_id}")
 
-        url = f'http://{self.host}:{self.port}/api/v1/admin/services/{service_id}'
+        url = f"http://{self.host}:{self.port}/api/v1/admin/services/{service_id}"
         response = self.session.get(url)
         res_json = response.json()
         if response.status_code == 200:
-            res_data = res_json['data']
-            if 'status' in res_data and res_data['status'] == 'alive':
+            res_data = res_json["data"]
+            if "status" in res_data and res_data["status"] == "alive":
                 print(f"Service {res_data['service_name']} is alive, ")
-                if isinstance(res_data['message'], str):
-                    print(res_data['message'])
+                if isinstance(res_data["message"], str):
+                    print(res_data["message"])
                 else:
-                    self._print_table_simple(res_data['message'])
+                    data = self._format_service_detail_table(res_data["message"])
+                    self._print_table_simple(data)
             else:
                 print(f"Service {res_data['service_name']} is down, {res_data['message']}")
         else:
             print(f"Fail to show service, code: {res_json['code']}, message: {res_json['message']}")
 
     def _handle_restart_service(self, command):
-        service_id: int = command['number']
+        service_id: int = command["number"]
         print(f"Restart service {service_id}")
 
     def _handle_shutdown_service(self, command):
-        service_id: int = command['number']
+        service_id: int = command["number"]
         print(f"Shutdown service {service_id}")
 
     def _handle_startup_service(self, command):
-        service_id: int = command['number']
+        service_id: int = command["number"]
         print(f"Startup service {service_id}")
 
     def _handle_list_users(self, command):
         print("Listing all users")
 
-        url = f'http://{self.host}:{self.port}/api/v1/admin/users'
+        url = f"http://{self.host}:{self.port}/api/v1/admin/users"
         response = self.session.get(url)
         res_json = response.json()
         if response.status_code == 200:
-            self._print_table_simple(res_json['data'])
+            self._print_table_simple(res_json["data"])
         else:
             print(f"Fail to get all users, code: {res_json['code']}, message: {res_json['message']}")
 
     def _handle_show_user(self, command):
-        username_tree: Tree = command['user_name']
+        username_tree: Tree = command["user_name"]
         user_name: str = username_tree.children[0].strip("'\"")
         print(f"Showing user: {user_name}")
-        url = f'http://{self.host}:{self.port}/api/v1/admin/users/{user_name}'
+        url = f"http://{self.host}:{self.port}/api/v1/admin/users/{user_name}"
         response = self.session.get(url)
         res_json = response.json()
         if response.status_code == 200:
-            self._print_table_simple(res_json['data'])
+            table_data = res_json["data"]
+            table_data.pop("avatar")
+            self._print_table_simple(table_data)
         else:
             print(f"Fail to get user {user_name}, code: {res_json['code']}, message: {res_json['message']}")
 
     def _handle_drop_user(self, command):
-        username_tree: Tree = command['user_name']
+        username_tree: Tree = command["user_name"]
         user_name: str = username_tree.children[0].strip("'\"")
         print(f"Drop user: {user_name}")
-        url = f'http://{self.host}:{self.port}/api/v1/admin/users/{user_name}'
+        url = f"http://{self.host}:{self.port}/api/v1/admin/users/{user_name}"
         response = self.session.delete(url)
         res_json = response.json()
         if response.status_code == 200:
@@ -639,13 +653,13 @@ class AdminCLI(Cmd):
             print(f"Fail to drop user, code: {res_json['code']}, message: {res_json['message']}")
 
     def _handle_alter_user(self, command):
-        user_name_tree: Tree = command['user_name']
+        user_name_tree: Tree = command["user_name"]
         user_name: str = user_name_tree.children[0].strip("'\"")
-        password_tree: Tree = command['password']
+        password_tree: Tree = command["password"]
         password: str = password_tree.children[0].strip("'\"")
-        print(f"Alter user: {user_name}, password: {password}")
-        url = f'http://{self.host}:{self.port}/api/v1/admin/users/{user_name}/password'
-        response = self.session.put(url, json={'new_password': encrypt(password)})
+        print(f"Alter user: {user_name}, password: ******")
+        url = f"http://{self.host}:{self.port}/api/v1/admin/users/{user_name}/password"
+        response = self.session.put(url, json={"new_password": encrypt(password)})
         res_json = response.json()
         if response.status_code == 200:
             print(res_json["message"])
@@ -653,32 +667,29 @@ class AdminCLI(Cmd):
             print(f"Fail to alter password, code: {res_json['code']}, message: {res_json['message']}")
 
     def _handle_create_user(self, command):
-        user_name_tree: Tree = command['user_name']
+        user_name_tree: Tree = command["user_name"]
         user_name: str = user_name_tree.children[0].strip("'\"")
-        password_tree: Tree = command['password']
+        password_tree: Tree = command["password"]
         password: str = password_tree.children[0].strip("'\"")
-        role: str = command['role']
-        print(f"Create user: {user_name}, password: {password}, role: {role}")
-        url = f'http://{self.host}:{self.port}/api/v1/admin/users'
-        response = self.session.post(
-            url,
-            json={'user_name': user_name, 'password': encrypt(password), 'role': role}
-        )
+        role: str = command["role"]
+        print(f"Create user: {user_name}, password: ******, role: {role}")
+        url = f"http://{self.host}:{self.port}/api/v1/admin/users"
+        response = self.session.post(url, json={"user_name": user_name, "password": encrypt(password), "role": role})
         res_json = response.json()
         if response.status_code == 200:
-            self._print_table_simple(res_json['data'])
+            self._print_table_simple(res_json["data"])
         else:
             print(f"Fail to create user {user_name}, code: {res_json['code']}, message: {res_json['message']}")
 
     def _handle_activate_user(self, command):
-        user_name_tree: Tree = command['user_name']
+        user_name_tree: Tree = command["user_name"]
         user_name: str = user_name_tree.children[0].strip("'\"")
-        activate_tree: Tree = command['activate_status']
+        activate_tree: Tree = command["activate_status"]
         activate_status: str = activate_tree.children[0].strip("'\"")
-        if activate_status.lower() in ['on', 'off']:
+        if activate_status.lower() in ["on", "off"]:
             print(f"Alter user {user_name} activate status, turn {activate_status.lower()}.")
-            url = f'http://{self.host}:{self.port}/api/v1/admin/users/{user_name}/activate'
-            response = self.session.put(url, json={'activate_status': activate_status})
+            url = f"http://{self.host}:{self.port}/api/v1/admin/users/{user_name}/activate"
+            response = self.session.put(url, json={"activate_status": activate_status})
             res_json = response.json()
             if response.status_code == 200:
                 print(res_json["message"])
@@ -688,186 +699,182 @@ class AdminCLI(Cmd):
             print(f"Unknown activate status: {activate_status}.")
 
     def _handle_list_datasets(self, command):
-        username_tree: Tree = command['user_name']
+        username_tree: Tree = command["user_name"]
         user_name: str = username_tree.children[0].strip("'\"")
         print(f"Listing all datasets of user: {user_name}")
-        url = f'http://{self.host}:{self.port}/api/v1/admin/users/{user_name}/datasets'
+        url = f"http://{self.host}:{self.port}/api/v1/admin/users/{user_name}/datasets"
         response = self.session.get(url)
         res_json = response.json()
         if response.status_code == 200:
-            self._print_table_simple(res_json['data'])
+            table_data = res_json["data"]
+            for t in table_data:
+                t.pop("avatar")
+            self._print_table_simple(table_data)
         else:
             print(f"Fail to get all datasets of {user_name}, code: {res_json['code']}, message: {res_json['message']}")
 
     def _handle_list_agents(self, command):
-        username_tree: Tree = command['user_name']
+        username_tree: Tree = command["user_name"]
         user_name: str = username_tree.children[0].strip("'\"")
         print(f"Listing all agents of user: {user_name}")
-        url = f'http://{self.host}:{self.port}/api/v1/admin/users/{user_name}/agents'
+        url = f"http://{self.host}:{self.port}/api/v1/admin/users/{user_name}/agents"
         response = self.session.get(url)
         res_json = response.json()
         if response.status_code == 200:
-            self._print_table_simple(res_json['data'])
+            table_data = res_json["data"]
+            for t in table_data:
+                t.pop("avatar")
+            self._print_table_simple(table_data)
         else:
             print(f"Fail to get all agents of {user_name}, code: {res_json['code']}, message: {res_json['message']}")
 
     def _create_role(self, command):
-        role_name_tree: Tree = command['role_name']
+        role_name_tree: Tree = command["role_name"]
         role_name: str = role_name_tree.children[0].strip("'\"")
-        desc_str: str = ''
-        if 'description' in command:
-            desc_tree: Tree = command['description']
+        desc_str: str = ""
+        if "description" in command:
+            desc_tree: Tree = command["description"]
             desc_str = desc_tree.children[0].strip("'\"")
 
         print(f"create role name: {role_name}, description: {desc_str}")
-        url = f'http://{self.host}:{self.port}/api/v1/admin/roles'
-        response = self.session.post(
-            url,
-            json={'role_name': role_name, 'description': desc_str}
-        )
+        url = f"http://{self.host}:{self.port}/api/v1/admin/roles"
+        response = self.session.post(url, json={"role_name": role_name, "description": desc_str})
         res_json = response.json()
         if response.status_code == 200:
-            self._print_table_simple(res_json['data'])
+            self._print_table_simple(res_json["data"])
         else:
             print(f"Fail to create role {role_name}, code: {res_json['code']}, message: {res_json['message']}")
 
     def _drop_role(self, command):
-        role_name_tree: Tree = command['role_name']
+        role_name_tree: Tree = command["role_name"]
         role_name: str = role_name_tree.children[0].strip("'\"")
         print(f"drop role name: {role_name}")
-        url = f'http://{self.host}:{self.port}/api/v1/admin/roles/{role_name}'
+        url = f"http://{self.host}:{self.port}/api/v1/admin/roles/{role_name}"
         response = self.session.delete(url)
         res_json = response.json()
         if response.status_code == 200:
-            self._print_table_simple(res_json['data'])
+            self._print_table_simple(res_json["data"])
         else:
             print(f"Fail to drop role {role_name}, code: {res_json['code']}, message: {res_json['message']}")
 
     def _alter_role(self, command):
-        role_name_tree: Tree = command['role_name']
+        role_name_tree: Tree = command["role_name"]
         role_name: str = role_name_tree.children[0].strip("'\"")
-        desc_tree: Tree = command['description']
+        desc_tree: Tree = command["description"]
         desc_str: str = desc_tree.children[0].strip("'\"")
 
         print(f"alter role name: {role_name}, description: {desc_str}")
-        url = f'http://{self.host}:{self.port}/api/v1/admin/roles/{role_name}'
-        response = self.session.put(
-            url,
-            json={'description': desc_str}
-        )
+        url = f"http://{self.host}:{self.port}/api/v1/admin/roles/{role_name}"
+        response = self.session.put(url, json={"description": desc_str})
         res_json = response.json()
         if response.status_code == 200:
-            self._print_table_simple(res_json['data'])
+            self._print_table_simple(res_json["data"])
         else:
-            print(
-                f"Fail to update role {role_name} with description: {desc_str}, code: {res_json['code']}, message: {res_json['message']}")
+            print(f"Fail to update role {role_name} with description: {desc_str}, code: {res_json['code']}, message: {res_json['message']}")
 
     def _list_roles(self, command):
         print("Listing all roles")
-        url = f'http://{self.host}:{self.port}/api/v1/admin/roles'
+        url = f"http://{self.host}:{self.port}/api/v1/admin/roles"
         response = self.session.get(url)
         res_json = response.json()
         if response.status_code == 200:
-            self._print_table_simple(res_json['data'])
+            self._print_table_simple(res_json["data"])
         else:
             print(f"Fail to list roles, code: {res_json['code']}, message: {res_json['message']}")
 
     def _show_role(self, command):
-        role_name_tree: Tree = command['role_name']
+        role_name_tree: Tree = command["role_name"]
         role_name: str = role_name_tree.children[0].strip("'\"")
         print(f"show role: {role_name}")
-        url = f'http://{self.host}:{self.port}/api/v1/admin/roles/{role_name}/permission'
+        url = f"http://{self.host}:{self.port}/api/v1/admin/roles/{role_name}/permission"
         response = self.session.get(url)
         res_json = response.json()
         if response.status_code == 200:
-            self._print_table_simple(res_json['data'])
+            self._print_table_simple(res_json["data"])
         else:
             print(f"Fail to list roles, code: {res_json['code']}, message: {res_json['message']}")
 
     def _grant_permission(self, command):
-        role_name_tree: Tree = command['role_name']
+        role_name_tree: Tree = command["role_name"]
         role_name_str: str = role_name_tree.children[0].strip("'\"")
-        resource_tree: Tree = command['resource']
+        resource_tree: Tree = command["resource"]
         resource_str: str = resource_tree.children[0].strip("'\"")
-        action_tree_list: list = command['actions']
+        action_tree_list: list = command["actions"]
         actions: list = []
         for action_tree in action_tree_list:
             action_str: str = action_tree.children[0].strip("'\"")
             actions.append(action_str)
         print(f"grant role_name: {role_name_str}, resource: {resource_str}, actions: {actions}")
-        url = f'http://{self.host}:{self.port}/api/v1/admin/roles/{role_name_str}/permission'
-        response = self.session.post(
-            url,
-            json={'actions': actions, 'resource': resource_str}
-        )
+        url = f"http://{self.host}:{self.port}/api/v1/admin/roles/{role_name_str}/permission"
+        response = self.session.post(url, json={"actions": actions, "resource": resource_str})
         res_json = response.json()
         if response.status_code == 200:
-            self._print_table_simple(res_json['data'])
+            self._print_table_simple(res_json["data"])
         else:
-            print(
-                f"Fail to grant role {role_name_str} with {actions} on {resource_str}, code: {res_json['code']}, message: {res_json['message']}")
+            print(f"Fail to grant role {role_name_str} with {actions} on {resource_str}, code: {res_json['code']}, message: {res_json['message']}")
 
     def _revoke_permission(self, command):
-        role_name_tree: Tree = command['role_name']
+        role_name_tree: Tree = command["role_name"]
         role_name_str: str = role_name_tree.children[0].strip("'\"")
-        resource_tree: Tree = command['resource']
+        resource_tree: Tree = command["resource"]
         resource_str: str = resource_tree.children[0].strip("'\"")
-        action_tree_list: list = command['actions']
+        action_tree_list: list = command["actions"]
         actions: list = []
         for action_tree in action_tree_list:
             action_str: str = action_tree.children[0].strip("'\"")
             actions.append(action_str)
         print(f"revoke role_name: {role_name_str}, resource: {resource_str}, actions: {actions}")
-        url = f'http://{self.host}:{self.port}/api/v1/admin/roles/{role_name_str}/permission'
-        response = self.session.delete(
-            url,
-            json={'actions': actions, 'resource': resource_str}
-        )
+        url = f"http://{self.host}:{self.port}/api/v1/admin/roles/{role_name_str}/permission"
+        response = self.session.delete(url, json={"actions": actions, "resource": resource_str})
         res_json = response.json()
         if response.status_code == 200:
-            self._print_table_simple(res_json['data'])
+            self._print_table_simple(res_json["data"])
         else:
-            print(
-                f"Fail to revoke role {role_name_str} with {actions} on {resource_str}, code: {res_json['code']}, message: {res_json['message']}")
+            print(f"Fail to revoke role {role_name_str} with {actions} on {resource_str}, code: {res_json['code']}, message: {res_json['message']}")
 
     def _alter_user_role(self, command):
-        role_name_tree: Tree = command['role_name']
+        role_name_tree: Tree = command["role_name"]
         role_name_str: str = role_name_tree.children[0].strip("'\"")
-        user_name_tree: Tree = command['user_name']
+        user_name_tree: Tree = command["user_name"]
         user_name_str: str = user_name_tree.children[0].strip("'\"")
         print(f"alter_user_role user_name: {user_name_str}, role_name: {role_name_str}")
-        url = f'http://{self.host}:{self.port}/api/v1/admin/users/{user_name_str}/role'
-        response = self.session.put(
-            url,
-            json={'role_name': role_name_str}
-        )
+        url = f"http://{self.host}:{self.port}/api/v1/admin/users/{user_name_str}/role"
+        response = self.session.put(url, json={"role_name": role_name_str})
         res_json = response.json()
         if response.status_code == 200:
-            self._print_table_simple(res_json['data'])
+            self._print_table_simple(res_json["data"])
         else:
-            print(
-                f"Fail to alter user: {user_name_str} to role {role_name_str}, code: {res_json['code']}, message: {res_json['message']}")
+            print(f"Fail to alter user: {user_name_str} to role {role_name_str}, code: {res_json['code']}, message: {res_json['message']}")
 
     def _show_user_permission(self, command):
-        user_name_tree: Tree = command['user_name']
+        user_name_tree: Tree = command["user_name"]
         user_name_str: str = user_name_tree.children[0].strip("'\"")
         print(f"show_user_permission user_name: {user_name_str}")
-        url = f'http://{self.host}:{self.port}/api/v1/admin/users/{user_name_str}/permission'
+        url = f"http://{self.host}:{self.port}/api/v1/admin/users/{user_name_str}/permission"
         response = self.session.get(url)
         res_json = response.json()
         if response.status_code == 200:
-            self._print_table_simple(res_json['data'])
+            self._print_table_simple(res_json["data"])
         else:
-            print(
-                f"Fail to show user: {user_name_str} permission, code: {res_json['code']}, message: {res_json['message']}")
+            print(f"Fail to show user: {user_name_str} permission, code: {res_json['code']}, message: {res_json['message']}")
+
+    def _show_version(self, command):
+        print("show_version")
+        url = f"http://{self.host}:{self.port}/api/v1/admin/version"
+        response = self.session.get(url)
+        res_json = response.json()
+        if response.status_code == 200:
+            self._print_table_simple(res_json["data"])
+        else:
+            print(f"Fail to show version, code: {res_json['code']}, message: {res_json['message']}")
 
     def _handle_meta_command(self, command):
-        meta_command = command['command']
-        args = command.get('args', [])
+        meta_command = command["command"]
+        args = command.get("args", [])
 
-        if meta_command in ['?', 'h', 'help']:
+        if meta_command in ["?", "h", "help"]:
             self.show_help()
-        elif meta_command in ['q', 'quit', 'exit']:
+        elif meta_command in ["q", "quit", "exit"]:
             print("Goodbye!")
         else:
             print(f"Meta command '{meta_command}' with args {args}")
@@ -903,17 +910,17 @@ def main():
     cli = AdminCLI()
 
     args = cli.parse_connection_args(sys.argv)
-    if 'error' in args:
-        print(f"Error: {args['error']}")
+    if "error" in args:
+        print("Error: Invalid connection arguments")
         return
 
-    if 'command' in args:
-        if 'password' not in args:
+    if "command" in args:
+        if "password" not in args:
             print("Error: password is missing")
             return
         if cli.verify_admin(args, single_command=True):
-            command: str = args['command']
-            print(f"Run single command: {command}")
+            command: str = args["command"]
+            # print(f"Run single command: {command}")
             cli.run_single_command(command)
     else:
         if cli.verify_admin(args, single_command=False):
@@ -927,5 +934,5 @@ def main():
             cli.cmdloop()
 
 
-if __name__ == '__main__':
+if __name__ == "__main__":
     main()

admin/client/pyproject.toml

@@ -1,14 +1,14 @@
 [project]
 name = "ragflow-cli"
-version = "0.21.1"
+version = "0.23.1"
 description = "Admin Service's client of [RAGFlow](https://github.com/infiniflow/ragflow). The Admin Service provides user management and system monitoring. "
 authors = [{ name = "Lynn", email = "lynn_inf@hotmail.com" }]
 license = { text = "Apache License, Version 2.0" }
 readme = "README.md"
-requires-python = ">=3.10,<3.13"
+requires-python = ">=3.12,<3.15"
 dependencies = [
     "requests>=2.30.0,<3.0.0",
-    "beartype>=0.18.5,<0.19.0",
+    "beartype>=0.20.0,<1.0.0",
     "pycryptodomex>=3.10.0",
     "lark>=1.1.0",
 ]

admin/client/uv.lock

@@ -0,0 +1,298 @@
+version = 1
+revision = 3
+requires-python = ">=3.10, <3.13"
+
+[[package]]
+name = "beartype"
+version = "0.22.6"
+source = { registry = "https://pypi.tuna.tsinghua.edu.cn/simple" }
+sdist = { url = "https://pypi.tuna.tsinghua.edu.cn/packages/88/e2/105ceb1704cb80fe4ab3872529ab7b6f365cf7c74f725e6132d0efcf1560/beartype-0.22.6.tar.gz", hash = "sha256:97fbda69c20b48c5780ac2ca60ce3c1bb9af29b3a1a0216898ffabdd523e48f4", size = 1588975, upload-time = "2025-11-20T04:47:14.736Z" }
+wheels = [
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/98/c9/ceecc71fe2c9495a1d8e08d44f5f31f5bca1350d5b2e27a4b6265424f59e/beartype-0.22.6-py3-none-any.whl", hash = "sha256:0584bc46a2ea2a871509679278cda992eadde676c01356ab0ac77421f3c9a093", size = 1324807, upload-time = "2025-11-20T04:47:11.837Z" },
+]
+
+[[package]]
+name = "certifi"
+version = "2025.11.12"
+source = { registry = "https://pypi.tuna.tsinghua.edu.cn/simple" }
+sdist = { url = "https://pypi.tuna.tsinghua.edu.cn/packages/a2/8c/58f469717fa48465e4a50c014a0400602d3c437d7c0c468e17ada824da3a/certifi-2025.11.12.tar.gz", hash = "sha256:d8ab5478f2ecd78af242878415affce761ca6bc54a22a27e026d7c25357c3316", size = 160538, upload-time = "2025-11-12T02:54:51.517Z" }
+wheels = [
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/70/7d/9bc192684cea499815ff478dfcdc13835ddf401365057044fb721ec6bddb/certifi-2025.11.12-py3-none-any.whl", hash = "sha256:97de8790030bbd5c2d96b7ec782fc2f7820ef8dba6db909ccf95449f2d062d4b", size = 159438, upload-time = "2025-11-12T02:54:49.735Z" },
+]
+
+[[package]]
+name = "charset-normalizer"
+version = "3.4.4"
+source = { registry = "https://pypi.tuna.tsinghua.edu.cn/simple" }
+sdist = { url = "https://pypi.tuna.tsinghua.edu.cn/packages/13/69/33ddede1939fdd074bce5434295f38fae7136463422fe4fd3e0e89b98062/charset_normalizer-3.4.4.tar.gz", hash = "sha256:94537985111c35f28720e43603b8e7b43a6ecfb2ce1d3058bbe955b73404e21a", size = 129418, upload-time = "2025-10-14T04:42:32.879Z" }
+wheels = [
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/1f/b8/6d51fc1d52cbd52cd4ccedd5b5b2f0f6a11bbf6765c782298b0f3e808541/charset_normalizer-3.4.4-cp310-cp310-macosx_10_9_universal2.whl", hash = "sha256:e824f1492727fa856dd6eda4f7cee25f8518a12f3c4a56a74e8095695089cf6d", size = 209709, upload-time = "2025-10-14T04:40:11.385Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/5c/af/1f9d7f7faafe2ddfb6f72a2e07a548a629c61ad510fe60f9630309908fef/charset_normalizer-3.4.4-cp310-cp310-manylinux2014_aarch64.manylinux_2_17_aarch64.manylinux_2_28_aarch64.whl", hash = "sha256:4bd5d4137d500351a30687c2d3971758aac9a19208fc110ccb9d7188fbe709e8", size = 148814, upload-time = "2025-10-14T04:40:13.135Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/79/3d/f2e3ac2bbc056ca0c204298ea4e3d9db9b4afe437812638759db2c976b5f/charset_normalizer-3.4.4-cp310-cp310-manylinux2014_armv7l.manylinux_2_17_armv7l.manylinux_2_31_armv7l.whl", hash = "sha256:027f6de494925c0ab2a55eab46ae5129951638a49a34d87f4c3eda90f696b4ad", size = 144467, upload-time = "2025-10-14T04:40:14.728Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/ec/85/1bf997003815e60d57de7bd972c57dc6950446a3e4ccac43bc3070721856/charset_normalizer-3.4.4-cp310-cp310-manylinux2014_ppc64le.manylinux_2_17_ppc64le.manylinux_2_28_ppc64le.whl", hash = "sha256:f820802628d2694cb7e56db99213f930856014862f3fd943d290ea8438d07ca8", size = 162280, upload-time = "2025-10-14T04:40:16.14Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/3e/8e/6aa1952f56b192f54921c436b87f2aaf7c7a7c3d0d1a765547d64fd83c13/charset_normalizer-3.4.4-cp310-cp310-manylinux2014_s390x.manylinux_2_17_s390x.manylinux_2_28_s390x.whl", hash = "sha256:798d75d81754988d2565bff1b97ba5a44411867c0cf32b77a7e8f8d84796b10d", size = 159454, upload-time = "2025-10-14T04:40:17.567Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/36/3b/60cbd1f8e93aa25d1c669c649b7a655b0b5fb4c571858910ea9332678558/charset_normalizer-3.4.4-cp310-cp310-manylinux2014_x86_64.manylinux_2_17_x86_64.manylinux_2_28_x86_64.whl", hash = "sha256:9d1bb833febdff5c8927f922386db610b49db6e0d4f4ee29601d71e7c2694313", size = 153609, upload-time = "2025-10-14T04:40:19.08Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/64/91/6a13396948b8fd3c4b4fd5bc74d045f5637d78c9675585e8e9fbe5636554/charset_normalizer-3.4.4-cp310-cp310-manylinux_2_31_riscv64.manylinux_2_39_riscv64.whl", hash = "sha256:9cd98cdc06614a2f768d2b7286d66805f94c48cde050acdbbb7db2600ab3197e", size = 151849, upload-time = "2025-10-14T04:40:20.607Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/b7/7a/59482e28b9981d105691e968c544cc0df3b7d6133152fb3dcdc8f135da7a/charset_normalizer-3.4.4-cp310-cp310-musllinux_1_2_aarch64.whl", hash = "sha256:077fbb858e903c73f6c9db43374fd213b0b6a778106bc7032446a8e8b5b38b93", size = 151586, upload-time = "2025-10-14T04:40:21.719Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/92/59/f64ef6a1c4bdd2baf892b04cd78792ed8684fbc48d4c2afe467d96b4df57/charset_normalizer-3.4.4-cp310-cp310-musllinux_1_2_armv7l.whl", hash = "sha256:244bfb999c71b35de57821b8ea746b24e863398194a4014e4c76adc2bbdfeff0", size = 145290, upload-time = "2025-10-14T04:40:23.069Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/6b/63/3bf9f279ddfa641ffa1962b0db6a57a9c294361cc2f5fcac997049a00e9c/charset_normalizer-3.4.4-cp310-cp310-musllinux_1_2_ppc64le.whl", hash = "sha256:64b55f9dce520635f018f907ff1b0df1fdc31f2795a922fb49dd14fbcdf48c84", size = 163663, upload-time = "2025-10-14T04:40:24.17Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/ed/09/c9e38fc8fa9e0849b172b581fd9803bdf6e694041127933934184e19f8c3/charset_normalizer-3.4.4-cp310-cp310-musllinux_1_2_riscv64.whl", hash = "sha256:faa3a41b2b66b6e50f84ae4a68c64fcd0c44355741c6374813a800cd6695db9e", size = 151964, upload-time = "2025-10-14T04:40:25.368Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/d2/d1/d28b747e512d0da79d8b6a1ac18b7ab2ecfd81b2944c4c710e166d8dd09c/charset_normalizer-3.4.4-cp310-cp310-musllinux_1_2_s390x.whl", hash = "sha256:6515f3182dbe4ea06ced2d9e8666d97b46ef4c75e326b79bb624110f122551db", size = 161064, upload-time = "2025-10-14T04:40:26.806Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/bb/9a/31d62b611d901c3b9e5500c36aab0ff5eb442043fb3a1c254200d3d397d9/charset_normalizer-3.4.4-cp310-cp310-musllinux_1_2_x86_64.whl", hash = "sha256:cc00f04ed596e9dc0da42ed17ac5e596c6ccba999ba6bd92b0e0aef2f170f2d6", size = 155015, upload-time = "2025-10-14T04:40:28.284Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/1f/f3/107e008fa2bff0c8b9319584174418e5e5285fef32f79d8ee6a430d0039c/charset_normalizer-3.4.4-cp310-cp310-win32.whl", hash = "sha256:f34be2938726fc13801220747472850852fe6b1ea75869a048d6f896838c896f", size = 99792, upload-time = "2025-10-14T04:40:29.613Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/eb/66/e396e8a408843337d7315bab30dbf106c38966f1819f123257f5520f8a96/charset_normalizer-3.4.4-cp310-cp310-win_amd64.whl", hash = "sha256:a61900df84c667873b292c3de315a786dd8dac506704dea57bc957bd31e22c7d", size = 107198, upload-time = "2025-10-14T04:40:30.644Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/b5/58/01b4f815bf0312704c267f2ccb6e5d42bcc7752340cd487bc9f8c3710597/charset_normalizer-3.4.4-cp310-cp310-win_arm64.whl", hash = "sha256:cead0978fc57397645f12578bfd2d5ea9138ea0fac82b2f63f7f7c6877986a69", size = 100262, upload-time = "2025-10-14T04:40:32.108Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/ed/27/c6491ff4954e58a10f69ad90aca8a1b6fe9c5d3c6f380907af3c37435b59/charset_normalizer-3.4.4-cp311-cp311-macosx_10_9_universal2.whl", hash = "sha256:6e1fcf0720908f200cd21aa4e6750a48ff6ce4afe7ff5a79a90d5ed8a08296f8", size = 206988, upload-time = "2025-10-14T04:40:33.79Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/94/59/2e87300fe67ab820b5428580a53cad894272dbb97f38a7a814a2a1ac1011/charset_normalizer-3.4.4-cp311-cp311-manylinux2014_aarch64.manylinux_2_17_aarch64.manylinux_2_28_aarch64.whl", hash = "sha256:5f819d5fe9234f9f82d75bdfa9aef3a3d72c4d24a6e57aeaebba32a704553aa0", size = 147324, upload-time = "2025-10-14T04:40:34.961Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/07/fb/0cf61dc84b2b088391830f6274cb57c82e4da8bbc2efeac8c025edb88772/charset_normalizer-3.4.4-cp311-cp311-manylinux2014_armv7l.manylinux_2_17_armv7l.manylinux_2_31_armv7l.whl", hash = "sha256:a59cb51917aa591b1c4e6a43c132f0cdc3c76dbad6155df4e28ee626cc77a0a3", size = 142742, upload-time = "2025-10-14T04:40:36.105Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/62/8b/171935adf2312cd745d290ed93cf16cf0dfe320863ab7cbeeae1dcd6535f/charset_normalizer-3.4.4-cp311-cp311-manylinux2014_ppc64le.manylinux_2_17_ppc64le.manylinux_2_28_ppc64le.whl", hash = "sha256:8ef3c867360f88ac904fd3f5e1f902f13307af9052646963ee08ff4f131adafc", size = 160863, upload-time = "2025-10-14T04:40:37.188Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/09/73/ad875b192bda14f2173bfc1bc9a55e009808484a4b256748d931b6948442/charset_normalizer-3.4.4-cp311-cp311-manylinux2014_s390x.manylinux_2_17_s390x.manylinux_2_28_s390x.whl", hash = "sha256:d9e45d7faa48ee908174d8fe84854479ef838fc6a705c9315372eacbc2f02897", size = 157837, upload-time = "2025-10-14T04:40:38.435Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/6d/fc/de9cce525b2c5b94b47c70a4b4fb19f871b24995c728e957ee68ab1671ea/charset_normalizer-3.4.4-cp311-cp311-manylinux2014_x86_64.manylinux_2_17_x86_64.manylinux_2_28_x86_64.whl", hash = "sha256:840c25fb618a231545cbab0564a799f101b63b9901f2569faecd6b222ac72381", size = 151550, upload-time = "2025-10-14T04:40:40.053Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/55/c2/43edd615fdfba8c6f2dfbd459b25a6b3b551f24ea21981e23fb768503ce1/charset_normalizer-3.4.4-cp311-cp311-manylinux_2_31_riscv64.manylinux_2_39_riscv64.whl", hash = "sha256:ca5862d5b3928c4940729dacc329aa9102900382fea192fc5e52eb69d6093815", size = 149162, upload-time = "2025-10-14T04:40:41.163Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/03/86/bde4ad8b4d0e9429a4e82c1e8f5c659993a9a863ad62c7df05cf7b678d75/charset_normalizer-3.4.4-cp311-cp311-musllinux_1_2_aarch64.whl", hash = "sha256:d9c7f57c3d666a53421049053eaacdd14bbd0a528e2186fcb2e672effd053bb0", size = 150019, upload-time = "2025-10-14T04:40:42.276Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/1f/86/a151eb2af293a7e7bac3a739b81072585ce36ccfb4493039f49f1d3cae8c/charset_normalizer-3.4.4-cp311-cp311-musllinux_1_2_armv7l.whl", hash = "sha256:277e970e750505ed74c832b4bf75dac7476262ee2a013f5574dd49075879e161", size = 143310, upload-time = "2025-10-14T04:40:43.439Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/b5/fe/43dae6144a7e07b87478fdfc4dbe9efd5defb0e7ec29f5f58a55aeef7bf7/charset_normalizer-3.4.4-cp311-cp311-musllinux_1_2_ppc64le.whl", hash = "sha256:31fd66405eaf47bb62e8cd575dc621c56c668f27d46a61d975a249930dd5e2a4", size = 162022, upload-time = "2025-10-14T04:40:44.547Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/80/e6/7aab83774f5d2bca81f42ac58d04caf44f0cc2b65fc6db2b3b2e8a05f3b3/charset_normalizer-3.4.4-cp311-cp311-musllinux_1_2_riscv64.whl", hash = "sha256:0d3d8f15c07f86e9ff82319b3d9ef6f4bf907608f53fe9d92b28ea9ae3d1fd89", size = 149383, upload-time = "2025-10-14T04:40:46.018Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/4f/e8/b289173b4edae05c0dde07f69f8db476a0b511eac556dfe0d6bda3c43384/charset_normalizer-3.4.4-cp311-cp311-musllinux_1_2_s390x.whl", hash = "sha256:9f7fcd74d410a36883701fafa2482a6af2ff5ba96b9a620e9e0721e28ead5569", size = 159098, upload-time = "2025-10-14T04:40:47.081Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/d8/df/fe699727754cae3f8478493c7f45f777b17c3ef0600e28abfec8619eb49c/charset_normalizer-3.4.4-cp311-cp311-musllinux_1_2_x86_64.whl", hash = "sha256:ebf3e58c7ec8a8bed6d66a75d7fb37b55e5015b03ceae72a8e7c74495551e224", size = 152991, upload-time = "2025-10-14T04:40:48.246Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/1a/86/584869fe4ddb6ffa3bd9f491b87a01568797fb9bd8933f557dba9771beaf/charset_normalizer-3.4.4-cp311-cp311-win32.whl", hash = "sha256:eecbc200c7fd5ddb9a7f16c7decb07b566c29fa2161a16cf67b8d068bd21690a", size = 99456, upload-time = "2025-10-14T04:40:49.376Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/65/f6/62fdd5feb60530f50f7e38b4f6a1d5203f4d16ff4f9f0952962c044e919a/charset_normalizer-3.4.4-cp311-cp311-win_amd64.whl", hash = "sha256:5ae497466c7901d54b639cf42d5b8c1b6a4fead55215500d2f486d34db48d016", size = 106978, upload-time = "2025-10-14T04:40:50.844Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/7a/9d/0710916e6c82948b3be62d9d398cb4fcf4e97b56d6a6aeccd66c4b2f2bd5/charset_normalizer-3.4.4-cp311-cp311-win_arm64.whl", hash = "sha256:65e2befcd84bc6f37095f5961e68a6f077bf44946771354a28ad434c2cce0ae1", size = 99969, upload-time = "2025-10-14T04:40:52.272Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/f3/85/1637cd4af66fa687396e757dec650f28025f2a2f5a5531a3208dc0ec43f2/charset_normalizer-3.4.4-cp312-cp312-macosx_10_13_universal2.whl", hash = "sha256:0a98e6759f854bd25a58a73fa88833fba3b7c491169f86ce1180c948ab3fd394", size = 208425, upload-time = "2025-10-14T04:40:53.353Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/9d/6a/04130023fef2a0d9c62d0bae2649b69f7b7d8d24ea5536feef50551029df/charset_normalizer-3.4.4-cp312-cp312-manylinux2014_aarch64.manylinux_2_17_aarch64.manylinux_2_28_aarch64.whl", hash = "sha256:b5b290ccc2a263e8d185130284f8501e3e36c5e02750fc6b6bdeb2e9e96f1e25", size = 148162, upload-time = "2025-10-14T04:40:54.558Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/78/29/62328d79aa60da22c9e0b9a66539feae06ca0f5a4171ac4f7dc285b83688/charset_normalizer-3.4.4-cp312-cp312-manylinux2014_armv7l.manylinux_2_17_armv7l.manylinux_2_31_armv7l.whl", hash = "sha256:74bb723680f9f7a6234dcf67aea57e708ec1fbdf5699fb91dfd6f511b0a320ef", size = 144558, upload-time = "2025-10-14T04:40:55.677Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/86/bb/b32194a4bf15b88403537c2e120b817c61cd4ecffa9b6876e941c3ee38fe/charset_normalizer-3.4.4-cp312-cp312-manylinux2014_ppc64le.manylinux_2_17_ppc64le.manylinux_2_28_ppc64le.whl", hash = "sha256:f1e34719c6ed0b92f418c7c780480b26b5d9c50349e9a9af7d76bf757530350d", size = 161497, upload-time = "2025-10-14T04:40:57.217Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/19/89/a54c82b253d5b9b111dc74aca196ba5ccfcca8242d0fb64146d4d3183ff1/charset_normalizer-3.4.4-cp312-cp312-manylinux2014_s390x.manylinux_2_17_s390x.manylinux_2_28_s390x.whl", hash = "sha256:2437418e20515acec67d86e12bf70056a33abdacb5cb1655042f6538d6b085a8", size = 159240, upload-time = "2025-10-14T04:40:58.358Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/c0/10/d20b513afe03acc89ec33948320a5544d31f21b05368436d580dec4e234d/charset_normalizer-3.4.4-cp312-cp312-manylinux2014_x86_64.manylinux_2_17_x86_64.manylinux_2_28_x86_64.whl", hash = "sha256:11d694519d7f29d6cd09f6ac70028dba10f92f6cdd059096db198c283794ac86", size = 153471, upload-time = "2025-10-14T04:40:59.468Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/61/fa/fbf177b55bdd727010f9c0a3c49eefa1d10f960e5f09d1d887bf93c2e698/charset_normalizer-3.4.4-cp312-cp312-manylinux_2_31_riscv64.manylinux_2_39_riscv64.whl", hash = "sha256:ac1c4a689edcc530fc9d9aa11f5774b9e2f33f9a0c6a57864e90908f5208d30a", size = 150864, upload-time = "2025-10-14T04:41:00.623Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/05/12/9fbc6a4d39c0198adeebbde20b619790e9236557ca59fc40e0e3cebe6f40/charset_normalizer-3.4.4-cp312-cp312-musllinux_1_2_aarch64.whl", hash = "sha256:21d142cc6c0ec30d2efee5068ca36c128a30b0f2c53c1c07bd78cb6bc1d3be5f", size = 150647, upload-time = "2025-10-14T04:41:01.754Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/ad/1f/6a9a593d52e3e8c5d2b167daf8c6b968808efb57ef4c210acb907c365bc4/charset_normalizer-3.4.4-cp312-cp312-musllinux_1_2_armv7l.whl", hash = "sha256:5dbe56a36425d26d6cfb40ce79c314a2e4dd6211d51d6d2191c00bed34f354cc", size = 145110, upload-time = "2025-10-14T04:41:03.231Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/30/42/9a52c609e72471b0fc54386dc63c3781a387bb4fe61c20231a4ebcd58bdd/charset_normalizer-3.4.4-cp312-cp312-musllinux_1_2_ppc64le.whl", hash = "sha256:5bfbb1b9acf3334612667b61bd3002196fe2a1eb4dd74d247e0f2a4d50ec9bbf", size = 162839, upload-time = "2025-10-14T04:41:04.715Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/c4/5b/c0682bbf9f11597073052628ddd38344a3d673fda35a36773f7d19344b23/charset_normalizer-3.4.4-cp312-cp312-musllinux_1_2_riscv64.whl", hash = "sha256:d055ec1e26e441f6187acf818b73564e6e6282709e9bcb5b63f5b23068356a15", size = 150667, upload-time = "2025-10-14T04:41:05.827Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/e4/24/a41afeab6f990cf2daf6cb8c67419b63b48cf518e4f56022230840c9bfb2/charset_normalizer-3.4.4-cp312-cp312-musllinux_1_2_s390x.whl", hash = "sha256:af2d8c67d8e573d6de5bc30cdb27e9b95e49115cd9baad5ddbd1a6207aaa82a9", size = 160535, upload-time = "2025-10-14T04:41:06.938Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/2a/e5/6a4ce77ed243c4a50a1fecca6aaaab419628c818a49434be428fe24c9957/charset_normalizer-3.4.4-cp312-cp312-musllinux_1_2_x86_64.whl", hash = "sha256:780236ac706e66881f3b7f2f32dfe90507a09e67d1d454c762cf642e6e1586e0", size = 154816, upload-time = "2025-10-14T04:41:08.101Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/a8/ef/89297262b8092b312d29cdb2517cb1237e51db8ecef2e9af5edbe7b683b1/charset_normalizer-3.4.4-cp312-cp312-win32.whl", hash = "sha256:5833d2c39d8896e4e19b689ffc198f08ea58116bee26dea51e362ecc7cd3ed26", size = 99694, upload-time = "2025-10-14T04:41:09.23Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/3d/2d/1e5ed9dd3b3803994c155cd9aacb60c82c331bad84daf75bcb9c91b3295e/charset_normalizer-3.4.4-cp312-cp312-win_amd64.whl", hash = "sha256:a79cfe37875f822425b89a82333404539ae63dbdddf97f84dcbc3d339aae9525", size = 107131, upload-time = "2025-10-14T04:41:10.467Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/d0/d9/0ed4c7098a861482a7b6a95603edce4c0d9db2311af23da1fb2b75ec26fc/charset_normalizer-3.4.4-cp312-cp312-win_arm64.whl", hash = "sha256:376bec83a63b8021bb5c8ea75e21c4ccb86e7e45ca4eb81146091b56599b80c3", size = 100390, upload-time = "2025-10-14T04:41:11.915Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/0a/4c/925909008ed5a988ccbb72dcc897407e5d6d3bd72410d69e051fc0c14647/charset_normalizer-3.4.4-py3-none-any.whl", hash = "sha256:7a32c560861a02ff789ad905a2fe94e3f840803362c84fecf1851cb4cf3dc37f", size = 53402, upload-time = "2025-10-14T04:42:31.76Z" },
+]
+
+[[package]]
+name = "colorama"
+version = "0.4.6"
+source = { registry = "https://pypi.tuna.tsinghua.edu.cn/simple" }
+sdist = { url = "https://pypi.tuna.tsinghua.edu.cn/packages/d8/53/6f443c9a4a8358a93a6792e2acffb9d9d5cb0a5cfd8802644b7b1c9a02e4/colorama-0.4.6.tar.gz", hash = "sha256:08695f5cb7ed6e0531a20572697297273c47b8cae5a63ffc6d6ed5c201be6e44", size = 27697, upload-time = "2022-10-25T02:36:22.414Z" }
+wheels = [
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/d1/d6/3965ed04c63042e047cb6a3e6ed1a63a35087b6a609aa3a15ed8ac56c221/colorama-0.4.6-py2.py3-none-any.whl", hash = "sha256:4f1d9991f5acc0ca119f9d443620b77f9d6b33703e51011c16baf57afb285fc6", size = 25335, upload-time = "2022-10-25T02:36:20.889Z" },
+]
+
+[[package]]
+name = "exceptiongroup"
+version = "1.3.1"
+source = { registry = "https://pypi.tuna.tsinghua.edu.cn/simple" }
+dependencies = [
+    { name = "typing-extensions" },
+]
+sdist = { url = "https://pypi.tuna.tsinghua.edu.cn/packages/50/79/66800aadf48771f6b62f7eb014e352e5d06856655206165d775e675a02c9/exceptiongroup-1.3.1.tar.gz", hash = "sha256:8b412432c6055b0b7d14c310000ae93352ed6754f70fa8f7c34141f91c4e3219", size = 30371, upload-time = "2025-11-21T23:01:54.787Z" }
+wheels = [
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/8a/0e/97c33bf5009bdbac74fd2beace167cab3f978feb69cc36f1ef79360d6c4e/exceptiongroup-1.3.1-py3-none-any.whl", hash = "sha256:a7a39a3bd276781e98394987d3a5701d0c4edffb633bb7a5144577f82c773598", size = 16740, upload-time = "2025-11-21T23:01:53.443Z" },
+]
+
+[[package]]
+name = "idna"
+version = "3.11"
+source = { registry = "https://pypi.tuna.tsinghua.edu.cn/simple" }
+sdist = { url = "https://pypi.tuna.tsinghua.edu.cn/packages/6f/6d/0703ccc57f3a7233505399edb88de3cbd678da106337b9fcde432b65ed60/idna-3.11.tar.gz", hash = "sha256:795dafcc9c04ed0c1fb032c2aa73654d8e8c5023a7df64a53f39190ada629902", size = 194582, upload-time = "2025-10-12T14:55:20.501Z" }
+wheels = [
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/0e/61/66938bbb5fc52dbdf84594873d5b51fb1f7c7794e9c0f5bd885f30bc507b/idna-3.11-py3-none-any.whl", hash = "sha256:771a87f49d9defaf64091e6e6fe9c18d4833f140bd19464795bc32d966ca37ea", size = 71008, upload-time = "2025-10-12T14:55:18.883Z" },
+]
+
+[[package]]
+name = "iniconfig"
+version = "2.3.0"
+source = { registry = "https://pypi.tuna.tsinghua.edu.cn/simple" }
+sdist = { url = "https://pypi.tuna.tsinghua.edu.cn/packages/72/34/14ca021ce8e5dfedc35312d08ba8bf51fdd999c576889fc2c24cb97f4f10/iniconfig-2.3.0.tar.gz", hash = "sha256:c76315c77db068650d49c5b56314774a7804df16fee4402c1f19d6d15d8c4730", size = 20503, upload-time = "2025-10-18T21:55:43.219Z" }
+wheels = [
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/cb/b1/3846dd7f199d53cb17f49cba7e651e9ce294d8497c8c150530ed11865bb8/iniconfig-2.3.0-py3-none-any.whl", hash = "sha256:f631c04d2c48c52b84d0d0549c99ff3859c98df65b3101406327ecc7d53fbf12", size = 7484, upload-time = "2025-10-18T21:55:41.639Z" },
+]
+
+[[package]]
+name = "lark"
+version = "1.3.1"
+source = { registry = "https://pypi.tuna.tsinghua.edu.cn/simple" }
+sdist = { url = "https://pypi.tuna.tsinghua.edu.cn/packages/da/34/28fff3ab31ccff1fd4f6c7c7b0ceb2b6968d8ea4950663eadcb5720591a0/lark-1.3.1.tar.gz", hash = "sha256:b426a7a6d6d53189d318f2b6236ab5d6429eaf09259f1ca33eb716eed10d2905", size = 382732, upload-time = "2025-10-27T18:25:56.653Z" }
+wheels = [
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/82/3d/14ce75ef66813643812f3093ab17e46d3a206942ce7376d31ec2d36229e7/lark-1.3.1-py3-none-any.whl", hash = "sha256:c629b661023a014c37da873b4ff58a817398d12635d3bbb2c5a03be7fe5d1e12", size = 113151, upload-time = "2025-10-27T18:25:54.882Z" },
+]
+
+[[package]]
+name = "packaging"
+version = "25.0"
+source = { registry = "https://pypi.tuna.tsinghua.edu.cn/simple" }
+sdist = { url = "https://pypi.tuna.tsinghua.edu.cn/packages/a1/d4/1fc4078c65507b51b96ca8f8c3ba19e6a61c8253c72794544580a7b6c24d/packaging-25.0.tar.gz", hash = "sha256:d443872c98d677bf60f6a1f2f8c1cb748e8fe762d2bf9d3148b5599295b0fc4f", size = 165727, upload-time = "2025-04-19T11:48:59.673Z" }
+wheels = [
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/20/12/38679034af332785aac8774540895e234f4d07f7545804097de4b666afd8/packaging-25.0-py3-none-any.whl", hash = "sha256:29572ef2b1f17581046b3a2227d5c611fb25ec70ca1ba8554b24b0e69331a484", size = 66469, upload-time = "2025-04-19T11:48:57.875Z" },
+]
+
+[[package]]
+name = "pluggy"
+version = "1.6.0"
+source = { registry = "https://pypi.tuna.tsinghua.edu.cn/simple" }
+sdist = { url = "https://pypi.tuna.tsinghua.edu.cn/packages/f9/e2/3e91f31a7d2b083fe6ef3fa267035b518369d9511ffab804f839851d2779/pluggy-1.6.0.tar.gz", hash = "sha256:7dcc130b76258d33b90f61b658791dede3486c3e6bfb003ee5c9bfb396dd22f3", size = 69412, upload-time = "2025-05-15T12:30:07.975Z" }
+wheels = [
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/54/20/4d324d65cc6d9205fabedc306948156824eb9f0ee1633355a8f7ec5c66bf/pluggy-1.6.0-py3-none-any.whl", hash = "sha256:e920276dd6813095e9377c0bc5566d94c932c33b27a3e3945d8389c374dd4746", size = 20538, upload-time = "2025-05-15T12:30:06.134Z" },
+]
+
+[[package]]
+name = "pycryptodomex"
+version = "3.23.0"
+source = { registry = "https://pypi.tuna.tsinghua.edu.cn/simple" }
+sdist = { url = "https://pypi.tuna.tsinghua.edu.cn/packages/c9/85/e24bf90972a30b0fcd16c73009add1d7d7cd9140c2498a68252028899e41/pycryptodomex-3.23.0.tar.gz", hash = "sha256:71909758f010c82bc99b0abf4ea12012c98962fbf0583c2164f8b84533c2e4da", size = 4922157, upload-time = "2025-05-17T17:23:41.434Z" }
+wheels = [
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/dd/9c/1a8f35daa39784ed8adf93a694e7e5dc15c23c741bbda06e1d45f8979e9e/pycryptodomex-3.23.0-cp37-abi3-macosx_10_9_universal2.whl", hash = "sha256:06698f957fe1ab229a99ba2defeeae1c09af185baa909a31a5d1f9d42b1aaed6", size = 2499240, upload-time = "2025-05-17T17:22:46.953Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/7a/62/f5221a191a97157d240cf6643747558759126c76ee92f29a3f4aee3197a5/pycryptodomex-3.23.0-cp37-abi3-macosx_10_9_x86_64.whl", hash = "sha256:b2c2537863eccef2d41061e82a881dcabb04944c5c06c5aa7110b577cc487545", size = 1644042, upload-time = "2025-05-17T17:22:49.098Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/8c/fd/5a054543c8988d4ed7b612721d7e78a4b9bf36bc3c5ad45ef45c22d0060e/pycryptodomex-3.23.0-cp37-abi3-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:43c446e2ba8df8889e0e16f02211c25b4934898384c1ec1ec04d7889c0333587", size = 2186227, upload-time = "2025-05-17T17:22:51.139Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/c8/a9/8862616a85cf450d2822dbd4fff1fcaba90877907a6ff5bc2672cafe42f8/pycryptodomex-3.23.0-cp37-abi3-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:f489c4765093fb60e2edafdf223397bc716491b2b69fe74367b70d6999257a5c", size = 2272578, upload-time = "2025-05-17T17:22:53.676Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/46/9f/bda9c49a7c1842820de674ab36c79f4fbeeee03f8ff0e4f3546c3889076b/pycryptodomex-3.23.0-cp37-abi3-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:bdc69d0d3d989a1029df0eed67cc5e8e5d968f3724f4519bd03e0ec68df7543c", size = 2312166, upload-time = "2025-05-17T17:22:56.585Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/03/cc/870b9bf8ca92866ca0186534801cf8d20554ad2a76ca959538041b7a7cf4/pycryptodomex-3.23.0-cp37-abi3-musllinux_1_2_aarch64.whl", hash = "sha256:6bbcb1dd0f646484939e142462d9e532482bc74475cecf9c4903d4e1cd21f003", size = 2185467, upload-time = "2025-05-17T17:22:59.237Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/96/e3/ce9348236d8e669fea5dd82a90e86be48b9c341210f44e25443162aba187/pycryptodomex-3.23.0-cp37-abi3-musllinux_1_2_i686.whl", hash = "sha256:8a4fcd42ccb04c31268d1efeecfccfd1249612b4de6374205376b8f280321744", size = 2346104, upload-time = "2025-05-17T17:23:02.112Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/a5/e9/e869bcee87beb89040263c416a8a50204f7f7a83ac11897646c9e71e0daf/pycryptodomex-3.23.0-cp37-abi3-musllinux_1_2_x86_64.whl", hash = "sha256:55ccbe27f049743a4caf4f4221b166560d3438d0b1e5ab929e07ae1702a4d6fd", size = 2271038, upload-time = "2025-05-17T17:23:04.872Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/8d/67/09ee8500dd22614af5fbaa51a4aee6e342b5fa8aecf0a6cb9cbf52fa6d45/pycryptodomex-3.23.0-cp37-abi3-win32.whl", hash = "sha256:189afbc87f0b9f158386bf051f720e20fa6145975f1e76369303d0f31d1a8d7c", size = 1771969, upload-time = "2025-05-17T17:23:07.115Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/69/96/11f36f71a865dd6df03716d33bd07a67e9d20f6b8d39820470b766af323c/pycryptodomex-3.23.0-cp37-abi3-win_amd64.whl", hash = "sha256:52e5ca58c3a0b0bd5e100a9fbc8015059b05cffc6c66ce9d98b4b45e023443b9", size = 1803124, upload-time = "2025-05-17T17:23:09.267Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/f9/93/45c1cdcbeb182ccd2e144c693eaa097763b08b38cded279f0053ed53c553/pycryptodomex-3.23.0-cp37-abi3-win_arm64.whl", hash = "sha256:02d87b80778c171445d67e23d1caef279bf4b25c3597050ccd2e13970b57fd51", size = 1707161, upload-time = "2025-05-17T17:23:11.414Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/f3/b8/3e76d948c3c4ac71335bbe75dac53e154b40b0f8f1f022dfa295257a0c96/pycryptodomex-3.23.0-pp310-pypy310_pp73-macosx_10_15_x86_64.whl", hash = "sha256:ebfff755c360d674306e5891c564a274a47953562b42fb74a5c25b8fc1fb1cb5", size = 1627695, upload-time = "2025-05-17T17:23:17.38Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/6a/cf/80f4297a4820dfdfd1c88cf6c4666a200f204b3488103d027b5edd9176ec/pycryptodomex-3.23.0-pp310-pypy310_pp73-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:eca54f4bb349d45afc17e3011ed4264ef1cc9e266699874cdd1349c504e64798", size = 1675772, upload-time = "2025-05-17T17:23:19.202Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/d1/42/1e969ee0ad19fe3134b0e1b856c39bd0b70d47a4d0e81c2a8b05727394c9/pycryptodomex-3.23.0-pp310-pypy310_pp73-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:4f2596e643d4365e14d0879dc5aafe6355616c61c2176009270f3048f6d9a61f", size = 1668083, upload-time = "2025-05-17T17:23:21.867Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/6e/c3/1de4f7631fea8a992a44ba632aa40e0008764c0fb9bf2854b0acf78c2cf2/pycryptodomex-3.23.0-pp310-pypy310_pp73-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:fdfac7cda115bca3a5abb2f9e43bc2fb66c2b65ab074913643803ca7083a79ea", size = 1706056, upload-time = "2025-05-17T17:23:24.031Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/f2/5f/af7da8e6f1e42b52f44a24d08b8e4c726207434e2593732d39e7af5e7256/pycryptodomex-3.23.0-pp310-pypy310_pp73-win_amd64.whl", hash = "sha256:14c37aaece158d0ace436f76a7bb19093db3b4deade9797abfc39ec6cd6cc2fe", size = 1806478, upload-time = "2025-05-17T17:23:26.066Z" },
+]
+
+[[package]]
+name = "pygments"
+version = "2.19.2"
+source = { registry = "https://pypi.tuna.tsinghua.edu.cn/simple" }
+sdist = { url = "https://pypi.tuna.tsinghua.edu.cn/packages/b0/77/a5b8c569bf593b0140bde72ea885a803b82086995367bf2037de0159d924/pygments-2.19.2.tar.gz", hash = "sha256:636cb2477cec7f8952536970bc533bc43743542f70392ae026374600add5b887", size = 4968631, upload-time = "2025-06-21T13:39:12.283Z" }
+wheels = [
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/c7/21/705964c7812476f378728bdf590ca4b771ec72385c533964653c68e86bdc/pygments-2.19.2-py3-none-any.whl", hash = "sha256:86540386c03d588bb81d44bc3928634ff26449851e99741617ecb9037ee5ec0b", size = 1225217, upload-time = "2025-06-21T13:39:07.939Z" },
+]
+
+[[package]]
+name = "pytest"
+version = "9.0.1"
+source = { registry = "https://pypi.tuna.tsinghua.edu.cn/simple" }
+dependencies = [
+    { name = "colorama", marker = "sys_platform == 'win32'" },
+    { name = "exceptiongroup", marker = "python_full_version < '3.11'" },
+    { name = "iniconfig" },
+    { name = "packaging" },
+    { name = "pluggy" },
+    { name = "pygments" },
+    { name = "tomli", marker = "python_full_version < '3.11'" },
+]
+sdist = { url = "https://pypi.tuna.tsinghua.edu.cn/packages/07/56/f013048ac4bc4c1d9be45afd4ab209ea62822fb1598f40687e6bf45dcea4/pytest-9.0.1.tar.gz", hash = "sha256:3e9c069ea73583e255c3b21cf46b8d3c56f6e3a1a8f6da94ccb0fcf57b9d73c8", size = 1564125, upload-time = "2025-11-12T13:05:09.333Z" }
+wheels = [
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/0b/8b/6300fb80f858cda1c51ffa17075df5d846757081d11ab4aa35cef9e6258b/pytest-9.0.1-py3-none-any.whl", hash = "sha256:67be0030d194df2dfa7b556f2e56fb3c3315bd5c8822c6951162b92b32ce7dad", size = 373668, upload-time = "2025-11-12T13:05:07.379Z" },
+]
+
+[[package]]
+name = "ragflow-cli"
+version = "0.23.1"
+source = { virtual = "." }
+dependencies = [
+    { name = "beartype" },
+    { name = "lark" },
+    { name = "pycryptodomex" },
+    { name = "requests" },
+]
+
+[package.dev-dependencies]
+test = [
+    { name = "pytest" },
+    { name = "requests" },
+    { name = "requests-toolbelt" },
+]
+
+[package.metadata]
+requires-dist = [
+    { name = "beartype", specifier = ">=0.20.0,<1.0.0" },
+    { name = "lark", specifier = ">=1.1.0" },
+    { name = "pycryptodomex", specifier = ">=3.10.0" },
+    { name = "requests", specifier = ">=2.30.0,<3.0.0" },
+]
+
+[package.metadata.requires-dev]
+test = [
+    { name = "pytest", specifier = ">=8.3.5" },
+    { name = "requests", specifier = ">=2.32.3" },
+    { name = "requests-toolbelt", specifier = ">=1.0.0" },
+]
+
+[[package]]
+name = "requests"
+version = "2.32.5"
+source = { registry = "https://pypi.tuna.tsinghua.edu.cn/simple" }
+dependencies = [
+    { name = "certifi" },
+    { name = "charset-normalizer" },
+    { name = "idna" },
+    { name = "urllib3" },
+]
+sdist = { url = "https://pypi.tuna.tsinghua.edu.cn/packages/c9/74/b3ff8e6c8446842c3f5c837e9c3dfcfe2018ea6ecef224c710c85ef728f4/requests-2.32.5.tar.gz", hash = "sha256:dbba0bac56e100853db0ea71b82b4dfd5fe2bf6d3754a8893c3af500cec7d7cf", size = 134517, upload-time = "2025-08-18T20:46:02.573Z" }
+wheels = [
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/1e/db/4254e3eabe8020b458f1a747140d32277ec7a271daf1d235b70dc0b4e6e3/requests-2.32.5-py3-none-any.whl", hash = "sha256:2462f94637a34fd532264295e186976db0f5d453d1cdd31473c85a6a161affb6", size = 64738, upload-time = "2025-08-18T20:46:00.542Z" },
+]
+
+[[package]]
+name = "requests-toolbelt"
+version = "1.0.0"
+source = { registry = "https://pypi.tuna.tsinghua.edu.cn/simple" }
+dependencies = [
+    { name = "requests" },
+]
+sdist = { url = "https://pypi.tuna.tsinghua.edu.cn/packages/f3/61/d7545dafb7ac2230c70d38d31cbfe4cc64f7144dc41f6e4e4b78ecd9f5bb/requests-toolbelt-1.0.0.tar.gz", hash = "sha256:7681a0a3d047012b5bdc0ee37d7f8f07ebe76ab08caeccfc3921ce23c88d5bc6", size = 206888, upload-time = "2023-05-01T04:11:33.229Z" }
+wheels = [
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/3f/51/d4db610ef29373b879047326cbf6fa98b6c1969d6f6dc423279de2b1be2c/requests_toolbelt-1.0.0-py2.py3-none-any.whl", hash = "sha256:cccfdd665f0a24fcf4726e690f65639d272bb0637b9b92dfd91a5568ccf6bd06", size = 54481, upload-time = "2023-05-01T04:11:28.427Z" },
+]
+
+[[package]]
+name = "tomli"
+version = "2.3.0"
+source = { registry = "https://pypi.tuna.tsinghua.edu.cn/simple" }
+sdist = { url = "https://pypi.tuna.tsinghua.edu.cn/packages/52/ed/3f73f72945444548f33eba9a87fc7a6e969915e7b1acc8260b30e1f76a2f/tomli-2.3.0.tar.gz", hash = "sha256:64be704a875d2a59753d80ee8a533c3fe183e3f06807ff7dc2232938ccb01549", size = 17392, upload-time = "2025-10-08T22:01:47.119Z" }
+wheels = [
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/b3/2e/299f62b401438d5fe1624119c723f5d877acc86a4c2492da405626665f12/tomli-2.3.0-cp311-cp311-macosx_10_9_x86_64.whl", hash = "sha256:88bd15eb972f3664f5ed4b57c1634a97153b4bac4479dcb6a495f41921eb7f45", size = 153236, upload-time = "2025-10-08T22:01:00.137Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/86/7f/d8fffe6a7aefdb61bced88fcb5e280cfd71e08939da5894161bd71bea022/tomli-2.3.0-cp311-cp311-macosx_11_0_arm64.whl", hash = "sha256:883b1c0d6398a6a9d29b508c331fa56adbcdff647f6ace4dfca0f50e90dfd0ba", size = 148084, upload-time = "2025-10-08T22:01:01.63Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/47/5c/24935fb6a2ee63e86d80e4d3b58b222dafaf438c416752c8b58537c8b89a/tomli-2.3.0-cp311-cp311-manylinux2014_aarch64.manylinux_2_17_aarch64.manylinux_2_28_aarch64.whl", hash = "sha256:d1381caf13ab9f300e30dd8feadb3de072aeb86f1d34a8569453ff32a7dea4bf", size = 234832, upload-time = "2025-10-08T22:01:02.543Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/89/da/75dfd804fc11e6612846758a23f13271b76d577e299592b4371a4ca4cd09/tomli-2.3.0-cp311-cp311-manylinux2014_x86_64.manylinux_2_17_x86_64.manylinux_2_28_x86_64.whl", hash = "sha256:a0e285d2649b78c0d9027570d4da3425bdb49830a6156121360b3f8511ea3441", size = 242052, upload-time = "2025-10-08T22:01:03.836Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/70/8c/f48ac899f7b3ca7eb13af73bacbc93aec37f9c954df3c08ad96991c8c373/tomli-2.3.0-cp311-cp311-musllinux_1_2_aarch64.whl", hash = "sha256:0a154a9ae14bfcf5d8917a59b51ffd5a3ac1fd149b71b47a3a104ca4edcfa845", size = 239555, upload-time = "2025-10-08T22:01:04.834Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/ba/28/72f8afd73f1d0e7829bfc093f4cb98ce0a40ffc0cc997009ee1ed94ba705/tomli-2.3.0-cp311-cp311-musllinux_1_2_x86_64.whl", hash = "sha256:74bf8464ff93e413514fefd2be591c3b0b23231a77f901db1eb30d6f712fc42c", size = 245128, upload-time = "2025-10-08T22:01:05.84Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/b6/eb/a7679c8ac85208706d27436e8d421dfa39d4c914dcf5fa8083a9305f58d9/tomli-2.3.0-cp311-cp311-win32.whl", hash = "sha256:00b5f5d95bbfc7d12f91ad8c593a1659b6387b43f054104cda404be6bda62456", size = 96445, upload-time = "2025-10-08T22:01:06.896Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/0a/fe/3d3420c4cb1ad9cb462fb52967080575f15898da97e21cb6f1361d505383/tomli-2.3.0-cp311-cp311-win_amd64.whl", hash = "sha256:4dc4ce8483a5d429ab602f111a93a6ab1ed425eae3122032db7e9acf449451be", size = 107165, upload-time = "2025-10-08T22:01:08.107Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/ff/b7/40f36368fcabc518bb11c8f06379a0fd631985046c038aca08c6d6a43c6e/tomli-2.3.0-cp312-cp312-macosx_10_13_x86_64.whl", hash = "sha256:d7d86942e56ded512a594786a5ba0a5e521d02529b3826e7761a05138341a2ac", size = 154891, upload-time = "2025-10-08T22:01:09.082Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/f9/3f/d9dd692199e3b3aab2e4e4dd948abd0f790d9ded8cd10cbaae276a898434/tomli-2.3.0-cp312-cp312-macosx_11_0_arm64.whl", hash = "sha256:73ee0b47d4dad1c5e996e3cd33b8a76a50167ae5f96a2607cbe8cc773506ab22", size = 148796, upload-time = "2025-10-08T22:01:10.266Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/60/83/59bff4996c2cf9f9387a0f5a3394629c7efa5ef16142076a23a90f1955fa/tomli-2.3.0-cp312-cp312-manylinux2014_aarch64.manylinux_2_17_aarch64.manylinux_2_28_aarch64.whl", hash = "sha256:792262b94d5d0a466afb5bc63c7daa9d75520110971ee269152083270998316f", size = 242121, upload-time = "2025-10-08T22:01:11.332Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/45/e5/7c5119ff39de8693d6baab6c0b6dcb556d192c165596e9fc231ea1052041/tomli-2.3.0-cp312-cp312-manylinux2014_x86_64.manylinux_2_17_x86_64.manylinux_2_28_x86_64.whl", hash = "sha256:4f195fe57ecceac95a66a75ac24d9d5fbc98ef0962e09b2eddec5d39375aae52", size = 250070, upload-time = "2025-10-08T22:01:12.498Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/45/12/ad5126d3a278f27e6701abde51d342aa78d06e27ce2bb596a01f7709a5a2/tomli-2.3.0-cp312-cp312-musllinux_1_2_aarch64.whl", hash = "sha256:e31d432427dcbf4d86958c184b9bfd1e96b5b71f8eb17e6d02531f434fd335b8", size = 245859, upload-time = "2025-10-08T22:01:13.551Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/fb/a1/4d6865da6a71c603cfe6ad0e6556c73c76548557a8d658f9e3b142df245f/tomli-2.3.0-cp312-cp312-musllinux_1_2_x86_64.whl", hash = "sha256:7b0882799624980785240ab732537fcfc372601015c00f7fc367c55308c186f6", size = 250296, upload-time = "2025-10-08T22:01:14.614Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/a0/b7/a7a7042715d55c9ba6e8b196d65d2cb662578b4d8cd17d882d45322b0d78/tomli-2.3.0-cp312-cp312-win32.whl", hash = "sha256:ff72b71b5d10d22ecb084d345fc26f42b5143c5533db5e2eaba7d2d335358876", size = 97124, upload-time = "2025-10-08T22:01:15.629Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/06/1e/f22f100db15a68b520664eb3328fb0ae4e90530887928558112c8d1f4515/tomli-2.3.0-cp312-cp312-win_amd64.whl", hash = "sha256:1cb4ed918939151a03f33d4242ccd0aa5f11b3547d0cf30f7c74a408a5b99878", size = 107698, upload-time = "2025-10-08T22:01:16.51Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/77/b8/0135fadc89e73be292b473cb820b4f5a08197779206b33191e801feeae40/tomli-2.3.0-py3-none-any.whl", hash = "sha256:e95b1af3c5b07d9e643909b5abbec77cd9f1217e6d0bca72b0234736b9fb1f1b", size = 14408, upload-time = "2025-10-08T22:01:46.04Z" },
+]
+
+[[package]]
+name = "typing-extensions"
+version = "4.15.0"
+source = { registry = "https://pypi.tuna.tsinghua.edu.cn/simple" }
+sdist = { url = "https://pypi.tuna.tsinghua.edu.cn/packages/72/94/1a15dd82efb362ac84269196e94cf00f187f7ed21c242792a923cdb1c61f/typing_extensions-4.15.0.tar.gz", hash = "sha256:0cea48d173cc12fa28ecabc3b837ea3cf6f38c6d1136f85cbaaf598984861466", size = 109391, upload-time = "2025-08-25T13:49:26.313Z" }
+wheels = [
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/18/67/36e9267722cc04a6b9f15c7f3441c2363321a3ea07da7ae0c0707beb2a9c/typing_extensions-4.15.0-py3-none-any.whl", hash = "sha256:f0fa19c6845758ab08074a0cfa8b7aecb71c999ca73d62883bc25cc018c4e548", size = 44614, upload-time = "2025-08-25T13:49:24.86Z" },
+]
+
+[[package]]
+name = "urllib3"
+version = "2.5.0"
+source = { registry = "https://pypi.tuna.tsinghua.edu.cn/simple" }
+sdist = { url = "https://pypi.tuna.tsinghua.edu.cn/packages/15/22/9ee70a2574a4f4599c47dd506532914ce044817c7752a79b6a51286319bc/urllib3-2.5.0.tar.gz", hash = "sha256:3fc47733c7e419d4bc3f6b3dc2b4f890bb743906a30d56ba4a5bfa4bbff92760", size = 393185, upload-time = "2025-06-18T14:07:41.644Z" }
+wheels = [
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/a7/c2/fe1e52489ae3122415c51f387e221dd0773709bad6c6cdaa599e8a2c5185/urllib3-2.5.0-py3-none-any.whl", hash = "sha256:e6b01673c0fa6a13e374b50871808eb3bf7046c4b125b216f6bf1cc604cff0dc", size = 129795, upload-time = "2025-06-18T14:07:40.39Z" },
+]

admin/server/admin_server.py

@@ -20,21 +20,25 @@ import logging
 import time
 import threading
 import traceback
-from werkzeug.serving import run_simple
+import faulthandler
+
 from flask import Flask
+from flask_login import LoginManager
+from werkzeug.serving import run_simple
 from routes import admin_bp
 from common.log_utils import init_root_logger
 from common.constants import SERVICE_CONF
 from common.config_utils import show_configs
-from api import settings
+from common import settings
 from config import load_configurations, SERVICE_CONFIGS
 from auth import init_default_admin, setup_auth
 from flask_session import Session
-from flask_login import LoginManager
+from common.versions import get_ragflow_version
 
 stop_event = threading.Event()
 
 if __name__ == '__main__':
+    faulthandler.enable()
     init_root_logger("admin_service")
     logging.info(r"""
         ____  ___   ______________                 ___       __          _     
@@ -52,6 +56,7 @@ if __name__ == '__main__':
         os.environ.get("MAX_CONTENT_LENGTH", 1024 * 1024 * 1024)
     )
     Session(app)
+    logging.info(f'RAGFlow version: {get_ragflow_version()}')
     show_configs()
     login_manager = LoginManager()
     login_manager.init_app(app)
@@ -67,7 +72,7 @@ if __name__ == '__main__':
             port=9381,
             application=app,
             threaded=True,
-            use_reloader=True,
+            use_reloader=False,
             use_debugger=True,
         )
     except Exception:

admin/server/auth.py

@@ -19,19 +19,20 @@ import logging
 import uuid
 from functools import wraps
 from datetime import datetime
-from flask import request, jsonify
+
+from flask import jsonify, request
 from flask_login import current_user, login_user
 from itsdangerous.url_safe import URLSafeTimedSerializer as Serializer
 
-from api import settings
 from api.common.exceptions import AdminException, UserNotFoundError
-from api.db.init_data import encode_to_base64
+from api.common.base64 import encode_to_base64
 from api.db.services import UserService
 from common.constants import ActiveEnum, StatusEnum
 from api.utils.crypt import decrypt
 from common.misc_utils import get_uuid
 from common.time_utils import current_timestamp, datetime_format, get_format_time
-from common.connection_utils import construct_response
+from common.connection_utils import sync_construct_response
+from common import settings
 
 
 def setup_auth(login_manager):
@@ -129,7 +130,7 @@ def login_admin(email: str, password: str):
     user.last_login_time = get_format_time()
     user.save()
     msg = "Welcome back!"
-    return construct_response(data=resp, auth=user.get_id(), message=msg)
+    return sync_construct_response(data=resp, auth=user.get_id(), message=msg)
 
 
 def check_admin(username: str, password: str):
@@ -169,17 +170,17 @@ def login_verify(f):
         username = auth.parameters['username']
         password = auth.parameters['password']
         try:
-            if check_admin(username, password) is False:
+            if not check_admin(username, password):
                 return jsonify({
                     "code": 500,
                     "message": "Access denied",
                     "data": None
                 }), 200
-        except Exception as e:
-            error_msg = str(e)
+        except Exception:
+            logging.exception("An error occurred during admin login verification.")
             return jsonify({
                 "code": 500,
-                "message": error_msg
+                "message": "An internal server error occurred."
             }), 200
 
         return f(*args, **kwargs)

admin/server/config.py

@@ -25,8 +25,21 @@ from common.config_utils import read_config
 from urllib.parse import urlparse
 
 
+class BaseConfig(BaseModel):
+    id: int
+    name: str
+    host: str
+    port: int
+    service_type: str
+    detail_func_name: str
+
+    def to_dict(self) -> dict[str, Any]:
+        return {'id': self.id, 'name': self.name, 'host': self.host, 'port': self.port,
+                'service_type': self.service_type}
+
+
 class ServiceConfigs:
-    configs = dict
+    configs = list[BaseConfig]
 
     def __init__(self):
         self.configs = []
@@ -45,19 +58,6 @@ class ServiceType(Enum):
     FILE_STORE = "file_store"
 
 
-class BaseConfig(BaseModel):
-    id: int
-    name: str
-    host: str
-    port: int
-    service_type: str
-    detail_func_name: str
-
-    def to_dict(self) -> dict[str, Any]:
-        return {'id': self.id, 'name': self.name, 'host': self.host, 'port': self.port,
-                'service_type': self.service_type}
-
-
 class MetaConfig(BaseConfig):
     meta_type: str
 
@@ -183,11 +183,13 @@ class RAGFlowServerConfig(BaseConfig):
 
 
 class TaskExecutorConfig(BaseConfig):
+    message_queue_type: str
 
     def to_dict(self) -> dict[str, Any]:
         result = super().to_dict()
         if 'extra' not in result:
             result['extra'] = dict()
+        result['extra']['message_queue_type'] = self.message_queue_type
         return result
 
 
@@ -225,7 +227,7 @@ def load_configurations(config_path: str) -> list[BaseConfig]:
     ragflow_count = 0
     id_count = 0
     for k, v in raw_configs.items():
-        match (k):
+        match k:
             case "ragflow":
                 name: str = f'ragflow_{ragflow_count}'
                 host: str = v['host']
@@ -299,6 +301,15 @@ def load_configurations(config_path: str) -> list[BaseConfig]:
                 id_count += 1
             case "admin":
                 pass
+            case "task_executor":
+                name: str = 'task_executor'
+                host: str = v.get('host', '')
+                port: int = v.get('port', 0)
+                message_queue_type: str = v.get('message_queue_type')
+                config = TaskExecutorConfig(id=id_count, name=name, host=host, port=port, message_queue_type=message_queue_type,
+                                            service_type="task_executor", detail_func_name="check_task_executor_alive")
+                configurations.append(config)
+                id_count += 1
             case _:
                 logging.warning(f"Unknown configuration key: {k}")
                 continue

admin/server/responses.py

@@ -13,8 +13,6 @@
 #  See the License for the specific language governing permissions and
 #  limitations under the License.
 #
-
-
 from flask import jsonify
 
 

admin/server/routes.py

@@ -17,17 +17,23 @@
 import secrets
 
 from flask import Blueprint, request
-from flask_login import current_user, logout_user, login_required
+from flask_login import current_user, login_required, logout_user
 
 from auth import login_verify, login_admin, check_admin_auth
 from responses import success_response, error_response
 from services import UserMgr, ServiceMgr, UserServiceMgr
 from roles import RoleMgr
 from api.common.exceptions import AdminException
+from common.versions import get_ragflow_version
 
 admin_bp = Blueprint('admin', __name__, url_prefix='/api/v1/admin')
 
 
+@admin_bp.route('/ping', methods=['GET'])
+def ping():
+    return success_response('PONG')
+
+
 @admin_bp.route('/login', methods=['POST'])
 def login():
     if not request.json:
@@ -369,3 +375,13 @@ def get_user_permission(user_name: str):
         return success_response(res)
     except Exception as e:
         return error_response(str(e), 500)
+
+@admin_bp.route('/version', methods=['GET'])
+@login_required
+@check_admin_auth
+def show_version():
+    try:
+        res = {"version": get_ragflow_version()}
+        return success_response(res)
+    except Exception as e:
+        return error_response(str(e), 500)

admin/server/services.py

@@ -14,7 +14,8 @@
 #  limitations under the License.
 #
 
-
+import os
+import logging
 import re
 from werkzeug.security import check_password_hash
 from common.constants import ActiveEnum
@@ -52,6 +53,7 @@ class UserMgr:
         result = []
         for user in users:
             result.append({
+                'avatar': user.avatar,
                 'email': user.email,
                 'language': user.language,
                 'last_login_time': user.last_login_time,
@@ -170,7 +172,8 @@ class UserServiceMgr:
         return [{
             'title': r['title'],
             'permission': r['permission'],
-            'canvas_category': r['canvas_category'].split('_')[0]
+            'canvas_category': r['canvas_category'].split('_')[0],
+            'avatar': r['avatar']
         } for r in res]
 
 
@@ -178,18 +181,27 @@ class ServiceMgr:
 
     @staticmethod
     def get_all_services():
+        doc_engine = os.getenv('DOC_ENGINE', 'elasticsearch')
         result = []
         configs = SERVICE_CONFIGS.configs
         for service_id, config in enumerate(configs):
             config_dict = config.to_dict()
+            if config_dict['service_type'] == 'retrieval':
+                if config_dict['extra']['retrieval_type'] != doc_engine:
+                    continue
             try:
                 service_detail = ServiceMgr.get_service_details(service_id)
                 if "status" in service_detail:
                     config_dict['status'] = service_detail['status']
                 else:
                     config_dict['status'] = 'timeout'
-            except Exception:
+            except Exception as e:
+                logging.warning(f"Can't get service details, error: {e}")
                 config_dict['status'] = 'timeout'
+            if not config_dict['host']:
+                config_dict['host'] = '-'
+            if not config_dict['port']:
+                config_dict['port'] = '-'
             result.append(config_dict)
         return result
 
@@ -199,17 +211,13 @@ class ServiceMgr:
 
     @staticmethod
     def get_service_details(service_id: int):
-        service_id = int(service_id)
+        service_idx = int(service_id)
         configs = SERVICE_CONFIGS.configs
-        service_config_mapping = {
-            c.id: {
-                'name': c.name,
-                'detail_func_name': c.detail_func_name
-            } for c in configs
-        }
-        service_info = service_config_mapping.get(service_id, {})
-        if not service_info:
-            raise AdminException(f"invalid service_id: {service_id}")
+        if service_idx < 0 or service_idx >= len(configs):
+            raise AdminException(f"invalid service_index: {service_idx}")
+
+        service_config = configs[service_idx]
+        service_info = {'name': service_config.name, 'detail_func_name': service_config.detail_func_name}
 
         detail_func = getattr(health_utils, service_info.get('detail_func_name'))
         res = detail_func()

agent/__init__.py

@@ -13,6 +13,3 @@
 #  See the License for the specific language governing permissions and
 #  limitations under the License.
 #
-
-from beartype.claw import beartype_this_package
-beartype_this_package()

agent/canvas.py

@@ -13,7 +13,10 @@
 #  See the License for the specific language governing permissions and
 #  limitations under the License.
 #
+import asyncio
 import base64
+import inspect
+import binascii
 import json
 import logging
 import re
@@ -26,7 +29,11 @@ from typing import Any, Union, Tuple
 from agent.component import component_class
 from agent.component.base import ComponentBase
 from api.db.services.file_service import FileService
+from api.db.services.llm_service import LLMBundle
+from api.db.services.task_service import has_canceled
+from common.constants import LLMType
 from common.misc_utils import get_uuid, hash_str2int
+from common.exceptions import TaskCanceledException
 from rag.prompts.generator import chunks_format
 from rag.utils.redis_conn import REDIS_CONN
 
@@ -78,14 +85,12 @@ class Graph:
         self.dsl = json.loads(dsl)
         self._tenant_id = tenant_id
         self.task_id = task_id if task_id else get_uuid()
+        self._thread_pool = ThreadPoolExecutor(max_workers=5)
         self.load()
 
     def load(self):
         self.components = self.dsl["components"]
         cpn_nms = set([])
-        for k, cpn in self.components.items():
-            cpn_nms.add(cpn["obj"]["component_name"])
-
         for k, cpn in self.components.items():
             cpn_nms.add(cpn["obj"]["component_name"])
             param = component_class(cpn["obj"]["component_name"] + "Param")()
@@ -126,6 +131,7 @@ class Graph:
             self.components[k]["obj"].reset()
         try:
             REDIS_CONN.delete(f"{self.task_id}-logs")
+            REDIS_CONN.delete(f"{self.task_id}-cancel")
         except Exception as e:
             logging.exception(e)
 
@@ -153,6 +159,33 @@ class Graph:
     def get_tenant_id(self):
         return self._tenant_id
 
+    def get_value_with_variable(self,value: str) -> Any:
+        pat = re.compile(r"\{* *\{([a-zA-Z:0-9]+@[A-Za-z0-9_.-]+|sys\.[A-Za-z0-9_.]+|env\.[A-Za-z0-9_.]+)\} *\}*")
+        out_parts = []
+        last = 0
+
+        for m in pat.finditer(value):
+            out_parts.append(value[last:m.start()])
+            key = m.group(1)
+            v = self.get_variable_value(key)
+            if v is None:
+                rep = ""
+            elif isinstance(v, partial):
+                buf = []
+                for chunk in v():
+                    buf.append(chunk)
+                rep = "".join(buf)
+            elif isinstance(v, str):
+                rep = v
+            else:
+                rep = json.dumps(v, ensure_ascii=False)
+
+            out_parts.append(rep)
+            last = m.end()
+
+        out_parts.append(value[last:])
+        return("".join(out_parts))
+
     def get_variable_value(self, exp: str) -> Any:
         exp = exp.strip("{").strip("}").strip(" ").strip("{").strip("}")
         if exp.find("@") < 0:
@@ -169,7 +202,7 @@ class Graph:
         if not rest:
             return root_val
         return self.get_variable_param_value(root_val,rest)
-    
+
     def get_variable_param_value(self, obj: Any, path: str) -> Any:
         cur = obj
         if not path:
@@ -177,28 +210,84 @@ class Graph:
         for key in path.split('.'):
             if cur is None:
                 return None
+
             if isinstance(cur, str):
                 try:
                     cur = json.loads(cur)
                 except Exception:
                     return None
+
             if isinstance(cur, dict):
                 cur = cur.get(key)
-            else:
-                cur = getattr(cur, key, None)
+                continue
+
+            if isinstance(cur, (list, tuple)):
+                try:
+                    idx = int(key)
+                    cur = cur[idx]
+                except Exception:
+                    return None
+                continue
+
+            cur = getattr(cur, key, None)
         return cur
 
+    def set_variable_value(self, exp: str,value):
+        exp = exp.strip("{").strip("}").strip(" ").strip("{").strip("}")
+        if exp.find("@") < 0:
+            self.globals[exp] = value
+            return
+        cpn_id, var_nm = exp.split("@")
+        cpn = self.get_component(cpn_id)
+        if not cpn:
+            raise Exception(f"Can't find variable: '{cpn_id}@{var_nm}'")
+        parts = var_nm.split(".", 1)
+        root_key = parts[0]
+        rest = parts[1] if len(parts) > 1 else ""
+        if not rest:
+            cpn["obj"].set_output(root_key, value)
+            return
+        root_val = cpn["obj"].output(root_key)
+        if not root_val:
+            root_val = {}
+        cpn["obj"].set_output(root_key, self.set_variable_param_value(root_val,rest,value))
+
+    def set_variable_param_value(self, obj: Any, path: str, value) -> Any:
+        cur = obj
+        keys = path.split('.')
+        if not path:
+            return value
+        for key in keys:
+            if key not in cur or not isinstance(cur[key], dict):
+                cur[key] = {}
+            cur = cur[key]
+        cur[keys[-1]] = value
+        return obj
+
+    def is_canceled(self) -> bool:
+        return has_canceled(self.task_id)
+
+    def cancel_task(self) -> bool:
+        try:
+            REDIS_CONN.set(f"{self.task_id}-cancel", "x")
+        except Exception as e:
+            logging.exception(e)
+            return False
+        return True
+
 
 class Canvas(Graph):
 
-    def __init__(self, dsl: str, tenant_id=None, task_id=None):
+    def __init__(self, dsl: str, tenant_id=None, task_id=None, canvas_id=None):
         self.globals = {
             "sys.query": "",
             "sys.user_id": tenant_id,
             "sys.conversation_turns": 0,
             "sys.files": []
         }
+        self.variables = {}
         super().__init__(dsl, tenant_id, task_id)
+        self._id = canvas_id
 
     def load(self):
         super().load()
@@ -212,7 +301,11 @@ class Canvas(Graph):
             "sys.conversation_turns": 0,
             "sys.files": []
         }
-            
+        if "variables" in self.dsl:
+            self.variables = self.dsl["variables"]
+        else:
+            self.variables = {}
+
         self.retrieval = self.dsl["retrieval"]
         self.memory = self.dsl.get("memory", [])
 
@@ -228,32 +321,67 @@ class Canvas(Graph):
             self.history = []
             self.retrieval = []
             self.memory = []
+        print(self.variables)
         for k in self.globals.keys():
-            if isinstance(self.globals[k], str):
-                self.globals[k] = ""
-            elif isinstance(self.globals[k], int):
-                self.globals[k] = 0
-            elif isinstance(self.globals[k], float):
-                self.globals[k] = 0
-            elif isinstance(self.globals[k], list):
-                self.globals[k] = []
-            elif isinstance(self.globals[k], dict):
-                self.globals[k] = {}
-            else:
-                self.globals[k] = None
+            if k.startswith("sys."):
+                if isinstance(self.globals[k], str):
+                    self.globals[k] = ""
+                elif isinstance(self.globals[k], int):
+                    self.globals[k] = 0
+                elif isinstance(self.globals[k], float):
+                    self.globals[k] = 0
+                elif isinstance(self.globals[k], list):
+                    self.globals[k] = []
+                elif isinstance(self.globals[k], dict):
+                    self.globals[k] = {}
+                else:
+                    self.globals[k] = None
+            if k.startswith("env."):
+                key = k[4:]
+                if key in self.variables:
+                    variable = self.variables[key]
+                    if variable["value"]:
+                        self.globals[k] = variable["value"]
+                    else:
+                        if variable["type"] == "string":
+                            self.globals[k] = ""
+                        elif variable["type"] == "number":
+                            self.globals[k] = 0
+                        elif variable["type"] == "boolean":
+                            self.globals[k] = False
+                        elif variable["type"] == "object":
+                            self.globals[k] = {}
+                        elif variable["type"].startswith("array"):
+                            self.globals[k] = []
+                        else:
+                            self.globals[k] = ""
+                else:
+                    self.globals[k] = ""
 
-    def run(self, **kwargs):
+    async def run(self, **kwargs):
         st = time.perf_counter()
+        self._loop = asyncio.get_running_loop()
         self.message_id = get_uuid()
         created_at = int(time.time())
         self.add_user_input(kwargs.get("query"))
         for k, cpn in self.components.items():
             self.components[k]["obj"].reset(True)
 
+        if kwargs.get("webhook_payload"):
+            for k, cpn in self.components.items():
+                if self.components[k]["obj"].component_name.lower() == "begin"  and self.components[k]["obj"]._param.mode == "Webhook":
+                    payload = kwargs.get("webhook_payload", {})
+                    if "input" in payload:
+                        self.components[k]["obj"].set_input_value("request", payload["input"])
+                    for kk, vv in payload.items():
+                        if kk == "input":
+                            continue
+                        self.components[k]["obj"].set_output(kk, vv)
+
         for k in kwargs.keys():
             if k in ["query", "user_id", "files"] and kwargs[k]:
                 if k == "files":
-                    self.globals[f"sys.{k}"] = self.get_files(kwargs[k])
+                    self.globals[f"sys.{k}"] = await self.get_files_async(kwargs[k])
                 else:
                     self.globals[f"sys.{k}"] = kwargs[k]
         if not self.globals["sys.conversation_turns"] :
@@ -275,20 +403,58 @@ class Canvas(Graph):
             self.path.append("begin")
             self.retrieval.append({"chunks": [], "doc_aggs": []})
 
+        if self.is_canceled():
+            msg = f"Task {self.task_id} has been canceled before starting."
+            logging.info(msg)
+            raise TaskCanceledException(msg)
+
         yield decorate("workflow_started", {"inputs": kwargs.get("inputs")})
         self.retrieval.append({"chunks": {}, "doc_aggs": {}})
 
-        def _run_batch(f, t):
-            with ThreadPoolExecutor(max_workers=5) as executor:
-                thr = []
-                for i in range(f, t):
-                    cpn = self.get_component_obj(self.path[i])
-                    if cpn.component_name.lower() in ["begin", "userfillup"]:
-                        thr.append(executor.submit(cpn.invoke, inputs=kwargs.get("inputs", {})))
+        async def _run_batch(f, t):
+            if self.is_canceled():
+                msg = f"Task {self.task_id} has been canceled during batch execution."
+                logging.info(msg)
+                raise TaskCanceledException(msg)
+
+            loop = asyncio.get_running_loop()
+            tasks = []
+
+            def _run_async_in_thread(coro_func, **call_kwargs):
+                return asyncio.run(coro_func(**call_kwargs))
+
+            i = f
+            while i < t:
+                cpn = self.get_component_obj(self.path[i])
+                task_fn = None
+                call_kwargs = None
+
+                if cpn.component_name.lower() in ["begin", "userfillup"]:
+                    call_kwargs = {"inputs": kwargs.get("inputs", {})}
+                    task_fn = cpn.invoke
+                    i += 1
+                else:
+                    for _, ele in cpn.get_input_elements().items():
+                        if isinstance(ele, dict) and ele.get("_cpn_id") and ele.get("_cpn_id") not in self.path[:i] and self.path[0].lower().find("userfillup") < 0:
+                            self.path.pop(i)
+                            t -= 1
+                            break
                     else:
-                        thr.append(executor.submit(cpn.invoke, **cpn.get_input()))
-                for t in thr:
-                    t.result()
+                        call_kwargs = cpn.get_input()
+                        task_fn = cpn.invoke
+                        i += 1
+
+                if task_fn is None:
+                    continue
+
+                invoke_async = getattr(cpn, "invoke_async", None)
+                if invoke_async and asyncio.iscoroutinefunction(invoke_async):
+                    tasks.append(loop.run_in_executor(self._thread_pool, partial(_run_async_in_thread, invoke_async, **(call_kwargs or {}))))
+                else:
+                    tasks.append(loop.run_in_executor(self._thread_pool, partial(task_fn, **(call_kwargs or {}))))
+
+            if tasks:
+                await asyncio.gather(*tasks)
 
         def _node_finished(cpn_obj):
             return decorate("node_finished",{
@@ -305,6 +471,7 @@ class Canvas(Graph):
         self.error = ""
         idx = len(self.path) - 1
         partials = []
+        tts_mdl = None
         while idx < len(self.path):
             to = len(self.path)
             for i in range(idx, to):
@@ -315,30 +482,72 @@ class Canvas(Graph):
                     "component_type": self.get_component_type(self.path[i]),
                     "thoughts": self.get_component_thoughts(self.path[i])
                 })
-            _run_batch(idx, to)
-            # post processing of components invocation
+            await _run_batch(idx, to)
+            to = len(self.path)
+            # post-processing of components invocation
             for i in range(idx, to):
                 cpn = self.get_component(self.path[i])
                 cpn_obj = self.get_component_obj(self.path[i])
                 if cpn_obj.component_name.lower() == "message":
+                    if cpn_obj.get_param("auto_play"):
+                        tts_mdl = LLMBundle(self._tenant_id, LLMType.TTS)
                     if isinstance(cpn_obj.output("content"), partial):
                         _m = ""
-                        for m in cpn_obj.output("content")():
+                        buff_m = ""
+                        stream = cpn_obj.output("content")()
+                        async def _process_stream(m):
+                            nonlocal buff_m, _m, tts_mdl
                             if not m:
-                                continue
+                                return
                             if m == "<think>":
-                                yield decorate("message", {"content": "", "start_to_think": True})
+                                return decorate("message", {"content": "", "start_to_think": True})
+
                             elif m == "</think>":
-                                yield decorate("message", {"content": "", "end_to_think": True})
-                            else:
-                                yield decorate("message", {"content": m})
-                                _m += m
+                                return decorate("message", {"content": "", "end_to_think": True})
+
+                            buff_m += m
+                            _m += m
+
+                            if len(buff_m) > 16:
+                                ev = decorate(
+                                    "message",
+                                    {
+                                        "content": m,
+                                        "audio_binary": self.tts(tts_mdl, buff_m)
+                                    }
+                                )
+                                buff_m = ""
+                                return ev
+
+                            return decorate("message", {"content": m})
+
+                        if inspect.isasyncgen(stream):
+                            async for m in stream:
+                                ev= await _process_stream(m)
+                                if ev:
+                                    yield ev
+                        else:
+                            for m in stream:
+                                ev= await _process_stream(m)
+                                if ev:
+                                    yield ev
+                        if buff_m:
+                            yield decorate("message", {"content": "", "audio_binary": self.tts(tts_mdl, buff_m)})
+                            buff_m = ""
                         cpn_obj.set_output("content", _m)
                         cite = re.search(r"\[ID:[ 0-9]+\]", _m)
                     else:
                         yield decorate("message", {"content": cpn_obj.output("content")})
                         cite = re.search(r"\[ID:[ 0-9]+\]",  cpn_obj.output("content"))
-                    yield decorate("message_end", {"reference": self.get_reference() if cite else None})
+
+                    message_end = {}
+                    if cpn_obj.get_param("status"):
+                        message_end["status"] = cpn_obj.get_param("status")
+                    if isinstance(cpn_obj.output("attachment"), dict):
+                        message_end["attachment"] = cpn_obj.output("attachment")
+                    if cite:
+                        message_end["reference"] = self.get_reference()
+                    yield decorate("message_end", message_end)
 
                     while partials:
                         _cpn_obj = self.get_component_obj(partials[0])
@@ -359,7 +568,7 @@ class Canvas(Graph):
                     else:
                         self.error = cpn_obj.error()
 
-                if cpn_obj.component_name.lower() != "iteration":
+                if cpn_obj.component_name.lower() not in ("iteration","loop"):
                     if isinstance(cpn_obj.output("content"), partial):
                         if self.error:
                             cpn_obj.set_output("content", None)
@@ -384,14 +593,16 @@ class Canvas(Graph):
                     for cpn_id in cpn_ids:
                         _append_path(cpn_id)
 
-                if cpn_obj.component_name.lower() == "iterationitem" and cpn_obj.end():
+                if cpn_obj.component_name.lower() in ("iterationitem","loopitem") and cpn_obj.end():
                     iter = cpn_obj.get_parent()
                     yield _node_finished(iter)
                     _extend_path(self.get_component(cpn["parent_id"])["downstream"])
                 elif cpn_obj.component_name.lower() in ["categorize", "switch"]:
                     _extend_path(cpn_obj.output("_next"))
-                elif cpn_obj.component_name.lower() == "iteration":
+                elif cpn_obj.component_name.lower() in ("iteration", "loop"):
                     _append_path(cpn_obj.get_start())
+                elif cpn_obj.component_name.lower() == "exitloop" and cpn_obj.get_parent().component_name.lower() == "loop":
+                    _extend_path(self.get_component(cpn["parent_id"])["downstream"])
                 elif not cpn["downstream"] and cpn_obj.get_parent():
                     _append_path(cpn_obj.get_parent().get_start())
                 else:
@@ -410,9 +621,10 @@ class Canvas(Graph):
                 for c in path:
                     o = self.get_component_obj(c)
                     if o.component_name.lower() == "userfillup":
+                        o.invoke()
                         another_inputs.update(o.get_input_elements())
                         if o.get_param("enable_tips"):
-                            tips = o.get_param("tips")
+                            tips = o.output("tips")
                 self.path = path
                 yield decorate("user_inputs", {"inputs": another_inputs, "tips": tips})
                 return
@@ -426,6 +638,14 @@ class Canvas(Graph):
                            "created_at": st,
                        })
             self.history.append(("assistant", self.get_component_obj(self.path[-1]).output()))
+        elif "Task has been canceled" in self.error:
+            yield decorate("workflow_finished",
+                       {
+                           "inputs": kwargs.get("inputs"),
+                           "outputs": "Task has been canceled",
+                           "elapsed_time": time.perf_counter() - st,
+                           "created_at": st,
+                       })
 
     def is_reff(self, exp: str) -> bool:
         exp = exp.strip("{").strip("}")
@@ -438,6 +658,50 @@ class Canvas(Graph):
             return False
         return True
 
+
+    def tts(self,tts_mdl, text):
+        def clean_tts_text(text: str) -> str:
+            if not text:
+                return ""
+
+            text = text.encode("utf-8", "ignore").decode("utf-8", "ignore")
+
+            text = re.sub(r"[\x00-\x08\x0B-\x0C\x0E-\x1F\x7F]", "", text)
+
+            emoji_pattern = re.compile(
+                "[\U0001F600-\U0001F64F"
+                "\U0001F300-\U0001F5FF"
+                "\U0001F680-\U0001F6FF"
+                "\U0001F1E0-\U0001F1FF"
+                "\U00002700-\U000027BF"
+                "\U0001F900-\U0001F9FF"
+                "\U0001FA70-\U0001FAFF"
+                "\U0001FAD0-\U0001FAFF]+",
+                flags=re.UNICODE
+            )
+            text = emoji_pattern.sub("", text)
+
+            text = re.sub(r"\s+", " ", text).strip()
+
+            MAX_LEN = 500
+            if len(text) > MAX_LEN:
+                text = text[:MAX_LEN]
+
+            return text
+        if not tts_mdl or not text:
+            return None
+        text = clean_tts_text(text)
+        if not text:
+            return None
+        bin = b""
+        try:
+            for chunk in tts_mdl.tts(text):
+                bin += chunk
+        except Exception as e:
+            logging.error(f"TTS failed: {e}, text={text!r}")
+            return None
+        return binascii.hexlify(bin).decode("utf-8")
+
     def get_history(self, window_size):
         convs = []
         if window_size <= 0:
@@ -458,6 +722,9 @@ class Canvas(Graph):
     def get_mode(self):
         return self.components["begin"]["obj"]._param.mode
 
+    def get_sys_query(self):
+        return self.globals.get("sys.query", "")
+
     def set_global_param(self, **kwargs):
         self.globals.update(kwargs)
 
@@ -467,20 +734,30 @@ class Canvas(Graph):
     def get_component_input_elements(self, cpnnm):
         return self.components[cpnnm]["obj"].get_input_elements()
 
-    def get_files(self, files: Union[None, list[dict]]) -> list[str]:
+    async def get_files_async(self, files: Union[None, list[dict]]) -> list[str]:
         if not files:
             return  []
         def image_to_base64(file):
             return "data:{};base64,{}".format(file["mime_type"],
                                         base64.b64encode(FileService.get_blob(file["created_by"], file["id"])).decode("utf-8"))
-        exe = ThreadPoolExecutor(max_workers=5)
-        threads = []
+        loop = asyncio.get_running_loop()
+        tasks = []
         for file in files:
             if file["mime_type"].find("image") >=0:
-                threads.append(exe.submit(image_to_base64, file))
+                tasks.append(loop.run_in_executor(self._thread_pool, image_to_base64, file))
                 continue
-            threads.append(exe.submit(FileService.parse, file["name"], FileService.get_blob(file["created_by"], file["id"]), True, file["created_by"]))
-        return [th.result() for th in threads]
+            tasks.append(loop.run_in_executor(self._thread_pool, FileService.parse, file["name"], FileService.get_blob(file["created_by"], file["id"]), True, file["created_by"]))
+        return await asyncio.gather(*tasks)
+
+    def get_files(self, files: Union[None, list[dict]]) -> list[str]:
+        """
+        Synchronous wrapper for get_files_async, used by sync component invoke paths.
+        """
+        loop = getattr(self, "_loop", None)
+        if loop and loop.is_running():
+            return asyncio.run_coroutine_threadsafe(self.get_files_async(files), loop).result()
+
+        return asyncio.run(self.get_files_async(files))
 
     def tool_use_callback(self, agent_id: str, func_name: str, params: dict, result: Any, elapsed_time=None):
         agent_ids = agent_id.split("-->")
@@ -534,4 +811,3 @@ class Canvas(Graph):
 
     def get_component_thoughts(self, cpn_id) -> str:
         return self.components.get(cpn_id)["obj"].thoughts()
-

agent/component/__init__.py

@@ -13,7 +13,6 @@
 #  See the License for the specific language governing permissions and
 #  limitations under the License.
 #
-
 import os
 import importlib
 import inspect
@@ -50,9 +49,10 @@ del _package_path, _import_submodules, _extract_classes_from_module
 
 
 def component_class(class_name):
-    for mdl in ["agent.component", "agent.tools", "rag.flow"]:
+    for module_name in ["agent.component", "agent.tools", "rag.flow"]:
         try:
-            return getattr(importlib.import_module(mdl), class_name)
+            return getattr(importlib.import_module(module_name), class_name)
         except Exception:
+            # logging.warning(f"Can't import module: {module_name}, error: {e}")
             pass
     assert False, f"Can't import {class_name}"

agent/component/agent_with_tools.py

@@ -13,10 +13,11 @@
 #  See the License for the specific language governing permissions and
 #  limitations under the License.
 #
+import asyncio
+import json
 import logging
 import os
 import re
-from concurrent.futures import ThreadPoolExecutor
 from copy import deepcopy
 from functools import partial
 from typing import Any
@@ -28,9 +29,9 @@ from api.db.services.llm_service import LLMBundle
 from api.db.services.tenant_llm_service import TenantLLMService
 from api.db.services.mcp_server_service import MCPServerService
 from common.connection_utils import timeout
-from rag.prompts.generator import next_step, COMPLETE_TASK, analyze_task, \
-    citation_prompt, reflect, rank_memories, kb_prompt, citation_plus, full_question, message_fit_in
-from rag.utils.mcp_tool_call_conn import MCPToolCallSession, mcp_tool_metadata_to_openai_tool
+from rag.prompts.generator import next_step_async, COMPLETE_TASK, \
+    citation_prompt, kb_prompt, citation_plus, full_question, message_fit_in, structured_output_prompt
+from common.mcp_tool_call_conn import MCPToolCallSession, mcp_tool_metadata_to_openai_tool
 from agent.component.llm import LLMParam, LLM
 
 
@@ -83,9 +84,11 @@ class Agent(LLM, ToolBase):
     def __init__(self, canvas, id, param: LLMParam):
         LLM.__init__(self, canvas, id, param)
         self.tools = {}
-        for cpn in self._param.tools:
+        for idx, cpn in enumerate(self._param.tools):
             cpn = self._load_tool_obj(cpn)
-            self.tools[cpn.get_meta()["function"]["name"]] = cpn
+            original_name = cpn.get_meta()["function"]["name"]
+            indexed_name = f"{original_name}_{idx}"
+            self.tools[indexed_name] = cpn
 
         self.chat_mdl = LLMBundle(self._canvas.get_tenant_id(), TenantLLMService.llm_id2llm_type(self._param.llm_id), self._param.llm_id,
                                   max_retries=self._param.max_retries,
@@ -93,7 +96,12 @@ class Agent(LLM, ToolBase):
                                   max_rounds=self._param.max_rounds,
                                   verbose_tool_use=True
                                   )
-        self.tool_meta = [v.get_meta() for _,v in self.tools.items()]
+        self.tool_meta = []
+        for indexed_name, tool_obj in self.tools.items():
+            original_meta = tool_obj.get_meta()
+            indexed_meta = deepcopy(original_meta)
+            indexed_meta["function"]["name"] = indexed_name
+            self.tool_meta.append(indexed_meta)
 
         for mcp in self._param.mcp:
             _, mcp_server = MCPServerService.get_by_id(mcp["mcp_id"])
@@ -107,7 +115,8 @@ class Agent(LLM, ToolBase):
 
     def _load_tool_obj(self, cpn: dict) -> object:
         from agent.component import component_class
-        param = component_class(cpn["component_name"] + "Param")()
+        tool_name = cpn["component_name"]
+        param = component_class(tool_name + "Param")()
         param.update(cpn["params"])
         try:
             param.check()
@@ -137,8 +146,37 @@ class Agent(LLM, ToolBase):
             res.update(cpn.get_input_form())
         return res
 
-    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 20*60)))
+    def _get_output_schema(self):
+        try:
+            cand = self._param.outputs.get("structured")
+        except Exception:
+            return None
+
+        if isinstance(cand, dict):
+            if isinstance(cand.get("properties"), dict) and len(cand["properties"]) > 0:
+                return cand
+            for k in ("schema", "structured"):
+                if isinstance(cand.get(k), dict) and isinstance(cand[k].get("properties"), dict) and len(cand[k]["properties"]) > 0:
+                    return cand[k]
+
+        return None
+
+    async def _force_format_to_schema_async(self, text: str, schema_prompt: str) -> str:
+        fmt_msgs = [
+            {"role": "system", "content": schema_prompt + "\nIMPORTANT: Output ONLY valid JSON. No markdown, no extra text."},
+            {"role": "user", "content": text},
+        ]
+        _, fmt_msgs = message_fit_in(fmt_msgs, int(self.chat_mdl.max_length * 0.97))
+        return await self._generate_async(fmt_msgs)
+
     def _invoke(self, **kwargs):
+        return asyncio.run(self._invoke_async(**kwargs))
+
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 20*60)))
+    async def _invoke_async(self, **kwargs):
+        if self.check_if_canceled("Agent processing"):
+            return
+
         if kwargs.get("user_prompt"):
             usr_pmt = ""
             if kwargs.get("reasoning"):
@@ -152,25 +190,29 @@ class Agent(LLM, ToolBase):
             self._param.prompts = [{"role": "user", "content": usr_pmt}]
 
         if not self.tools:
-            return LLM._invoke(self, **kwargs)
+            if self.check_if_canceled("Agent processing"):
+                return
+            return await LLM._invoke_async(self, **kwargs)
 
         prompt, msg, user_defined_prompt = self._prepare_prompt_variables()
+        output_schema = self._get_output_schema()
+        schema_prompt = ""
+        if output_schema:
+            schema = json.dumps(output_schema, ensure_ascii=False, indent=2)
+            schema_prompt = structured_output_prompt(schema)
 
         downstreams = self._canvas.get_component(self._id)["downstream"] if self._canvas.get_component(self._id) else []
         ex = self.exception_handler()
-        output_structure=None
-        try:
-            output_structure=self._param.outputs['structured']
-        except Exception:
-            pass
-        if any([self._canvas.get_component_obj(cid).component_name.lower()=="message" for cid in downstreams]) and not output_structure and not (ex and ex["goto"]):
-            self.set_output("content", partial(self.stream_output_with_tools, prompt, msg, user_defined_prompt))
+        if any([self._canvas.get_component_obj(cid).component_name.lower()=="message" for cid in downstreams]) and not (ex and ex["goto"]) and not output_schema:
+            self.set_output("content", partial(self.stream_output_with_tools_async, prompt, deepcopy(msg), user_defined_prompt))
             return
 
         _, msg = message_fit_in([{"role": "system", "content": prompt}, *msg], int(self.chat_mdl.max_length * 0.97))
         use_tools = []
         ans = ""
-        for delta_ans, tk in self._react_with_tools_streamly(prompt, msg, use_tools, user_defined_prompt):
+        async for delta_ans, _tk in self._react_with_tools_streamly_async_simple(prompt, msg, use_tools, user_defined_prompt,schema_prompt=schema_prompt):
+            if self.check_if_canceled("Agent processing"):
+                return
             ans += delta_ans
 
         if ans.find("**ERROR**") >= 0:
@@ -181,22 +223,48 @@ class Agent(LLM, ToolBase):
                 self.set_output("_ERROR", ans)
             return
 
+        if output_schema:
+            error = ""
+            for _ in range(self._param.max_retries + 1):
+                try:
+                    def clean_formated_answer(ans: str) -> str:
+                        ans = re.sub(r"^.*</think>", "", ans, flags=re.DOTALL)
+                        ans = re.sub(r"^.*```json", "", ans, flags=re.DOTALL)
+                        return re.sub(r"```\n*$", "", ans, flags=re.DOTALL)
+                    obj = json_repair.loads(clean_formated_answer(ans))
+                    self.set_output("structured", obj)
+                    if use_tools:
+                        self.set_output("use_tools", use_tools)
+                    return obj
+                except Exception:
+                    error = "The answer cannot be parsed as JSON"
+                    ans = await self._force_format_to_schema_async(ans, schema_prompt)
+                    if ans.find("**ERROR**") >= 0:
+                        continue
+
+            self.set_output("_ERROR", error)
+            return
+
         self.set_output("content", ans)
         if use_tools:
             self.set_output("use_tools", use_tools)
         return ans
 
-    def stream_output_with_tools(self, prompt, msg, user_defined_prompt={}):
+    async def stream_output_with_tools_async(self, prompt, msg, user_defined_prompt={}):
         _, msg = message_fit_in([{"role": "system", "content": prompt}, *msg], int(self.chat_mdl.max_length * 0.97))
         answer_without_toolcall = ""
         use_tools = []
-        for delta_ans,_ in self._react_with_tools_streamly(prompt, msg, use_tools, user_defined_prompt):
+        async for delta_ans, _ in self._react_with_tools_streamly_async_simple(prompt, msg, use_tools, user_defined_prompt):
+            if self.check_if_canceled("Agent streaming"):
+                return
+
             if delta_ans.find("**ERROR**") >= 0:
                 if self.get_exception_default_value():
                     self.set_output("content", self.get_exception_default_value())
                     yield self.get_exception_default_value()
                 else:
                     self.set_output("_ERROR", delta_ans)
+                    return
             answer_without_toolcall += delta_ans
             yield delta_ans
 
@@ -204,55 +272,58 @@ class Agent(LLM, ToolBase):
         if use_tools:
             self.set_output("use_tools", use_tools)
 
-    def _gen_citations(self, text):
-        retrievals = self._canvas.get_reference()
-        retrievals = {"chunks": list(retrievals["chunks"].values()), "doc_aggs": list(retrievals["doc_aggs"].values())}
-        formated_refer = kb_prompt(retrievals, self.chat_mdl.max_length, True)
-        for delta_ans in self._generate_streamly([{"role": "system", "content": citation_plus("\n\n".join(formated_refer))},
-                                                  {"role": "user", "content": text}
-                                                  ]):
-            yield delta_ans
-
-    def _react_with_tools_streamly(self, prompt, history: list[dict], use_tools, user_defined_prompt={}):
+    async def _react_with_tools_streamly_async_simple(self, prompt, history: list[dict], use_tools, user_defined_prompt={}, schema_prompt: str = ""):
         token_count = 0
         tool_metas = self.tool_meta
         hist = deepcopy(history)
         last_calling = ""
         if len(hist) > 3:
             st = timer()
-            user_request = full_question(messages=history, chat_mdl=self.chat_mdl)
+            user_request = await full_question(messages=history, chat_mdl=self.chat_mdl)
             self.callback("Multi-turn conversation optimization", {}, user_request, elapsed_time=timer()-st)
         else:
             user_request = history[-1]["content"]
 
-        def use_tool(name, args):
-            nonlocal hist, use_tools, token_count,last_calling,user_request
+        def build_task_desc(prompt: str, user_request: str, user_defined_prompt: dict | None = None) -> str:
+            """Build a minimal task_desc by concatenating prompt, query, and tool schemas."""
+            user_defined_prompt = user_defined_prompt or {}
+
+            task_desc = (
+                "### Agent Prompt\n"
+                f"{prompt}\n\n"
+                "### User Request\n"
+                f"{user_request}\n\n"
+            )
+
+            if user_defined_prompt:
+                udp_json = json.dumps(user_defined_prompt, ensure_ascii=False, indent=2)
+                task_desc += "\n### User Defined Prompts\n" + udp_json + "\n"
+
+            return task_desc
+
+
+        async def use_tool_async(name, args):
+            nonlocal hist, use_tools, last_calling
             logging.info(f"{last_calling=} == {name=}")
-            # Summarize of function calling
-            #if all([
-            #    isinstance(self.toolcall_session.get_tool_obj(name), Agent),
-            #    last_calling,
-            #    last_calling != name
-            #]):
-            #    self.toolcall_session.get_tool_obj(name).add2system_prompt(f"The chat history with other agents are as following: \n" + self.get_useful_memory(user_request, str(args["user_prompt"]),user_defined_prompt))
             last_calling = name
-            tool_response = self.toolcall_session.tool_call(name, args)
+            tool_response = await self.toolcall_session.tool_call_async(name, args)
             use_tools.append({
                 "name": name,
                 "arguments": args,
                 "results": tool_response
             })
-            # self.callback("add_memory", {}, "...")
-            #self.add_memory(hist[-2]["content"], hist[-1]["content"], name, args, str(tool_response), user_defined_prompt)
-
             return name, tool_response
 
-        def complete():
+        async def complete():
             nonlocal hist
             need2cite = self._param.cite and self._canvas.get_reference()["chunks"] and self._id.find("-->") < 0
+            if schema_prompt:
+                need2cite = False
             cited = False
-            if hist[0]["role"] == "system" and need2cite:
-                if len(hist) < 7:
+            if hist and hist[0]["role"] == "system":
+                if schema_prompt:
+                    hist[0]["content"] += "\n" + schema_prompt
+                if need2cite and len(hist) < 7:
                     hist[0]["content"] += citation_prompt()
                     cited = True
             yield "", token_count
@@ -261,7 +332,7 @@ class Agent(LLM, ToolBase):
             if len(hist) > 12:
                 _hist = [hist[0], hist[1], *hist[-10:]]
             entire_txt = ""
-            for delta_ans in self._generate_streamly(_hist):
+            async for delta_ans in self._generate_streamly(_hist):
                 if not need2cite or cited:
                     yield delta_ans, 0
                 entire_txt += delta_ans
@@ -270,12 +341,29 @@ class Agent(LLM, ToolBase):
 
             st = timer()
             txt = ""
-            for delta_ans in self._gen_citations(entire_txt):
+            async for delta_ans in self._gen_citations_async(entire_txt):
+                if self.check_if_canceled("Agent streaming"):
+                    return
                 yield delta_ans, 0
                 txt += delta_ans
 
             self.callback("gen_citations", {}, txt, elapsed_time=timer()-st)
 
+        def build_observation(tool_call_res: list[tuple]) -> str:
+            """
+            Build a Observation from tool call results.
+            No LLM involved.
+            """
+            if not tool_call_res:
+                return ""
+
+            lines = ["Observation:"]
+            for name, result in tool_call_res:
+                lines.append(f"[{name} result]")
+                lines.append(str(result))
+
+            return "\n".join(lines)
+
         def append_user_content(hist, content):
             if hist[-1]["role"] == "user":
                 hist[-1]["content"] += content
@@ -283,12 +371,14 @@ class Agent(LLM, ToolBase):
                 hist.append({"role": "user", "content": content})
 
         st = timer()
-        task_desc = analyze_task(self.chat_mdl, prompt, user_request, tool_metas, user_defined_prompt)
+        task_desc = build_task_desc(prompt, user_request, user_defined_prompt)
         self.callback("analyze_task", {}, task_desc, elapsed_time=timer()-st)
         for _ in range(self._param.max_rounds + 1):
-            response, tk = next_step(self.chat_mdl, hist, tool_metas, task_desc, user_defined_prompt)
+            if self.check_if_canceled("Agent streaming"):
+                return
+            response, tk = await next_step_async(self.chat_mdl, hist, tool_metas, task_desc, user_defined_prompt)
             # self.callback("next_step", {}, str(response)[:256]+"...")
-            token_count += tk
+            token_count += tk or 0
             hist.append({"role": "assistant", "content": response})
             try:
                 functions = json_repair.loads(re.sub(r"```.*", "", response))
@@ -297,23 +387,24 @@ class Agent(LLM, ToolBase):
                 for f in functions:
                     if not isinstance(f, dict):
                         raise TypeError(f"An object type should be returned, but `{f}`")
-                with ThreadPoolExecutor(max_workers=5) as executor:
-                    thr = []
-                    for func in functions:
-                        name = func["name"]
-                        args = func["arguments"]
-                        if name == COMPLETE_TASK:
-                            append_user_content(hist, f"Respond with a formal answer. FORGET(DO NOT mention) about `{COMPLETE_TASK}`. The language for the response MUST be as the same as the first user request.\n")
-                            for txt, tkcnt in complete():
-                                yield txt, tkcnt
-                            return
 
-                        thr.append(executor.submit(use_tool, name, args))
+                tool_tasks = []
+                for func in functions:
+                    name = func["name"]
+                    args = func["arguments"]
+                    if name == COMPLETE_TASK:
+                        append_user_content(hist, f"Respond with a formal answer. FORGET(DO NOT mention) about `{COMPLETE_TASK}`. The language for the response MUST be as the same as the first user request.\n")
+                        async for txt, tkcnt in complete():
+                            yield txt, tkcnt
+                        return
 
-                    st = timer()
-                    reflection = reflect(self.chat_mdl, hist, [th.result() for th in thr], user_defined_prompt)
-                    append_user_content(hist, reflection)
-                    self.callback("reflection", {}, str(reflection), elapsed_time=timer()-st)
+                    tool_tasks.append(asyncio.create_task(use_tool_async(name, args)))
+
+                results = await asyncio.gather(*tool_tasks) if tool_tasks else []
+                st = timer()
+                reflection = build_observation(results)
+                append_user_content(hist, reflection)
+                self.callback("reflection", {}, str(reflection), elapsed_time=timer()-st)
 
             except Exception as e:
                 logging.exception(msg=f"Wrong JSON argument format in LLM ReAct response: {e}")
@@ -333,29 +424,163 @@ Instructions:
 6. Focus on delivering VALUE with the information already gathered
 Respond immediately with your final comprehensive answer.
         """
+        if self.check_if_canceled("Agent final instruction"):
+            return
         append_user_content(hist, final_instruction)
 
-        for txt, tkcnt in complete():
+        async for txt, tkcnt in complete():
             yield txt, tkcnt
 
-    def get_useful_memory(self, goal: str, sub_goal:str, topn=3, user_defined_prompt:dict={}) -> str:
-        # self.callback("get_useful_memory", {"topn": 3}, "...")
-        mems = self._canvas.get_memory()
-        rank = rank_memories(self.chat_mdl, goal, sub_goal, [summ for (user, assist, summ) in mems], user_defined_prompt)
-        try:
-            rank = json_repair.loads(re.sub(r"```.*", "", rank))[:topn]
-            mems = [mems[r] for r in rank]
-            return "\n\n".join([f"User: {u}\nAgent: {a}" for u, a,_ in mems])
-        except Exception as e:
-            logging.exception(e)
+#     async def _react_with_tools_streamly_async(self, prompt, history: list[dict], use_tools, user_defined_prompt={}, schema_prompt: str = ""):
+#         token_count = 0
+#         tool_metas = self.tool_meta
+#         hist = deepcopy(history)
+#         last_calling = ""
+#         if len(hist) > 3:
+#             st = timer()
+#             user_request = await full_question(messages=history, chat_mdl=self.chat_mdl)
+#             self.callback("Multi-turn conversation optimization", {}, user_request, elapsed_time=timer()-st)
+#         else:
+#             user_request = history[-1]["content"]
 
-        return "Error occurred."
+#         async def use_tool_async(name, args):
+#             nonlocal hist, use_tools, last_calling
+#             logging.info(f"{last_calling=} == {name=}")
+#             last_calling = name
+#             tool_response = await self.toolcall_session.tool_call_async(name, args)
+#             use_tools.append({
+#                 "name": name,
+#                 "arguments": args,
+#                 "results": tool_response
+#             })
+#             # self.callback("add_memory", {}, "...")
+#             #self.add_memory(hist[-2]["content"], hist[-1]["content"], name, args, str(tool_response), user_defined_prompt)
 
-    def reset(self, temp=False):
+#             return name, tool_response
+
+#         async def complete():
+#             nonlocal hist
+#             need2cite = self._param.cite and self._canvas.get_reference()["chunks"] and self._id.find("-->") < 0
+#             if schema_prompt:
+#                 need2cite = False
+#             cited = False
+#             if hist and hist[0]["role"] == "system":
+#                 if schema_prompt:
+#                     hist[0]["content"] += "\n" + schema_prompt
+#                 if need2cite and len(hist) < 7:
+#                     hist[0]["content"] += citation_prompt()
+#                     cited = True
+#             yield "", token_count
+
+#             _hist = hist
+#             if len(hist) > 12:
+#                 _hist = [hist[0], hist[1], *hist[-10:]]
+#             entire_txt = ""
+#             async for delta_ans in self._generate_streamly(_hist):
+#                 if not need2cite or cited:
+#                     yield delta_ans, 0
+#                 entire_txt += delta_ans
+#             if not need2cite or cited:
+#                 return
+
+#             st = timer()
+#             txt = ""
+#             async for delta_ans in self._gen_citations_async(entire_txt):
+#                 if self.check_if_canceled("Agent streaming"):
+#                     return
+#                 yield delta_ans, 0
+#                 txt += delta_ans
+
+#             self.callback("gen_citations", {}, txt, elapsed_time=timer()-st)
+
+#         def append_user_content(hist, content):
+#             if hist[-1]["role"] == "user":
+#                 hist[-1]["content"] += content
+#             else:
+#                 hist.append({"role": "user", "content": content})
+
+#         st = timer()
+#         task_desc = await analyze_task_async(self.chat_mdl, prompt, user_request, tool_metas, user_defined_prompt)
+#         self.callback("analyze_task", {}, task_desc, elapsed_time=timer()-st)
+#         for _ in range(self._param.max_rounds + 1):
+#             if self.check_if_canceled("Agent streaming"):
+#                 return
+#             response, tk = await next_step_async(self.chat_mdl, hist, tool_metas, task_desc, user_defined_prompt)
+#             # self.callback("next_step", {}, str(response)[:256]+"...")
+#             token_count += tk or 0
+#             hist.append({"role": "assistant", "content": response})
+#             try:
+#                 functions = json_repair.loads(re.sub(r"```.*", "", response))
+#                 if not isinstance(functions, list):
+#                     raise TypeError(f"List should be returned, but `{functions}`")
+#                 for f in functions:
+#                     if not isinstance(f, dict):
+#                         raise TypeError(f"An object type should be returned, but `{f}`")
+
+#                 tool_tasks = []
+#                 for func in functions:
+#                     name = func["name"]
+#                     args = func["arguments"]
+#                     if name == COMPLETE_TASK:
+#                         append_user_content(hist, f"Respond with a formal answer. FORGET(DO NOT mention) about `{COMPLETE_TASK}`. The language for the response MUST be as the same as the first user request.\n")
+#                         async for txt, tkcnt in complete():
+#                             yield txt, tkcnt
+#                         return
+
+#                     tool_tasks.append(asyncio.create_task(use_tool_async(name, args)))
+
+#                 results = await asyncio.gather(*tool_tasks) if tool_tasks else []
+#                 st = timer()
+#                 reflection = await reflect_async(self.chat_mdl, hist, results, user_defined_prompt)
+#                 append_user_content(hist, reflection)
+#                 self.callback("reflection", {}, str(reflection), elapsed_time=timer()-st)
+
+#             except Exception as e:
+#                 logging.exception(msg=f"Wrong JSON argument format in LLM ReAct response: {e}")
+#                 e = f"\nTool call error, please correct the input parameter of response format and call it again.\n *** Exception ***\n{e}"
+#                 append_user_content(hist, str(e))
+
+#         logging.warning( f"Exceed max rounds: {self._param.max_rounds}")
+#         final_instruction = f"""
+# {user_request}
+# IMPORTANT: You have reached the conversation limit. Based on ALL the information and research you have gathered so far, please provide a DIRECT and COMPREHENSIVE final answer to the original request.
+# Instructions:
+# 1. SYNTHESIZE all information collected during this conversation
+# 2. Provide a COMPLETE response using existing data - do not suggest additional research
+# 3. Structure your response as a FINAL DELIVERABLE, not a plan
+# 4. If information is incomplete, state what you found and provide the best analysis possible with available data
+# 5. DO NOT mention conversation limits or suggest further steps
+# 6. Focus on delivering VALUE with the information already gathered
+# Respond immediately with your final comprehensive answer.
+#         """
+#         if self.check_if_canceled("Agent final instruction"):
+#             return
+#         append_user_content(hist, final_instruction)
+
+#         async for txt, tkcnt in complete():
+#             yield txt, tkcnt
+
+    async def _gen_citations_async(self, text):
+        retrievals = self._canvas.get_reference()
+        retrievals = {"chunks": list(retrievals["chunks"].values()), "doc_aggs": list(retrievals["doc_aggs"].values())}
+        formated_refer = kb_prompt(retrievals, self.chat_mdl.max_length, True)
+        async for delta_ans in self._generate_streamly([{"role": "system", "content": citation_plus("\n\n".join(formated_refer))},
+                                                  {"role": "user", "content": text}
+                                                  ]):
+            yield delta_ans
+
+    def reset(self, only_output=False):
         """
         Reset all tools if they have a reset method. This avoids errors for tools like MCPToolCallSession.
         """
+        for k in self._param.outputs.keys():
+            self._param.outputs[k]["value"] = None
+
         for k, cpn in self.tools.items():
             if hasattr(cpn, "reset") and callable(cpn.reset):
                 cpn.reset()
-
+        if only_output:
+            return
+        for k in self._param.inputs.keys():
+            self._param.inputs[k]["value"] = None
+        self._param.debug_inputs = {}

agent/component/base.py

@@ -14,6 +14,7 @@
 #  limitations under the License.
 #
 
+import asyncio
 import re
 import time
 from abc import ABC
@@ -23,11 +24,9 @@ import os
 import logging
 from typing import Any, List, Union
 import pandas as pd
-import trio
 from agent import settings
 from common.connection_utils import timeout
 
-
 _FEEDED_DEPRECATED_PARAMS = "_feeded_deprecated_params"
 _DEPRECATED_PARAMS = "_deprecated_params"
 _USER_FEEDED_PARAMS = "_user_feeded_params"
@@ -97,7 +96,7 @@ class ComponentParamBase(ABC):
         def _recursive_convert_obj_to_dict(obj):
             ret_dict = {}
             if isinstance(obj, dict):
-                for k,v in obj.items():
+                for k, v in obj.items():
                     if isinstance(v, dict) or (v and type(v).__name__ not in dir(builtins)):
                         ret_dict[k] = _recursive_convert_obj_to_dict(v)
                     else:
@@ -253,96 +252,65 @@ class ComponentParamBase(ABC):
                 self._validate_param(attr, validation_json)
 
     @staticmethod
-    def check_string(param, descr):
+    def check_string(param, description):
         if type(param).__name__ not in ["str"]:
-            raise ValueError(
-                descr + " {} not supported, should be string type".format(param)
-            )
+            raise ValueError(description + " {} not supported, should be string type".format(param))
 
     @staticmethod
-    def check_empty(param, descr):
+    def check_empty(param, description):
         if not param:
-            raise ValueError(
-                descr + " does not support empty value."
-            )
+            raise ValueError(description + " does not support empty value.")
 
     @staticmethod
-    def check_positive_integer(param, descr):
+    def check_positive_integer(param, description):
         if type(param).__name__ not in ["int", "long"] or param <= 0:
-            raise ValueError(
-                descr + " {} not supported, should be positive integer".format(param)
-            )
+            raise ValueError(description + " {} not supported, should be positive integer".format(param))
 
     @staticmethod
-    def check_positive_number(param, descr):
+    def check_positive_number(param, description):
         if type(param).__name__ not in ["float", "int", "long"] or param <= 0:
-            raise ValueError(
-                descr + " {} not supported, should be positive numeric".format(param)
-            )
+            raise ValueError(description + " {} not supported, should be positive numeric".format(param))
 
     @staticmethod
-    def check_nonnegative_number(param, descr):
+    def check_nonnegative_number(param, description):
         if type(param).__name__ not in ["float", "int", "long"] or param < 0:
-            raise ValueError(
-                descr
-                + " {} not supported, should be non-negative numeric".format(param)
-            )
+            raise ValueError(description + " {} not supported, should be non-negative numeric".format(param))
 
     @staticmethod
-    def check_decimal_float(param, descr):
+    def check_decimal_float(param, description):
         if type(param).__name__ not in ["float", "int"] or param < 0 or param > 1:
-            raise ValueError(
-                descr
-                + " {} not supported, should be a float number in range [0, 1]".format(
-                    param
-                )
-            )
+            raise ValueError(description + " {} not supported, should be a float number in range [0, 1]".format(param))
 
     @staticmethod
-    def check_boolean(param, descr):
+    def check_boolean(param, description):
         if type(param).__name__ != "bool":
-            raise ValueError(
-                descr + " {} not supported, should be bool type".format(param)
-            )
+            raise ValueError(description + " {} not supported, should be bool type".format(param))
 
     @staticmethod
-    def check_open_unit_interval(param, descr):
+    def check_open_unit_interval(param, description):
         if type(param).__name__ not in ["float"] or param <= 0 or param >= 1:
-            raise ValueError(
-                descr + " should be a numeric number between 0 and 1 exclusively"
-            )
+            raise ValueError(description + " should be a numeric number between 0 and 1 exclusively")
 
     @staticmethod
-    def check_valid_value(param, descr, valid_values):
+    def check_valid_value(param, description, valid_values):
         if param not in valid_values:
-            raise ValueError(
-                descr
-                + " {} is not supported, it should be in {}".format(param, valid_values)
-            )
+            raise ValueError(description + " {} is not supported, it should be in {}".format(param, valid_values))
 
     @staticmethod
-    def check_defined_type(param, descr, types):
+    def check_defined_type(param, description, types):
         if type(param).__name__ not in types:
-            raise ValueError(
-                descr + " {} not supported, should be one of {}".format(param, types)
-            )
+            raise ValueError(description + " {} not supported, should be one of {}".format(param, types))
 
     @staticmethod
-    def check_and_change_lower(param, valid_list, descr=""):
+    def check_and_change_lower(param, valid_list, description=""):
         if type(param).__name__ != "str":
-            raise ValueError(
-                descr
-                + " {} not supported, should be one of {}".format(param, valid_list)
-            )
+            raise ValueError(description + " {} not supported, should be one of {}".format(param, valid_list))
 
         lower_param = param.lower()
         if lower_param in valid_list:
             return lower_param
         else:
-            raise ValueError(
-                descr
-                + " {} not supported, should be one of {}".format(param, valid_list)
-            )
+            raise ValueError(description + " {} not supported, should be one of {}".format(param, valid_list))
 
     @staticmethod
     def _greater_equal_than(value, limit):
@@ -374,16 +342,16 @@ class ComponentParamBase(ABC):
     def _not_in(value, wrong_value_list):
         return value not in wrong_value_list
 
-    def _warn_deprecated_param(self, param_name, descr):
+    def _warn_deprecated_param(self, param_name, description):
         if self._deprecated_params_set.get(param_name):
             logging.warning(
-                f"{descr} {param_name} is deprecated and ignored in this version."
+                f"{description} {param_name} is deprecated and ignored in this version."
             )
 
-    def _warn_to_deprecate_param(self, param_name, descr, new_param):
+    def _warn_to_deprecate_param(self, param_name, description, new_param):
         if self._deprecated_params_set.get(param_name):
             logging.warning(
-                f"{descr} {param_name} will be deprecated in future release; "
+                f"{description} {param_name} will be deprecated in future release; "
                 f"please use {new_param} instead."
             )
             return True
@@ -392,8 +360,8 @@ class ComponentParamBase(ABC):
 
 class ComponentBase(ABC):
     component_name: str
-    thread_limiter = trio.CapacityLimiter(int(os.environ.get('MAX_CONCURRENT_CHATS', 10)))
-    variable_ref_patt = r"\{* *\{([a-zA-Z:0-9]+@[A-Za-z:0-9_.-]+|sys\.[a-z_]+)\} *\}*"
+    thread_limiter = asyncio.Semaphore(int(os.environ.get("MAX_CONCURRENT_CHATS", 10)))
+    variable_ref_patt = r"\{* *\{([a-zA-Z:0-9]+@[A-Za-z0-9_.-]+|sys\.[A-Za-z0-9_.]+|env\.[A-Za-z0-9_.]+)\} *\}*"
 
     def __str__(self):
         """
@@ -407,7 +375,7 @@ class ComponentBase(ABC):
             "params": {}
         }}""".format(self.component_name,
                      self._param
-        )
+                     )
 
     def __init__(self, canvas, id, param: ComponentParamBase):
         from agent.canvas import Graph  # Local import to avoid cyclic dependency
@@ -417,6 +385,20 @@ class ComponentBase(ABC):
         self._param = param
         self._param.check()
 
+    def is_canceled(self) -> bool:
+        return self._canvas.is_canceled()
+
+    def check_if_canceled(self, message: str = "") -> bool:
+        if self.is_canceled():
+            task_id = getattr(self._canvas, 'task_id', 'unknown')
+            log_message = f"Task {task_id} has been canceled"
+            if message:
+                log_message += f" during {message}"
+            logging.info(log_message)
+            self.set_output("_ERROR", "Task has been canceled")
+            return True
+        return False
+
     def invoke(self, **kwargs) -> dict[str, Any]:
         self.set_output("_created_time", time.perf_counter())
         try:
@@ -431,14 +413,42 @@ class ComponentBase(ABC):
         self.set_output("_elapsed_time", time.perf_counter() - self.output("_created_time"))
         return self.output()
 
-    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60)))
+    async def invoke_async(self, **kwargs) -> dict[str, Any]:
+        """
+        Async wrapper for component invocation.
+        Prefers coroutine `_invoke_async` if present; otherwise falls back to `_invoke`.
+        Handles timing and error recording consistently with `invoke`.
+        """
+        self.set_output("_created_time", time.perf_counter())
+        try:
+            if self.check_if_canceled("Component processing"):
+                return
+
+            fn_async = getattr(self, "_invoke_async", None)
+            if fn_async and asyncio.iscoroutinefunction(fn_async):
+                await fn_async(**kwargs)
+            elif asyncio.iscoroutinefunction(self._invoke):
+                await self._invoke(**kwargs)
+            else:
+                await asyncio.to_thread(self._invoke, **kwargs)
+        except Exception as e:
+            if self.get_exception_default_value():
+                self.set_exception_default_value()
+            else:
+                self.set_output("_ERROR", str(e))
+            logging.exception(e)
+        self._param.debug_inputs = {}
+        self.set_output("_elapsed_time", time.perf_counter() - self.output("_created_time"))
+        return self.output()
+
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10 * 60)))
     def _invoke(self, **kwargs):
         raise NotImplementedError()
 
-    def output(self, var_nm: str=None) -> Union[dict[str, Any], Any]:
+    def output(self, var_nm: str = None) -> Union[dict[str, Any], Any]:
         if var_nm:
             return self._param.outputs.get(var_nm, {}).get("value", "")
-        return {k: o.get("value") for k,o in self._param.outputs.items()}
+        return {k: o.get("value") for k, o in self._param.outputs.items()}
 
     def set_output(self, key: str, value: Any):
         if key not in self._param.outputs:
@@ -449,15 +459,18 @@ class ComponentBase(ABC):
         return self._param.outputs.get("_ERROR", {}).get("value")
 
     def reset(self, only_output=False):
-        for k in self._param.outputs.keys():
-            self._param.outputs[k]["value"] = None
+        outputs: dict = self._param.outputs  # for better performance
+        for k in outputs.keys():
+            outputs[k]["value"] = None
         if only_output:
             return
-        for k in self._param.inputs.keys():
-            self._param.inputs[k]["value"] = None
+
+        inputs: dict = self._param.inputs  # for better performance
+        for k in inputs.keys():
+            inputs[k]["value"] = None
         self._param.debug_inputs = {}
 
-    def get_input(self, key: str=None) -> Union[Any, dict[str, Any]]:
+    def get_input(self, key: str = None) -> Union[Any, dict[str, Any]]:
         if key:
             return self._param.inputs.get(key, {}).get("value")
 
@@ -481,13 +494,13 @@ class ComponentBase(ABC):
 
     def get_input_elements_from_text(self, txt: str) -> dict[str, dict[str, str]]:
         res = {}
-        for r in re.finditer(self.variable_ref_patt, txt, flags=re.IGNORECASE|re.DOTALL):
+        for r in re.finditer(self.variable_ref_patt, txt, flags=re.IGNORECASE | re.DOTALL):
             exp = r.group(1)
-            cpn_id, var_nm = exp.split("@") if exp.find("@")>0 else ("", exp)
+            cpn_id, var_nm = exp.split("@") if exp.find("@") > 0 else ("", exp)
             res[exp] = {
-                "name": (self._canvas.get_component_name(cpn_id) +f"@{var_nm}") if cpn_id else exp,
+                "name": (self._canvas.get_component_name(cpn_id) + f"@{var_nm}") if cpn_id else exp,
                 "value": self._canvas.get_variable_value(exp),
-                "_retrival": self._canvas.get_variable_value(f"{cpn_id}@_references") if cpn_id else None,
+                "_retrieval": self._canvas.get_variable_value(f"{cpn_id}@_references") if cpn_id else None,
                 "_cpn_id": cpn_id
             }
         return res
@@ -514,6 +527,7 @@ class ComponentBase(ABC):
     def get_param(self, name):
         if hasattr(self._param, name):
             return getattr(self._param, name)
+        return None
 
     def debug(self, **kwargs):
         return self._invoke(**kwargs)
@@ -521,7 +535,7 @@ class ComponentBase(ABC):
     def get_parent(self) -> Union[object, None]:
         pid = self._canvas.get_component(self._id).get("parent_id")
         if not pid:
-            return
+            return None
         return self._canvas.get_component(pid)["obj"]
 
     def get_upstream(self) -> List[str]:
@@ -537,6 +551,7 @@ class ComponentBase(ABC):
         for n, v in kv.items():
             def repl(_match, val=v):
                 return str(val) if val is not None else ""
+
             content = re.sub(
                 r"\{%s\}" % re.escape(n),
                 repl,
@@ -546,7 +561,7 @@ class ComponentBase(ABC):
 
     def exception_handler(self):
         if not self._param.exception_method:
-            return
+            return None
         return {
             "goto": self._param.exception_goto,
             "default_value": self._param.exception_default_value

agent/component/begin.py

@@ -14,6 +14,7 @@
 #  limitations under the License.
 #
 from agent.component.fillup import UserFillUpParam, UserFillUp
+from api.db.services.file_service import FileService
 
 
 class BeginParam(UserFillUpParam):
@@ -27,7 +28,7 @@ class BeginParam(UserFillUpParam):
         self.prologue = "Hi! I'm your smart assistant. What can I do for you?"
 
     def check(self):
-        self.check_valid_value(self.mode, "The 'mode' should be either `conversational` or `task`", ["conversational", "task"])
+        self.check_valid_value(self.mode, "The 'mode' should be either `conversational` or `task`", ["conversational", "task","Webhook"])
 
     def get_input_form(self) -> dict[str, dict]:
         return getattr(self, "inputs")
@@ -37,12 +38,18 @@ class Begin(UserFillUp):
     component_name = "Begin"
 
     def _invoke(self, **kwargs):
+        if self.check_if_canceled("Begin processing"):
+            return
+
         for k, v in kwargs.get("inputs", {}).items():
+            if self.check_if_canceled("Begin processing"):
+                return
+
             if isinstance(v, dict) and v.get("type", "").lower().find("file") >=0:
                 if v.get("optional") and v.get("value", None) is None:
                     v = None
                 else:
-                    v = self._canvas.get_files([v["value"]])
+                    v = FileService.get_files([v["value"]])
             else:
                 v = v.get("value")
             self.set_output(k, v)

agent/component/categorize.py

@@ -13,6 +13,7 @@
 #  See the License for the specific language governing permissions and
 #  limitations under the License.
 #
+import asyncio
 import logging
 import os
 import re
@@ -97,7 +98,10 @@ class Categorize(LLM, ABC):
     component_name = "Categorize"
 
     @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60)))
-    def _invoke(self, **kwargs):
+    async def _invoke_async(self, **kwargs):
+        if self.check_if_canceled("Categorize processing"):
+            return
+
         msg = self._canvas.get_history(self._param.message_history_window_size)
         if not msg:
             msg = [{"role": "user", "content": ""}]
@@ -114,10 +118,18 @@ class Categorize(LLM, ABC):
 ---- Real Data ----
 {} →
 """.format(" | ".join(["{}: \"{}\"".format(c["role"].upper(), re.sub(r"\n", "", c["content"], flags=re.DOTALL)) for c in msg]))
-        ans = chat_mdl.chat(self._param.sys_prompt, [{"role": "user", "content": user_prompt}], self._param.gen_conf())
+
+        if self.check_if_canceled("Categorize processing"):
+            return
+
+        ans = await chat_mdl.async_chat(self._param.sys_prompt, [{"role": "user", "content": user_prompt}], self._param.gen_conf())
         logging.info(f"input: {user_prompt}, answer: {str(ans)}")
         if ERROR_PREFIX in ans:
             raise Exception(ans)
+
+        if self.check_if_canceled("Categorize processing"):
+            return
+
         # Count the number of times each category appears in the answer.
         category_counts = {}
         for c in self._param.category_description.keys():
@@ -133,5 +145,9 @@ class Categorize(LLM, ABC):
         self.set_output("category_name", max_category)
         self.set_output("_next", cpn_ids)
 
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60)))
+    def _invoke(self, **kwargs):
+        return asyncio.run(self._invoke_async(**kwargs))
+
     def thoughts(self) -> str:
         return "Which should it falls into {}? ...".format(",".join([f"`{c}`" for c, _ in self._param.category_description.items()]))

agent/component/data_operations.py

@@ -1,3 +1,18 @@
+#
+#  Copyright 2024 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+#
 from abc import ABC
 import ast
 import os
@@ -10,7 +25,7 @@ class DataOperationsParam(ComponentParamBase):
     """
     def __init__(self):
         super().__init__()
-        self.inputs = []
+        self.query = []
         self.operations = "literal_eval"
         self.select_keys = []
         self.filter_values=[]
@@ -35,18 +50,19 @@ class DataOperations(ComponentBase,ABC):
     def get_input_form(self) -> dict[str, dict]:
         return {
             k: {"name": o.get("name", ""), "type": "line"}
-            for input_item in (self._param.inputs or [])
+            for input_item in (self._param.query or [])
             for k, o in self.get_input_elements_from_text(input_item).items()
         }
 
     @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60)))
     def _invoke(self, **kwargs):
         self.input_objects=[]
-        inputs = getattr(self._param, "inputs", None)
+        inputs = getattr(self._param, "query", None)
         if not isinstance(inputs, (list, tuple)):
             inputs = [inputs]
-        for input_ref in self._param.inputs:
+        for input_ref in inputs:
             input_object=self._canvas.get_variable_value(input_ref)
+            self.set_input_value(input_ref, input_object)
             if input_object is None:
                 continue
             if isinstance(input_object,dict):
@@ -57,7 +73,7 @@ class DataOperations(ComponentBase,ABC):
                 continue
         if self._param.operations == "select_keys":
             self._select_keys()
-        elif self._param.operations == "literal_eval":
+        elif self._param.operations == "recursive_eval":
             self._literal_eval()
         elif self._param.operations == "combine":
             self._combine()
@@ -100,7 +116,7 @@ class DataOperations(ComponentBase,ABC):
 
     def _combine(self):
         result={}
-        for obj in self.input_objects():
+        for obj in self.input_objects:
             for key, value in obj.items():
                 if key not in result:
                     result[key] = value
@@ -123,6 +139,7 @@ class DataOperations(ComponentBase,ABC):
         key = rule.get("key")
         op = (rule.get("operator") or "equals").lower()
         target = self.norm(rule.get("value"))
+        target = self._canvas.get_value_with_variable(target) or target
         if key not in obj:
             return False
         val = obj.get(key, None)
@@ -142,7 +159,7 @@ class DataOperations(ComponentBase,ABC):
     def _filter_values(self):
         results=[]
         rules = (getattr(self._param, "filter_values", None) or [])
-        for obj in self.input_objects():
+        for obj in self.input_objects:
             if not rules:
                 results.append(obj)
                 continue
@@ -154,7 +171,7 @@ class DataOperations(ComponentBase,ABC):
     def _append_or_update(self):
         results=[]
         updates = getattr(self._param, "updates", []) or [] 
-        for obj in self.input_objects():
+        for obj in self.input_objects:
             new_obj = dict(obj)
             for item in updates:
                 if not isinstance(item, dict):
@@ -162,7 +179,7 @@ class DataOperations(ComponentBase,ABC):
                 k = (item.get("key") or "").strip()
                 if not k:
                     continue
-                new_obj[k] = item.get("value")
+                new_obj[k] = self._canvas.get_value_with_variable(item.get("value")) or item.get("value")
             results.append(new_obj)
         self.set_output("result", results)
 

agent/component/excel_processor.py

@@ -0,0 +1,401 @@
+#
+#  Copyright 2025 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+#
+
+"""
+ExcelProcessor Component
+
+A component for reading, processing, and generating Excel files in RAGFlow agents.
+Supports multiple Excel file inputs, data transformation, and Excel output generation.
+"""
+
+import logging
+import os
+from abc import ABC
+from io import BytesIO
+
+import pandas as pd
+
+from agent.component.base import ComponentBase, ComponentParamBase
+from api.db.services.file_service import FileService
+from api.utils.api_utils import timeout
+from common import settings
+from common.misc_utils import get_uuid
+
+
+class ExcelProcessorParam(ComponentParamBase):
+    """
+    Define the ExcelProcessor component parameters.
+    """
+    def __init__(self):
+        super().__init__()
+        # Input configuration
+        self.input_files = []  # Variable references to uploaded files
+        self.operation = "read"  # read, merge, transform, output
+        
+        # Processing options
+        self.sheet_selection = "all"  # all, first, or comma-separated sheet names
+        self.merge_strategy = "concat"  # concat, join
+        self.join_on = ""  # Column name for join operations
+        
+        # Transform options (for LLM-guided transformations)
+        self.transform_instructions = ""
+        self.transform_data = ""  # Variable reference to transformation data
+        
+        # Output options
+        self.output_format = "xlsx"  # xlsx, csv
+        self.output_filename = "output"
+        
+        # Component outputs
+        self.outputs = {
+            "data": {
+                "type": "object",
+                "value": {}
+            },
+            "summary": {
+                "type": "str",
+                "value": ""
+            },
+            "markdown": {
+                "type": "str",
+                "value": ""
+            }
+        }
+    
+    def check(self):
+        self.check_valid_value(
+            self.operation, 
+            "[ExcelProcessor] Operation", 
+            ["read", "merge", "transform", "output"]
+        )
+        self.check_valid_value(
+            self.output_format,
+            "[ExcelProcessor] Output format",
+            ["xlsx", "csv"]
+        )
+        return True
+
+
+class ExcelProcessor(ComponentBase, ABC):
+    """
+    Excel processing component for RAGFlow agents.
+    
+    Operations:
+    - read: Parse Excel files into structured data
+    - merge: Combine multiple Excel files
+    - transform: Apply data transformations based on instructions
+    - output: Generate Excel file output
+    """
+    component_name = "ExcelProcessor"
+
+    def get_input_form(self) -> dict[str, dict]:
+        """Define input form for the component."""
+        res = {}
+        for ref in (self._param.input_files or []):
+            for k, o in self.get_input_elements_from_text(ref).items():
+                res[k] = {"name": o.get("name", ""), "type": "file"}
+        if self._param.transform_data:
+            for k, o in self.get_input_elements_from_text(self._param.transform_data).items():
+                res[k] = {"name": o.get("name", ""), "type": "object"}
+        return res
+
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60)))
+    def _invoke(self, **kwargs):
+        if self.check_if_canceled("ExcelProcessor processing"):
+            return
+
+        operation = self._param.operation.lower()
+        
+        if operation == "read":
+            self._read_excels()
+        elif operation == "merge":
+            self._merge_excels()
+        elif operation == "transform":
+            self._transform_data()
+        elif operation == "output":
+            self._output_excel()
+        else:
+            self.set_output("summary", f"Unknown operation: {operation}")
+
+    def _get_file_content(self, file_ref: str) -> tuple[bytes, str]:
+        """
+        Get file content from a variable reference.
+        Returns (content_bytes, filename).
+        """
+        value = self._canvas.get_variable_value(file_ref)
+        if value is None:
+            return None, None
+            
+        # Handle different value formats
+        if isinstance(value, dict):
+            # File reference from Begin/UserFillUp component
+            file_id = value.get("id") or value.get("file_id")
+            created_by = value.get("created_by") or self._canvas.get_tenant_id()
+            filename = value.get("name") or value.get("filename", "unknown.xlsx")
+            if file_id:
+                content = FileService.get_blob(created_by, file_id)
+                return content, filename
+        elif isinstance(value, list) and len(value) > 0:
+            # List of file references - return first
+            return self._get_file_content_from_list(value[0])
+        elif isinstance(value, str):
+            # Could be base64 encoded or a path
+            if value.startswith("data:"):
+                import base64
+                # Extract base64 content
+                _, encoded = value.split(",", 1)
+                return base64.b64decode(encoded), "uploaded.xlsx"
+                
+        return None, None
+    
+    def _get_file_content_from_list(self, item) -> tuple[bytes, str]:
+        """Extract file content from a list item."""
+        if isinstance(item, dict):
+            return self._get_file_content(item)
+        return None, None
+
+    def _parse_excel_to_dataframes(self, content: bytes, filename: str) -> dict[str, pd.DataFrame]:
+        """Parse Excel content into a dictionary of DataFrames (one per sheet)."""
+        try:
+            excel_file = BytesIO(content)
+            
+            if filename.lower().endswith(".csv"):
+                df = pd.read_csv(excel_file)
+                return {"Sheet1": df}
+            else:
+                # Read all sheets
+                xlsx = pd.ExcelFile(excel_file, engine='openpyxl')
+                sheet_selection = self._param.sheet_selection
+                
+                if sheet_selection == "all":
+                    sheets_to_read = xlsx.sheet_names
+                elif sheet_selection == "first":
+                    sheets_to_read = [xlsx.sheet_names[0]] if xlsx.sheet_names else []
+                else:
+                    # Comma-separated sheet names
+                    requested = [s.strip() for s in sheet_selection.split(",")]
+                    sheets_to_read = [s for s in requested if s in xlsx.sheet_names]
+                
+                dfs = {}
+                for sheet in sheets_to_read:
+                    dfs[sheet] = pd.read_excel(xlsx, sheet_name=sheet)
+                return dfs
+                
+        except Exception as e:
+            logging.error(f"Error parsing Excel file {filename}: {e}")
+            return {}
+
+    def _read_excels(self):
+        """Read and parse Excel files into structured data."""
+        all_data = {}
+        summaries = []
+        markdown_parts = []
+        
+        for file_ref in (self._param.input_files or []):
+            if self.check_if_canceled("ExcelProcessor reading"):
+                return
+                
+            # Get variable value
+            value = self._canvas.get_variable_value(file_ref)
+            self.set_input_value(file_ref, str(value)[:200] if value else "")
+            
+            if value is None:
+                continue
+            
+            # Handle file content
+            content, filename = self._get_file_content(file_ref)
+            if content is None:
+                continue
+                
+            # Parse Excel
+            dfs = self._parse_excel_to_dataframes(content, filename)
+            
+            for sheet_name, df in dfs.items():
+                key = f"{filename}_{sheet_name}" if len(dfs) > 1 else filename
+                all_data[key] = df.to_dict(orient="records")
+                
+                # Build summary
+                summaries.append(f"**{key}**: {len(df)} rows, {len(df.columns)} columns ({', '.join(df.columns.tolist()[:5])}{'...' if len(df.columns) > 5 else ''})")
+                
+                # Build markdown table
+                markdown_parts.append(f"### {key}\n\n{df.head(10).to_markdown(index=False)}\n")
+        
+        # Set outputs
+        self.set_output("data", all_data)
+        self.set_output("summary", "\n".join(summaries) if summaries else "No Excel files found")
+        self.set_output("markdown", "\n\n".join(markdown_parts) if markdown_parts else "No data")
+
+    def _merge_excels(self):
+        """Merge multiple Excel files/sheets into one."""
+        all_dfs = []
+        
+        for file_ref in (self._param.input_files or []):
+            if self.check_if_canceled("ExcelProcessor merging"):
+                return
+                
+            value = self._canvas.get_variable_value(file_ref)
+            self.set_input_value(file_ref, str(value)[:200] if value else "")
+            
+            if value is None:
+                continue
+                
+            content, filename = self._get_file_content(file_ref)
+            if content is None:
+                continue
+                
+            dfs = self._parse_excel_to_dataframes(content, filename)
+            all_dfs.extend(dfs.values())
+        
+        if not all_dfs:
+            self.set_output("data", {})
+            self.set_output("summary", "No data to merge")
+            return
+        
+        # Merge strategy
+        if self._param.merge_strategy == "concat":
+            merged_df = pd.concat(all_dfs, ignore_index=True)
+        elif self._param.merge_strategy == "join" and self._param.join_on:
+            # Join on specified column
+            merged_df = all_dfs[0]
+            for df in all_dfs[1:]:
+                merged_df = merged_df.merge(df, on=self._param.join_on, how="outer")
+        else:
+            merged_df = pd.concat(all_dfs, ignore_index=True)
+        
+        self.set_output("data", {"merged": merged_df.to_dict(orient="records")})
+        self.set_output("summary", f"Merged {len(all_dfs)} sources into {len(merged_df)} rows, {len(merged_df.columns)} columns")
+        self.set_output("markdown", merged_df.head(20).to_markdown(index=False))
+
+    def _transform_data(self):
+        """Apply transformations to data based on instructions or input data."""
+        # Get the data to transform
+        transform_ref = self._param.transform_data
+        if not transform_ref:
+            self.set_output("summary", "No transform data reference provided")
+            return
+            
+        data = self._canvas.get_variable_value(transform_ref)
+        self.set_input_value(transform_ref, str(data)[:300] if data else "")
+        
+        if data is None:
+            self.set_output("summary", "Transform data is empty")
+            return
+        
+        # Convert to DataFrame
+        if isinstance(data, dict):
+            # Could be {"sheet": [rows]} format
+            if all(isinstance(v, list) for v in data.values()):
+                # Multiple sheets
+                all_markdown = []
+                for sheet_name, rows in data.items():
+                    df = pd.DataFrame(rows)
+                    all_markdown.append(f"### {sheet_name}\n\n{df.to_markdown(index=False)}")
+                self.set_output("data", data)
+                self.set_output("markdown", "\n\n".join(all_markdown))
+            else:
+                df = pd.DataFrame([data])
+                self.set_output("data", df.to_dict(orient="records"))
+                self.set_output("markdown", df.to_markdown(index=False))
+        elif isinstance(data, list):
+            df = pd.DataFrame(data)
+            self.set_output("data", df.to_dict(orient="records"))
+            self.set_output("markdown", df.to_markdown(index=False))
+        else:
+            self.set_output("data", {"raw": str(data)})
+            self.set_output("markdown", str(data))
+        
+        self.set_output("summary", "Transformed data ready for processing")
+
+    def _output_excel(self):
+        """Generate Excel file output from data."""
+        # Get data from transform_data reference
+        transform_ref = self._param.transform_data
+        if not transform_ref:
+            self.set_output("summary", "No data reference for output")
+            return
+            
+        data = self._canvas.get_variable_value(transform_ref)
+        self.set_input_value(transform_ref, str(data)[:300] if data else "")
+        
+        if data is None:
+            self.set_output("summary", "No data to output")
+            return
+        
+        try:
+            # Prepare DataFrames
+            if isinstance(data, dict):
+                if all(isinstance(v, list) for v in data.values()):
+                    # Multi-sheet format
+                    dfs = {k: pd.DataFrame(v) for k, v in data.items()}
+                else:
+                    dfs = {"Sheet1": pd.DataFrame([data])}
+            elif isinstance(data, list):
+                dfs = {"Sheet1": pd.DataFrame(data)}
+            else:
+                self.set_output("summary", "Invalid data format for Excel output")
+                return
+            
+            # Generate output
+            doc_id = get_uuid()
+            
+            if self._param.output_format == "csv":
+                # For CSV, only output first sheet
+                first_df = list(dfs.values())[0]
+                binary_content = first_df.to_csv(index=False).encode("utf-8")
+                filename = f"{self._param.output_filename}.csv"
+            else:
+                # Excel output
+                excel_io = BytesIO()
+                with pd.ExcelWriter(excel_io, engine='openpyxl') as writer:
+                    for sheet_name, df in dfs.items():
+                        # Sanitize sheet name (max 31 chars, no special chars)
+                        safe_name = sheet_name[:31].replace("/", "_").replace("\\", "_")
+                        df.to_excel(writer, sheet_name=safe_name, index=False)
+                excel_io.seek(0)
+                binary_content = excel_io.read()
+                filename = f"{self._param.output_filename}.xlsx"
+            
+            # Store file
+            settings.STORAGE_IMPL.put(self._canvas._tenant_id, doc_id, binary_content)
+            
+            # Set attachment output
+            self.set_output("attachment", {
+                "doc_id": doc_id,
+                "format": self._param.output_format,
+                "file_name": filename
+            })
+            
+            total_rows = sum(len(df) for df in dfs.values())
+            self.set_output("summary", f"Generated {filename} with {len(dfs)} sheet(s), {total_rows} total rows")
+            self.set_output("data", {k: v.to_dict(orient="records") for k, v in dfs.items()})
+            
+            logging.info(f"ExcelProcessor: Generated {filename} as {doc_id}")
+            
+        except Exception as e:
+            logging.error(f"ExcelProcessor output error: {e}")
+            self.set_output("summary", f"Error generating output: {str(e)}")
+
+    def thoughts(self) -> str:
+        """Return component thoughts for UI display."""
+        op = self._param.operation
+        if op == "read":
+            return "Reading Excel files..."
+        elif op == "merge":
+            return "Merging Excel data..."
+        elif op == "transform":
+            return "Transforming data..."
+        elif op == "output":
+            return "Generating Excel output..."
+        return "Processing Excel..."

agent/component/exit_loop.py

@@ -0,0 +1,32 @@
+#
+#  Copyright 2024 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+#
+from abc import ABC
+from agent.component.base import ComponentBase, ComponentParamBase
+
+
+class ExitLoopParam(ComponentParamBase, ABC):
+    def check(self):
+        return True
+
+
+class ExitLoop(ComponentBase, ABC):
+    component_name = "ExitLoop"
+
+    def _invoke(self, **kwargs):
+        pass
+
+    def thoughts(self) -> str:
+        return ""

agent/component/fillup.py

@@ -13,7 +13,12 @@
 #  See the License for the specific language governing permissions and
 #  limitations under the License.
 #
-from agent.component.base import ComponentBase, ComponentParamBase
+import json
+import re
+from functools import partial
+
+from agent.component.base import ComponentParamBase, ComponentBase
+from api.db.services.file_service import FileService
 
 
 class UserFillUpParam(ComponentParamBase):
@@ -31,10 +36,42 @@ class UserFillUp(ComponentBase):
     component_name = "UserFillUp"
 
     def _invoke(self, **kwargs):
+        if self.check_if_canceled("UserFillUp processing"):
+            return
+
+        if self._param.enable_tips:
+            content = self._param.tips
+            for k, v in self.get_input_elements_from_text(self._param.tips).items():
+                v = v["value"]
+                ans = ""
+                if isinstance(v, partial):
+                    for t in v():
+                        ans += t
+                elif isinstance(v, list):
+                    ans = ",".join([str(vv) for vv in v])
+                elif not isinstance(v, str):
+                    try:
+                        ans = json.dumps(v, ensure_ascii=False)
+                    except Exception:
+                        pass
+                else:
+                    ans = v
+                if not ans:
+                    ans = ""
+                content = re.sub(r"\{%s\}"%k, ans, content)
+
+            self.set_output("tips", content)
         for k, v in kwargs.get("inputs", {}).items():
+            if self.check_if_canceled("UserFillUp processing"):
+                return
+            if isinstance(v, dict) and v.get("type", "").lower().find("file") >=0:
+                if v.get("optional") and v.get("value", None) is None:
+                    v = None
+                else:
+                    v = FileService.get_files([v["value"]])
+            else:
+                v = v.get("value")
             self.set_output(k, v)
 
     def thoughts(self) -> str:
         return "Waiting for your input..."
-
-

agent/component/invoke.py

@@ -56,6 +56,9 @@ class Invoke(ComponentBase, ABC):
 
     @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 3)))
     def _invoke(self, **kwargs):
+        if self.check_if_canceled("Invoke processing"):
+            return
+
         args = {}
         for para in self._param.variables:
             if para.get("value"):
@@ -89,6 +92,9 @@ class Invoke(ComponentBase, ABC):
 
         last_e = ""
         for _ in range(self._param.max_retries + 1):
+            if self.check_if_canceled("Invoke processing"):
+                return
+
             try:
                 if method == "get":
                     response = requests.get(url=url, params=args, headers=headers, proxies=proxies, timeout=self._param.timeout)
@@ -121,6 +127,9 @@ class Invoke(ComponentBase, ABC):
 
                 return self.output("result")
             except Exception as e:
+                if self.check_if_canceled("Invoke processing"):
+                    return
+
                 last_e = e
                 logging.exception(f"Http request error: {e}")
                 time.sleep(self._param.delay_after_error)

agent/component/iteration.py

@@ -16,6 +16,13 @@
 from abc import ABC
 from agent.component.base import ComponentBase, ComponentParamBase
 
+"""
+class VariableModel(BaseModel):
+    data_type: Annotated[Literal["string", "number", "Object", "Boolean", "Array<string>", "Array<number>", "Array<object>", "Array<boolean>"], Field(default="Array<string>")]
+    input_mode: Annotated[Literal["constant", "variable"], Field(default="constant")]
+    value: Annotated[Any, Field(default=None)]
+    model_config = ConfigDict(extra="forbid")
+"""
 
 class IterationParam(ComponentParamBase):
     """
@@ -25,6 +32,7 @@ class IterationParam(ComponentParamBase):
     def __init__(self):
         super().__init__()
         self.items_ref = ""
+        self.variable={}
 
     def get_input_form(self) -> dict[str, dict]:
         return {
@@ -49,6 +57,9 @@ class Iteration(ComponentBase, ABC):
                 return cid
 
     def _invoke(self, **kwargs):
+        if self.check_if_canceled("Iteration processing"):
+            return
+
         arr = self._canvas.get_variable_value(self._param.items_ref)
         if not isinstance(arr, list):
             self.set_output("_ERROR", self._param.items_ref + " must be an array, but its type is "+str(type(arr)))

agent/component/iterationitem.py

@@ -33,6 +33,9 @@ class IterationItem(ComponentBase, ABC):
         self._idx = 0
 
     def _invoke(self, **kwargs):
+        if self.check_if_canceled("IterationItem processing"):
+            return
+
         parent = self.get_parent()
         arr = self._canvas.get_variable_value(parent._param.items_ref)
         if not isinstance(arr, list):
@@ -40,12 +43,17 @@ class IterationItem(ComponentBase, ABC):
             raise Exception(parent._param.items_ref + " must be an array, but its type is "+str(type(arr)))
 
         if self._idx > 0:
+            if self.check_if_canceled("IterationItem processing"):
+                return
             self.output_collation()
 
         if self._idx >= len(arr):
             self._idx = -1
             return
 
+        if self.check_if_canceled("IterationItem processing"):
+            return
+
         self.set_output("item", arr[self._idx])
         self.set_output("index", self._idx)
 
@@ -80,4 +88,4 @@ class IterationItem(ComponentBase, ABC):
         return self._idx == -1
 
     def thoughts(self) -> str:
-        return "Next turn..."
+        return "Next turn..."

agent/component/list_operations.py

@@ -0,0 +1,168 @@
+from abc import ABC
+import os
+from agent.component.base import ComponentBase, ComponentParamBase
+from api.utils.api_utils import timeout
+
+class ListOperationsParam(ComponentParamBase):
+    """
+    Define the List Operations component parameters.
+    """
+    def __init__(self):
+        super().__init__()
+        self.query = ""
+        self.operations = "topN"
+        self.n=0
+        self.sort_method = "asc"
+        self.filter = {
+            "operator": "=",
+            "value": ""
+        }
+        self.outputs = {
+            "result": {
+                "value": [],
+                "type": "Array of ?"
+            },
+            "first": {
+                "value": "",
+                "type": "?"
+            },
+            "last": {
+                "value": "",
+                "type": "?"
+            }
+        }
+    
+    def check(self):
+        self.check_empty(self.query, "query")
+        self.check_valid_value(self.operations, "Support operations", ["topN","head","tail","filter","sort","drop_duplicates"])
+
+    def get_input_form(self) -> dict[str, dict]:
+        return {}
+    
+
+class ListOperations(ComponentBase,ABC):
+    component_name = "ListOperations"
+
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60)))
+    def _invoke(self, **kwargs):
+        self.input_objects=[]
+        inputs = getattr(self._param, "query", None)
+        self.inputs = self._canvas.get_variable_value(inputs)
+        if not isinstance(self.inputs, list):
+            raise TypeError("The input of List Operations should be an array.")
+        self.set_input_value(inputs, self.inputs)
+        if self._param.operations == "topN":
+            self._topN()
+        elif self._param.operations == "head":
+            self._head()
+        elif self._param.operations == "tail":
+            self._tail()
+        elif self._param.operations == "filter":
+            self._filter()
+        elif self._param.operations == "sort":
+            self._sort()
+        elif self._param.operations == "drop_duplicates":
+            self._drop_duplicates()
+
+
+    def _coerce_n(self):
+        try:
+            return int(getattr(self._param, "n", 0))
+        except Exception:
+            return 0
+        
+    def _set_outputs(self, outputs):
+        self._param.outputs["result"]["value"] = outputs
+        self._param.outputs["first"]["value"] = outputs[0] if outputs else None
+        self._param.outputs["last"]["value"]  = outputs[-1] if outputs else None
+        
+    def _topN(self):
+        n = self._coerce_n()
+        if n < 1:
+            outputs = []
+        else:
+            n = min(n, len(self.inputs))
+            outputs = self.inputs[:n]
+        self._set_outputs(outputs)
+
+    def _head(self):
+        n = self._coerce_n()
+        if 1 <= n <= len(self.inputs):
+            outputs = [self.inputs[n - 1]]
+        else:
+            outputs = []
+        self._set_outputs(outputs)
+
+    def _tail(self):
+        n = self._coerce_n()
+        if 1 <= n <= len(self.inputs):
+            outputs = [self.inputs[-n]]
+        else:
+            outputs = []
+        self._set_outputs(outputs)
+
+    def _filter(self):
+        self._set_outputs([i for i in self.inputs if self._eval(self._norm(i),self._param.filter["operator"],self._param.filter["value"])])
+
+    def _norm(self,v):
+        s = "" if v is None else str(v)
+        return s
+    
+    def _eval(self, v, operator, value):
+        if operator == "=":
+            return v == value
+        elif operator == "≠":
+            return v != value
+        elif operator == "contains":
+            return value in v
+        elif operator == "start with":
+            return v.startswith(value)
+        elif operator == "end with":
+            return v.endswith(value)
+        else:
+            return False
+
+    def _sort(self):
+        items = self.inputs or []
+        method = getattr(self._param, "sort_method", "asc") or "asc"
+        reverse = method == "desc"
+
+        if not items:
+            self._set_outputs([])
+            return
+
+        first = items[0]
+
+        if isinstance(first, dict):
+            outputs = sorted(
+                items,
+                key=lambda x: self._hashable(x),
+                reverse=reverse,
+            )
+        else:
+            outputs = sorted(items, reverse=reverse)
+
+        self._set_outputs(outputs)
+
+    def _drop_duplicates(self):
+        seen = set()
+        outs = []
+        for item in self.inputs:
+            k = self._hashable(item)
+            if k in seen:
+                continue
+            seen.add(k)
+            outs.append(item)
+        self._set_outputs(outs)
+
+    def _hashable(self,x):
+        if isinstance(x, dict):
+            return tuple(sorted((k, self._hashable(v)) for k, v in x.items()))
+        if isinstance(x, (list, tuple)):
+            return tuple(self._hashable(v) for v in x)
+        if isinstance(x, set):
+            return tuple(sorted(self._hashable(v) for v in x))
+        return x
+    
+    def thoughts(self) -> str:
+        return "ListOperation in progress"

agent/component/llm.py

@@ -13,12 +13,13 @@
 #  See the License for the specific language governing permissions and
 #  limitations under the License.
 #
+import asyncio
 import json
 import logging
 import os
 import re
 from copy import deepcopy
-from typing import Any, Generator
+from typing import Any, AsyncGenerator
 import json_repair
 from functools import partial
 from common.constants import LLMType
@@ -55,7 +56,6 @@ class LLMParam(ComponentParamBase):
         self.check_nonnegative_number(int(self.max_tokens), "[Agent] Max tokens")
         self.check_decimal_float(float(self.top_p), "[Agent] Top P")
         self.check_empty(self.llm_id, "[Agent] LLM")
-        self.check_empty(self.sys_prompt, "[Agent] System prompt")
         self.check_empty(self.prompts, "[Agent] User prompt")
 
     def gen_conf(self):
@@ -166,25 +166,67 @@ class LLM(ComponentBase):
             sys_prompt = re.sub(rf"<{tag}>(.*?)</{tag}>", "", sys_prompt, flags=re.DOTALL|re.IGNORECASE)
         return pts, sys_prompt
 
-    def _generate(self, msg:list[dict], **kwargs) -> str:
+    async def _generate_async(self, msg: list[dict], **kwargs) -> str:
         if not self.imgs:
-            return self.chat_mdl.chat(msg[0]["content"], msg[1:], self._param.gen_conf(), **kwargs)
-        return self.chat_mdl.chat(msg[0]["content"], msg[1:], self._param.gen_conf(), images=self.imgs, **kwargs)
+            return await self.chat_mdl.async_chat(msg[0]["content"], msg[1:], self._param.gen_conf(), **kwargs)
+        return await self.chat_mdl.async_chat(msg[0]["content"], msg[1:], self._param.gen_conf(), images=self.imgs, **kwargs)
 
-    def _generate_streamly(self, msg:list[dict], **kwargs) -> Generator[str, None, None]:
-        ans = ""
+    async def _generate_streamly(self, msg: list[dict], **kwargs) -> AsyncGenerator[str, None]:
+        async def delta_wrapper(txt_iter):
+            ans = ""
+            last_idx = 0
+            endswith_think = False
+
+            def delta(txt):
+                nonlocal ans, last_idx, endswith_think
+                delta_ans = txt[last_idx:]
+                ans = txt
+
+                if delta_ans.find("<think>") == 0:
+                    last_idx += len("<think>")
+                    return "<think>"
+                elif delta_ans.find("<think>") > 0:
+                    delta_ans = txt[last_idx:last_idx + delta_ans.find("<think>")]
+                    last_idx += delta_ans.find("<think>")
+                    return delta_ans
+                elif delta_ans.endswith("</think>"):
+                    endswith_think = True
+                elif endswith_think:
+                    endswith_think = False
+                    return "</think>"
+
+                last_idx = len(ans)
+                if ans.endswith("</think>"):
+                    last_idx -= len("</think>")
+                return re.sub(r"(<think>|</think>)", "", delta_ans)
+
+            async for t in txt_iter:
+                yield delta(t)
+
+        if not self.imgs:
+            async for t in delta_wrapper(self.chat_mdl.async_chat_streamly(msg[0]["content"], msg[1:], self._param.gen_conf(), **kwargs)):
+                yield t
+            return
+
+        async for t in delta_wrapper(self.chat_mdl.async_chat_streamly(msg[0]["content"], msg[1:], self._param.gen_conf(), images=self.imgs, **kwargs)):
+            yield t
+
+    async def _stream_output_async(self, prompt, msg):
+        _, msg = message_fit_in([{"role": "system", "content": prompt}, *msg], int(self.chat_mdl.max_length * 0.97))
+        answer = ""
         last_idx = 0
         endswith_think = False
+
         def delta(txt):
-            nonlocal ans, last_idx, endswith_think
+            nonlocal answer, last_idx, endswith_think
             delta_ans = txt[last_idx:]
-            ans = txt
+            answer = txt
 
             if delta_ans.find("<think>") == 0:
                 last_idx += len("<think>")
                 return "<think>"
             elif delta_ans.find("<think>") > 0:
-                delta_ans = txt[last_idx:last_idx+delta_ans.find("<think>")]
+                delta_ans = txt[last_idx:last_idx + delta_ans.find("<think>")]
                 last_idx += delta_ans.find("<think>")
                 return delta_ans
             elif delta_ans.endswith("</think>"):
@@ -193,20 +235,36 @@ class LLM(ComponentBase):
                 endswith_think = False
                 return "</think>"
 
-            last_idx = len(ans)
-            if ans.endswith("</think>"):
+            last_idx = len(answer)
+            if answer.endswith("</think>"):
                 last_idx -= len("</think>")
             return re.sub(r"(<think>|</think>)", "", delta_ans)
 
-        if not self.imgs:
-            for txt in self.chat_mdl.chat_streamly(msg[0]["content"], msg[1:], self._param.gen_conf(), **kwargs):
-                yield delta(txt)
-        else:
-            for txt in self.chat_mdl.chat_streamly(msg[0]["content"], msg[1:], self._param.gen_conf(), images=self.imgs, **kwargs):
-                yield delta(txt)
+        stream_kwargs = {"images": self.imgs} if self.imgs else {}
+        async for ans in self.chat_mdl.async_chat_streamly(msg[0]["content"], msg[1:], self._param.gen_conf(), **stream_kwargs):
+            if self.check_if_canceled("LLM streaming"):
+                return
+
+            if isinstance(ans, int):
+                continue
+
+            if ans.find("**ERROR**") >= 0:
+                if self.get_exception_default_value():
+                    self.set_output("content", self.get_exception_default_value())
+                    yield self.get_exception_default_value()
+                else:
+                    self.set_output("_ERROR", ans)
+                return
+
+            yield delta(ans)
+
+        self.set_output("content", answer)
 
     @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60)))
-    def _invoke(self, **kwargs):
+    async def _invoke_async(self, **kwargs):
+        if self.check_if_canceled("LLM processing"):
+            return
+
         def clean_formated_answer(ans: str) -> str:
             ans = re.sub(r"^.*</think>", "", ans, flags=re.DOTALL)
             ans = re.sub(r"^.*```json", "", ans, flags=re.DOTALL)
@@ -214,19 +272,25 @@ class LLM(ComponentBase):
 
         prompt, msg, _ = self._prepare_prompt_variables()
         error: str = ""
-        output_structure=None
+        output_structure = None
         try:
-            output_structure=self._param.outputs['structured']
+            output_structure = self._param.outputs["structured"]
         except Exception:
             pass
-        if output_structure:
-            schema=json.dumps(output_structure, ensure_ascii=False, indent=2)
-            prompt += structured_output_prompt(schema)
-            for _ in range(self._param.max_retries+1):
-                _, msg = message_fit_in([{"role": "system", "content": prompt}, *msg], int(self.chat_mdl.max_length * 0.97))
+        if output_structure and isinstance(output_structure, dict) and output_structure.get("properties") and len(output_structure["properties"]) > 0:
+            schema = json.dumps(output_structure, ensure_ascii=False, indent=2)
+            prompt_with_schema = prompt + structured_output_prompt(schema)
+            for _ in range(self._param.max_retries + 1):
+                if self.check_if_canceled("LLM processing"):
+                    return
+
+                _, msg_fit = message_fit_in(
+                    [{"role": "system", "content": prompt_with_schema}, *deepcopy(msg)],
+                    int(self.chat_mdl.max_length * 0.97),
+                )
                 error = ""
-                ans = self._generate(msg)
-                msg.pop(0)
+                ans = await self._generate_async(msg_fit)
+                msg_fit.pop(0)
                 if ans.find("**ERROR**") >= 0:
                     logging.error(f"LLM response error: {ans}")
                     error = ans
@@ -235,7 +299,7 @@ class LLM(ComponentBase):
                     self.set_output("structured", json_repair.loads(clean_formated_answer(ans)))
                     return
                 except Exception:
-                    msg.append({"role": "user", "content": "The answer can't not be parsed as JSON"})
+                    msg_fit.append({"role": "user", "content": "The answer can't not be parsed as JSON"})
                     error = "The answer can't not be parsed as JSON"
             if error:
                 self.set_output("_ERROR", error)
@@ -243,15 +307,23 @@ class LLM(ComponentBase):
 
         downstreams = self._canvas.get_component(self._id)["downstream"] if self._canvas.get_component(self._id) else []
         ex = self.exception_handler()
-        if any([self._canvas.get_component_obj(cid).component_name.lower()=="message" for cid in downstreams]) and not output_structure and not (ex and ex["goto"]):
-            self.set_output("content", partial(self._stream_output, prompt, msg))
+        if any([self._canvas.get_component_obj(cid).component_name.lower() == "message" for cid in downstreams]) and not (
+            ex and ex["goto"]
+        ):
+            self.set_output("content", partial(self._stream_output_async, prompt, deepcopy(msg)))
             return
 
-        for _ in range(self._param.max_retries+1):
-            _, msg = message_fit_in([{"role": "system", "content": prompt}, *msg], int(self.chat_mdl.max_length * 0.97))
+        error = ""
+        for _ in range(self._param.max_retries + 1):
+            if self.check_if_canceled("LLM processing"):
+                return
+
+            _, msg_fit = message_fit_in(
+                [{"role": "system", "content": prompt}, *deepcopy(msg)], int(self.chat_mdl.max_length * 0.97)
+            )
             error = ""
-            ans = self._generate(msg)
-            msg.pop(0)
+            ans = await self._generate_async(msg_fit)
+            msg_fit.pop(0)
             if ans.find("**ERROR**") >= 0:
                 logging.error(f"LLM response error: {ans}")
                 error = ans
@@ -265,26 +337,15 @@ class LLM(ComponentBase):
             else:
                 self.set_output("_ERROR", error)
 
-    def _stream_output(self, prompt, msg):
-        _, msg = message_fit_in([{"role": "system", "content": prompt}, *msg], int(self.chat_mdl.max_length * 0.97))
-        answer = ""
-        for ans in self._generate_streamly(msg):
-            if ans.find("**ERROR**") >= 0:
-                if self.get_exception_default_value():
-                    self.set_output("content", self.get_exception_default_value())
-                    yield self.get_exception_default_value()
-                else:
-                    self.set_output("_ERROR", ans)
-                return
-            yield ans
-            answer += ans
-        self.set_output("content", answer)
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60)))
+    def _invoke(self, **kwargs):
+        return asyncio.run(self._invoke_async(**kwargs))
 
-    def add_memory(self, user:str, assist:str, func_name: str, params: dict, results: str, user_defined_prompt:dict={}):
-        summ = tool_call_summary(self.chat_mdl, func_name, params, results, user_defined_prompt)
+    async def add_memory(self, user:str, assist:str, func_name: str, params: dict, results: str, user_defined_prompt:dict={}):
+        summ = await tool_call_summary(self.chat_mdl, func_name, params, results, user_defined_prompt)
         logging.info(f"[MEMORY]: {summ}")
         self._canvas.add_memory(user, assist, summ)
 
     def thoughts(self) -> str:
         _, msg,_ = self._prepare_prompt_variables()
-        return "⌛Give me a moment—starting from: \n\n" + re.sub(r"(User's query:|[\\]+)", '', msg[-1]['content'], flags=re.DOTALL) + "\n\nI’ll figure out our best next move."
+        return "⌛Give me a moment—starting from: \n\n" + re.sub(r"(User's query:|[\\]+)", '', msg[-1]['content'], flags=re.DOTALL) + "\n\nI’ll figure out our best next move."

agent/component/loop.py

@@ -0,0 +1,80 @@
+#
+#  Copyright 2024 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+#
+from abc import ABC
+from agent.component.base import ComponentBase, ComponentParamBase
+
+
+class LoopParam(ComponentParamBase):
+    """
+    Define the Loop component parameters.
+    """
+
+    def __init__(self):
+        super().__init__()
+        self.loop_variables = []
+        self.loop_termination_condition=[]
+        self.maximum_loop_count = 0
+
+    def get_input_form(self) -> dict[str, dict]:
+        return {
+            "items": {
+                "type": "json",
+                "name": "Items"
+            }
+        }
+
+    def check(self):
+        return True
+
+
+class Loop(ComponentBase, ABC):
+    component_name = "Loop"
+
+    def get_start(self):
+        for cid in self._canvas.components.keys():
+            if self._canvas.get_component(cid)["obj"].component_name.lower() != "loopitem":
+                continue
+            if self._canvas.get_component(cid)["parent_id"] == self._id:
+                return cid
+
+    def _invoke(self, **kwargs):
+        if self.check_if_canceled("Loop processing"):
+            return
+
+        for item in self._param.loop_variables:
+            if any([not item.get("variable"), not item.get("input_mode"), not item.get("value"),not item.get("type")]):
+                assert "Loop Variable is not complete."
+            if item["input_mode"]=="variable":
+                self.set_output(item["variable"],self._canvas.get_variable_value(item["value"]))
+            elif item["input_mode"]=="constant":
+                self.set_output(item["variable"],item["value"])
+            else:
+                if item["type"] == "number":
+                    self.set_output(item["variable"], 0)
+                elif item["type"] == "string":
+                    self.set_output(item["variable"], "")
+                elif item["type"] == "boolean":
+                    self.set_output(item["variable"], False)
+                elif item["type"].startswith("object"):
+                    self.set_output(item["variable"], {})
+                elif item["type"].startswith("array"):
+                    self.set_output(item["variable"], [])
+                else:
+                    self.set_output(item["variable"], "")
+
+
+    def thoughts(self) -> str:
+        return "Loop from canvas."

agent/component/loopitem.py

@@ -0,0 +1,167 @@
+#
+#  Copyright 2024 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+#
+from abc import ABC
+from agent.component.base import ComponentBase, ComponentParamBase
+
+
+class LoopItemParam(ComponentParamBase):
+    """
+    Define the LoopItem component parameters.
+    """
+    def check(self):
+        return True
+
+class LoopItem(ComponentBase, ABC):
+    component_name = "LoopItem"
+
+    def __init__(self, canvas, id, param: ComponentParamBase):
+        super().__init__(canvas, id, param)
+        self._idx = 0
+
+
+    def _invoke(self, **kwargs):
+        if self.check_if_canceled("LoopItem processing"):
+            return
+        parent = self.get_parent()
+        maximum_loop_count = parent._param.maximum_loop_count
+        if self._idx >= maximum_loop_count:
+            self._idx = -1
+            return
+        if self._idx > 0:
+            if self.check_if_canceled("LoopItem processing"):
+                return
+        self._idx += 1
+
+    def evaluate_condition(self,var, operator, value):
+        if isinstance(var, str):
+            if operator == "contains":
+                return value in var
+            elif operator == "not contains":
+                return value not in var
+            elif operator == "start with":
+                return var.startswith(value)
+            elif operator == "end with":
+                return var.endswith(value)
+            elif operator == "is":
+                return var == value
+            elif operator == "is not":
+                return var != value
+            elif operator == "empty":
+                return var == ""
+            elif operator == "not empty":
+                return var != ""
+
+        elif isinstance(var, (int, float)):
+            if operator == "=":
+                return var == value
+            elif operator == "≠":
+                return var != value
+            elif operator == ">":
+                return var > value
+            elif operator == "<":
+                return var < value
+            elif operator == "≥":
+                return var >= value
+            elif operator == "≤":
+                return var <= value
+            elif operator == "empty":
+                return var is None
+            elif operator == "not empty":
+                return var is not None
+
+        elif isinstance(var, bool):
+            if operator == "is":
+                return var is value
+            elif operator == "is not":
+                return var is not value
+            elif operator == "empty":
+                return var is None
+            elif operator == "not empty":
+                return var is not None
+
+        elif isinstance(var, dict):
+            if operator == "empty":
+                return len(var) == 0
+            elif operator == "not empty":
+                return len(var) > 0
+
+        elif isinstance(var, list):
+            if operator == "contains":
+                return value in var
+            elif operator == "not contains":
+                return value not in var
+
+            elif operator == "is":
+                return var == value
+            elif operator == "is not":
+                return var != value
+
+            elif operator == "empty":
+                return len(var) == 0
+            elif operator == "not empty":
+                return len(var) > 0
+        elif var is None:
+            if operator == "empty":
+                return True
+            return False
+
+        raise Exception(f"Invalid operator: {operator}")
+
+    def end(self):
+        if self._idx == -1:
+            return True
+        parent = self.get_parent()
+        logical_operator = parent._param.logical_operator if hasattr(parent._param, "logical_operator") else "and"
+        conditions = []
+        for item in parent._param.loop_termination_condition:
+            if not item.get("variable") or not item.get("operator"):
+                raise ValueError("Loop condition is incomplete.")
+            var = self._canvas.get_variable_value(item["variable"])
+            operator = item["operator"]
+            input_mode = item.get("input_mode", "constant")
+
+            if input_mode == "variable":
+                value = self._canvas.get_variable_value(item.get("value", ""))
+            elif input_mode == "constant":
+                value = item.get("value", "")
+            else:
+                raise ValueError("Invalid input mode.")
+            conditions.append(self.evaluate_condition(var, operator, value))
+        should_end = (
+            all(conditions) if logical_operator == "and"
+            else any(conditions) if logical_operator == "or"
+            else None
+        )
+        if should_end is None:
+            raise ValueError("Invalid logical operator,should be 'and' or 'or'.")
+
+        if should_end:
+            self._idx = -1
+            return True
+
+        return False
+
+    def next(self):
+        if self._idx == -1:
+            self._idx = 0
+        else:
+            self._idx += 1
+            if self._idx >= len(self._items):
+                self._idx = -1
+        return False
+
+    def thoughts(self) -> str:
+        return "Next turn..."

agent/component/message.py

@@ -13,10 +13,16 @@
 #  See the License for the specific language governing permissions and
 #  limitations under the License.
 #
+import asyncio
+import nest_asyncio
+nest_asyncio.apply()
+import inspect
 import json
 import os
 import random
 import re
+import logging
+import tempfile
 from functools import partial
 from typing import Any
 
@@ -24,6 +30,10 @@ from agent.component.base import ComponentBase, ComponentParamBase
 from jinja2 import Template as Jinja2Template
 
 from common.connection_utils import timeout
+from common.misc_utils import get_uuid
+from common import settings
+
+from api.db.joint_services.memory_message_service import queue_save_to_memory_task
 
 
 class MessageParam(ComponentParamBase):
@@ -34,6 +44,8 @@ class MessageParam(ComponentParamBase):
         super().__init__()
         self.content = []
         self.stream = True
+        self.output_format = None  # default output format
+        self.auto_play = False
         self.outputs = {
             "content": {
                 "type": "str"
@@ -49,6 +61,9 @@ class MessageParam(ComponentParamBase):
 class Message(ComponentBase):
     component_name = "Message"
 
+    def get_input_elements(self) -> dict[str, Any]:
+        return self.get_input_elements_from_text("".join(self._param.content))
+
     def get_kwargs(self, script:str, kwargs:dict = {}, delimiter:str=None) -> tuple[str, dict[str, str | list | Any]]:
         for k,v in self.get_input_elements_from_text(script).items():
             if k in kwargs:
@@ -58,8 +73,12 @@ class Message(ComponentBase):
                 v = ""
             ans = ""
             if isinstance(v, partial):
-                for t in v():
-                    ans += t
+                iter_obj = v()
+                if inspect.isasyncgen(iter_obj):
+                    ans = asyncio.run(self._consume_async_gen(iter_obj))
+                else:
+                    for t in iter_obj:
+                        ans += t
             elif isinstance(v, list) and delimiter:
                 ans = delimiter.join([str(vv) for vv in v])
             elif not isinstance(v, str):
@@ -81,11 +100,20 @@ class Message(ComponentBase):
             _kwargs[_n] = v
         return script, _kwargs
 
-    def _stream(self, rand_cnt:str):
+    async def _consume_async_gen(self, agen):
+        buf = ""
+        async for t in agen:
+            buf += t
+        return buf
+
+    async def _stream(self, rand_cnt:str):
         s = 0
         all_content = ""
         cache = {}
         for r in re.finditer(self.variable_ref_patt, rand_cnt, flags=re.DOTALL):
+            if self.check_if_canceled("Message streaming"):
+                return
+
             all_content += rand_cnt[s: r.start()]
             yield rand_cnt[s: r.start()]
             s = r.end()
@@ -96,30 +124,51 @@ class Message(ComponentBase):
                 continue
 
             v = self._canvas.get_variable_value(exp)
-            if not v:
+            if v is None:
                 v = ""
             if isinstance(v, partial):
                 cnt = ""
-                for t in v():
-                    all_content += t
-                    cnt += t
-                    yield t
+                iter_obj = v()
+                if inspect.isasyncgen(iter_obj):
+                    async for t in iter_obj:
+                        if self.check_if_canceled("Message streaming"):
+                            return
 
+                        all_content += t
+                        cnt += t
+                        yield t
+                else:
+                    for t in iter_obj:
+                        if self.check_if_canceled("Message streaming"):
+                            return
+
+                        all_content += t
+                        cnt += t
+                        yield t
+                self.set_input_value(exp, cnt)
                 continue
+            elif inspect.isawaitable(v):
+                v = await v
             elif not isinstance(v, str):
                 try:
-                    v = json.dumps(v, ensure_ascii=False, indent=2)
+                    v = json.dumps(v, ensure_ascii=False)
                 except Exception:
                     v = str(v)
             yield v
+            self.set_input_value(exp, v)
             all_content += v
             cache[exp] = v
 
         if s < len(rand_cnt):
+            if self.check_if_canceled("Message streaming"):
+                return
+
             all_content += rand_cnt[s: ]
             yield rand_cnt[s: ]
 
         self.set_output("content", all_content)
+        self._convert_content(all_content)
+        await self._save_to_memory(all_content)
 
     def _is_jinjia2(self, content:str) -> bool:
         patt = [
@@ -129,6 +178,9 @@ class Message(ComponentBase):
 
     @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60)))
     def _invoke(self, **kwargs):
+        if self.check_if_canceled("Message processing"):
+            return
+
         rand_cnt = random.choice(self._param.content)
         if self._param.stream and not self._is_jinjia2(rand_cnt):
             self.set_output("content", partial(self._stream, rand_cnt))
@@ -141,10 +193,248 @@ class Message(ComponentBase):
         except Exception:
             pass
 
+        if self.check_if_canceled("Message processing"):
+            return
+
         for n, v in kwargs.items():
             content = re.sub(n, v, content)
 
         self.set_output("content", content)
+        self._convert_content(content)
+        self._save_to_memory(content)
 
     def thoughts(self) -> str:
         return ""
+
+    def _parse_markdown_table_lines(self, table_lines: list):
+        """
+        Parse a list of Markdown table lines into a pandas DataFrame.
+        
+        Args:
+            table_lines: List of strings, each representing a row in the Markdown table
+                        (excluding separator lines like |---|---|)
+        
+        Returns:
+            pandas DataFrame with the table data, or None if parsing fails
+        """
+        import pandas as pd
+        
+        if not table_lines:
+            return None
+        
+        rows = []
+        headers = None
+        
+        for line in table_lines:
+            # Split by | and clean up
+            cells = [cell.strip() for cell in line.split('|')]
+            # Remove empty first and last elements from split (caused by leading/trailing |)
+            cells = [c for c in cells if c]
+            
+            if headers is None:
+                headers = cells
+            else:
+                rows.append(cells)
+        
+        if headers and rows:
+            # Ensure all rows have same number of columns as headers
+            normalized_rows = []
+            for row in rows:
+                while len(row) < len(headers):
+                    row.append('')
+                normalized_rows.append(row[:len(headers)])
+            
+            return pd.DataFrame(normalized_rows, columns=headers)
+        
+        return None
+
+    def _convert_content(self, content):
+        if not self._param.output_format:
+            return
+
+        import pypandoc
+        doc_id = get_uuid()
+
+        if self._param.output_format.lower() not in {"markdown", "html", "pdf", "docx", "xlsx"}:
+            self._param.output_format = "markdown"
+
+        try:
+            if self._param.output_format in {"markdown", "html"}:
+                if isinstance(content, str):
+                    converted = pypandoc.convert_text(
+                        content,
+                        to=self._param.output_format,
+                        format="markdown",
+                    )
+                else:
+                    converted = pypandoc.convert_file(
+                        content,
+                        to=self._param.output_format,
+                        format="markdown",
+                    )
+
+                binary_content = converted.encode("utf-8")
+
+            elif self._param.output_format == "xlsx":
+                import pandas as pd
+                from io import BytesIO
+
+                # Debug: log the content being parsed
+                logging.info(f"XLSX Parser: Content length={len(content) if content else 0}, first 500 chars: {content[:500] if content else 'None'}")
+                
+                # Try to parse ALL Markdown tables from the content
+                # Each table will be written to a separate sheet
+                tables = []  # List of (sheet_name, dataframe)
+                
+                if isinstance(content, str):
+                    lines = content.strip().split('\n')
+                    logging.info(f"XLSX Parser: Total lines={len(lines)}, lines starting with '|': {sum(1 for line in lines if line.strip().startswith('|'))}")
+                    current_table_lines = []
+                    current_table_title = None
+                    pending_title = None
+                    in_table = False
+                    table_count = 0
+                    
+                    for i, line in enumerate(lines):
+                        stripped = line.strip()
+                        
+                        # Check for potential table title (lines before a table)
+                        # Look for patterns like "Table 1:", "## Table", or markdown headers
+                        if not in_table and stripped and not stripped.startswith('|'):
+                            # Check if this could be a table title
+                            lower_stripped = stripped.lower()
+                            if (lower_stripped.startswith('table') or 
+                                stripped.startswith('#') or
+                                ':' in stripped):
+                                pending_title = stripped.lstrip('#').strip()
+                        
+                        if stripped.startswith('|') and '|' in stripped[1:]:
+                            # Check if this is a separator line (|---|---|)
+                            cleaned = stripped.replace(' ', '').replace('|', '').replace('-', '').replace(':', '')
+                            if cleaned == '':
+                                continue  # Skip separator line
+                            
+                            if not in_table:
+                                # Starting a new table
+                                in_table = True
+                                current_table_lines = []
+                                current_table_title = pending_title
+                                pending_title = None
+                            
+                            current_table_lines.append(stripped)
+                        
+                        elif in_table and not stripped.startswith('|'):
+                            # End of current table - save it
+                            if current_table_lines:
+                                df = self._parse_markdown_table_lines(current_table_lines)
+                                if df is not None and not df.empty:
+                                    table_count += 1
+                                    # Generate sheet name
+                                    if current_table_title:
+                                        # Clean and truncate title for sheet name
+                                        sheet_name = current_table_title[:31]
+                                        sheet_name = sheet_name.replace('/', '_').replace('\\', '_').replace('*', '').replace('?', '').replace('[', '').replace(']', '').replace(':', '')
+                                    else:
+                                        sheet_name = f"Table_{table_count}"
+                                    tables.append((sheet_name, df))
+                            
+                            # Reset for next table
+                            in_table = False
+                            current_table_lines = []
+                            current_table_title = None
+                            
+                            # Check if this line could be a title for the next table
+                            if stripped:
+                                lower_stripped = stripped.lower()
+                                if (lower_stripped.startswith('table') or 
+                                    stripped.startswith('#') or
+                                    ':' in stripped):
+                                    pending_title = stripped.lstrip('#').strip()
+                    
+                    # Don't forget the last table if content ends with a table
+                    if in_table and current_table_lines:
+                        df = self._parse_markdown_table_lines(current_table_lines)
+                        if df is not None and not df.empty:
+                            table_count += 1
+                            if current_table_title:
+                                sheet_name = current_table_title[:31]
+                                sheet_name = sheet_name.replace('/', '_').replace('\\', '_').replace('*', '').replace('?', '').replace('[', '').replace(']', '').replace(':', '')
+                            else:
+                                sheet_name = f"Table_{table_count}"
+                            tables.append((sheet_name, df))
+                
+                # Fallback: if no tables found, create single sheet with content
+                if not tables:
+                    df = pd.DataFrame({"Content": [content if content else ""]})
+                    tables = [("Data", df)]
+
+                # Write all tables to Excel, each in a separate sheet
+                excel_io = BytesIO()
+                with pd.ExcelWriter(excel_io, engine='openpyxl') as writer:
+                    used_names = set()
+                    for sheet_name, df in tables:
+                        # Ensure unique sheet names
+                        original_name = sheet_name
+                        counter = 1
+                        while sheet_name in used_names:
+                            suffix = f"_{counter}"
+                            sheet_name = original_name[:31-len(suffix)] + suffix
+                            counter += 1
+                        used_names.add(sheet_name)
+                        df.to_excel(writer, sheet_name=sheet_name, index=False)
+                
+                excel_io.seek(0)
+                binary_content = excel_io.read()
+                
+                logging.info(f"Generated Excel with {len(tables)} sheet(s): {[t[0] for t in tables]}")
+
+            else:  # pdf, docx
+                with tempfile.NamedTemporaryFile(suffix=f".{self._param.output_format}", delete=False) as tmp:
+                    tmp_name = tmp.name
+
+                try:
+                    if isinstance(content, str):
+                        pypandoc.convert_text(
+                            content,
+                            to=self._param.output_format,
+                            format="markdown",
+                            outputfile=tmp_name,
+                        )
+                    else:
+                        pypandoc.convert_file(
+                            content,
+                            to=self._param.output_format,
+                            format="markdown",
+                            outputfile=tmp_name,
+                        )
+
+                    with open(tmp_name, "rb") as f:
+                        binary_content = f.read()
+
+                finally:
+                    if os.path.exists(tmp_name):
+                        os.remove(tmp_name)
+
+            settings.STORAGE_IMPL.put(self._canvas._tenant_id, doc_id, binary_content)
+            self.set_output("attachment", {
+                "doc_id":doc_id,
+                "format":self._param.output_format,
+                "file_name":f"{doc_id[:8]}.{self._param.output_format}"})
+
+            logging.info(f"Converted content uploaded as {doc_id} (format={self._param.output_format})")
+
+        except Exception as e:
+            logging.error(f"Error converting content to {self._param.output_format}: {e}")
+
+    async def _save_to_memory(self, content):
+        if not hasattr(self._param, "memory_ids") or not self._param.memory_ids:
+            return True, "No memory selected."
+
+        message_dict = {
+            "user_id": self._canvas._tenant_id,
+            "agent_id": self._canvas._id,
+            "session_id": self._canvas.task_id,
+            "user_input": self._canvas.get_sys_query(),
+            "agent_response": content
+        }
+        return await queue_save_to_memory_task(self._param.memory_ids, message_dict)

agent/component/string_transform.py

@@ -16,6 +16,8 @@
 import os
 import re
 from abc import ABC
+from typing import Any
+
 from jinja2 import Template as Jinja2Template
 from agent.component.base import ComponentParamBase
 from common.connection_utils import timeout
@@ -43,6 +45,9 @@ class StringTransformParam(ComponentParamBase):
 class StringTransform(Message, ABC):
     component_name = "StringTransform"
 
+    def get_input_elements(self) -> dict[str, Any]:
+        return self.get_input_elements_from_text(self._param.script)
+
     def get_input_form(self) -> dict[str, dict]:
         if self._param.method == "split":
             return {
@@ -58,17 +63,24 @@ class StringTransform(Message, ABC):
 
     @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60)))
     def _invoke(self, **kwargs):
+        if self.check_if_canceled("StringTransform processing"):
+            return
+
         if self._param.method == "split":
             self._split(kwargs.get("line"))
         else:
             self._merge(kwargs)
 
     def _split(self, line:str|None = None):
+        if self.check_if_canceled("StringTransform split processing"):
+            return
+
         var = self._canvas.get_variable_value(self._param.split_ref) if not line else line
         if not var:
             var = ""
         assert isinstance(var, str), "The input variable is not a string: {}".format(type(var))
         self.set_input_value(self._param.split_ref, var)
+
         res = []
         for i,s in enumerate(re.split(r"(%s)"%("|".join([re.escape(d) for d in self._param.delimiters])), var, flags=re.DOTALL)):
             if i % 2 == 1:
@@ -77,6 +89,9 @@ class StringTransform(Message, ABC):
         self.set_output("result", res)
 
     def _merge(self, kwargs:dict[str, str] = {}):
+        if self.check_if_canceled("StringTransform merge processing"):
+            return
+
         script = self._param.script
         script, kwargs = self.get_kwargs(script, kwargs, self._param.delimiters[0])
 

agent/component/switch.py

@@ -63,9 +63,18 @@ class Switch(ComponentBase, ABC):
 
     @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 3)))
     def _invoke(self, **kwargs):
+        if self.check_if_canceled("Switch processing"):
+            return
+
         for cond in self._param.conditions:
+            if self.check_if_canceled("Switch processing"):
+                return
+
             res = []
             for item in cond["items"]:
+                if self.check_if_canceled("Switch processing"):
+                    return
+
                 if not item["cpn_id"]:
                     continue
                 cpn_v = self._canvas.get_variable_value(item["cpn_id"])
@@ -128,4 +137,4 @@ class Switch(ComponentBase, ABC):
         raise ValueError('Not supported operator' + operator)
 
     def thoughts(self) -> str:
-        return "I’m weighing a few options and will pick the next step shortly."
+        return "I’m weighing a few options and will pick the next step shortly."

agent/component/varaiable_aggregator.py

@@ -0,0 +1,84 @@
+#
+#  Copyright 2024 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+
+from typing import Any
+import os
+
+from common.connection_utils import timeout
+from agent.component.base import ComponentBase, ComponentParamBase
+
+
+class VariableAggregatorParam(ComponentParamBase):
+    """
+    Parameters for VariableAggregator
+
+    - groups: list of dicts {"group_name": str, "variables": [variable selectors]}
+    """
+
+    def __init__(self):
+        super().__init__()
+        # each group expects: {"group_name": str, "variables": List[str]}
+        self.groups = []
+
+    def check(self):
+        self.check_empty(self.groups, "[VariableAggregator] groups")
+        for g in self.groups:
+            if not g.get("group_name"):
+                raise ValueError("[VariableAggregator] group_name can not be empty!")
+            if not g.get("variables"):
+                raise ValueError(
+                    f"[VariableAggregator] variables of group `{g.get('group_name')}` can not be empty"
+                )
+            if not isinstance(g.get("variables"), list):
+                raise ValueError(
+                    f"[VariableAggregator] variables of group `{g.get('group_name')}` should be a list of strings"
+                )
+
+    def get_input_form(self) -> dict[str, dict]:
+        return {
+            "variables": {
+                "name": "Variables",
+                "type": "list",
+            }
+        }
+
+
+class VariableAggregator(ComponentBase):
+    component_name = "VariableAggregator"
+
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 3)))
+    def _invoke(self, **kwargs):
+        # Group mode: for each group, pick the first available variable
+        for group in self._param.groups:
+            gname = group.get("group_name")
+
+            # record candidate selectors within this group
+            self.set_input_value(f"{gname}.variables", list(group.get("variables", [])))
+            for selector in group.get("variables", []):
+                val = self._canvas.get_variable_value(selector['value'])
+                if val:
+                    self.set_output(gname, val)
+                    break
+            
+    @staticmethod
+    def _to_object(value: Any) -> Any:
+        # Try to convert value to serializable object if it has to_object()
+        try:
+            return value.to_object()  # type: ignore[attr-defined]
+        except Exception:
+            return value
+
+    def thoughts(self) -> str:
+        return "Aggregating variables from canvas and grouping as configured."

agent/component/variable_assigner.py

@@ -0,0 +1,192 @@
+#
+#  Copyright 2024 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+#
+from abc import ABC
+import os
+import numbers
+from agent.component.base import ComponentBase, ComponentParamBase
+from api.utils.api_utils import timeout
+
+class VariableAssignerParam(ComponentParamBase):
+    """
+    Define the Variable Assigner component parameters.
+    """
+    def __init__(self):
+        super().__init__()
+        self.variables=[]
+
+    def check(self):
+        return True
+    
+    def get_input_form(self) -> dict[str, dict]:
+        return {
+            "items": {
+                "type": "json",
+                "name": "Items"
+            }
+        }
+
+class VariableAssigner(ComponentBase,ABC):
+    component_name = "VariableAssigner"
+
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60)))
+    def _invoke(self, **kwargs):
+        if not isinstance(self._param.variables,list):
+            return
+        else:
+            for item in self._param.variables:
+                if any([not item.get("variable"), not item.get("operator"), not item.get("parameter")]):
+                    assert "Variable is not complete."
+                variable=item["variable"]
+                operator=item["operator"]
+                parameter=item["parameter"]
+                variable_value=self._canvas.get_variable_value(variable)
+                new_variable=self._operate(variable_value,operator,parameter)
+                self._canvas.set_variable_value(variable, new_variable)
+
+    def _operate(self,variable,operator,parameter):
+        if operator == "overwrite":
+            return self._overwrite(parameter)
+        elif operator == "clear":
+            return self._clear(variable)
+        elif operator == "set":
+            return self._set(variable,parameter)
+        elif operator == "append":
+            return self._append(variable,parameter)
+        elif operator == "extend":
+            return self._extend(variable,parameter)
+        elif operator == "remove_first":
+            return self._remove_first(variable)
+        elif operator == "remove_last":
+            return self._remove_last(variable)
+        elif operator == "+=":
+            return self._add(variable,parameter)
+        elif operator == "-=":
+            return self._subtract(variable,parameter)
+        elif operator == "*=":
+            return self._multiply(variable,parameter)
+        elif operator == "/=":
+            return self._divide(variable,parameter)
+        else:
+            return
+    
+    def _overwrite(self,parameter):
+        return self._canvas.get_variable_value(parameter)
+
+    def _clear(self,variable):
+        if isinstance(variable,list):
+            return []
+        elif isinstance(variable,str):
+            return ""
+        elif isinstance(variable,dict):
+            return {}
+        elif isinstance(variable,int):
+            return 0
+        elif isinstance(variable,float):
+            return 0.0
+        elif isinstance(variable,bool):
+            return False
+        else:
+            return None
+
+    def _set(self,variable,parameter):
+        if variable is None:
+            return self._canvas.get_value_with_variable(parameter)
+        elif isinstance(variable,str):
+            return self._canvas.get_value_with_variable(parameter)
+        elif isinstance(variable,bool):
+            return parameter
+        elif isinstance(variable,int):
+            return parameter
+        elif isinstance(variable,float):
+            return parameter
+        else:
+            return parameter
+
+    def _append(self,variable,parameter):
+        parameter=self._canvas.get_variable_value(parameter)
+        if variable is None:
+            variable=[]
+        if not isinstance(variable,list):
+            return "ERROR:VARIABLE_NOT_LIST"
+        elif len(variable)!=0 and not isinstance(parameter,type(variable[0])):
+            return "ERROR:PARAMETER_NOT_LIST_ELEMENT_TYPE"
+        else:
+            variable.append(parameter)
+            return variable
+
+    def _extend(self,variable,parameter):
+        parameter=self._canvas.get_variable_value(parameter)
+        if variable is None:
+            variable=[]
+        if not isinstance(variable,list):
+            return "ERROR:VARIABLE_NOT_LIST"
+        elif not isinstance(parameter,list):
+            return "ERROR:PARAMETER_NOT_LIST"
+        elif len(variable)!=0 and len(parameter)!=0 and not isinstance(parameter[0],type(variable[0])):
+            return "ERROR:PARAMETER_NOT_LIST_ELEMENT_TYPE"
+        else:
+            return variable + parameter
+
+    def _remove_first(self,variable):
+        if len(variable)==0:
+            return variable
+        if not isinstance(variable,list):
+            return "ERROR:VARIABLE_NOT_LIST"
+        else:
+            return variable[1:]
+
+    def _remove_last(self,variable):
+        if len(variable)==0:
+            return variable
+        if not isinstance(variable,list):
+            return "ERROR:VARIABLE_NOT_LIST"
+        else:
+            return variable[:-1]
+
+    def is_number(self, value):
+        if isinstance(value, bool):
+            return False
+        return isinstance(value, numbers.Number)
+
+    def _add(self,variable,parameter):
+        if self.is_number(variable) and self.is_number(parameter):
+            return variable + parameter
+        else:
+            return "ERROR:VARIABLE_NOT_NUMBER or PARAMETER_NOT_NUMBER"
+
+    def _subtract(self,variable,parameter):
+        if self.is_number(variable) and self.is_number(parameter):
+            return variable - parameter
+        else:
+            return "ERROR:VARIABLE_NOT_NUMBER or PARAMETER_NOT_NUMBER"
+
+    def _multiply(self,variable,parameter):
+        if self.is_number(variable) and self.is_number(parameter):
+            return variable * parameter
+        else:
+            return "ERROR:VARIABLE_NOT_NUMBER or PARAMETER_NOT_NUMBER"
+
+    def _divide(self,variable,parameter):
+        if self.is_number(variable) and self.is_number(parameter):
+            if  parameter==0:
+                return "ERROR:DIVIDE_BY_ZERO"
+            else:
+                return variable/parameter
+        else:
+            return  "ERROR:VARIABLE_NOT_NUMBER or PARAMETER_NOT_NUMBER"
+
+    def thoughts(self) -> str:
+        return "Assign variables from canvas."

agent/templates/advanced_ingestion_pipeline.json

@@ -193,7 +193,7 @@
                         "presence_penalty": 0.4,
                         "prompts": [
                             {
-                                "content": "Text Content:\n{Splitter:KindDingosJam@chunks}\n",
+                                "content": "Text Content:\n{Splitter:NineTiesSin@chunks}\n",
                                 "role": "user"
                             }
                         ],
@@ -226,7 +226,7 @@
                         "presence_penalty": 0.4,
                         "prompts": [
                             {
-                                "content": "Text Content:\n\n{Splitter:KindDingosJam@chunks}\n",
+                                "content": "Text Content:\n\n{Splitter:TastyPointsLay@chunks}\n",
                                 "role": "user"
                             }
                         ],
@@ -259,7 +259,7 @@
                         "presence_penalty": 0.4,
                         "prompts": [
                             {
-                                "content": "Content: \n\n{Splitter:KindDingosJam@chunks}",
+                                "content": "Content: \n\n{Splitter:CuteBusesBet@chunks}",
                                 "role": "user"
                             }
                         ],
@@ -485,7 +485,7 @@
                             "outputs": {},
                             "presencePenaltyEnabled": false,
                             "presence_penalty": 0.4,
-                            "prompts": "Text Content:\n{Splitter:KindDingosJam@chunks}\n",
+                            "prompts": "Text Content:\n{Splitter:NineTiesSin@chunks}\n",
                             "sys_prompt": "Role\nYou are a text analyzer.\n\nTask\nExtract the most important keywords/phrases of a given piece of text content.\n\nRequirements\n- Summarize the text content, and give the top 5 important keywords/phrases.\n- The keywords MUST be in the same language as the given piece of text content.\n- The keywords are delimited by ENGLISH COMMA.\n- Output keywords ONLY.",
                             "temperature": 0.1,
                             "temperatureEnabled": false,
@@ -522,7 +522,7 @@
                             "outputs": {},
                             "presencePenaltyEnabled": false,
                             "presence_penalty": 0.4,
-                            "prompts": "Text Content:\n\n{Splitter:KindDingosJam@chunks}\n",
+                            "prompts": "Text Content:\n\n{Splitter:TastyPointsLay@chunks}\n",
                             "sys_prompt": "Role\nYou are a text analyzer.\n\nTask\nPropose 3 questions about a given piece of text content.\n\nRequirements\n- Understand and summarize the text content, and propose the top 3 important questions.\n- The questions SHOULD NOT have overlapping meanings.\n- The questions SHOULD cover the main content of the text as much as possible.\n- The questions MUST be in the same language as the given piece of text content.\n- One question per line.\n- Output questions ONLY.",
                             "temperature": 0.1,
                             "temperatureEnabled": false,
@@ -559,7 +559,7 @@
                             "outputs": {},
                             "presencePenaltyEnabled": false,
                             "presence_penalty": 0.4,
-                            "prompts": "Content: \n\n{Splitter:KindDingosJam@chunks}",
+                            "prompts": "Content: \n\n{Splitter:BlueResultsWink@chunks}",
                             "sys_prompt": "Extract important structured information from the given content. Output ONLY a valid JSON string with no additional text. If no important structured information is found, output an empty JSON object: {}.\n\nImportant structured information may include: names, dates, locations, events, key facts, numerical data, or other extractable entities.",
                             "temperature": 0.1,
                             "temperatureEnabled": false,

agent/templates/sql_assistant.json

@@ -83,10 +83,10 @@
                                     "value": []
                                 }
                             },
-                            "password": "20010812Yy!",
+                            "password": "",
                             "port": 3306,
                             "sql": "{Agent:WickedGoatsDivide@content}",
-                            "username": "13637682833@163.com"
+                            "username": ""
                         }
                     },
                     "upstream": [
@@ -527,10 +527,10 @@
                                         "value": []
                                     }
                                 },
-                                "password": "20010812Yy!",
+                                "password": "",
                                 "port": 3306,
                                 "sql": "{Agent:WickedGoatsDivide@content}",
-                                "username": "13637682833@163.com"
+                                "username": ""
                             },
                             "label": "ExeSQL",
                             "name": "ExeSQL"
@@ -578,7 +578,7 @@
                     {
                         "data": {
                             "form": {
-                                "text": "Searches for relevant database creation statements.\n\nIt should label with a knowledgebase to which the schema is dumped in. You could use \" General \" as parsing method, \" 2 \" as chunk size and \" ; \"  as delimiter."
+                                "text": "Searches for relevant database creation statements.\n\nIt should label with a dataset to which the schema is dumped in. You could use \" General \" as parsing method, \" 2 \" as chunk size and \" ; \"  as delimiter."
                             },
                             "label": "Note",
                             "name": "Note Schema"

agent/templates/user_interaction.json

@@ -0,0 +1,519 @@
+{
+    "id": 27,
+    "title": {
+        "en": "Interactive Agent",
+        "zh": "可交互的 Agent"
+    },
+    "description": {
+        "en": "During the Agent’s execution, users can actively intervene and interact with the Agent to adjust or guide its output, ensuring the final result aligns with their intentions.",
+        "zh": "在 Agent 的运行过程中，用户可以随时介入，与 Agent 进行交互，以调整或引导生成结果，使最终输出更符合预期。"
+    },
+    "canvas_type": "Agent",
+    "dsl": {
+        "components": {
+            "Agent:LargeFliesMelt": {
+            "downstream": [
+                "UserFillUp:GoldBroomsRelate"
+            ],
+            "obj": {
+                "component_name": "Agent",
+                "params": {
+                "cite": true,
+                "delay_after_error": 1,
+                "description": "",
+                "exception_default_value": "",
+                "exception_goto": [],
+                "exception_method": "",
+                "frequencyPenaltyEnabled": false,
+                "frequency_penalty": 0.7,
+                "llm_id": "qwen-turbo@Tongyi-Qianwen",
+                "maxTokensEnabled": false,
+                "max_retries": 3,
+                "max_rounds": 1,
+                "max_tokens": 256,
+                "mcp": [],
+                "message_history_window_size": 12,
+                "outputs": {
+                    "content": {
+                    "type": "string",
+                    "value": ""
+                    },
+                    "structured": {}
+                },
+                "presencePenaltyEnabled": false,
+                "presence_penalty": 0.4,
+                "prompts": [
+                    {
+                    "content": "User query:{sys.query}",
+                    "role": "user"
+                    }
+                ],
+                "sys_prompt": "<role>\nYou are the Planning Agent in a multi-agent RAG workflow.\nYour sole job is to design a crisp, executable Search Plan for the next agent. Do not search or answer the user’s question.\n</role>\n<objectives>\nUnderstand the user’s task and decompose it into evidence-seeking steps.\nProduce high-quality queries and retrieval settings tailored to the task type (fact lookup, multi-hop reasoning, comparison, statistics, how-to, etc.).\nIdentify missing information that would materially change the plan (≤3 concise questions).\nOptimize for source trustworthiness, diversity, and recency; define stopping criteria to avoid over-searching.\nAnswer in 150 words.\n<objectives>",
+                "temperature": 0.1,
+                "temperatureEnabled": false,
+                "tools": [],
+                "topPEnabled": false,
+                "top_p": 0.3,
+                "user_prompt": "",
+                "visual_files_var": ""
+                }
+            },
+            "upstream": [
+                "begin"
+            ]
+            },
+            "Agent:TangyWordsType": {
+            "downstream": [
+                "Message:FreshWallsStudy"
+            ],
+            "obj": {
+                "component_name": "Agent",
+                "params": {
+                "cite": true,
+                "delay_after_error": 1,
+                "description": "",
+                "exception_default_value": "",
+                "exception_goto": [],
+                "exception_method": "",
+                "frequencyPenaltyEnabled": false,
+                "frequency_penalty": 0.7,
+                "llm_id": "qwen-turbo@Tongyi-Qianwen",
+                "maxTokensEnabled": false,
+                "max_retries": 3,
+                "max_rounds": 1,
+                "max_tokens": 256,
+                "mcp": [],
+                "message_history_window_size": 12,
+                "outputs": {
+                    "content": {
+                    "type": "string",
+                    "value": ""
+                    },
+                    "structured": {}
+                },
+                "presencePenaltyEnabled": false,
+                "presence_penalty": 0.4,
+                "prompts": [
+                    {
+                    "content": "Search Plan: {Agent:LargeFliesMelt@content}\n\n\n\nAwait Response feedback:{UserFillUp:GoldBroomsRelate@instructions}\n",
+                    "role": "user"
+                    }
+                ],
+                "sys_prompt": "<role>\nYou are the Search Agent.\nYour job is to execute the approved Search Plan, integrate the Await Response feedback, retrieve evidence, and produce a well-grounded answer.\n</role>\n<objectives>\nTranslate the plan + feedback into concrete searches.\nCollect diverse, trustworthy, and recent evidence meeting the plan’s evidence bar.\nSynthesize a concise answer; include citations next to claims they support.\nIf evidence is insufficient or conflicting, clearly state limitations and propose next steps.\n</objectives>\n <tools>\nRetrieval: You must use Retrieval to do the search.\n </tools>\n",
+                "temperature": 0.1,
+                "temperatureEnabled": false,
+                "tools": [
+                    {
+                    "component_name": "Retrieval",
+                    "name": "Retrieval",
+                    "params": {
+                        "cross_languages": [],
+                        "description": "",
+                        "empty_response": "",
+                        "kb_ids": [],
+                        "keywords_similarity_weight": 0.7,
+                        "outputs": {
+                        "formalized_content": {
+                            "type": "string",
+                            "value": ""
+                        },
+                        "json": {
+                            "type": "Array<Object>",
+                            "value": []
+                        }
+                        },
+                        "rerank_id": "",
+                        "similarity_threshold": 0.2,
+                        "toc_enhance": false,
+                        "top_k": 1024,
+                        "top_n": 8,
+                        "use_kg": false
+                    }
+                    }
+                ],
+                "topPEnabled": false,
+                "top_p": 0.3,
+                "user_prompt": "",
+                "visual_files_var": ""
+                }
+            },
+            "upstream": [
+                "UserFillUp:GoldBroomsRelate"
+            ]
+            },
+            "Message:FreshWallsStudy": {
+            "downstream": [],
+            "obj": {
+                "component_name": "Message",
+                "params": {
+                "content": [
+                    "{Agent:TangyWordsType@content}"
+                ]
+                }
+            },
+            "upstream": [
+                "Agent:TangyWordsType"
+            ]
+            },
+            "UserFillUp:GoldBroomsRelate": {
+            "downstream": [
+                "Agent:TangyWordsType"
+            ],
+            "obj": {
+                "component_name": "UserFillUp",
+                "params": {
+                "enable_tips": true,
+                "inputs": {
+                    "instructions": {
+                    "name": "instructions",
+                    "optional": false,
+                    "options": [],
+                    "type": "paragraph"
+                    }
+                },
+                "outputs": {
+                    "instructions": {
+                    "name": "instructions",
+                    "optional": false,
+                    "options": [],
+                    "type": "paragraph"
+                    }
+                },
+                "tips": "Here is my search plan:\n{Agent:LargeFliesMelt@content}\nAre you okay with it?"
+                }
+            },
+            "upstream": [
+                "Agent:LargeFliesMelt"
+            ]
+            },
+            "begin": {
+            "downstream": [
+                "Agent:LargeFliesMelt"
+            ],
+            "obj": {
+                "component_name": "Begin",
+                "params": {}
+            },
+            "upstream": []
+            }
+        },
+        "globals": {
+            "sys.conversation_turns": 0,
+            "sys.files": [],
+            "sys.query": "",
+            "sys.user_id": ""
+        },
+        "graph": {
+            "edges": [
+            {
+                "data": {
+                "isHovered": false
+                },
+                "id": "xy-edge__beginstart-Agent:LargeFliesMeltend",
+                "source": "begin",
+                "sourceHandle": "start",
+                "target": "Agent:LargeFliesMelt",
+                "targetHandle": "end"
+            },
+            {
+                "data": {
+                "isHovered": false
+                },
+                "id": "xy-edge__Agent:LargeFliesMeltstart-UserFillUp:GoldBroomsRelateend",
+                "source": "Agent:LargeFliesMelt",
+                "sourceHandle": "start",
+                "target": "UserFillUp:GoldBroomsRelate",
+                "targetHandle": "end"
+            },
+            {
+                "data": {
+                "isHovered": false
+                },
+                "id": "xy-edge__UserFillUp:GoldBroomsRelatestart-Agent:TangyWordsTypeend",
+                "source": "UserFillUp:GoldBroomsRelate",
+                "sourceHandle": "start",
+                "target": "Agent:TangyWordsType",
+                "targetHandle": "end"
+            },
+            {
+                "id": "xy-edge__Agent:TangyWordsTypetool-Tool:NastyBatsGoend",
+                "source": "Agent:TangyWordsType",
+                "sourceHandle": "tool",
+                "target": "Tool:NastyBatsGo",
+                "targetHandle": "end"
+            },
+            {
+                "id": "xy-edge__Agent:TangyWordsTypestart-Message:FreshWallsStudyend",
+                "source": "Agent:TangyWordsType",
+                "sourceHandle": "start",
+                "target": "Message:FreshWallsStudy",
+                "targetHandle": "end"
+            }
+            ],
+            "nodes": [
+            {
+                "data": {
+                "label": "Begin",
+                "name": "begin"
+                },
+                "dragging": false,
+                "id": "begin",
+                "measured": {
+                "height": 50,
+                "width": 200
+                },
+                "position": {
+                "x": 154.9008789064451,
+                "y": 119.51001744285344
+                },
+                "selected": false,
+                "sourcePosition": "left",
+                "targetPosition": "right",
+                "type": "beginNode"
+            },
+            {
+                "data": {
+                "form": {
+                    "cite": true,
+                    "delay_after_error": 1,
+                    "description": "",
+                    "exception_default_value": "",
+                    "exception_goto": [],
+                    "exception_method": "",
+                    "frequencyPenaltyEnabled": false,
+                    "frequency_penalty": 0.7,
+                    "llm_id": "qwen-turbo@Tongyi-Qianwen",
+                    "maxTokensEnabled": false,
+                    "max_retries": 3,
+                    "max_rounds": 1,
+                    "max_tokens": 256,
+                    "mcp": [],
+                    "message_history_window_size": 12,
+                    "outputs": {
+                    "content": {
+                        "type": "string",
+                        "value": ""
+                    },
+                    "structured": {}
+                    },
+                    "presencePenaltyEnabled": false,
+                    "presence_penalty": 0.4,
+                    "prompts": [
+                    {
+                        "content": "User query:{sys.query}",
+                        "role": "user"
+                    }
+                    ],
+                    "sys_prompt": "<role>\nYou are the Planning Agent in a multi-agent RAG workflow.\nYour sole job is to design a crisp, executable Search Plan for the next agent. Do not search or answer the user’s question.\n</role>\n<objectives>\nUnderstand the user’s task and decompose it into evidence-seeking steps.\nProduce high-quality queries and retrieval settings tailored to the task type (fact lookup, multi-hop reasoning, comparison, statistics, how-to, etc.).\nIdentify missing information that would materially change the plan (≤3 concise questions).\nOptimize for source trustworthiness, diversity, and recency; define stopping criteria to avoid over-searching.\nAnswer in 150 words.\n<objectives>",
+                    "temperature": 0.1,
+                    "temperatureEnabled": false,
+                    "tools": [],
+                    "topPEnabled": false,
+                    "top_p": 0.3,
+                    "user_prompt": "",
+                    "visual_files_var": ""
+                },
+                "label": "Agent",
+                "name": "Planning Agent"
+                },
+                "dragging": false,
+                "id": "Agent:LargeFliesMelt",
+                "measured": {
+                "height": 90,
+                "width": 200
+                },
+                "position": {
+                "x": 443.96309330796714,
+                "y": 104.61370811205677
+                },
+                "selected": false,
+                "sourcePosition": "right",
+                "targetPosition": "left",
+                "type": "agentNode"
+            },
+            {
+                "data": {
+                "form": {
+                    "enable_tips": true,
+                    "inputs": {
+                    "instructions": {
+                        "name": "instructions",
+                        "optional": false,
+                        "options": [],
+                        "type": "paragraph"
+                    }
+                    },
+                    "outputs": {
+                    "instructions": {
+                        "name": "instructions",
+                        "optional": false,
+                        "options": [],
+                        "type": "paragraph"
+                    }
+                    },
+                    "tips": "Here is my search plan:\n{Agent:LargeFliesMelt@content}\nAre you okay with it?"
+                },
+                "label": "UserFillUp",
+                "name": "Await Response"
+                },
+                "dragging": false,
+                "id": "UserFillUp:GoldBroomsRelate",
+                "measured": {
+                "height": 50,
+                "width": 200
+                },
+                "position": {
+                "x": 683.3409492927474,
+                "y": 116.76274137645598
+                },
+                "selected": false,
+                "sourcePosition": "right",
+                "targetPosition": "left",
+                "type": "ragNode"
+            },
+            {
+                "data": {
+                "form": {
+                    "cite": true,
+                    "delay_after_error": 1,
+                    "description": "",
+                    "exception_default_value": "",
+                    "exception_goto": [],
+                    "exception_method": "",
+                    "frequencyPenaltyEnabled": false,
+                    "frequency_penalty": 0.7,
+                    "llm_id": "qwen-turbo@Tongyi-Qianwen",
+                    "maxTokensEnabled": false,
+                    "max_retries": 3,
+                    "max_rounds": 1,
+                    "max_tokens": 256,
+                    "mcp": [],
+                    "message_history_window_size": 12,
+                    "outputs": {
+                    "content": {
+                        "type": "string",
+                        "value": ""
+                    },
+                    "structured": {}
+                    },
+                    "presencePenaltyEnabled": false,
+                    "presence_penalty": 0.4,
+                    "prompts": [
+                    {
+                        "content": "Search Plan: {Agent:LargeFliesMelt@content}\n\n\n\nAwait Response feedback:{UserFillUp:GoldBroomsRelate@instructions}\n",
+                        "role": "user"
+                    }
+                    ],
+                    "sys_prompt": "<role>\nYou are the Search Agent.\nYour job is to execute the approved Search Plan, integrate the Await Response feedback, retrieve evidence, and produce a well-grounded answer.\n</role>\n<objectives>\nTranslate the plan + feedback into concrete searches.\nCollect diverse, trustworthy, and recent evidence meeting the plan’s evidence bar.\nSynthesize a concise answer; include citations next to claims they support.\nIf evidence is insufficient or conflicting, clearly state limitations and propose next steps.\n</objectives>\n <tools>\nRetrieval: You must use Retrieval to do the search.\n </tools>\n",
+                    "temperature": 0.1,
+                    "temperatureEnabled": false,
+                    "tools": [
+                    {
+                        "component_name": "Retrieval",
+                        "name": "Retrieval",
+                        "params": {
+                        "cross_languages": [],
+                        "description": "",
+                        "empty_response": "",
+                        "kb_ids": [],
+                        "keywords_similarity_weight": 0.7,
+                        "outputs": {
+                            "formalized_content": {
+                            "type": "string",
+                            "value": ""
+                            },
+                            "json": {
+                            "type": "Array<Object>",
+                            "value": []
+                            }
+                        },
+                        "rerank_id": "",
+                        "similarity_threshold": 0.2,
+                        "toc_enhance": false,
+                        "top_k": 1024,
+                        "top_n": 8,
+                        "use_kg": false
+                        }
+                    }
+                    ],
+                    "topPEnabled": false,
+                    "top_p": 0.3,
+                    "user_prompt": "",
+                    "visual_files_var": ""
+                },
+                "label": "Agent",
+                "name": "Search Agent"
+                },
+                "dragging": false,
+                "id": "Agent:TangyWordsType",
+                "measured": {
+                "height": 90,
+                "width": 200
+                },
+                "position": {
+                "x": 944.6411255659472,
+                "y": 99.84499066368488
+                },
+                "selected": true,
+                "sourcePosition": "right",
+                "targetPosition": "left",
+                "type": "agentNode"
+            },
+            {
+                "data": {
+                "form": {
+                    "description": "This is an agent for a specific task.",
+                    "user_prompt": "This is the order you need to send to the agent."
+                },
+                "label": "Tool",
+                "name": "flow.tool_0"
+                },
+                "id": "Tool:NastyBatsGo",
+                "measured": {
+                "height": 50,
+                "width": 200
+                },
+                "position": {
+                "x": 862.6411255659472,
+                "y": 239.84499066368488
+                },
+                "sourcePosition": "right",
+                "targetPosition": "left",
+                "type": "toolNode"
+            },
+            {
+                "data": {
+                "form": {
+                    "content": [
+                    "{Agent:TangyWordsType@content}"
+                    ]
+                },
+                "label": "Message",
+                "name": "Message"
+                },
+                "dragging": false,
+                "id": "Message:FreshWallsStudy",
+                "measured": {
+                "height": 50,
+                "width": 200
+                },
+                "position": {
+                "x": 1216.7057997987163,
+                "y": 120.48541298149814
+                },
+                "selected": false,
+                "sourcePosition": "right",
+                "targetPosition": "left",
+                "type": "messageNode"
+            }
+            ]
+        },
+        "history": [],
+        "messages": [],
+        "path": [],
+        "retrieval": [],
+        "variables": {}
+        },
+        "avatar":
+    "data:image/png;base64,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"
+}

agent/test/client.py

@@ -16,7 +16,7 @@
 import argparse
 import os
 from agent.canvas import Canvas
-from api import settings
+from common import settings
 
 if __name__ == '__main__':
     parser = argparse.ArgumentParser()

agent/test/dsl_examples/categorize_and_agent_with_tavily.json

@@ -75,7 +75,7 @@
   },
   "history": [],
   "path": [],
-  "retrival": {"chunks": [], "doc_aggs": []},
+  "retrieval": {"chunks": [], "doc_aggs": []},
   "globals": {
     "sys.query": "",
     "sys.user_id": "",

agent/test/dsl_examples/iteration.json

@@ -82,7 +82,7 @@
   },
   "history": [],
   "path": [],
-  "retrival": {"chunks": [], "doc_aggs": []},
+  "retrieval": {"chunks": [], "doc_aggs": []},
   "globals": {
     "sys.query": "",
     "sys.user_id": "",

agent/test/dsl_examples/retrieval_and_generate.json

@@ -31,7 +31,7 @@
                     "component_name": "LLM",
                     "params": {
                       "llm_id": "deepseek-chat",
-                      "sys_prompt": "You are an intelligent assistant. Please summarize the content of the knowledge base to answer the question. Please list the data in the knowledge base and answer in detail. When all knowledge base content is irrelevant to the question, your answer must include the sentence \"The answer you are looking for is not found in the knowledge base!\" Answers need to consider chat history.\n      Here is the knowledge base:\n      {retrieval:0@formalized_content}\n      The above is the knowledge base.",
+                      "sys_prompt": "You are an intelligent assistant. Please summarize the content of the knowledge base to answer the question. Please list the data in the knowledge base and answer in detail. When all knowledge base content is irrelevant to the question, your answer must include the sentence \"The answer you are looking for is not found in the knowledge base!\" Answers need to consider chat history.\n      Here is the knowledge base:\n      {retrieval:0@formalized_content}\n      Above is the knowledge base.",
                       "temperature": 0.2
                     }
                 },
@@ -51,7 +51,7 @@
   },
   "history": [],
   "path": [],
-  "retrival": {"chunks": [], "doc_aggs": []},
+  "retrieval": {"chunks": [], "doc_aggs": []},
   "globals": {
     "sys.query": "",
     "sys.user_id": "",

agent/test/dsl_examples/retrieval_categorize_and_generate.json

@@ -65,7 +65,7 @@
                     "component_name": "Agent",
                     "params": {
                       "llm_id": "deepseek-chat",
-                      "sys_prompt": "You are an intelligent assistant. Please summarize the content of the knowledge base to answer the question. Please list the data in the knowledge base and answer in detail. When all knowledge base content is irrelevant to the question, your answer must include the sentence \"The answer you are looking for is not found in the knowledge base!\" Answers need to consider chat history.\n      Here is the knowledge base:\n      {retrieval:0@formalized_content}\n      The above is the knowledge base.",
+                      "sys_prompt": "You are an intelligent assistant. Please summarize the content of the dataset to answer the question. Please list the data in the knowledge base and answer in detail. When all knowledge base content is irrelevant to the question, your answer must include the sentence \"The answer you are looking for is not found in the knowledge base!\" Answers need to consider chat history.\n      Here is the knowledge base:\n      {retrieval:0@formalized_content}\n      The above is the knowledge base.",
                       "temperature": 0.2
                     }
                 },
@@ -85,7 +85,7 @@
   },
   "history": [],
   "path": [],
-  "retrival": {"chunks": [], "doc_aggs": []},
+  "retrieval": {"chunks": [], "doc_aggs": []},
   "globals": {
     "sys.query": "",
     "sys.user_id": "",

agent/test/dsl_examples/tavily_and_generate.json

@@ -25,7 +25,7 @@
                     "component_name": "LLM",
                     "params": {
                       "llm_id": "deepseek-chat",
-                      "sys_prompt": "You are an intelligent assistant. Please summarize the content of the knowledge base to answer the question. Please list the data in the knowledge base and answer in detail. When all knowledge base content is irrelevant to the question, your answer must include the sentence \"The answer you are looking for is not found in the knowledge base!\" Answers need to consider chat history.\n      Here is the knowledge base:\n      {tavily:0@formalized_content}\n      The above is the knowledge base.",
+                      "sys_prompt": "You are an intelligent assistant. Please summarize the content of the knowledge base to answer the question. Please list the data in the knowledge base and answer in detail. When all knowledge base content is irrelevant to the question, your answer must include the sentence \"The answer you are looking for is not found in the knowledge base!\" Answers need to consider chat history.\n      Here is the knowledge base:\n      {tavily:0@formalized_content}\n      Above is the knowledge base.",
                       "temperature": 0.2
                     }
                 },
@@ -45,7 +45,7 @@
   },
   "history": [],
   "path": [],
-  "retrival": {"chunks": [], "doc_aggs": []},
+  "retrieval": {"chunks": [], "doc_aggs": []},
   "globals": {
     "sys.query": "",
     "sys.user_id": "",

agent/tools/arxiv.py

@@ -63,12 +63,18 @@ class ArXiv(ToolBase, ABC):
 
     @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12)))
     def _invoke(self, **kwargs):
+        if self.check_if_canceled("ArXiv processing"):
+            return
+
         if not kwargs.get("query"):
             self.set_output("formalized_content", "")
             return ""
 
         last_e = ""
         for _ in range(self._param.max_retries+1):
+            if self.check_if_canceled("ArXiv processing"):
+                return
+
             try:
                 sort_choices = {"relevance": arxiv.SortCriterion.Relevance,
                                 "lastUpdatedDate": arxiv.SortCriterion.LastUpdatedDate,
@@ -79,12 +85,20 @@ class ArXiv(ToolBase, ABC):
                     max_results=self._param.top_n,
                     sort_by=sort_choices[self._param.sort_by]
                 )
-                self._retrieve_chunks(list(arxiv_client.results(search)),
+                results = list(arxiv_client.results(search))
+
+                if self.check_if_canceled("ArXiv processing"):
+                    return
+
+                self._retrieve_chunks(results,
                                       get_title=lambda r: r.title,
                                       get_url=lambda r: r.pdf_url,
                                       get_content=lambda r: r.summary)
                 return self.output("formalized_content")
             except Exception as e:
+                if self.check_if_canceled("ArXiv processing"):
+                    return
+
                 last_e = e
                 logging.exception(f"ArXiv error: {e}")
                 time.sleep(self._param.delay_after_error)

agent/tools/base.py

@@ -17,13 +17,13 @@ import logging
 import re
 import time
 from copy import deepcopy
+import asyncio
 from functools import partial
 from typing import TypedDict, List, Any
 from agent.component.base import ComponentParamBase, ComponentBase
 from common.misc_utils import hash_str2int
-from rag.llm.chat_model import ToolCallSession
 from rag.prompts.generator import kb_prompt
-from rag.utils.mcp_tool_call_conn import MCPToolCallSession
+from common.mcp_tool_call_conn import MCPToolCallSession, ToolCallSession
 from timeit import default_timer as timer
 
 
@@ -49,12 +49,19 @@ class LLMToolPluginCallSession(ToolCallSession):
         self.callback = callback
 
     def tool_call(self, name: str, arguments: dict[str, Any]) -> Any:
+        return asyncio.run(self.tool_call_async(name, arguments))
+
+    async def tool_call_async(self, name: str, arguments: dict[str, Any]) -> Any:
         assert name in self.tools_map, f"LLM tool {name} does not exist"
         st = timer()
-        if isinstance(self.tools_map[name], MCPToolCallSession):
-            resp = self.tools_map[name].tool_call(name, arguments, 60)
+        tool_obj = self.tools_map[name]
+        if isinstance(tool_obj, MCPToolCallSession):
+            resp = await asyncio.to_thread(tool_obj.tool_call, name, arguments, 60)
         else:
-            resp = self.tools_map[name].invoke(**arguments)
+            if hasattr(tool_obj, "invoke_async") and asyncio.iscoroutinefunction(tool_obj.invoke_async):
+                resp = await tool_obj.invoke_async(**arguments)
+            else:
+                resp = await asyncio.to_thread(tool_obj.invoke, **arguments)
 
         self.callback(name, arguments, resp, elapsed_time=timer()-st)
         return resp
@@ -125,6 +132,9 @@ class ToolBase(ComponentBase):
         return self._param.get_meta()
 
     def invoke(self, **kwargs):
+        if self.check_if_canceled("Tool processing"):
+            return
+
         self.set_output("_created_time", time.perf_counter())
         try:
             res = self._invoke(**kwargs)
@@ -137,6 +147,33 @@ class ToolBase(ComponentBase):
         self.set_output("_elapsed_time", time.perf_counter() - self.output("_created_time"))
         return res
 
+    async def invoke_async(self, **kwargs):
+        """
+        Async wrapper for tool invocation.
+        If `_invoke` is a coroutine, await it directly; otherwise run in a thread to avoid blocking.
+        Mirrors the exception handling of `invoke`.
+        """
+        if self.check_if_canceled("Tool processing"):
+            return
+
+        self.set_output("_created_time", time.perf_counter())
+        try:
+            fn_async = getattr(self, "_invoke_async", None)
+            if fn_async and asyncio.iscoroutinefunction(fn_async):
+                res = await fn_async(**kwargs)
+            elif asyncio.iscoroutinefunction(self._invoke):
+                res = await self._invoke(**kwargs)
+            else:
+                res = await asyncio.to_thread(self._invoke, **kwargs)
+        except Exception as e:
+            self._param.outputs["_ERROR"] = {"value": str(e)}
+            logging.exception(e)
+            res = str(e)
+        self._param.debug_inputs = []
+
+        self.set_output("_elapsed_time", time.perf_counter() - self.output("_created_time"))
+        return res
+
     def _retrieve_chunks(self, res_list: list, get_title, get_url, get_content, get_score=None):
         chunks = []
         aggs = []
@@ -170,4 +207,4 @@ class ToolBase(ComponentBase):
         self.set_output("formalized_content", "\n".join(kb_prompt({"chunks": chunks, "doc_aggs": aggs}, 200000, True)))
 
     def thoughts(self) -> str:
-        return self._canvas.get_component_name(self._id) + " is running..."
+        return self._canvas.get_component_name(self._id) + " is running..."

agent/tools/code_exec.py

@@ -13,15 +13,19 @@
 #  See the License for the specific language governing permissions and
 #  limitations under the License.
 #
+import ast
 import base64
+import json
 import logging
 import os
 from abc import ABC
-from strenum import StrEnum
 from typing import Optional
+
 from pydantic import BaseModel, Field, field_validator
-from agent.tools.base import ToolParamBase, ToolBase, ToolMeta
-from api import settings
+from strenum import StrEnum
+
+from agent.tools.base import ToolBase, ToolMeta, ToolParamBase
+from common import settings
 from common.connection_utils import timeout
 
 
@@ -62,10 +66,10 @@ class CodeExecParam(ToolParamBase):
     """
 
     def __init__(self):
-        self.meta:ToolMeta = {
+        self.meta: ToolMeta = {
             "name": "execute_code",
             "description": """
-This tool has a sandbox that can execute code written in 'Python'/'Javascript'. It recieves a piece of code and return a Json string.
+This tool has a sandbox that can execute code written in 'Python'/'Javascript'. It receives a piece of code and return a Json string.
 Here's a code example for Python(`main` function MUST be included):
 def main() -> dict:
     \"\"\"
@@ -99,16 +103,12 @@ module.exports = { main };
                     "enum": ["python", "javascript"],
                     "required": True,
                 },
-                "script": {
-                    "type": "string",
-                    "description": "A piece of code in right format. There MUST be main function.",
-                    "required": True
-                }
-            }
+                "script": {"type": "string", "description": "A piece of code in right format. There MUST be main function.", "required": True},
+            },
         }
         super().__init__()
         self.lang = Language.PYTHON.value
-        self.script = "def main(arg1: str, arg2: str) -> dict: return {\"result\": arg1 + arg2}"
+        self.script = 'def main(arg1: str, arg2: str) -> dict: return {"result": arg1 + arg2}'
         self.arguments = {}
         self.outputs = {"result": {"value": "", "type": "string"}}
 
@@ -119,18 +119,18 @@ module.exports = { main };
     def get_input_form(self) -> dict[str, dict]:
         res = {}
         for k, v in self.arguments.items():
-            res[k] = {
-                "type": "line",
-                "name": k
-            }
+            res[k] = {"type": "line", "name": k}
         return res
 
 
 class CodeExec(ToolBase, ABC):
     component_name = "CodeExec"
 
-    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60)))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10 * 60)))
     def _invoke(self, **kwargs):
+        if self.check_if_canceled("CodeExec processing"):
+            return
+
         lang = kwargs.get("lang", self._param.lang)
         script = kwargs.get("script", self._param.script)
         arguments = {}
@@ -140,24 +140,33 @@ class CodeExec(ToolBase, ABC):
                 continue
             arguments[k] = self._canvas.get_variable_value(v) if v else None
 
-        self._execute_code(
-            language=lang,
-            code=script,
-            arguments=arguments
-        )
+        self._execute_code(language=lang, code=script, arguments=arguments)
 
     def _execute_code(self, language: str, code: str, arguments: dict):
         import requests
 
+        if self.check_if_canceled("CodeExec execution"):
+            return
+
         try:
             code_b64 = self._encode_code(code)
             code_req = CodeExecutionRequest(code_b64=code_b64, language=language, arguments=arguments).model_dump()
         except Exception as e:
+            if self.check_if_canceled("CodeExec execution"):
+                return
+
             self.set_output("_ERROR", "construct code request error: " + str(e))
 
         try:
-            resp = requests.post(url=f"http://{settings.SANDBOX_HOST}:9385/run", json=code_req, timeout=int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60)))
+            if self.check_if_canceled("CodeExec execution"):
+                return "Task has been canceled"
+
+            resp = requests.post(url=f"http://{settings.SANDBOX_HOST}:9385/run", json=code_req, timeout=int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10 * 60)))
             logging.info(f"http://{settings.SANDBOX_HOST}:9385/run,  code_req: {code_req}, resp.status_code {resp.status_code}:")
+
+            if self.check_if_canceled("CodeExec execution"):
+                return "Task has been canceled"
+
             if resp.status_code != 200:
                 resp.raise_for_status()
             body = resp.json()
@@ -166,30 +175,17 @@ class CodeExec(ToolBase, ABC):
                 if stderr:
                     self.set_output("_ERROR", stderr)
                     return
-                try:
-                    rt = eval(body.get("stdout", ""))
-                except Exception:
-                    rt = body.get("stdout", "")
-                logging.info(f"http://{settings.SANDBOX_HOST}:9385/run -> {rt}")
-                if isinstance(rt, tuple):
-                    for i, (k, o) in enumerate(self._param.outputs.items()):
-                        if k.find("_") == 0:
-                            continue
-                        o["value"] = rt[i]
-                elif isinstance(rt, dict):
-                    for i, (k, o) in enumerate(self._param.outputs.items()):
-                        if k not in rt or k.find("_") == 0:
-                            continue
-                        o["value"] = rt[k]
-                else:
-                    for i, (k, o) in enumerate(self._param.outputs.items()):
-                        if k.find("_") == 0:
-                            continue
-                        o["value"] = rt
+                raw_stdout = body.get("stdout", "")
+                parsed_stdout = self._deserialize_stdout(raw_stdout)
+                logging.info(f"[CodeExec]: http://{settings.SANDBOX_HOST}:9385/run -> {parsed_stdout}")
+                self._populate_outputs(parsed_stdout, raw_stdout)
             else:
                 self.set_output("_ERROR", "There is no response from sandbox")
 
         except Exception as e:
+            if self.check_if_canceled("CodeExec execution"):
+                return
+
             self.set_output("_ERROR", "Exception executing code: " + str(e))
 
         return self.output()
@@ -199,3 +195,149 @@ class CodeExec(ToolBase, ABC):
 
     def thoughts(self) -> str:
         return "Running a short script to process data."
+
+    def _deserialize_stdout(self, stdout: str):
+        text = str(stdout).strip()
+        if not text:
+            return ""
+        for loader in (json.loads, ast.literal_eval):
+            try:
+                return loader(text)
+            except Exception:
+                continue
+        return text
+
+    def _coerce_output_value(self, value, expected_type: Optional[str]):
+        if expected_type is None:
+            return value
+
+        etype = expected_type.strip().lower()
+        inner_type = None
+        if etype.startswith("array<") and etype.endswith(">"):
+            inner_type = etype[6:-1].strip()
+            etype = "array"
+
+        try:
+            if etype == "string":
+                return "" if value is None else str(value)
+
+            if etype == "number":
+                if value is None or value == "":
+                    return None
+                if isinstance(value, (int, float)):
+                    return value
+                if isinstance(value, str):
+                    try:
+                        return float(value)
+                    except Exception:
+                        return value
+                return float(value)
+
+            if etype == "boolean":
+                if isinstance(value, bool):
+                    return value
+                if isinstance(value, str):
+                    lv = value.lower()
+                    if lv in ("true", "1", "yes", "y", "on"):
+                        return True
+                    if lv in ("false", "0", "no", "n", "off"):
+                        return False
+                return bool(value)
+
+            if etype == "array":
+                candidate = value
+                if isinstance(candidate, str):
+                    parsed = self._deserialize_stdout(candidate)
+                    candidate = parsed
+                if isinstance(candidate, tuple):
+                    candidate = list(candidate)
+                if not isinstance(candidate, list):
+                    candidate = [] if candidate is None else [candidate]
+
+                if inner_type == "string":
+                    return ["" if v is None else str(v) for v in candidate]
+                if inner_type == "number":
+                    coerced = []
+                    for v in candidate:
+                        try:
+                            if v is None or v == "":
+                                coerced.append(None)
+                            elif isinstance(v, (int, float)):
+                                coerced.append(v)
+                            else:
+                                coerced.append(float(v))
+                        except Exception:
+                            coerced.append(v)
+                    return coerced
+                return candidate
+
+            if etype == "object":
+                if isinstance(value, dict):
+                    return value
+                if isinstance(value, str):
+                    parsed = self._deserialize_stdout(value)
+                    if isinstance(parsed, dict):
+                        return parsed
+                return value
+        except Exception:
+            return value
+
+        return value
+
+    def _populate_outputs(self, parsed_stdout, raw_stdout: str):
+        outputs_items = list(self._param.outputs.items())
+        logging.info(f"[CodeExec]: outputs schema keys: {[k for k, _ in outputs_items]}")
+        if not outputs_items:
+            return
+
+        if isinstance(parsed_stdout, dict):
+            for key, meta in outputs_items:
+                if key.startswith("_"):
+                    continue
+                val = self._get_by_path(parsed_stdout, key)
+                coerced = self._coerce_output_value(val, meta.get("type"))
+                logging.info(f"[CodeExec]: populate dict key='{key}' raw='{val}' coerced='{coerced}'")
+                self.set_output(key, coerced)
+            return
+
+        if isinstance(parsed_stdout, (list, tuple)):
+            for idx, (key, meta) in enumerate(outputs_items):
+                if key.startswith("_"):
+                    continue
+                val = parsed_stdout[idx] if idx < len(parsed_stdout) else None
+                coerced = self._coerce_output_value(val, meta.get("type"))
+                logging.info(f"[CodeExec]: populate list key='{key}' raw='{val}' coerced='{coerced}'")
+                self.set_output(key, coerced)
+            return
+
+        default_val = parsed_stdout if parsed_stdout is not None else raw_stdout
+        for idx, (key, meta) in enumerate(outputs_items):
+            if key.startswith("_"):
+                continue
+            val = default_val if idx == 0 else None
+            coerced = self._coerce_output_value(val, meta.get("type"))
+            logging.info(f"[CodeExec]: populate scalar key='{key}' raw='{val}' coerced='{coerced}'")
+            self.set_output(key, coerced)
+
+    def _get_by_path(self, data, path: str):
+        if not path:
+            return None
+        cur = data
+        for part in path.split("."):
+            part = part.strip()
+            if not part:
+                return None
+            if isinstance(cur, dict):
+                cur = cur.get(part)
+            elif isinstance(cur, list):
+                try:
+                    idx = int(part)
+                    cur = cur[idx]
+                except Exception:
+                    return None
+            else:
+                return None
+            if cur is None:
+                return None
+        logging.info(f"[CodeExec]: resolve path '{path}' -> {cur}")
+        return cur

agent/tools/crawler.py

@@ -29,7 +29,7 @@ class CrawlerParam(ToolParamBase):
         super().__init__()
         self.proxy = None
         self.extract_type = "markdown"
-    
+
     def check(self):
         self.check_valid_value(self.extract_type, "Type of content from the crawler", ['html', 'markdown', 'content'])
 
@@ -47,18 +47,24 @@ class Crawler(ToolBase, ABC):
             result = asyncio.run(self.get_web(ans))
 
             return Crawler.be_output(result)
-            
+
         except Exception as e:
             return Crawler.be_output(f"An unexpected error occurred: {str(e)}")
 
     async def get_web(self, url):
+        if self.check_if_canceled("Crawler async operation"):
+            return
+
         proxy = self._param.proxy if self._param.proxy else None
         async with AsyncWebCrawler(verbose=True, proxy=proxy) as crawler:
             result = await crawler.arun(
                 url=url,
                 bypass_cache=True
             )
-            
+
+            if self.check_if_canceled("Crawler async operation"):
+                return
+
             if self._param.extract_type == 'html':
                 return result.cleaned_html
             elif self._param.extract_type == 'markdown':

agent/tools/deepl.py

@@ -46,11 +46,16 @@ class DeepL(ComponentBase, ABC):
     component_name = "DeepL"
 
     def _run(self, history, **kwargs):
+        if self.check_if_canceled("DeepL processing"):
+            return
         ans = self.get_input()
         ans = " - ".join(ans["content"]) if "content" in ans else ""
         if not ans:
             return DeepL.be_output("")
 
+        if self.check_if_canceled("DeepL processing"):
+            return
+
         try:
             translator = deepl.Translator(self._param.auth_key)
             result = translator.translate_text(ans, source_lang=self._param.source_lang,
@@ -58,4 +63,6 @@ class DeepL(ComponentBase, ABC):
 
             return DeepL.be_output(result.text)
         except Exception as e:
+            if self.check_if_canceled("DeepL processing"):
+                return
             DeepL.be_output("**Error**:" + str(e))

agent/tools/duckduckgo.py

@@ -75,17 +75,30 @@ class DuckDuckGo(ToolBase, ABC):
 
     @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12)))
     def _invoke(self, **kwargs):
+        if self.check_if_canceled("DuckDuckGo processing"):
+            return
+
         if not kwargs.get("query"):
             self.set_output("formalized_content", "")
             return ""
 
         last_e = ""
         for _ in range(self._param.max_retries+1):
+            if self.check_if_canceled("DuckDuckGo processing"):
+                return
+
             try:
                 if kwargs.get("topic", "general") == "general":
                     with DDGS() as ddgs:
+                        if self.check_if_canceled("DuckDuckGo processing"):
+                            return
+
                         # {'title': '', 'href': '', 'body': ''}
                         duck_res = ddgs.text(kwargs["query"], max_results=self._param.top_n)
+
+                        if self.check_if_canceled("DuckDuckGo processing"):
+                            return
+
                         self._retrieve_chunks(duck_res,
                                               get_title=lambda r: r["title"],
                                               get_url=lambda r: r.get("href", r.get("url")),
@@ -94,8 +107,15 @@ class DuckDuckGo(ToolBase, ABC):
                         return self.output("formalized_content")
                 else:
                     with DDGS() as ddgs:
+                        if self.check_if_canceled("DuckDuckGo processing"):
+                            return
+
                         # {'date': '', 'title': '', 'body': '', 'url': '', 'image': '', 'source': ''}
                         duck_res = ddgs.news(kwargs["query"], max_results=self._param.top_n)
+
+                        if self.check_if_canceled("DuckDuckGo processing"):
+                            return
+
                         self._retrieve_chunks(duck_res,
                                               get_title=lambda r: r["title"],
                                               get_url=lambda r: r.get("href", r.get("url")),
@@ -103,6 +123,9 @@ class DuckDuckGo(ToolBase, ABC):
                         self.set_output("json", duck_res)
                         return self.output("formalized_content")
             except Exception as e:
+                if self.check_if_canceled("DuckDuckGo processing"):
+                    return
+
                 last_e = e
                 logging.exception(f"DuckDuckGo error: {e}")
                 time.sleep(self._param.delay_after_error)

agent/tools/email.py

@@ -101,19 +101,27 @@ class Email(ToolBase, ABC):
 
     @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 60)))
     def _invoke(self, **kwargs):
+        if self.check_if_canceled("Email processing"):
+            return
+
         if not kwargs.get("to_email"):
             self.set_output("success", False)
             return ""
 
         last_e = ""
         for _ in range(self._param.max_retries+1):
+            if self.check_if_canceled("Email processing"):
+                return
+
             try:
                 # Parse JSON string passed from upstream
                 email_data = kwargs
 
                 # Validate required fields
                 if "to_email" not in email_data:
-                    return Email.be_output("Missing required field: to_email")
+                    self.set_output("_ERROR", "Missing required field: to_email")
+                    self.set_output("success", False)
+                    return False
 
                 # Create email object
                 msg = MIMEMultipart('alternative')
@@ -133,6 +141,9 @@ class Email(ToolBase, ABC):
                 # Connect to SMTP server and send
                 logging.info(f"Connecting to SMTP server {self._param.smtp_server}:{self._param.smtp_port}")
 
+                if self.check_if_canceled("Email processing"):
+                    return
+
                 context = smtplib.ssl.create_default_context()
                 with smtplib.SMTP(self._param.smtp_server, self._param.smtp_port) as server:
                     server.ehlo()
@@ -149,6 +160,10 @@ class Email(ToolBase, ABC):
 
                     # Send email
                     logging.info(f"Sending email to recipients: {recipients}")
+
+                    if self.check_if_canceled("Email processing"):
+                        return
+
                     try:
                         server.send_message(msg, self._param.email, recipients)
                         success = True

agent/tools/exesql.py

@@ -81,6 +81,8 @@ class ExeSQL(ToolBase, ABC):
 
     @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 60)))
     def _invoke(self, **kwargs):
+        if self.check_if_canceled("ExeSQL processing"):
+            return
 
         def convert_decimals(obj):
             from decimal import Decimal
@@ -96,6 +98,9 @@ class ExeSQL(ToolBase, ABC):
         if not sql:
             raise Exception("SQL for `ExeSQL` MUST not be empty.")
 
+        if self.check_if_canceled("ExeSQL processing"):
+            return
+
         vars = self.get_input_elements_from_text(sql)
         args = {}
         for k, o in vars.items():
@@ -108,6 +113,9 @@ class ExeSQL(ToolBase, ABC):
             self.set_input_value(k, args[k])
         sql = self.string_format(sql, args)
 
+        if self.check_if_canceled("ExeSQL processing"):
+            return
+
         sqls = sql.split(";")
         if self._param.db_type in ["mysql", "mariadb"]:
             db = pymysql.connect(db=self._param.database, user=self._param.username, host=self._param.host,
@@ -181,6 +189,10 @@ class ExeSQL(ToolBase, ABC):
             sql_res = []
             formalized_content = []
             for single_sql in sqls:
+                if self.check_if_canceled("ExeSQL processing"):
+                    ibm_db.close(conn)
+                    return
+
                 single_sql = single_sql.replace("```", "").strip()
                 if not single_sql:
                     continue
@@ -190,6 +202,9 @@ class ExeSQL(ToolBase, ABC):
                 rows = []
                 row = ibm_db.fetch_assoc(stmt)
                 while row and len(rows) < self._param.max_records:
+                    if self.check_if_canceled("ExeSQL processing"):
+                        ibm_db.close(conn)
+                        return
                     rows.append(row)
                     row = ibm_db.fetch_assoc(stmt)
 
@@ -220,6 +235,11 @@ class ExeSQL(ToolBase, ABC):
         sql_res = []
         formalized_content = []
         for single_sql in sqls:
+            if self.check_if_canceled("ExeSQL processing"):
+                cursor.close()
+                db.close()
+                return
+
             single_sql = single_sql.replace('```','')
             if not single_sql:
                 continue
@@ -244,6 +264,9 @@ class ExeSQL(ToolBase, ABC):
             sql_res.append(convert_decimals(single_res.to_dict(orient='records')))
             formalized_content.append(single_res.to_markdown(index=False, floatfmt=".6f"))
 
+        cursor.close()
+        db.close()
+
         self.set_output("json", sql_res)
         self.set_output("formalized_content", "\n\n".join(formalized_content))
         return self.output("formalized_content")

agent/tools/github.py

@@ -59,17 +59,27 @@ class GitHub(ToolBase, ABC):
 
     @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12)))
     def _invoke(self, **kwargs):
+        if self.check_if_canceled("GitHub processing"):
+            return
+
         if not kwargs.get("query"):
             self.set_output("formalized_content", "")
             return ""
 
         last_e = ""
         for _ in range(self._param.max_retries+1):
+            if self.check_if_canceled("GitHub processing"):
+                return
+
             try:
                 url = 'https://api.github.com/search/repositories?q=' + kwargs["query"] + '&sort=stars&order=desc&per_page=' + str(
                     self._param.top_n)
                 headers = {"Content-Type": "application/vnd.github+json", "X-GitHub-Api-Version": '2022-11-28'}
                 response = requests.get(url=url, headers=headers).json()
+
+                if self.check_if_canceled("GitHub processing"):
+                    return
+
                 self._retrieve_chunks(response['items'],
                                       get_title=lambda r: r["name"],
                                       get_url=lambda r: r["html_url"],
@@ -77,6 +87,9 @@ class GitHub(ToolBase, ABC):
                 self.set_output("json", response['items'])
                 return self.output("formalized_content")
             except Exception as e:
+                if self.check_if_canceled("GitHub processing"):
+                    return
+
                 last_e = e
                 logging.exception(f"GitHub error: {e}")
                 time.sleep(self._param.delay_after_error)

agent/tools/google.py

@@ -118,6 +118,9 @@ class Google(ToolBase, ABC):
 
     @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12)))
     def _invoke(self, **kwargs):
+        if self.check_if_canceled("Google processing"):
+            return
+
         if not kwargs.get("q"):
             self.set_output("formalized_content", "")
             return ""
@@ -132,8 +135,15 @@ class Google(ToolBase, ABC):
         }
         last_e = ""
         for _ in range(self._param.max_retries+1):
+            if self.check_if_canceled("Google processing"):
+                return
+
             try:
                 search = GoogleSearch(params).get_dict()
+
+                if self.check_if_canceled("Google processing"):
+                    return
+
                 self._retrieve_chunks(search["organic_results"],
                                       get_title=lambda r: r["title"],
                                       get_url=lambda r: r["link"],
@@ -142,6 +152,9 @@ class Google(ToolBase, ABC):
                 self.set_output("json", search["organic_results"])
                 return self.output("formalized_content")
             except Exception as e:
+                if self.check_if_canceled("Google processing"):
+                    return
+
                 last_e = e
                 logging.exception(f"Google error: {e}")
                 time.sleep(self._param.delay_after_error)

agent/tools/googlescholar.py

@@ -65,15 +65,25 @@ class GoogleScholar(ToolBase, ABC):
 
     @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12)))
     def _invoke(self, **kwargs):
+        if self.check_if_canceled("GoogleScholar processing"):
+            return
+
         if not kwargs.get("query"):
             self.set_output("formalized_content", "")
             return ""
 
         last_e = ""
         for _ in range(self._param.max_retries+1):
+            if self.check_if_canceled("GoogleScholar processing"):
+                return
+
             try:
                 scholar_client = scholarly.search_pubs(kwargs["query"], patents=self._param.patents, year_low=self._param.year_low,
                                                        year_high=self._param.year_high, sort_by=self._param.sort_by)
+
+                if self.check_if_canceled("GoogleScholar processing"):
+                    return
+
                 self._retrieve_chunks(scholar_client,
                                       get_title=lambda r: r['bib']['title'],
                                       get_url=lambda r: r["pub_url"],
@@ -82,6 +92,9 @@ class GoogleScholar(ToolBase, ABC):
                 self.set_output("json", list(scholar_client))
                 return self.output("formalized_content")
             except Exception as e:
+                if self.check_if_canceled("GoogleScholar processing"):
+                    return
+
                 last_e = e
                 logging.exception(f"GoogleScholar error: {e}")
                 time.sleep(self._param.delay_after_error)

agent/tools/jin10.py

@@ -50,6 +50,9 @@ class Jin10(ComponentBase, ABC):
     component_name = "Jin10"
 
     def _run(self, history, **kwargs):
+        if self.check_if_canceled("Jin10 processing"):
+            return
+
         ans = self.get_input()
         ans = " - ".join(ans["content"]) if "content" in ans else ""
         if not ans:
@@ -58,6 +61,9 @@ class Jin10(ComponentBase, ABC):
         jin10_res = []
         headers = {'secret-key': self._param.secret_key}
         try:
+            if self.check_if_canceled("Jin10 processing"):
+                return
+
             if self._param.type == "flash":
                 params = {
                     'category': self._param.flash_type,
@@ -69,6 +75,8 @@ class Jin10(ComponentBase, ABC):
                     headers=headers, data=json.dumps(params))
                 response = response.json()
                 for i in response['data']:
+                    if self.check_if_canceled("Jin10 processing"):
+                        return
                     jin10_res.append({"content": i['data']['content']})
             if self._param.type == "calendar":
                 params = {
@@ -79,6 +87,8 @@ class Jin10(ComponentBase, ABC):
                     headers=headers, data=json.dumps(params))
 
                 response = response.json()
+                if self.check_if_canceled("Jin10 processing"):
+                    return
                 jin10_res.append({"content": pd.DataFrame(response['data']).to_markdown()})
             if self._param.type == "symbols":
                 params = {
@@ -90,8 +100,12 @@ class Jin10(ComponentBase, ABC):
                     url='https://open-data-api.jin10.com/data-api/' + self._param.symbols_datatype + '?type=' + self._param.symbols_type,
                     headers=headers, data=json.dumps(params))
                 response = response.json()
+                if self.check_if_canceled("Jin10 processing"):
+                    return
                 if self._param.symbols_datatype == "symbols":
                     for i in response['data']:
+                        if self.check_if_canceled("Jin10 processing"):
+                            return
                         i['Commodity Code'] = i['c']
                         i['Stock Exchange'] = i['e']
                         i['Commodity Name'] = i['n']
@@ -99,6 +113,8 @@ class Jin10(ComponentBase, ABC):
                         del i['c'], i['e'], i['n'], i['t']
                 if self._param.symbols_datatype == "quotes":
                     for i in response['data']:
+                        if self.check_if_canceled("Jin10 processing"):
+                            return
                         i['Selling Price'] = i['a']
                         i['Buying Price'] = i['b']
                         i['Commodity Code'] = i['c']
@@ -120,8 +136,12 @@ class Jin10(ComponentBase, ABC):
                     url='https://open-data-api.jin10.com/data-api/news',
                     headers=headers, data=json.dumps(params))
                 response = response.json()
+                if self.check_if_canceled("Jin10 processing"):
+                    return
                 jin10_res.append({"content": pd.DataFrame(response['data']).to_markdown()})
         except Exception as e:
+            if self.check_if_canceled("Jin10 processing"):
+                return
             return Jin10.be_output("**ERROR**: " + str(e))
 
         if not jin10_res:

agent/tools/pubmed.py

@@ -71,23 +71,40 @@ class PubMed(ToolBase, ABC):
 
     @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12)))
     def _invoke(self, **kwargs):
+        if self.check_if_canceled("PubMed processing"):
+            return
+
         if not kwargs.get("query"):
             self.set_output("formalized_content", "")
             return ""
 
         last_e = ""
         for _ in range(self._param.max_retries+1):
+            if self.check_if_canceled("PubMed processing"):
+                return
+
             try:
                 Entrez.email = self._param.email
                 pubmedids = Entrez.read(Entrez.esearch(db='pubmed', retmax=self._param.top_n, term=kwargs["query"]))['IdList']
+
+                if self.check_if_canceled("PubMed processing"):
+                    return
+
                 pubmedcnt = ET.fromstring(re.sub(r'<(/?)b>|<(/?)i>', '', Entrez.efetch(db='pubmed', id=",".join(pubmedids),
                                                                                        retmode="xml").read().decode("utf-8")))
+
+                if self.check_if_canceled("PubMed processing"):
+                    return
+
                 self._retrieve_chunks(pubmedcnt.findall("PubmedArticle"),
                                       get_title=lambda child: child.find("MedlineCitation").find("Article").find("ArticleTitle").text,
                                       get_url=lambda child: "https://pubmed.ncbi.nlm.nih.gov/" + child.find("MedlineCitation").find("PMID").text,
                                       get_content=lambda child: self._format_pubmed_content(child),)
                 return self.output("formalized_content")
             except Exception as e:
+                if self.check_if_canceled("PubMed processing"):
+                    return
+
                 last_e = e
                 logging.exception(f"PubMed error: {e}")
                 time.sleep(self._param.delay_after_error)

agent/tools/qweather.py

@@ -58,12 +58,18 @@ class QWeather(ComponentBase, ABC):
     component_name = "QWeather"
 
     def _run(self, history, **kwargs):
+        if self.check_if_canceled("Qweather processing"):
+            return
+
         ans = self.get_input()
         ans = "".join(ans["content"]) if "content" in ans else ""
         if not ans:
             return QWeather.be_output("")
 
         try:
+            if self.check_if_canceled("Qweather processing"):
+                return
+
             response = requests.get(
                 url="https://geoapi.qweather.com/v2/city/lookup?location=" + ans + "&key=" + self._param.web_apikey).json()
             if response["code"] == "200":
@@ -71,16 +77,23 @@ class QWeather(ComponentBase, ABC):
             else:
                 return QWeather.be_output("**Error**" + self._param.error_code[response["code"]])
 
+            if self.check_if_canceled("Qweather processing"):
+                return
+
             base_url = "https://api.qweather.com/v7/" if self._param.user_type == 'paid' else "https://devapi.qweather.com/v7/"
 
             if self._param.type == "weather":
                 url = base_url + "weather/" + self._param.time_period + "?location=" + location_id + "&key=" + self._param.web_apikey + "&lang=" + self._param.lang
                 response = requests.get(url=url).json()
+                if self.check_if_canceled("Qweather processing"):
+                    return
                 if response["code"] == "200":
                     if self._param.time_period == "now":
                         return QWeather.be_output(str(response["now"]))
                     else:
                         qweather_res = [{"content": str(i) + "\n"} for i in response["daily"]]
+                        if self.check_if_canceled("Qweather processing"):
+                            return
                         if not qweather_res:
                             return QWeather.be_output("")
 
@@ -92,6 +105,8 @@ class QWeather(ComponentBase, ABC):
             elif self._param.type == "indices":
                 url = base_url + "indices/1d?type=0&location=" + location_id + "&key=" + self._param.web_apikey + "&lang=" + self._param.lang
                 response = requests.get(url=url).json()
+                if self.check_if_canceled("Qweather processing"):
+                    return
                 if response["code"] == "200":
                     indices_res = response["daily"][0]["date"] + "\n" + "\n".join(
                         [i["name"] + ": " + i["category"] + ", " + i["text"] for i in response["daily"]])
@@ -103,9 +118,13 @@ class QWeather(ComponentBase, ABC):
             elif self._param.type == "airquality":
                 url = base_url + "air/now?location=" + location_id + "&key=" + self._param.web_apikey + "&lang=" + self._param.lang
                 response = requests.get(url=url).json()
+                if self.check_if_canceled("Qweather processing"):
+                    return
                 if response["code"] == "200":
                     return QWeather.be_output(str(response["now"]))
                 else:
                     return QWeather.be_output("**Error**" + self._param.error_code[response["code"]])
         except Exception as e:
+            if self.check_if_canceled("Qweather processing"):
+                return
             return QWeather.be_output("**Error**" + str(e))

agent/tools/retrieval.py

@@ -13,6 +13,7 @@
 #  See the License for the specific language governing permissions and
 #  limitations under the License.
 #
+import asyncio
 from functools import partial
 import json
 import os
@@ -21,13 +22,15 @@ from abc import ABC
 from agent.tools.base import ToolParamBase, ToolBase, ToolMeta
 from common.constants import LLMType
 from api.db.services.document_service import DocumentService
-from api.db.services.dialog_service import meta_filter
+from common.metadata_utils import apply_meta_data_filter
 from api.db.services.knowledgebase_service import KnowledgebaseService
 from api.db.services.llm_service import LLMBundle
-from api import settings
+from api.db.services.memory_service import MemoryService
+from api.db.joint_services import memory_message_service
+from common import settings
 from common.connection_utils import timeout
 from rag.app.tag import label_question
-from rag.prompts.generator import cross_languages, kb_prompt, gen_meta_filter
+from rag.prompts.generator import cross_languages, kb_prompt, memory_prompt
 
 
 class RetrievalParam(ToolParamBase):
@@ -56,6 +59,7 @@ class RetrievalParam(ToolParamBase):
         self.top_n = 8
         self.top_k = 1024
         self.kb_ids = []
+        self.memory_ids = []
         self.kb_vars = []
         self.rerank_id = ""
         self.empty_response = ""
@@ -80,11 +84,7 @@ class RetrievalParam(ToolParamBase):
 class Retrieval(ToolBase, ABC):
     component_name = "Retrieval"
 
-    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12)))
-    def _invoke(self, **kwargs):
-        if not kwargs.get("query"):
-            self.set_output("formalized_content", self._param.empty_response)
-
+    async def _retrieve_kb(self, query_text: str):
         kb_ids: list[str] = []
         for id in self._param.kb_ids:
             if id.find("@") < 0:
@@ -119,54 +119,58 @@ class Retrieval(ToolBase, ABC):
         if self._param.rerank_id:
             rerank_mdl = LLMBundle(kbs[0].tenant_id, LLMType.RERANK, self._param.rerank_id)
 
-        vars = self.get_input_elements_from_text(kwargs["query"])
-        vars = {k:o["value"] for k,o in vars.items()}
-        query = self.string_format(kwargs["query"], vars)
-        
-        doc_ids=[]
-        if self._param.meta_data_filter!={}:
+        vars = self.get_input_elements_from_text(query_text)
+        vars = {k: o["value"] for k, o in vars.items()}
+        query = self.string_format(query_text, vars)
+
+        doc_ids = []
+        if self._param.meta_data_filter != {}:
             metas = DocumentService.get_meta_by_kbs(kb_ids)
-            if self._param.meta_data_filter.get("method") == "auto":
+
+            def _resolve_manual_filter(flt: dict) -> dict:
+                pat = re.compile(self.variable_ref_patt)
+                s = flt.get("value", "")
+                out_parts = []
+                last = 0
+
+                for m in pat.finditer(s):
+                    out_parts.append(s[last:m.start()])
+                    key = m.group(1)
+                    v = self._canvas.get_variable_value(key)
+                    if v is None:
+                        rep = ""
+                    elif isinstance(v, partial):
+                        buf = []
+                        for chunk in v():
+                            buf.append(chunk)
+                        rep = "".join(buf)
+                    elif isinstance(v, str):
+                        rep = v
+                    else:
+                        rep = json.dumps(v, ensure_ascii=False)
+
+                    out_parts.append(rep)
+                    last = m.end()
+
+                out_parts.append(s[last:])
+                flt["value"] = "".join(out_parts)
+                return flt
+
+            chat_mdl = None
+            if self._param.meta_data_filter.get("method") in ["auto", "semi_auto"]:
                 chat_mdl = LLMBundle(self._canvas.get_tenant_id(), LLMType.CHAT)
-                filters = gen_meta_filter(chat_mdl, metas, query)
-                doc_ids.extend(meta_filter(metas, filters))
-                if not doc_ids:
-                    doc_ids = None
-            elif self._param.meta_data_filter.get("method") == "manual":
-                filters=self._param.meta_data_filter["manual"]
-                for flt in filters:
-                    pat = re.compile(r"\{* *\{([a-zA-Z:0-9]+@[A-Za-z:0-9_.-]+|sys\.[a-z_]+)\} *\}*")
-                    s = flt["value"]
-                    out_parts = []
-                    last = 0
 
-                    for m in pat.finditer(s):
-                        out_parts.append(s[last:m.start()])
-                        key = m.group(1)
-                        v = self._canvas.get_variable_value(key)
-                        if v is None:
-                            rep = ""
-                        elif isinstance(v, partial):
-                            buf = []
-                            for chunk in v():
-                                buf.append(chunk)
-                            rep = "".join(buf)
-                        elif isinstance(v, str):
-                            rep = v
-                        else:
-                            rep = json.dumps(v, ensure_ascii=False)
-
-                        out_parts.append(rep)
-                        last = m.end()
-
-                    out_parts.append(s[last:])
-                    flt["value"] = "".join(out_parts)
-                doc_ids.extend(meta_filter(metas, filters))
-                if not doc_ids:
-                    doc_ids = None
+            doc_ids = await apply_meta_data_filter(
+                self._param.meta_data_filter,
+                metas,
+                query,
+                chat_mdl,
+                doc_ids,
+                _resolve_manual_filter if self._param.meta_data_filter.get("method") == "manual" else None,
+            )
 
         if self._param.cross_languages:
-            query = cross_languages(kbs[0].tenant_id, None, query, self._param.cross_languages)
+            query = await cross_languages(kbs[0].tenant_id, None, query, self._param.cross_languages)
 
         if kbs:
             query = re.sub(r"^user[:：\s]*", "", query, flags=re.IGNORECASE)
@@ -184,24 +188,37 @@ class Retrieval(ToolBase, ABC):
                 rerank_mdl=rerank_mdl,
                 rank_feature=label_question(query, kbs),
             )
+            if self.check_if_canceled("Retrieval processing"):
+                return
+
             if self._param.toc_enhance:
                 chat_mdl = LLMBundle(self._canvas._tenant_id, LLMType.CHAT)
-                cks = settings.retriever.retrieval_by_toc(query, kbinfos["chunks"], [kb.tenant_id for kb in kbs], chat_mdl, self._param.top_n)
+                cks = settings.retriever.retrieval_by_toc(query, kbinfos["chunks"], [kb.tenant_id for kb in kbs],
+                                                          chat_mdl, self._param.top_n)
+                if self.check_if_canceled("Retrieval processing"):
+                    return
                 if cks:
                     kbinfos["chunks"] = cks
+            kbinfos["chunks"] = settings.retriever.retrieval_by_children(kbinfos["chunks"],
+                                                                         [kb.tenant_id for kb in kbs])
             if self._param.use_kg:
                 ck = settings.kg_retriever.retrieval(query,
-                                                       [kb.tenant_id for kb in kbs],
-                                                       kb_ids,
-                                                       embd_mdl,
-                                                       LLMBundle(self._canvas.get_tenant_id(), LLMType.CHAT))
+                                                     [kb.tenant_id for kb in kbs],
+                                                     kb_ids,
+                                                     embd_mdl,
+                                                     LLMBundle(self._canvas.get_tenant_id(), LLMType.CHAT))
+                if self.check_if_canceled("Retrieval processing"):
+                    return
                 if ck["content_with_weight"]:
                     kbinfos["chunks"].insert(0, ck)
         else:
             kbinfos = {"chunks": [], "doc_aggs": []}
 
         if self._param.use_kg and kbs:
-            ck = settings.kg_retriever.retrieval(query, [kb.tenant_id for kb in kbs], filtered_kb_ids, embd_mdl, LLMBundle(kbs[0].tenant_id, LLMType.CHAT))
+            ck = settings.kg_retriever.retrieval(query, [kb.tenant_id for kb in kbs], filtered_kb_ids, embd_mdl,
+                                                 LLMBundle(kbs[0].tenant_id, LLMType.CHAT))
+            if self.check_if_canceled("Retrieval processing"):
+                return
             if ck["content_with_weight"]:
                 ck["content"] = ck["content_with_weight"]
                 del ck["content_with_weight"]
@@ -229,6 +246,58 @@ class Retrieval(ToolBase, ABC):
 
         return form_cnt
 
+    async def _retrieve_memory(self, query_text: str):
+        memory_ids: list[str] = [memory_id for memory_id in self._param.memory_ids]
+        memory_list = MemoryService.get_by_ids(memory_ids)
+        if not memory_list:
+            raise Exception("No memory is selected.")
+
+        embd_names = list({memory.embd_id for memory in memory_list})
+        assert len(embd_names) == 1, "Memory use different embedding models."
+
+        vars = self.get_input_elements_from_text(query_text)
+        vars = {k: o["value"] for k, o in vars.items()}
+        query = self.string_format(query_text, vars)
+        # query message
+        message_list = memory_message_service.query_message({"memory_id": memory_ids}, {
+            "query": query,
+            "similarity_threshold": self._param.similarity_threshold,
+            "keywords_similarity_weight": self._param.keywords_similarity_weight,
+            "top_n": self._param.top_n
+        })
+        if not message_list:
+            self.set_output("formalized_content", self._param.empty_response)
+            return ""
+        formated_content = "\n".join(memory_prompt(message_list, 200000))
+        # set formalized_content output
+        self.set_output("formalized_content", formated_content)
+
+        return formated_content
+
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12)))
+    async def _invoke_async(self, **kwargs):
+        if self.check_if_canceled("Retrieval processing"):
+            return
+        if not kwargs.get("query"):
+            self.set_output("formalized_content", self._param.empty_response)
+            return
+
+        if hasattr(self._param, "retrieval_from") and self._param.retrieval_from == "dataset":
+            return await self._retrieve_kb(kwargs["query"])
+        elif hasattr(self._param, "retrieval_from") and self._param.retrieval_from == "memory":
+            return await self._retrieve_memory(kwargs["query"])
+        elif self._param.kb_ids:
+            return await self._retrieve_kb(kwargs["query"])
+        elif hasattr(self._param, "memory_ids") and self._param.memory_ids:
+            return await self._retrieve_memory(kwargs["query"])
+        else:
+            self.set_output("formalized_content", self._param.empty_response)
+            return
+
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12)))
+    def _invoke(self, **kwargs):
+        return asyncio.run(self._invoke_async(**kwargs))
+
     def thoughts(self) -> str:
         return """
 Keywords: {}

agent/tools/searxng.py

@@ -79,6 +79,9 @@ class SearXNG(ToolBase, ABC):
 
     @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12)))
     def _invoke(self, **kwargs):
+        if self.check_if_canceled("SearXNG processing"):
+            return
+
         # Gracefully handle try-run without inputs
         query = kwargs.get("query")
         if not query or not isinstance(query, str) or not query.strip():
@@ -93,6 +96,9 @@ class SearXNG(ToolBase, ABC):
 
         last_e = ""
         for _ in range(self._param.max_retries+1):
+            if self.check_if_canceled("SearXNG processing"):
+                return
+
             try:
                 search_params = {
                     'q': query,
@@ -110,6 +116,9 @@ class SearXNG(ToolBase, ABC):
                 )
                 response.raise_for_status()
 
+                if self.check_if_canceled("SearXNG processing"):
+                    return
+
                 data = response.json()
 
                 if not data or not isinstance(data, dict):
@@ -121,6 +130,9 @@ class SearXNG(ToolBase, ABC):
 
                 results = results[:self._param.top_n]
 
+                if self.check_if_canceled("SearXNG processing"):
+                    return
+
                 self._retrieve_chunks(results,
                                       get_title=lambda r: r.get("title", ""),
                                       get_url=lambda r: r.get("url", ""),
@@ -130,10 +142,16 @@ class SearXNG(ToolBase, ABC):
                 return self.output("formalized_content")
 
             except requests.RequestException as e:
+                if self.check_if_canceled("SearXNG processing"):
+                    return
+
                 last_e = f"Network error: {e}"
                 logging.exception(f"SearXNG network error: {e}")
                 time.sleep(self._param.delay_after_error)
             except Exception as e:
+                if self.check_if_canceled("SearXNG processing"):
+                    return
+
                 last_e = str(e)
                 logging.exception(f"SearXNG error: {e}")
                 time.sleep(self._param.delay_after_error)

agent/tools/tavily.py

@@ -103,6 +103,9 @@ class TavilySearch(ToolBase, ABC):
 
     @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12)))
     def _invoke(self, **kwargs):
+        if self.check_if_canceled("TavilySearch processing"):
+            return
+
         if not kwargs.get("query"):
             self.set_output("formalized_content", "")
             return ""
@@ -113,10 +116,16 @@ class TavilySearch(ToolBase, ABC):
             if fld not in kwargs:
                 kwargs[fld] = getattr(self._param, fld)
         for _ in range(self._param.max_retries+1):
+            if self.check_if_canceled("TavilySearch processing"):
+                return
+
             try:
                 kwargs["include_images"] = False
                 kwargs["include_raw_content"] = False
                 res = self.tavily_client.search(**kwargs)
+                if self.check_if_canceled("TavilySearch processing"):
+                    return
+
                 self._retrieve_chunks(res["results"],
                                       get_title=lambda r: r["title"],
                                       get_url=lambda r: r["url"],
@@ -125,6 +134,9 @@ class TavilySearch(ToolBase, ABC):
                 self.set_output("json", res["results"])
                 return self.output("formalized_content")
             except Exception as e:
+                if self.check_if_canceled("TavilySearch processing"):
+                    return
+
                 last_e = e
                 logging.exception(f"Tavily error: {e}")
                 time.sleep(self._param.delay_after_error)
@@ -201,6 +213,9 @@ class TavilyExtract(ToolBase, ABC):
 
     @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60)))
     def _invoke(self, **kwargs):
+        if self.check_if_canceled("TavilyExtract processing"):
+            return
+
         self.tavily_client = TavilyClient(api_key=self._param.api_key)
         last_e = None
         for fld in ["urls", "extract_depth", "format"]:
@@ -209,12 +224,21 @@ class TavilyExtract(ToolBase, ABC):
         if kwargs.get("urls") and isinstance(kwargs["urls"], str):
             kwargs["urls"] = kwargs["urls"].split(",")
         for _ in range(self._param.max_retries+1):
+            if self.check_if_canceled("TavilyExtract processing"):
+                return
+
             try:
                 kwargs["include_images"] = False
                 res = self.tavily_client.extract(**kwargs)
+                if self.check_if_canceled("TavilyExtract processing"):
+                    return
+
                 self.set_output("json", res["results"])
                 return self.output("json")
             except Exception as e:
+                if self.check_if_canceled("TavilyExtract processing"):
+                    return
+
                 last_e = e
                 logging.exception(f"Tavily error: {e}")
         if last_e:

agent/tools/tushare.py

@@ -43,12 +43,18 @@ class TuShare(ComponentBase, ABC):
     component_name = "TuShare"
 
     def _run(self, history, **kwargs):
+        if self.check_if_canceled("TuShare processing"):
+            return
+
         ans = self.get_input()
         ans = ",".join(ans["content"]) if "content" in ans else ""
         if not ans:
             return TuShare.be_output("")
 
         try:
+            if self.check_if_canceled("TuShare processing"):
+                return
+
             tus_res = []
             params = {
                 "api_name": "news",
@@ -58,12 +64,18 @@ class TuShare(ComponentBase, ABC):
             }
             response = requests.post(url="http://api.tushare.pro", data=json.dumps(params).encode('utf-8'))
             response = response.json()
+            if self.check_if_canceled("TuShare processing"):
+                return
             if response['code'] != 0:
                 return TuShare.be_output(response['msg'])
             df = pd.DataFrame(response['data']['items'])
             df.columns = response['data']['fields']
+            if self.check_if_canceled("TuShare processing"):
+                return
             tus_res.append({"content": (df[df['content'].str.contains(self._param.keyword, case=False)]).to_markdown()})
         except Exception as e:
+            if self.check_if_canceled("TuShare processing"):
+                return
             return TuShare.be_output("**ERROR**: " + str(e))
 
         if not tus_res:

agent/tools/wencai.py

@@ -70,19 +70,31 @@ class WenCai(ToolBase, ABC):
 
     @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12)))
     def _invoke(self, **kwargs):
+        if self.check_if_canceled("WenCai processing"):
+            return
+
         if not kwargs.get("query"):
             self.set_output("report", "")
             return ""
 
         last_e = ""
         for _ in range(self._param.max_retries+1):
+            if self.check_if_canceled("WenCai processing"):
+                return
+
             try:
                 wencai_res = []
                 res = pywencai.get(query=kwargs["query"], query_type=self._param.query_type, perpage=self._param.top_n)
+                if self.check_if_canceled("WenCai processing"):
+                    return
+
                 if isinstance(res, pd.DataFrame):
                     wencai_res.append(res.to_markdown())
                 elif isinstance(res, dict):
                     for item in res.items():
+                        if self.check_if_canceled("WenCai processing"):
+                            return
+
                         if isinstance(item[1], list):
                             wencai_res.append(item[0] + "\n" + pd.DataFrame(item[1]).to_markdown())
                         elif isinstance(item[1], str):
@@ -100,6 +112,9 @@ class WenCai(ToolBase, ABC):
                 self.set_output("report", "\n\n".join(wencai_res))
                 return self.output("report")
             except Exception as e:
+                if self.check_if_canceled("WenCai processing"):
+                    return
+
                 last_e = e
                 logging.exception(f"WenCai error: {e}")
                 time.sleep(self._param.delay_after_error)

agent/tools/wikipedia.py

@@ -66,17 +66,26 @@ class Wikipedia(ToolBase, ABC):
 
     @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 60)))
     def _invoke(self, **kwargs):
+        if self.check_if_canceled("Wikipedia processing"):
+            return
+
         if not kwargs.get("query"):
             self.set_output("formalized_content", "")
             return ""
 
         last_e = ""
         for _ in range(self._param.max_retries+1):
+            if self.check_if_canceled("Wikipedia processing"):
+                return
+
             try:
                 wikipedia.set_lang(self._param.language)
                 wiki_engine = wikipedia
                 pages = []
                 for p in wiki_engine.search(kwargs["query"], results=self._param.top_n):
+                    if self.check_if_canceled("Wikipedia processing"):
+                        return
+
                     try:
                         pages.append(wikipedia.page(p))
                     except Exception:
@@ -87,6 +96,9 @@ class Wikipedia(ToolBase, ABC):
                                       get_content=lambda r: r.summary)
                 return self.output("formalized_content")
             except Exception as e:
+                if self.check_if_canceled("Wikipedia processing"):
+                    return
+
                 last_e = e
                 logging.exception(f"Wikipedia error: {e}")
                 time.sleep(self._param.delay_after_error)

agent/tools/yahoofinance.py

@@ -74,32 +74,44 @@ class YahooFinance(ToolBase, ABC):
 
     @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 60)))
     def _invoke(self, **kwargs):
+        if self.check_if_canceled("YahooFinance processing"):
+            return None
+
         if not kwargs.get("stock_code"):
             self.set_output("report", "")
             return ""
 
         last_e = ""
         for _ in range(self._param.max_retries+1):
-            yohoo_res = []
+            if self.check_if_canceled("YahooFinance processing"):
+                return None
+
+            yahoo_res = []
             try:
                 msft = yf.Ticker(kwargs["stock_code"])
+                if self.check_if_canceled("YahooFinance processing"):
+                    return None
+
                 if self._param.info:
-                    yohoo_res.append("# Information:\n" + pd.Series(msft.info).to_markdown() + "\n")
+                    yahoo_res.append("# Information:\n" + pd.Series(msft.info).to_markdown() + "\n")
                 if self._param.history:
-                    yohoo_res.append("# History:\n" + msft.history().to_markdown() + "\n")
+                    yahoo_res.append("# History:\n" + msft.history().to_markdown() + "\n")
                 if self._param.financials:
-                    yohoo_res.append("# Calendar:\n" + pd.DataFrame(msft.calendar).to_markdown() + "\n")
+                    yahoo_res.append("# Calendar:\n" + pd.DataFrame(msft.calendar).to_markdown() + "\n")
                 if self._param.balance_sheet:
-                    yohoo_res.append("# Balance sheet:\n" + msft.balance_sheet.to_markdown() + "\n")
-                    yohoo_res.append("# Quarterly balance sheet:\n" + msft.quarterly_balance_sheet.to_markdown() + "\n")
+                    yahoo_res.append("# Balance sheet:\n" + msft.balance_sheet.to_markdown() + "\n")
+                    yahoo_res.append("# Quarterly balance sheet:\n" + msft.quarterly_balance_sheet.to_markdown() + "\n")
                 if self._param.cash_flow_statement:
-                    yohoo_res.append("# Cash flow statement:\n" + msft.cashflow.to_markdown() + "\n")
-                    yohoo_res.append("# Quarterly cash flow statement:\n" + msft.quarterly_cashflow.to_markdown() + "\n")
+                    yahoo_res.append("# Cash flow statement:\n" + msft.cashflow.to_markdown() + "\n")
+                    yahoo_res.append("# Quarterly cash flow statement:\n" + msft.quarterly_cashflow.to_markdown() + "\n")
                 if self._param.news:
-                    yohoo_res.append("# News:\n" + pd.DataFrame(msft.news).to_markdown() + "\n")
-                self.set_output("report", "\n\n".join(yohoo_res))
+                    yahoo_res.append("# News:\n" + pd.DataFrame(msft.news).to_markdown() + "\n")
+                self.set_output("report", "\n\n".join(yahoo_res))
                 return self.output("report")
             except Exception as e:
+                if self.check_if_canceled("YahooFinance processing"):
+                    return None
+
                 last_e = e
                 logging.exception(f"YahooFinance error: {e}")
                 time.sleep(self._param.delay_after_error)

agentic_reasoning/deep_research.py

@@ -51,7 +51,7 @@ class DeepResearcher:
         """Remove Result Tags"""
         return DeepResearcher._remove_tags(text, BEGIN_SEARCH_RESULT, END_SEARCH_RESULT)
 
-    def _generate_reasoning(self, msg_history):
+    async def _generate_reasoning(self, msg_history):
         """Generate reasoning steps"""
         query_think = ""
         if msg_history[-1]["role"] != "user":
@@ -59,13 +59,14 @@ class DeepResearcher:
         else:
             msg_history[-1]["content"] += "\n\nContinues reasoning with the new information.\n"
             
-        for ans in self.chat_mdl.chat_streamly(REASON_PROMPT, msg_history, {"temperature": 0.7}):
+        async for ans in self.chat_mdl.async_chat_streamly(REASON_PROMPT, msg_history, {"temperature": 0.7}):
             ans = re.sub(r"^.*</think>", "", ans, flags=re.DOTALL)
             if not ans:
                 continue
             query_think = ans
             yield query_think
-        return query_think
+            query_think = ""
+        yield query_think
 
     def _extract_search_queries(self, query_think, question, step_index):
         """Extract search queries from thinking"""
@@ -143,10 +144,10 @@ class DeepResearcher:
                 if d["doc_id"] not in dids:
                     chunk_info["doc_aggs"].append(d)
 
-    def _extract_relevant_info(self, truncated_prev_reasoning, search_query, kbinfos):
+    async def _extract_relevant_info(self, truncated_prev_reasoning, search_query, kbinfos):
         """Extract and summarize relevant information"""
         summary_think = ""
-        for ans in self.chat_mdl.chat_streamly(
+        async for ans in self.chat_mdl.async_chat_streamly(
                 RELEVANT_EXTRACTION_PROMPT.format(
                     prev_reasoning=truncated_prev_reasoning,
                     search_query=search_query,
@@ -160,10 +161,11 @@ class DeepResearcher:
                 continue
             summary_think = ans
             yield summary_think
+            summary_think = ""
         
-        return summary_think
+        yield summary_think
 
-    def thinking(self, chunk_info: dict, question: str):
+    async def thinking(self, chunk_info: dict, question: str):
         executed_search_queries = []
         msg_history = [{"role": "user", "content": f'Question:\"{question}\"\n'}]
         all_reasoning_steps = []
@@ -180,7 +182,7 @@ class DeepResearcher:
 
             # Step 1: Generate reasoning
             query_think = ""
-            for ans in self._generate_reasoning(msg_history):
+            async for ans in self._generate_reasoning(msg_history):
                 query_think = ans
                 yield {"answer": think + self._remove_query_tags(query_think) + "</think>", "reference": {}, "audio_binary": None}
 
@@ -223,7 +225,7 @@ class DeepResearcher:
                 # Step 6: Extract relevant information
                 think += "\n\n"
                 summary_think = ""
-                for ans in self._extract_relevant_info(truncated_prev_reasoning, search_query, kbinfos):
+                async for ans in self._extract_relevant_info(truncated_prev_reasoning, search_query, kbinfos):
                     summary_think = ans
                     yield {"answer": think + self._remove_result_tags(summary_think) + "</think>", "reference": {}, "audio_binary": None}
 

api/__init__.py

@@ -14,5 +14,5 @@
 #  limitations under the License.
 #
 
-from beartype.claw import beartype_this_package
-beartype_this_package()
+# from beartype.claw import beartype_this_package
+# beartype_this_package()

api/apps/__init__.py

@@ -13,36 +13,33 @@
 #  See the License for the specific language governing permissions and
 #  limitations under the License.
 #
+import logging
 import os
 import sys
-import logging
 from importlib.util import module_from_spec, spec_from_file_location
 from pathlib import Path
-from flask import Blueprint, Flask
-from werkzeug.wrappers.request import Request
-from flask_cors import CORS
+from quart import Blueprint, Quart, request, g, current_app, session
 from flasgger import Swagger
 from itsdangerous.url_safe import URLSafeTimedSerializer as Serializer
-
+from quart_cors import cors
 from common.constants import StatusEnum
-from api.db.db_models import close_connection
+from api.db.db_models import close_connection, APIToken
 from api.db.services import UserService
 from api.utils.json_encode import CustomJSONEncoder
 from api.utils import commands
 
-from flask_mail import Mail
-from flask_session import Session
-from flask_login import LoginManager
-from api import settings
+from quart_auth import Unauthorized
+from common import settings
 from api.utils.api_utils import server_error_response
 from api.constants import API_VERSION
+from common.misc_utils import get_uuid
+
+settings.init_settings()
 
 __all__ = ["app"]
 
-Request.json = property(lambda self: self.get_json(force=True, silent=True))
-
-app = Flask(__name__)
-smtp_mail_server = Mail()
+app = Quart(__name__)
+app = cors(app, allow_origin="*")
 
 # Add this at the beginning of your file to configure Swagger UI
 swagger_config = {
@@ -76,32 +73,168 @@ swagger = Swagger(
     },
 )
 
-CORS(app, supports_credentials=True, max_age=2592000)
 app.url_map.strict_slashes = False
 app.json_encoder = CustomJSONEncoder
 app.errorhandler(Exception)(server_error_response)
 
+# Configure Quart timeouts for slow LLM responses (e.g., local Ollama on CPU)
+# Default Quart timeouts are 60 seconds which is too short for many LLM backends
+app.config["RESPONSE_TIMEOUT"] = int(os.environ.get("QUART_RESPONSE_TIMEOUT", 600))
+app.config["BODY_TIMEOUT"] = int(os.environ.get("QUART_BODY_TIMEOUT", 600))
+
 ## convince for dev and debug
 # app.config["LOGIN_DISABLED"] = True
 app.config["SESSION_PERMANENT"] = False
-app.config["SESSION_TYPE"] = "filesystem"
+app.config["SESSION_TYPE"] = "redis"
+app.config["SESSION_REDIS"] = settings.decrypt_database_config(name="redis")
 app.config["MAX_CONTENT_LENGTH"] = int(
     os.environ.get("MAX_CONTENT_LENGTH", 1024 * 1024 * 1024)
 )
-
-Session(app)
-login_manager = LoginManager()
-login_manager.init_app(app)
-
+app.config['SECRET_KEY'] = settings.SECRET_KEY
+app.secret_key = settings.SECRET_KEY
 commands.register_commands(app)
 
+from functools import wraps
+from typing import ParamSpec, TypeVar
+from collections.abc import Awaitable, Callable
+from werkzeug.local import LocalProxy
 
-def search_pages_path(pages_dir):
+T = TypeVar("T")
+P = ParamSpec("P")
+
+
+def _load_user():
+    jwt = Serializer(secret_key=settings.SECRET_KEY)
+    authorization = request.headers.get("Authorization")
+    g.user = None
+    if not authorization:
+        return None
+
+    try:
+        access_token = str(jwt.loads(authorization))
+
+        if not access_token or not access_token.strip():
+            logging.warning("Authentication attempt with empty access token")
+            return None
+
+        # Access tokens should be UUIDs (32 hex characters)
+        if len(access_token.strip()) < 32:
+            logging.warning(f"Authentication attempt with invalid token format: {len(access_token)} chars")
+            return None
+
+        user = UserService.query(
+            access_token=access_token, status=StatusEnum.VALID.value
+        )
+        if not user and len(authorization.split()) == 2:
+            objs = APIToken.query(token=authorization.split()[1])
+            if objs:
+                user = UserService.query(id=objs[0].tenant_id, status=StatusEnum.VALID.value)
+        if user:
+            if not user[0].access_token or not user[0].access_token.strip():
+                logging.warning(f"User {user[0].email} has empty access_token in database")
+                return None
+            g.user = user[0]
+            return user[0]
+    except Exception as e:
+        logging.warning(f"load_user got exception {e}")
+
+
+current_user = LocalProxy(_load_user)
+
+
+def login_required(func: Callable[P, Awaitable[T]]) -> Callable[P, Awaitable[T]]:
+    """A decorator to restrict route access to authenticated users.
+
+    This should be used to wrap a route handler (or view function) to
+    enforce that only authenticated requests can access it. Note that
+    it is important that this decorator be wrapped by the route
+    decorator and not vice, versa, as below.
+
+    .. code-block:: python
+
+        @app.route('/')
+        @login_required
+        async def index():
+            ...
+
+    If the request is not authenticated a
+    `quart.exceptions.Unauthorized` exception will be raised.
+
+    """
+
+    @wraps(func)
+    async def wrapper(*args: P.args, **kwargs: P.kwargs) -> T:
+        if not current_user:  # or not session.get("_user_id"):
+            raise Unauthorized()
+        else:
+            return await current_app.ensure_async(func)(*args, **kwargs)
+
+    return wrapper
+
+
+def login_user(user, remember=False, duration=None, force=False, fresh=True):
+    """
+    Logs a user in. You should pass the actual user object to this. If the
+    user's `is_active` property is ``False``, they will not be logged in
+    unless `force` is ``True``.
+
+    This will return ``True`` if the login attempt succeeds, and ``False`` if
+    it fails (i.e. because the user is inactive).
+
+    :param user: The user object to log in.
+    :type user: object
+    :param remember: Whether to remember the user after their session expires.
+        Defaults to ``False``.
+    :type remember: bool
+    :param duration: The amount of time before the remember cookie expires. If
+        ``None`` the value set in the settings is used. Defaults to ``None``.
+    :type duration: :class:`datetime.timedelta`
+    :param force: If the user is inactive, setting this to ``True`` will log
+        them in regardless. Defaults to ``False``.
+    :type force: bool
+    :param fresh: setting this to ``False`` will log in the user with a session
+        marked as not "fresh". Defaults to ``True``.
+    :type fresh: bool
+    """
+    if not force and not user.is_active:
+        return False
+
+    session["_user_id"] = user.id
+    session["_fresh"] = fresh
+    session["_id"] = get_uuid()
+    return True
+
+
+def logout_user():
+    """
+    Logs a user out. (You do not need to pass the actual user.) This will
+    also clean up the remember me cookie if it exists.
+    """
+    if "_user_id" in session:
+        session.pop("_user_id")
+
+    if "_fresh" in session:
+        session.pop("_fresh")
+
+    if "_id" in session:
+        session.pop("_id")
+
+    COOKIE_NAME = "remember_token"
+    cookie_name = current_app.config.get("REMEMBER_COOKIE_NAME", COOKIE_NAME)
+    if cookie_name in request.cookies:
+        session["_remember"] = "clear"
+        if "_remember_seconds" in session:
+            session.pop("_remember_seconds")
+
+    return True
+
+
+def search_pages_path(page_path):
     app_path_list = [
-        path for path in pages_dir.glob("*_app.py") if not path.name.startswith(".")
+        path for path in page_path.glob("*_app.py") if not path.name.startswith(".")
     ]
     api_path_list = [
-        path for path in pages_dir.glob("*sdk/*.py") if not path.name.startswith(".")
+        path for path in page_path.glob("*sdk/*.py") if not path.name.startswith(".")
     ]
     app_path_list.extend(api_path_list)
     return app_path_list
@@ -138,44 +271,22 @@ pages_dir = [
 ]
 
 client_urls_prefix = [
-    register_page(path) for dir in pages_dir for path in search_pages_path(dir)
+    register_page(path) for directory in pages_dir for path in search_pages_path(directory)
 ]
 
 
-@login_manager.request_loader
-def load_user(web_request):
-    jwt = Serializer(secret_key=settings.SECRET_KEY)
-    authorization = web_request.headers.get("Authorization")
-    if authorization:
-        try:
-            access_token = str(jwt.loads(authorization))
-
-            if not access_token or not access_token.strip():
-                logging.warning("Authentication attempt with empty access token")
-                return None
-
-            # Access tokens should be UUIDs (32 hex characters)
-            if len(access_token.strip()) < 32:
-                logging.warning(f"Authentication attempt with invalid token format: {len(access_token)} chars")
-                return None
-
-            user = UserService.query(
-                access_token=access_token, status=StatusEnum.VALID.value
-            )
-            if user:
-                if not user[0].access_token or not user[0].access_token.strip():
-                    logging.warning(f"User {user[0].email} has empty access_token in database")
-                    return None
-                return user[0]
-            else:
-                return None
-        except Exception as e:
-            logging.warning(f"load_user got exception {e}")
-            return None
-    else:
-        return None
+@app.errorhandler(404)
+async def not_found(error):
+    error_msg: str = f"The requested URL {request.path} was not found"
+    logging.error(error_msg)
+    return {
+        "error": "Not Found",
+        "message": error_msg,
+    }, 404
 
 
 @app.teardown_request
-def _db_close(exc):
+def _db_close(exception):
+    if exception:
+        logging.exception(f"Request failed: {exception}")
     close_connection()

api/apps/api_app.py

@@ -13,47 +13,20 @@
 #  See the License for the specific language governing permissions and
 #  limitations under the License.
 #
-import json
-import os
-import re
 from datetime import datetime, timedelta
-from flask import request, Response
-from api.db.services.llm_service import LLMBundle
-from flask_login import login_required, current_user
-
-from api.db import VALID_FILE_TYPES, FileType
-from api.db.db_models import APIToken, Task, File
-from api.db.services import duplicate_name
+from quart import request
+from api.db.db_models import APIToken
 from api.db.services.api_service import APITokenService, API4ConversationService
-from api.db.services.dialog_service import DialogService, chat
-from api.db.services.document_service import DocumentService, doc_upload_and_parse
-from api.db.services.file2document_service import File2DocumentService
-from api.db.services.file_service import FileService
-from api.db.services.knowledgebase_service import KnowledgebaseService
-from api.db.services.task_service import queue_tasks, TaskService
 from api.db.services.user_service import UserTenantService
-from api import settings
-from common.misc_utils import get_uuid
-from common.constants import RetCode, VALID_TASK_STATUS, LLMType, ParserType, FileSource
-from api.utils.api_utils import server_error_response, get_data_error_result, get_json_result, validate_request, \
-    generate_confirmation_token
-
-from api.utils.file_utils import filename_type, thumbnail
-from rag.app.tag import label_question
-from rag.prompts.generator import keyword_extraction
-from rag.utils.storage_factory import STORAGE_IMPL
+from api.utils.api_utils import generate_confirmation_token, get_data_error_result, get_json_result, get_request_json, server_error_response, validate_request
 from common.time_utils import current_timestamp, datetime_format
-
-from api.db.services.canvas_service import UserCanvasService
-from agent.canvas import Canvas
-from functools import partial
-from pathlib import Path
+from api.apps import login_required, current_user
 
 
 @manager.route('/new_token', methods=['POST'])  # noqa: F821
 @login_required
-def new_token():
-    req = request.json
+async def new_token():
+    req = await get_request_json()
     try:
         tenants = UserTenantService.query(user_id=current_user.id)
         if not tenants:
@@ -98,8 +71,8 @@ def token_list():
 @manager.route('/rm', methods=['POST'])  # noqa: F821
 @validate_request("tokens", "tenant_id")
 @login_required
-def rm():
-    req = request.json
+async def rm():
+    req = await get_request_json()
     try:
         for token in req["tokens"]:
             APITokenService.filter_delete(
@@ -127,774 +100,18 @@ def stats():
                 "to_date",
                 datetime.now().strftime("%Y-%m-%d %H:%M:%S")),
             "agent" if "canvas_id" in request.args else None)
-        res = {
-            "pv": [(o["dt"], o["pv"]) for o in objs],
-            "uv": [(o["dt"], o["uv"]) for o in objs],
-            "speed": [(o["dt"], float(o["tokens"]) / (float(o["duration"] + 0.1))) for o in objs],
-            "tokens": [(o["dt"], float(o["tokens"]) / 1000.) for o in objs],
-            "round": [(o["dt"], o["round"]) for o in objs],
-            "thumb_up": [(o["dt"], o["thumb_up"]) for o in objs]
-        }
+
+        res = {"pv": [], "uv": [], "speed": [], "tokens": [], "round": [], "thumb_up": []}
+
+        for obj in objs:
+            dt = obj["dt"]
+            res["pv"].append((dt, obj["pv"]))
+            res["uv"].append((dt, obj["uv"]))
+            res["speed"].append((dt, float(obj["tokens"]) / (float(obj["duration"]) + 0.1))) # +0.1 to avoid division by zero
+            res["tokens"].append((dt, float(obj["tokens"]) / 1000.0)) # convert to thousands
+            res["round"].append((dt, obj["round"]))
+            res["thumb_up"].append((dt, obj["thumb_up"]))
+
         return get_json_result(data=res)
     except Exception as e:
         return server_error_response(e)
-
-
-@manager.route('/new_conversation', methods=['GET'])  # noqa: F821
-def set_conversation():
-    token = request.headers.get('Authorization').split()[1]
-    objs = APIToken.query(token=token)
-    if not objs:
-        return get_json_result(
-            data=False, message='Authentication error: API key is invalid!"', code=RetCode.AUTHENTICATION_ERROR)
-    try:
-        if objs[0].source == "agent":
-            e, cvs = UserCanvasService.get_by_id(objs[0].dialog_id)
-            if not e:
-                return server_error_response("canvas not found.")
-            if not isinstance(cvs.dsl, str):
-                cvs.dsl = json.dumps(cvs.dsl, ensure_ascii=False)
-            canvas = Canvas(cvs.dsl, objs[0].tenant_id)
-            conv = {
-                "id": get_uuid(),
-                "dialog_id": cvs.id,
-                "user_id": request.args.get("user_id", ""),
-                "message": [{"role": "assistant", "content": canvas.get_prologue()}],
-                "source": "agent"
-            }
-            API4ConversationService.save(**conv)
-            return get_json_result(data=conv)
-        else:
-            e, dia = DialogService.get_by_id(objs[0].dialog_id)
-            if not e:
-                return get_data_error_result(message="Dialog not found")
-            conv = {
-                "id": get_uuid(),
-                "dialog_id": dia.id,
-                "user_id": request.args.get("user_id", ""),
-                "message": [{"role": "assistant", "content": dia.prompt_config["prologue"]}]
-            }
-            API4ConversationService.save(**conv)
-            return get_json_result(data=conv)
-    except Exception as e:
-        return server_error_response(e)
-
-
-@manager.route('/completion', methods=['POST'])  # noqa: F821
-@validate_request("conversation_id", "messages")
-def completion():
-    token = request.headers.get('Authorization').split()[1]
-    objs = APIToken.query(token=token)
-    if not objs:
-        return get_json_result(
-            data=False, message='Authentication error: API key is invalid!"', code=RetCode.AUTHENTICATION_ERROR)
-    req = request.json
-    e, conv = API4ConversationService.get_by_id(req["conversation_id"])
-    if not e:
-        return get_data_error_result(message="Conversation not found!")
-    if "quote" not in req:
-        req["quote"] = False
-
-    msg = []
-    for m in req["messages"]:
-        if m["role"] == "system":
-            continue
-        if m["role"] == "assistant" and not msg:
-            continue
-        msg.append(m)
-    if not msg[-1].get("id"):
-        msg[-1]["id"] = get_uuid()
-    message_id = msg[-1]["id"]
-
-    def fillin_conv(ans):
-        nonlocal conv, message_id
-        if not conv.reference:
-            conv.reference.append(ans["reference"])
-        else:
-            conv.reference[-1] = ans["reference"]
-        conv.message[-1] = {"role": "assistant", "content": ans["answer"], "id": message_id}
-        ans["id"] = message_id
-
-    def rename_field(ans):
-        reference = ans['reference']
-        if not isinstance(reference, dict):
-            return
-        for chunk_i in reference.get('chunks', []):
-            if 'docnm_kwd' in chunk_i:
-                chunk_i['doc_name'] = chunk_i['docnm_kwd']
-                chunk_i.pop('docnm_kwd')
-
-    try:
-        if conv.source == "agent":
-            stream = req.get("stream", True)
-            conv.message.append(msg[-1])
-            e, cvs = UserCanvasService.get_by_id(conv.dialog_id)
-            if not e:
-                return server_error_response("canvas not found.")
-            del req["conversation_id"]
-            del req["messages"]
-
-            if not isinstance(cvs.dsl, str):
-                cvs.dsl = json.dumps(cvs.dsl, ensure_ascii=False)
-
-            if not conv.reference:
-                conv.reference = []
-            conv.message.append({"role": "assistant", "content": "", "id": message_id})
-            conv.reference.append({"chunks": [], "doc_aggs": []})
-
-            final_ans = {"reference": [], "content": ""}
-            canvas = Canvas(cvs.dsl, objs[0].tenant_id)
-
-            canvas.messages.append(msg[-1])
-            canvas.add_user_input(msg[-1]["content"])
-            answer = canvas.run(stream=stream)
-
-            assert answer is not None, "Nothing. Is it over?"
-
-            if stream:
-                assert isinstance(answer, partial), "Nothing. Is it over?"
-
-                def sse():
-                    nonlocal answer, cvs, conv
-                    try:
-                        for ans in answer():
-                            for k in ans.keys():
-                                final_ans[k] = ans[k]
-                            ans = {"answer": ans["content"], "reference": ans.get("reference", [])}
-                            fillin_conv(ans)
-                            rename_field(ans)
-                            yield "data:" + json.dumps({"code": 0, "message": "", "data": ans},
-                                                       ensure_ascii=False) + "\n\n"
-
-                        canvas.messages.append({"role": "assistant", "content": final_ans["content"], "id": message_id})
-                        canvas.history.append(("assistant", final_ans["content"]))
-                        if final_ans.get("reference"):
-                            canvas.reference.append(final_ans["reference"])
-                        cvs.dsl = json.loads(str(canvas))
-                        API4ConversationService.append_message(conv.id, conv.to_dict())
-                    except Exception as e:
-                        yield "data:" + json.dumps({"code": 500, "message": str(e),
-                                                    "data": {"answer": "**ERROR**: " + str(e), "reference": []}},
-                                                   ensure_ascii=False) + "\n\n"
-                    yield "data:" + json.dumps({"code": 0, "message": "", "data": True}, ensure_ascii=False) + "\n\n"
-
-                resp = Response(sse(), mimetype="text/event-stream")
-                resp.headers.add_header("Cache-control", "no-cache")
-                resp.headers.add_header("Connection", "keep-alive")
-                resp.headers.add_header("X-Accel-Buffering", "no")
-                resp.headers.add_header("Content-Type", "text/event-stream; charset=utf-8")
-                return resp
-
-            final_ans["content"] = "\n".join(answer["content"]) if "content" in answer else ""
-            canvas.messages.append({"role": "assistant", "content": final_ans["content"], "id": message_id})
-            if final_ans.get("reference"):
-                canvas.reference.append(final_ans["reference"])
-            cvs.dsl = json.loads(str(canvas))
-
-            result = {"answer": final_ans["content"], "reference": final_ans.get("reference", [])}
-            fillin_conv(result)
-            API4ConversationService.append_message(conv.id, conv.to_dict())
-            rename_field(result)
-            return get_json_result(data=result)
-
-        # ******************For dialog******************
-        conv.message.append(msg[-1])
-        e, dia = DialogService.get_by_id(conv.dialog_id)
-        if not e:
-            return get_data_error_result(message="Dialog not found!")
-        del req["conversation_id"]
-        del req["messages"]
-
-        if not conv.reference:
-            conv.reference = []
-        conv.message.append({"role": "assistant", "content": "", "id": message_id})
-        conv.reference.append({"chunks": [], "doc_aggs": []})
-
-        def stream():
-            nonlocal dia, msg, req, conv
-            try:
-                for ans in chat(dia, msg, True, **req):
-                    fillin_conv(ans)
-                    rename_field(ans)
-                    yield "data:" + json.dumps({"code": 0, "message": "", "data": ans},
-                                               ensure_ascii=False) + "\n\n"
-                API4ConversationService.append_message(conv.id, conv.to_dict())
-            except Exception as e:
-                yield "data:" + json.dumps({"code": 500, "message": str(e),
-                                            "data": {"answer": "**ERROR**: " + str(e), "reference": []}},
-                                           ensure_ascii=False) + "\n\n"
-            yield "data:" + json.dumps({"code": 0, "message": "", "data": True}, ensure_ascii=False) + "\n\n"
-
-        if req.get("stream", True):
-            resp = Response(stream(), mimetype="text/event-stream")
-            resp.headers.add_header("Cache-control", "no-cache")
-            resp.headers.add_header("Connection", "keep-alive")
-            resp.headers.add_header("X-Accel-Buffering", "no")
-            resp.headers.add_header("Content-Type", "text/event-stream; charset=utf-8")
-            return resp
-
-        answer = None
-        for ans in chat(dia, msg, **req):
-            answer = ans
-            fillin_conv(ans)
-            API4ConversationService.append_message(conv.id, conv.to_dict())
-            break
-        rename_field(answer)
-        return get_json_result(data=answer)
-
-    except Exception as e:
-        return server_error_response(e)
-
-
-@manager.route('/conversation/<conversation_id>', methods=['GET'])  # noqa: F821
-# @login_required
-def get_conversation(conversation_id):
-    token = request.headers.get('Authorization').split()[1]
-    objs = APIToken.query(token=token)
-    if not objs:
-        return get_json_result(
-            data=False, message='Authentication error: API key is invalid!"', code=RetCode.AUTHENTICATION_ERROR)
-
-    try:
-        e, conv = API4ConversationService.get_by_id(conversation_id)
-        if not e:
-            return get_data_error_result(message="Conversation not found!")
-
-        conv = conv.to_dict()
-        if token != APIToken.query(dialog_id=conv['dialog_id'])[0].token:
-            return get_json_result(data=False, message='Authentication error: API key is invalid for this conversation_id!"',
-                                   code=RetCode.AUTHENTICATION_ERROR)
-
-        for referenct_i in conv['reference']:
-            if referenct_i is None or len(referenct_i) == 0:
-                continue
-            for chunk_i in referenct_i['chunks']:
-                if 'docnm_kwd' in chunk_i.keys():
-                    chunk_i['doc_name'] = chunk_i['docnm_kwd']
-                    chunk_i.pop('docnm_kwd')
-        return get_json_result(data=conv)
-    except Exception as e:
-        return server_error_response(e)
-
-
-@manager.route('/document/upload', methods=['POST'])  # noqa: F821
-@validate_request("kb_name")
-def upload():
-    token = request.headers.get('Authorization').split()[1]
-    objs = APIToken.query(token=token)
-    if not objs:
-        return get_json_result(
-            data=False, message='Authentication error: API key is invalid!"', code=RetCode.AUTHENTICATION_ERROR)
-
-    kb_name = request.form.get("kb_name").strip()
-    tenant_id = objs[0].tenant_id
-
-    try:
-        e, kb = KnowledgebaseService.get_by_name(kb_name, tenant_id)
-        if not e:
-            return get_data_error_result(
-                message="Can't find this knowledgebase!")
-        kb_id = kb.id
-    except Exception as e:
-        return server_error_response(e)
-
-    if 'file' not in request.files:
-        return get_json_result(
-            data=False, message='No file part!', code=RetCode.ARGUMENT_ERROR)
-
-    file = request.files['file']
-    if file.filename == '':
-        return get_json_result(
-            data=False, message='No file selected!', code=RetCode.ARGUMENT_ERROR)
-
-    root_folder = FileService.get_root_folder(tenant_id)
-    pf_id = root_folder["id"]
-    FileService.init_knowledgebase_docs(pf_id, tenant_id)
-    kb_root_folder = FileService.get_kb_folder(tenant_id)
-    kb_folder = FileService.new_a_file_from_kb(kb.tenant_id, kb.name, kb_root_folder["id"])
-
-    try:
-        if DocumentService.get_doc_count(kb.tenant_id) >= int(os.environ.get('MAX_FILE_NUM_PER_USER', 8192)):
-            return get_data_error_result(
-                message="Exceed the maximum file number of a free user!")
-
-        filename = duplicate_name(
-            DocumentService.query,
-            name=file.filename,
-            kb_id=kb_id)
-        filetype = filename_type(filename)
-        if not filetype:
-            return get_data_error_result(
-                message="This type of file has not been supported yet!")
-
-        location = filename
-        while STORAGE_IMPL.obj_exist(kb_id, location):
-            location += "_"
-        blob = request.files['file'].read()
-        STORAGE_IMPL.put(kb_id, location, blob)
-        doc = {
-            "id": get_uuid(),
-            "kb_id": kb.id,
-            "parser_id": kb.parser_id,
-            "parser_config": kb.parser_config,
-            "created_by": kb.tenant_id,
-            "type": filetype,
-            "name": filename,
-            "location": location,
-            "size": len(blob),
-            "thumbnail": thumbnail(filename, blob),
-            "suffix": Path(filename).suffix.lstrip("."),
-        }
-
-        form_data = request.form
-        if "parser_id" in form_data.keys():
-            if request.form.get("parser_id").strip() in list(vars(ParserType).values())[1:-3]:
-                doc["parser_id"] = request.form.get("parser_id").strip()
-        if doc["type"] == FileType.VISUAL:
-            doc["parser_id"] = ParserType.PICTURE.value
-        if doc["type"] == FileType.AURAL:
-            doc["parser_id"] = ParserType.AUDIO.value
-        if re.search(r"\.(ppt|pptx|pages)$", filename):
-            doc["parser_id"] = ParserType.PRESENTATION.value
-        if re.search(r"\.(eml)$", filename):
-            doc["parser_id"] = ParserType.EMAIL.value
-
-        doc_result = DocumentService.insert(doc)
-        FileService.add_file_from_kb(doc, kb_folder["id"], kb.tenant_id)
-    except Exception as e:
-        return server_error_response(e)
-
-    if "run" in form_data.keys():
-        if request.form.get("run").strip() == "1":
-            try:
-                info = {"run": 1, "progress": 0}
-                info["progress_msg"] = ""
-                info["chunk_num"] = 0
-                info["token_num"] = 0
-                DocumentService.update_by_id(doc["id"], info)
-                # if str(req["run"]) == TaskStatus.CANCEL.value:
-                tenant_id = DocumentService.get_tenant_id(doc["id"])
-                if not tenant_id:
-                    return get_data_error_result(message="Tenant not found!")
-
-                # e, doc = DocumentService.get_by_id(doc["id"])
-                TaskService.filter_delete([Task.doc_id == doc["id"]])
-                e, doc = DocumentService.get_by_id(doc["id"])
-                doc = doc.to_dict()
-                doc["tenant_id"] = tenant_id
-                bucket, name = File2DocumentService.get_storage_address(doc_id=doc["id"])
-                queue_tasks(doc, bucket, name, 0)
-            except Exception as e:
-                return server_error_response(e)
-
-    return get_json_result(data=doc_result.to_json())
-
-
-@manager.route('/document/upload_and_parse', methods=['POST'])  # noqa: F821
-@validate_request("conversation_id")
-def upload_parse():
-    token = request.headers.get('Authorization').split()[1]
-    objs = APIToken.query(token=token)
-    if not objs:
-        return get_json_result(
-            data=False, message='Authentication error: API key is invalid!"', code=RetCode.AUTHENTICATION_ERROR)
-
-    if 'file' not in request.files:
-        return get_json_result(
-            data=False, message='No file part!', code=RetCode.ARGUMENT_ERROR)
-
-    file_objs = request.files.getlist('file')
-    for file_obj in file_objs:
-        if file_obj.filename == '':
-            return get_json_result(
-                data=False, message='No file selected!', code=RetCode.ARGUMENT_ERROR)
-
-    doc_ids = doc_upload_and_parse(request.form.get("conversation_id"), file_objs, objs[0].tenant_id)
-    return get_json_result(data=doc_ids)
-
-
-@manager.route('/list_chunks', methods=['POST'])  # noqa: F821
-# @login_required
-def list_chunks():
-    token = request.headers.get('Authorization').split()[1]
-    objs = APIToken.query(token=token)
-    if not objs:
-        return get_json_result(
-            data=False, message='Authentication error: API key is invalid!"', code=RetCode.AUTHENTICATION_ERROR)
-
-    req = request.json
-
-    try:
-        if "doc_name" in req.keys():
-            tenant_id = DocumentService.get_tenant_id_by_name(req['doc_name'])
-            doc_id = DocumentService.get_doc_id_by_doc_name(req['doc_name'])
-
-        elif "doc_id" in req.keys():
-            tenant_id = DocumentService.get_tenant_id(req['doc_id'])
-            doc_id = req['doc_id']
-        else:
-            return get_json_result(
-                data=False, message="Can't find doc_name or doc_id"
-            )
-        kb_ids = KnowledgebaseService.get_kb_ids(tenant_id)
-
-        res = settings.retriever.chunk_list(doc_id, tenant_id, kb_ids)
-        res = [
-            {
-                "content": res_item["content_with_weight"],
-                "doc_name": res_item["docnm_kwd"],
-                "image_id": res_item["img_id"]
-            } for res_item in res
-        ]
-
-    except Exception as e:
-        return server_error_response(e)
-
-    return get_json_result(data=res)
-
-@manager.route('/get_chunk/<chunk_id>', methods=['GET'])  # noqa: F821
-# @login_required
-def get_chunk(chunk_id):
-    from rag.nlp import search
-    token = request.headers.get('Authorization').split()[1]
-    objs = APIToken.query(token=token)
-    if not objs:
-        return get_json_result(
-            data=False, message='Authentication error: API key is invalid!"', code=RetCode.AUTHENTICATION_ERROR)
-    try:
-        tenant_id = objs[0].tenant_id
-        kb_ids = KnowledgebaseService.get_kb_ids(tenant_id)
-        chunk = settings.docStoreConn.get(chunk_id, search.index_name(tenant_id), kb_ids)
-        if chunk is None:
-            return server_error_response(Exception("Chunk not found"))
-        k = []
-        for n in chunk.keys():
-            if re.search(r"(_vec$|_sm_|_tks|_ltks)", n):
-                k.append(n)
-        for n in k:
-            del chunk[n]
-
-        return get_json_result(data=chunk)
-    except Exception as e:
-        return server_error_response(e)
-
-@manager.route('/list_kb_docs', methods=['POST'])  # noqa: F821
-# @login_required
-def list_kb_docs():
-    token = request.headers.get('Authorization').split()[1]
-    objs = APIToken.query(token=token)
-    if not objs:
-        return get_json_result(
-            data=False, message='Authentication error: API key is invalid!"', code=RetCode.AUTHENTICATION_ERROR)
-
-    req = request.json
-    tenant_id = objs[0].tenant_id
-    kb_name = req.get("kb_name", "").strip()
-
-    try:
-        e, kb = KnowledgebaseService.get_by_name(kb_name, tenant_id)
-        if not e:
-            return get_data_error_result(
-                message="Can't find this knowledgebase!")
-        kb_id = kb.id
-
-    except Exception as e:
-        return server_error_response(e)
-
-    page_number = int(req.get("page", 1))
-    items_per_page = int(req.get("page_size", 15))
-    orderby = req.get("orderby", "create_time")
-    desc = req.get("desc", True)
-    keywords = req.get("keywords", "")
-    status = req.get("status", [])
-    if status:
-        invalid_status = {s for s in status if s not in VALID_TASK_STATUS}
-        if invalid_status:
-            return get_data_error_result(
-                message=f"Invalid filter status conditions: {', '.join(invalid_status)}"
-            )
-    types = req.get("types", [])
-    if types:
-        invalid_types = {t for t in types if t not in VALID_FILE_TYPES}
-        if invalid_types:
-            return get_data_error_result(
-                message=f"Invalid filter conditions: {', '.join(invalid_types)} type{'s' if len(invalid_types) > 1 else ''}"
-            )
-    try:
-        docs, tol = DocumentService.get_by_kb_id(
-            kb_id, page_number, items_per_page, orderby, desc, keywords, status, types)
-        docs = [{"doc_id": doc['id'], "doc_name": doc['name']} for doc in docs]
-
-        return get_json_result(data={"total": tol, "docs": docs})
-
-    except Exception as e:
-        return server_error_response(e)
-
-
-@manager.route('/document/infos', methods=['POST'])  # noqa: F821
-@validate_request("doc_ids")
-def docinfos():
-    token = request.headers.get('Authorization').split()[1]
-    objs = APIToken.query(token=token)
-    if not objs:
-        return get_json_result(
-            data=False, message='Authentication error: API key is invalid!"', code=RetCode.AUTHENTICATION_ERROR)
-    req = request.json
-    doc_ids = req["doc_ids"]
-    docs = DocumentService.get_by_ids(doc_ids)
-    return get_json_result(data=list(docs.dicts()))
-
-
-@manager.route('/document', methods=['DELETE'])  # noqa: F821
-# @login_required
-def document_rm():
-    token = request.headers.get('Authorization').split()[1]
-    objs = APIToken.query(token=token)
-    if not objs:
-        return get_json_result(
-            data=False, message='Authentication error: API key is invalid!"', code=RetCode.AUTHENTICATION_ERROR)
-
-    tenant_id = objs[0].tenant_id
-    req = request.json
-    try:
-        doc_ids = DocumentService.get_doc_ids_by_doc_names(req.get("doc_names", []))
-        for doc_id in req.get("doc_ids", []):
-            if doc_id not in doc_ids:
-                doc_ids.append(doc_id)
-
-        if not doc_ids:
-            return get_json_result(
-                data=False, message="Can't find doc_names or doc_ids"
-            )
-
-    except Exception as e:
-        return server_error_response(e)
-
-    root_folder = FileService.get_root_folder(tenant_id)
-    pf_id = root_folder["id"]
-    FileService.init_knowledgebase_docs(pf_id, tenant_id)
-
-    errors = ""
-    docs = DocumentService.get_by_ids(doc_ids)
-    doc_dic = {}
-    for doc in docs:
-        doc_dic[doc.id] = doc
-
-    for doc_id in doc_ids:
-        try:
-            if doc_id not in doc_dic:
-                return get_data_error_result(message="Document not found!")
-            doc = doc_dic[doc_id]
-            tenant_id = DocumentService.get_tenant_id(doc_id)
-            if not tenant_id:
-                return get_data_error_result(message="Tenant not found!")
-
-            b, n = File2DocumentService.get_storage_address(doc_id=doc_id)
-
-            if not DocumentService.remove_document(doc, tenant_id):
-                return get_data_error_result(
-                    message="Database error (Document removal)!")
-
-            f2d = File2DocumentService.get_by_document_id(doc_id)
-            FileService.filter_delete([File.source_type == FileSource.KNOWLEDGEBASE, File.id == f2d[0].file_id])
-            File2DocumentService.delete_by_document_id(doc_id)
-
-            STORAGE_IMPL.rm(b, n)
-        except Exception as e:
-            errors += str(e)
-
-    if errors:
-        return get_json_result(data=False, message=errors, code=RetCode.SERVER_ERROR)
-
-    return get_json_result(data=True)
-
-
-@manager.route('/completion_aibotk', methods=['POST'])  # noqa: F821
-@validate_request("Authorization", "conversation_id", "word")
-def completion_faq():
-    import base64
-    req = request.json
-
-    token = req["Authorization"]
-    objs = APIToken.query(token=token)
-    if not objs:
-        return get_json_result(
-            data=False, message='Authentication error: API key is invalid!"', code=RetCode.AUTHENTICATION_ERROR)
-
-    e, conv = API4ConversationService.get_by_id(req["conversation_id"])
-    if not e:
-        return get_data_error_result(message="Conversation not found!")
-    if "quote" not in req:
-        req["quote"] = True
-
-    msg = []
-    msg.append({"role": "user", "content": req["word"]})
-    if not msg[-1].get("id"):
-        msg[-1]["id"] = get_uuid()
-    message_id = msg[-1]["id"]
-
-    def fillin_conv(ans):
-        nonlocal conv, message_id
-        if not conv.reference:
-            conv.reference.append(ans["reference"])
-        else:
-            conv.reference[-1] = ans["reference"]
-        conv.message[-1] = {"role": "assistant", "content": ans["answer"], "id": message_id}
-        ans["id"] = message_id
-
-    try:
-        if conv.source == "agent":
-            conv.message.append(msg[-1])
-            e, cvs = UserCanvasService.get_by_id(conv.dialog_id)
-            if not e:
-                return server_error_response("canvas not found.")
-
-            if not isinstance(cvs.dsl, str):
-                cvs.dsl = json.dumps(cvs.dsl, ensure_ascii=False)
-
-            if not conv.reference:
-                conv.reference = []
-            conv.message.append({"role": "assistant", "content": "", "id": message_id})
-            conv.reference.append({"chunks": [], "doc_aggs": []})
-
-            final_ans = {"reference": [], "doc_aggs": []}
-            canvas = Canvas(cvs.dsl, objs[0].tenant_id)
-
-            canvas.messages.append(msg[-1])
-            canvas.add_user_input(msg[-1]["content"])
-            answer = canvas.run(stream=False)
-
-            assert answer is not None, "Nothing. Is it over?"
-
-            data_type_picture = {
-                "type": 3,
-                "url": "base64 content"
-            }
-            data = [
-                {
-                    "type": 1,
-                    "content": ""
-                }
-            ]
-            final_ans["content"] = "\n".join(answer["content"]) if "content" in answer else ""
-            canvas.messages.append({"role": "assistant", "content": final_ans["content"], "id": message_id})
-            if final_ans.get("reference"):
-                canvas.reference.append(final_ans["reference"])
-            cvs.dsl = json.loads(str(canvas))
-
-            ans = {"answer": final_ans["content"], "reference": final_ans.get("reference", [])}
-            data[0]["content"] += re.sub(r'##\d\$\$', '', ans["answer"])
-            fillin_conv(ans)
-            API4ConversationService.append_message(conv.id, conv.to_dict())
-
-            chunk_idxs = [int(match[2]) for match in re.findall(r'##\d\$\$', ans["answer"])]
-            for chunk_idx in chunk_idxs[:1]:
-                if ans["reference"]["chunks"][chunk_idx]["img_id"]:
-                    try:
-                        bkt, nm = ans["reference"]["chunks"][chunk_idx]["img_id"].split("-")
-                        response = STORAGE_IMPL.get(bkt, nm)
-                        data_type_picture["url"] = base64.b64encode(response).decode('utf-8')
-                        data.append(data_type_picture)
-                        break
-                    except Exception as e:
-                        return server_error_response(e)
-
-            response = {"code": 200, "msg": "success", "data": data}
-            return response
-
-        # ******************For dialog******************
-        conv.message.append(msg[-1])
-        e, dia = DialogService.get_by_id(conv.dialog_id)
-        if not e:
-            return get_data_error_result(message="Dialog not found!")
-        del req["conversation_id"]
-
-        if not conv.reference:
-            conv.reference = []
-        conv.message.append({"role": "assistant", "content": "", "id": message_id})
-        conv.reference.append({"chunks": [], "doc_aggs": []})
-
-        data_type_picture = {
-            "type": 3,
-            "url": "base64 content"
-        }
-        data = [
-            {
-                "type": 1,
-                "content": ""
-            }
-        ]
-        ans = ""
-        for a in chat(dia, msg, stream=False, **req):
-            ans = a
-            break
-        data[0]["content"] += re.sub(r'##\d\$\$', '', ans["answer"])
-        fillin_conv(ans)
-        API4ConversationService.append_message(conv.id, conv.to_dict())
-
-        chunk_idxs = [int(match[2]) for match in re.findall(r'##\d\$\$', ans["answer"])]
-        for chunk_idx in chunk_idxs[:1]:
-            if ans["reference"]["chunks"][chunk_idx]["img_id"]:
-                try:
-                    bkt, nm = ans["reference"]["chunks"][chunk_idx]["img_id"].split("-")
-                    response = STORAGE_IMPL.get(bkt, nm)
-                    data_type_picture["url"] = base64.b64encode(response).decode('utf-8')
-                    data.append(data_type_picture)
-                    break
-                except Exception as e:
-                    return server_error_response(e)
-
-        response = {"code": 200, "msg": "success", "data": data}
-        return response
-
-    except Exception as e:
-        return server_error_response(e)
-
-
-@manager.route('/retrieval', methods=['POST'])  # noqa: F821
-@validate_request("kb_id", "question")
-def retrieval():
-    token = request.headers.get('Authorization').split()[1]
-    objs = APIToken.query(token=token)
-    if not objs:
-        return get_json_result(
-            data=False, message='Authentication error: API key is invalid!"', code=RetCode.AUTHENTICATION_ERROR)
-
-    req = request.json
-    kb_ids = req.get("kb_id", [])
-    doc_ids = req.get("doc_ids", [])
-    question = req.get("question")
-    page = int(req.get("page", 1))
-    size = int(req.get("page_size", 30))
-    similarity_threshold = float(req.get("similarity_threshold", 0.2))
-    vector_similarity_weight = float(req.get("vector_similarity_weight", 0.3))
-    top = int(req.get("top_k", 1024))
-    highlight = bool(req.get("highlight", False))
-
-    try:
-        kbs = KnowledgebaseService.get_by_ids(kb_ids)
-        embd_nms = list(set([kb.embd_id for kb in kbs]))
-        if len(embd_nms) != 1:
-            return get_json_result(
-                data=False, message='Knowledge bases use different embedding models or does not exist."',
-                code=RetCode.AUTHENTICATION_ERROR)
-
-        embd_mdl = LLMBundle(kbs[0].tenant_id, LLMType.EMBEDDING, llm_name=kbs[0].embd_id)
-        rerank_mdl = None
-        if req.get("rerank_id"):
-            rerank_mdl = LLMBundle(kbs[0].tenant_id, LLMType.RERANK, llm_name=req["rerank_id"])
-        if req.get("keyword", False):
-            chat_mdl = LLMBundle(kbs[0].tenant_id, LLMType.CHAT)
-            question += keyword_extraction(chat_mdl, question)
-        ranks = settings.retriever.retrieval(question, embd_mdl, kbs[0].tenant_id, kb_ids, page, size,
-                                               similarity_threshold, vector_similarity_weight, top,
-                                               doc_ids, rerank_mdl=rerank_mdl, highlight= highlight,
-                                               rank_feature=label_question(question, kbs))
-        for c in ranks["chunks"]:
-            c.pop("vector", None)
-        return get_json_result(data=ranks)
-    except Exception as e:
-        if str(e).find("not_found") > 0:
-            return get_json_result(data=False, message='No chunk found! Check the chunk status please!',
-                                   code=RetCode.DATA_ERROR)
-        return server_error_response(e)

api/apps/auth/github.py

@@ -14,7 +14,7 @@
 #  limitations under the License.
 #
 
-import requests
+from common.http_client import async_request, sync_request
 from .oauth import OAuthClient, UserInfo
 
 
@@ -34,24 +34,49 @@ class GithubOAuthClient(OAuthClient):
 
     def fetch_user_info(self, access_token, **kwargs):
         """
-        Fetch github user info.
+        Fetch GitHub user info (synchronous).
         """
         user_info = {}
         try:
             headers = {"Authorization": f"Bearer {access_token}"}
-            # user info
-            response = requests.get(self.userinfo_url, headers=headers, timeout=self.http_request_timeout)
+            response = sync_request("GET", self.userinfo_url, headers=headers, timeout=self.http_request_timeout)
             response.raise_for_status()
             user_info.update(response.json())
-            # email info
-            response = requests.get(self.userinfo_url+"/emails", headers=headers, timeout=self.http_request_timeout)
-            response.raise_for_status()
-            email_info = response.json()
-            user_info["email"] = next(
-                (email for email in email_info if email["primary"]), None
-            )["email"]
+            email_response = sync_request(
+                "GET", self.userinfo_url + "/emails", headers=headers, timeout=self.http_request_timeout
+            )
+            email_response.raise_for_status()
+            email_info = email_response.json()
+            user_info["email"] = next((email for email in email_info if email["primary"]), None)["email"]
             return self.normalize_user_info(user_info)
-        except requests.exceptions.RequestException as e:
+        except Exception as e:
+            raise ValueError(f"Failed to fetch github user info: {e}")
+
+    async def async_fetch_user_info(self, access_token, **kwargs):
+        """Async variant of fetch_user_info using httpx."""
+        user_info = {}
+        headers = {"Authorization": f"Bearer {access_token}"}
+        try:
+            response = await async_request(
+                "GET",
+                self.userinfo_url,
+                headers=headers,
+                timeout=self.http_request_timeout,
+            )
+            response.raise_for_status()
+            user_info.update(response.json())
+
+            email_response = await async_request(
+                "GET",
+                self.userinfo_url + "/emails",
+                headers=headers,
+                timeout=self.http_request_timeout,
+            )
+            email_response.raise_for_status()
+            email_info = email_response.json()
+            user_info["email"] = next((email for email in email_info if email["primary"]), None)["email"]
+            return self.normalize_user_info(user_info)
+        except Exception as e:
             raise ValueError(f"Failed to fetch github user info: {e}")
 
 

api/apps/auth/oauth.py

@@ -14,8 +14,8 @@
 #  limitations under the License.
 #
 
-import requests
 import urllib.parse
+from common.http_client import async_request, sync_request
 
 
 class UserInfo:
@@ -74,15 +74,40 @@ class OAuthClient:
                 "redirect_uri": self.redirect_uri,
                 "grant_type": "authorization_code"
             }
-            response = requests.post(
+            response = sync_request(
+                "POST",
                 self.token_url,
                 data=payload,
                 headers={"Accept": "application/json"},
-                timeout=self.http_request_timeout
+                timeout=self.http_request_timeout,
             )
             response.raise_for_status()
             return response.json()
-        except requests.exceptions.RequestException as e:
+        except Exception as e:
+            raise ValueError(f"Failed to exchange authorization code for token: {e}")
+
+    async def async_exchange_code_for_token(self, code):
+        """
+        Async variant of exchange_code_for_token using httpx.
+        """
+        payload = {
+            "client_id": self.client_id,
+            "client_secret": self.client_secret,
+            "code": code,
+            "redirect_uri": self.redirect_uri,
+            "grant_type": "authorization_code",
+        }
+        try:
+            response = await async_request(
+                "POST",
+                self.token_url,
+                data=payload,
+                headers={"Accept": "application/json"},
+                timeout=self.http_request_timeout,
+            )
+            response.raise_for_status()
+            return response.json()
+        except Exception as e:
             raise ValueError(f"Failed to exchange authorization code for token: {e}")
 
 
@@ -92,11 +117,27 @@ class OAuthClient:
         """
         try:
             headers = {"Authorization": f"Bearer {access_token}"}
-            response = requests.get(self.userinfo_url, headers=headers, timeout=self.http_request_timeout)
+            response = sync_request("GET", self.userinfo_url, headers=headers, timeout=self.http_request_timeout)
             response.raise_for_status()
             user_info = response.json()
             return self.normalize_user_info(user_info)
-        except requests.exceptions.RequestException as e:
+        except Exception as e:
+            raise ValueError(f"Failed to fetch user info: {e}")
+
+    async def async_fetch_user_info(self, access_token, **kwargs):
+        """Async variant of fetch_user_info using httpx."""
+        headers = {"Authorization": f"Bearer {access_token}"}
+        try:
+            response = await async_request(
+                "GET",
+                self.userinfo_url,
+                headers=headers,
+                timeout=self.http_request_timeout,
+            )
+            response.raise_for_status()
+            user_info = response.json()
+            return self.normalize_user_info(user_info)
+        except Exception as e:
             raise ValueError(f"Failed to fetch user info: {e}")
 
 

api/apps/auth/oidc.py

@@ -15,7 +15,7 @@
 #
 
 import jwt
-import requests
+from common.http_client import sync_request
 from .oauth import OAuthClient
 
 
@@ -43,16 +43,17 @@ class OIDCClient(OAuthClient):
         self.jwks_uri = config['jwks_uri']
 
 
-    def _load_oidc_metadata(self, issuer):
+    @staticmethod
+    def _load_oidc_metadata(issuer):
         """
         Load OIDC metadata from `/.well-known/openid-configuration`.
         """
         try:
             metadata_url = f"{issuer}/.well-known/openid-configuration"
-            response = requests.get(metadata_url, timeout=7)
+            response = sync_request("GET", metadata_url, timeout=7)
             response.raise_for_status()
             return response.json()
-        except requests.exceptions.RequestException as e:
+        except Exception as e:
             raise ValueError(f"Failed to fetch OIDC metadata: {e}")
 
 
@@ -94,6 +95,13 @@ class OIDCClient(OAuthClient):
         user_info.update(super().fetch_user_info(access_token).to_dict())
         return self.normalize_user_info(user_info)
 
+    async def async_fetch_user_info(self, access_token, id_token=None, **kwargs):
+        user_info = {}
+        if id_token:
+            user_info = self.parse_id_token(id_token)
+        user_info.update((await super().async_fetch_user_info(access_token)).to_dict())
+        return self.normalize_user_info(user_info)
+
 
     def normalize_user_info(self, user_info):
         return super().normalize_user_info(user_info)

api/apps/canvas_app.py

@@ -13,20 +13,14 @@
 #  See the License for the specific language governing permissions and
 #  limitations under the License.
 #
+import asyncio
+import inspect
 import json
 import logging
-import re
-import sys
 from functools import partial
-
-import flask
-import trio
-from flask import request, Response
-from flask_login import login_required, current_user
-
+from quart import request, Response, make_response
 from agent.component import LLM
-from api import settings
-from api.db import CanvasCategory, FileType
+from api.db import CanvasCategory
 from api.db.services.canvas_service import CanvasTemplateService, UserCanvasService, API4ConversationService
 from api.db.services.document_service import DocumentService
 from api.db.services.file_service import FileService
@@ -36,16 +30,18 @@ from api.db.services.user_service import TenantService
 from api.db.services.user_canvas_version import UserCanvasVersionService
 from common.constants import RetCode
 from common.misc_utils import get_uuid
-from api.utils.api_utils import get_json_result, server_error_response, validate_request, get_data_error_result
+from api.utils.api_utils import get_json_result, server_error_response, validate_request, get_data_error_result, \
+    get_request_json
 from agent.canvas import Canvas
 from peewee import MySQLDatabase, PostgresqlDatabase
 from api.db.db_models import APIToken, Task
 import time
 
-from api.utils.file_utils import filename_type, read_potential_broken_pdf
 from rag.flow.pipeline import Pipeline
 from rag.nlp import search
 from rag.utils.redis_conn import REDIS_CONN
+from common import settings
+from api.apps import login_required, current_user
 
 
 @manager.route('/templates', methods=['GET'])  # noqa: F821
@@ -57,8 +53,9 @@ def templates():
 @manager.route('/rm', methods=['POST'])  # noqa: F821
 @validate_request("canvas_ids")
 @login_required
-def rm():
-    for i in request.json["canvas_ids"]:
+async def rm():
+    req = await get_request_json()
+    for i in req["canvas_ids"]:
         if not UserCanvasService.accessible(i, current_user.id):
             return get_json_result(
                 data=False, message='Only owner of canvas authorized for this operation.',
@@ -70,8 +67,8 @@ def rm():
 @manager.route('/set', methods=['POST'])  # noqa: F821
 @validate_request("dsl", "title")
 @login_required
-def save():
-    req = request.json
+async def save():
+    req = await get_request_json()
     if not isinstance(req["dsl"], str):
         req["dsl"] = json.dumps(req["dsl"], ensure_ascii=False)
     req["dsl"] = json.loads(req["dsl"])
@@ -129,18 +126,18 @@ def getsse(canvas_id):
 @manager.route('/completion', methods=['POST'])  # noqa: F821
 @validate_request("id")
 @login_required
-def run():
-    req = request.json
+async def run():
+    req = await get_request_json()
     query = req.get("query", "")
     files = req.get("files", [])
     inputs = req.get("inputs", {})
     user_id = req.get("user_id", current_user.id)
-    if not UserCanvasService.accessible(req["id"], current_user.id):
+    if not await asyncio.to_thread(UserCanvasService.accessible, req["id"], current_user.id):
         return get_json_result(
             data=False, message='Only owner of canvas authorized for this operation.',
             code=RetCode.OPERATING_ERROR)
 
-    e, cvs = UserCanvasService.get_by_id(req["id"])
+    e, cvs = await asyncio.to_thread(UserCanvasService.get_by_id, req["id"])
     if not e:
         return get_data_error_result(message="canvas not found.")
 
@@ -150,26 +147,28 @@ def run():
     if cvs.canvas_category == CanvasCategory.DataFlow:
         task_id = get_uuid()
         Pipeline(cvs.dsl, tenant_id=current_user.id, doc_id=CANVAS_DEBUG_DOC_ID, task_id=task_id, flow_id=req["id"])
-        ok, error_message = queue_dataflow(tenant_id=user_id, flow_id=req["id"], task_id=task_id, file=files[0], priority=0)
+        ok, error_message = await asyncio.to_thread(queue_dataflow, user_id, req["id"], task_id, CANVAS_DEBUG_DOC_ID, files[0], 0)
         if not ok:
             return get_data_error_result(message=error_message)
         return get_json_result(data={"message_id": task_id})
 
     try:
-        canvas = Canvas(cvs.dsl, current_user.id, req["id"])
+        canvas = Canvas(cvs.dsl, current_user.id, canvas_id=cvs.id)
     except Exception as e:
         return server_error_response(e)
 
-    def sse():
+    async def sse():
         nonlocal canvas, user_id
         try:
-            for ans in canvas.run(query=query, files=files, user_id=user_id, inputs=inputs):
+            async for ans in canvas.run(query=query, files=files, user_id=user_id, inputs=inputs):
                 yield "data:" + json.dumps(ans, ensure_ascii=False) + "\n\n"
 
             cvs.dsl = json.loads(str(canvas))
             UserCanvasService.update_by_id(req["id"], cvs.to_dict())
+
         except Exception as e:
             logging.exception(e)
+            canvas.cancel_task()
             yield "data:" + json.dumps({"code": 500, "message": str(e), "data": False}, ensure_ascii=False) + "\n\n"
 
     resp = Response(sse(), mimetype="text/event-stream")
@@ -177,14 +176,15 @@ def run():
     resp.headers.add_header("Connection", "keep-alive")
     resp.headers.add_header("X-Accel-Buffering", "no")
     resp.headers.add_header("Content-Type", "text/event-stream; charset=utf-8")
+    #resp.call_on_close(lambda: canvas.cancel_task())
     return resp
 
 
 @manager.route('/rerun', methods=['POST'])  # noqa: F821
 @validate_request("id", "dsl", "component_id")
 @login_required
-def rerun():
-    req = request.json
+async def rerun():
+    req = await get_request_json()
     doc = PipelineOperationLogService.get_documents_info(req["id"])
     if not doc:
         return get_data_error_result(message="Document not found.")
@@ -192,7 +192,7 @@ def rerun():
     if 0 < doc["progress"] < 1:
         return get_data_error_result(message=f"`{doc['name']}` is processing...")
 
-    if settings.docStoreConn.indexExist(search.index_name(current_user.id), doc["kb_id"]):
+    if settings.docStoreConn.index_exist(search.index_name(current_user.id), doc["kb_id"]):
         settings.docStoreConn.delete({"doc_id": doc["id"]}, search.index_name(current_user.id), doc["kb_id"])
     doc["progress_msg"] = ""
     doc["chunk_num"] = 0
@@ -221,8 +221,8 @@ def cancel(task_id):
 @manager.route('/reset', methods=['POST'])  # noqa: F821
 @validate_request("id")
 @login_required
-def reset():
-    req = request.json
+async def reset():
+    req = await get_request_json()
     if not UserCanvasService.accessible(req["id"], current_user.id):
         return get_json_result(
             data=False, message='Only owner of canvas authorized for this operation.',
@@ -232,7 +232,7 @@ def reset():
         if not e:
             return get_data_error_result(message="canvas not found.")
 
-        canvas = Canvas(json.dumps(user_canvas.dsl), current_user.id)
+        canvas = Canvas(json.dumps(user_canvas.dsl), current_user.id, canvas_id=user_canvas.id)
         canvas.reset()
         req["dsl"] = json.loads(str(canvas))
         UserCanvasService.update_by_id(req["id"], {"dsl": req["dsl"]})
@@ -242,76 +242,16 @@ def reset():
 
 
 @manager.route("/upload/<canvas_id>", methods=["POST"])  # noqa: F821
-def upload(canvas_id):
+async def upload(canvas_id):
     e, cvs = UserCanvasService.get_by_canvas_id(canvas_id)
     if not e:
         return get_data_error_result(message="canvas not found.")
 
     user_id = cvs["user_id"]
-    def structured(filename, filetype, blob, content_type):
-        nonlocal user_id
-        if filetype == FileType.PDF.value:
-            blob = read_potential_broken_pdf(blob)
-
-        location = get_uuid()
-        FileService.put_blob(user_id, location, blob)
-
-        return {
-            "id": location,
-            "name": filename,
-            "size": sys.getsizeof(blob),
-            "extension": filename.split(".")[-1].lower(),
-            "mime_type": content_type,
-            "created_by": user_id,
-            "created_at": time.time(),
-            "preview_url": None
-        }
-
-    if request.args.get("url"):
-        from crawl4ai import (
-            AsyncWebCrawler,
-            BrowserConfig,
-            CrawlerRunConfig,
-            DefaultMarkdownGenerator,
-            PruningContentFilter,
-            CrawlResult
-        )
-        try:
-            url = request.args.get("url")
-            filename = re.sub(r"\?.*", "", url.split("/")[-1])
-            async def adownload():
-                browser_config = BrowserConfig(
-                    headless=True,
-                    verbose=False,
-                )
-                async with AsyncWebCrawler(config=browser_config) as crawler:
-                    crawler_config = CrawlerRunConfig(
-                        markdown_generator=DefaultMarkdownGenerator(
-                            content_filter=PruningContentFilter()
-                        ),
-                        pdf=True,
-                        screenshot=False
-                    )
-                    result: CrawlResult = await crawler.arun(
-                        url=url,
-                        config=crawler_config
-                    )
-                    return result
-            page = trio.run(adownload())
-            if page.pdf:
-                if filename.split(".")[-1].lower() != "pdf":
-                    filename += ".pdf"
-                return get_json_result(data=structured(filename, "pdf", page.pdf, page.response_headers["content-type"]))
-
-            return get_json_result(data=structured(filename, "html", str(page.markdown).encode("utf-8"), page.response_headers["content-type"], user_id))
-
-        except Exception as e:
-            return  server_error_response(e)
-
-    file = request.files['file']
+    files = await request.files
+    file = files['file'] if files and files.get("file") else None
     try:
-        DocumentService.check_doc_health(user_id, file.filename)
-        return get_json_result(data=structured(file.filename, filename_type(file.filename), file.read(), file.content_type))
+        return get_json_result(data=FileService.upload_info(user_id, file, request.args.get("url")))
     except Exception as e:
         return  server_error_response(e)
 
@@ -330,7 +270,7 @@ def input_form():
                 data=False, message='Only owner of canvas authorized for this operation.',
                 code=RetCode.OPERATING_ERROR)
 
-        canvas = Canvas(json.dumps(user_canvas.dsl), current_user.id)
+        canvas = Canvas(json.dumps(user_canvas.dsl), current_user.id, canvas_id=user_canvas.id)
         return get_json_result(data=canvas.get_component_input_form(cpn_id))
     except Exception as e:
         return server_error_response(e)
@@ -339,15 +279,15 @@ def input_form():
 @manager.route('/debug', methods=['POST'])  # noqa: F821
 @validate_request("id", "component_id", "params")
 @login_required
-def debug():
-    req = request.json
+async def debug():
+    req = await get_request_json()
     if not UserCanvasService.accessible(req["id"], current_user.id):
         return get_json_result(
             data=False, message='Only owner of canvas authorized for this operation.',
             code=RetCode.OPERATING_ERROR)
     try:
         e, user_canvas = UserCanvasService.get_by_id(req["id"])
-        canvas = Canvas(json.dumps(user_canvas.dsl), current_user.id)
+        canvas = Canvas(json.dumps(user_canvas.dsl), current_user.id, canvas_id=user_canvas.id)
         canvas.reset()
         canvas.message_id = get_uuid()
         component = canvas.get_component(req["component_id"])["obj"]
@@ -360,8 +300,13 @@ def debug():
         for k in outputs.keys():
             if isinstance(outputs[k], partial):
                 txt = ""
-                for c in outputs[k]():
-                    txt += c
+                iter_obj = outputs[k]()
+                if inspect.isasyncgen(iter_obj):
+                    async for c in iter_obj:
+                        txt += c
+                else:
+                    for c in iter_obj:
+                        txt += c
                 outputs[k] = txt
         return get_json_result(data=outputs)
     except Exception as e:
@@ -371,8 +316,8 @@ def debug():
 @manager.route('/test_db_connect', methods=['POST'])  # noqa: F821
 @validate_request("db_type", "database", "username", "host", "port", "password")
 @login_required
-def test_db_connect():
-    req = request.json
+async def test_db_connect():
+    req = await get_request_json()
     try:
         if req["db_type"] in ["mysql", "mariadb"]:
             db = MySQLDatabase(req["database"], user=req["username"], host=req["host"], port=req["port"],
@@ -403,39 +348,46 @@ def test_db_connect():
                 f"UID={req['username']};"
                 f"PWD={req['password']};"
             )
-            logging.info(conn_str)
+            redacted_conn_str = (
+                f"DATABASE={req['database']};"
+                f"HOSTNAME={req['host']};"
+                f"PORT={req['port']};"
+                f"PROTOCOL=TCPIP;"
+                f"UID={req['username']};"
+                f"PWD=****;"
+            )
+            logging.info(redacted_conn_str)
             conn = ibm_db.connect(conn_str, "", "")
             stmt = ibm_db.exec_immediate(conn, "SELECT 1 FROM sysibm.sysdummy1")
             ibm_db.fetch_assoc(stmt)
             ibm_db.close(conn)
             return get_json_result(data="Database Connection Successful!")
         elif req["db_type"] == 'trino':
-            def _parse_catalog_schema(db: str):
-                if not db:
+            def _parse_catalog_schema(db_name: str):
+                if not db_name:
                     return None, None
-                if "." in db:
-                    c, s = db.split(".", 1)
-                elif "/" in db:
-                    c, s = db.split("/", 1)
+                if "." in db_name:
+                    catalog_name, schema_name = db_name.split(".", 1)
+                elif "/" in db_name:
+                    catalog_name, schema_name = db_name.split("/", 1)
                 else:
-                    c, s = db, "default"
-                return c, s
+                    catalog_name, schema_name = db_name, "default"
+                return catalog_name, schema_name
             try:
                 import trino
                 import os
-                from trino.auth import BasicAuthentication
-            except Exception:
-                return server_error_response("Missing dependency 'trino'. Please install: pip install trino")
+            except Exception as e:
+                return server_error_response(f"Missing dependency 'trino'. Please install: pip install trino, detail: {e}")
 
             catalog, schema = _parse_catalog_schema(req["database"])
             if not catalog:
                 return server_error_response("For Trino, 'database' must be 'catalog.schema' or at least 'catalog'.")
-            
+
             http_scheme = "https" if os.environ.get("TRINO_USE_TLS", "0") == "1" else "http"
 
             auth = None
             if http_scheme == "https" and req.get("password"):
-                auth = BasicAuthentication(req.get("username") or "ragflow", req["password"])
+                auth = trino.BasicAuthentication(req.get("username") or "ragflow", req["password"])
 
             conn = trino.dbapi.connect(
                 host=req["host"],
@@ -468,8 +420,8 @@ def test_db_connect():
 @login_required
 def getlistversion(canvas_id):
     try:
-        list =sorted([c.to_dict() for c in UserCanvasVersionService.list_by_canvas_id(canvas_id)], key=lambda x: x["update_time"]*-1)
-        return get_json_result(data=list)
+        versions =sorted([c.to_dict() for c in UserCanvasVersionService.list_by_canvas_id(canvas_id)], key=lambda x: x["update_time"]*-1)
+        return get_json_result(data=versions)
     except Exception as e:
         return get_data_error_result(message=f"Error getting history files: {e}")
 
@@ -479,7 +431,6 @@ def getlistversion(canvas_id):
 @login_required
 def getversion( version_id):
     try:
-
         e, version = UserCanvasVersionService.get_by_id(version_id)
         if version:
             return get_json_result(data=version.to_dict())
@@ -518,8 +469,8 @@ def list_canvas():
 @manager.route('/setting', methods=['POST'])  # noqa: F821
 @validate_request("id", "title", "permission")
 @login_required
-def setting():
-    req = request.json
+async def setting():
+    req = await get_request_json()
     req["user_id"] = current_user.id
 
     if not UserCanvasService.accessible(req["id"], current_user.id):
@@ -546,11 +497,11 @@ def trace():
     cvs_id = request.args.get("canvas_id")
     msg_id = request.args.get("message_id")
     try:
-        bin = REDIS_CONN.get(f"{cvs_id}-{msg_id}-logs")
-        if not bin:
+        binary = REDIS_CONN.get(f"{cvs_id}-{msg_id}-logs")
+        if not binary:
             return get_json_result(data={})
 
-        return get_json_result(data=json.loads(bin.encode("utf-8")))
+        return get_json_result(data=json.loads(binary.encode("utf-8")))
     except Exception as e:
         logging.exception(e)
 
@@ -600,8 +551,8 @@ def prompts():
 
 
 @manager.route('/download', methods=['GET'])  # noqa: F821
-def download():
+async def download():
     id = request.args.get("id")
     created_by = request.args.get("created_by")
     blob = FileService.get_blob(created_by, id)
-    return flask.make_response(blob)
+    return await make_response(blob)

api/apps/chunk_app.py

@@ -13,36 +13,37 @@
 #  See the License for the specific language governing permissions and
 #  limitations under the License.
 #
+import asyncio
 import datetime
 import json
 import re
-
+import base64
 import xxhash
-from flask import request
-from flask_login import current_user, login_required
+from quart import request
 
-from api import settings
-from api.db.services.dialog_service import meta_filter
 from api.db.services.document_service import DocumentService
 from api.db.services.knowledgebase_service import KnowledgebaseService
 from api.db.services.llm_service import LLMBundle
+from common.metadata_utils import apply_meta_data_filter
 from api.db.services.search_service import SearchService
 from api.db.services.user_service import UserTenantService
-from api.utils.api_utils import get_data_error_result, get_json_result, server_error_response, validate_request
+from api.utils.api_utils import get_data_error_result, get_json_result, server_error_response, validate_request, \
+    get_request_json
 from rag.app.qa import beAdoc, rmPrefix
 from rag.app.tag import label_question
 from rag.nlp import rag_tokenizer, search
-from rag.prompts.generator import gen_meta_filter, cross_languages, keyword_extraction
-from rag.settings import PAGERANK_FLD
+from rag.prompts.generator import cross_languages, keyword_extraction
 from common.string_utils import remove_redundant_spaces
-from common.constants import RetCode, LLMType, ParserType
+from common.constants import RetCode, LLMType, ParserType, PAGERANK_FLD
+from common import settings
+from api.apps import login_required, current_user
 
 
 @manager.route('/list', methods=['POST'])  # noqa: F821
 @login_required
 @validate_request("doc_id")
-def list_chunk():
-    req = request.json
+async def list_chunk():
+    req = await get_request_json()
     doc_id = req["doc_id"]
     page = int(req.get("page", 1))
     size = int(req.get("size", 30))
@@ -75,6 +76,7 @@ def list_chunk():
                 "image_id": sres.field[id].get("img_id", ""),
                 "available_int": int(sres.field[id].get("available_int", 1)),
                 "positions": sres.field[id].get("position_int", []),
+                "doc_type_kwd": sres.field[id].get("doc_type_kwd")
             }
             assert isinstance(d["positions"], list)
             assert len(d["positions"]) == 0 or (isinstance(d["positions"][0], list) and len(d["positions"][0]) == 5)
@@ -122,8 +124,8 @@ def get():
 @manager.route('/set', methods=['POST'])  # noqa: F821
 @login_required
 @validate_request("doc_id", "chunk_id", "content_with_weight")
-def set():
-    req = request.json
+async def set():
+    req = await get_request_json()
     d = {
         "id": req["chunk_id"],
         "content_with_weight": req["content_with_weight"]}
@@ -147,31 +149,42 @@ def set():
         d["available_int"] = req["available_int"]
 
     try:
-        tenant_id = DocumentService.get_tenant_id(req["doc_id"])
-        if not tenant_id:
-            return get_data_error_result(message="Tenant not found!")
+        def _set_sync():
+            tenant_id = DocumentService.get_tenant_id(req["doc_id"])
+            if not tenant_id:
+                return get_data_error_result(message="Tenant not found!")
 
-        embd_id = DocumentService.get_embd_id(req["doc_id"])
-        embd_mdl = LLMBundle(tenant_id, LLMType.EMBEDDING, embd_id)
+            embd_id = DocumentService.get_embd_id(req["doc_id"])
+            embd_mdl = LLMBundle(tenant_id, LLMType.EMBEDDING, embd_id)
 
-        e, doc = DocumentService.get_by_id(req["doc_id"])
-        if not e:
-            return get_data_error_result(message="Document not found!")
+            e, doc = DocumentService.get_by_id(req["doc_id"])
+            if not e:
+                return get_data_error_result(message="Document not found!")
 
-        if doc.parser_id == ParserType.QA:
-            arr = [
-                t for t in re.split(
-                    r"[\n\t]",
-                    req["content_with_weight"]) if len(t) > 1]
-            q, a = rmPrefix(arr[0]), rmPrefix("\n".join(arr[1:]))
-            d = beAdoc(d, q, a, not any(
-                [rag_tokenizer.is_chinese(t) for t in q + a]))
+            _d = d
+            if doc.parser_id == ParserType.QA:
+                arr = [
+                    t for t in re.split(
+                        r"[\n\t]",
+                        req["content_with_weight"]) if len(t) > 1]
+                q, a = rmPrefix(arr[0]), rmPrefix("\n".join(arr[1:]))
+                _d = beAdoc(d, q, a, not any(
+                    [rag_tokenizer.is_chinese(t) for t in q + a]))
 
-        v, c = embd_mdl.encode([doc.name, req["content_with_weight"] if not d.get("question_kwd") else "\n".join(d["question_kwd"])])
-        v = 0.1 * v[0] + 0.9 * v[1] if doc.parser_id != ParserType.QA else v[1]
-        d["q_%d_vec" % len(v)] = v.tolist()
-        settings.docStoreConn.update({"id": req["chunk_id"]}, d, search.index_name(tenant_id), doc.kb_id)
-        return get_json_result(data=True)
+            v, c = embd_mdl.encode([doc.name, req["content_with_weight"] if not _d.get("question_kwd") else "\n".join(_d["question_kwd"])])
+            v = 0.1 * v[0] + 0.9 * v[1] if doc.parser_id != ParserType.QA else v[1]
+            _d["q_%d_vec" % len(v)] = v.tolist()
+            settings.docStoreConn.update({"id": req["chunk_id"]}, _d, search.index_name(tenant_id), doc.kb_id)
+
+            # update image
+            image_base64 = req.get("image_base64", None)
+            if image_base64:
+                bkt, name = req.get("img_id", "-").split("-")
+                image_binary = base64.b64decode(image_base64)
+                settings.STORAGE_IMPL.put(bkt, name, image_binary)
+            return get_json_result(data=True)
+
+        return await asyncio.to_thread(_set_sync)
     except Exception as e:
         return server_error_response(e)
 
@@ -179,19 +192,22 @@ def set():
 @manager.route('/switch', methods=['POST'])  # noqa: F821
 @login_required
 @validate_request("chunk_ids", "available_int", "doc_id")
-def switch():
-    req = request.json
+async def switch():
+    req = await get_request_json()
     try:
-        e, doc = DocumentService.get_by_id(req["doc_id"])
-        if not e:
-            return get_data_error_result(message="Document not found!")
-        for cid in req["chunk_ids"]:
-            if not settings.docStoreConn.update({"id": cid},
-                                                {"available_int": int(req["available_int"])},
-                                                search.index_name(DocumentService.get_tenant_id(req["doc_id"])),
-                                                doc.kb_id):
-                return get_data_error_result(message="Index updating failure")
-        return get_json_result(data=True)
+        def _switch_sync():
+            e, doc = DocumentService.get_by_id(req["doc_id"])
+            if not e:
+                return get_data_error_result(message="Document not found!")
+            for cid in req["chunk_ids"]:
+                if not settings.docStoreConn.update({"id": cid},
+                                                    {"available_int": int(req["available_int"])},
+                                                    search.index_name(DocumentService.get_tenant_id(req["doc_id"])),
+                                                    doc.kb_id):
+                    return get_data_error_result(message="Index updating failure")
+            return get_json_result(data=True)
+
+        return await asyncio.to_thread(_switch_sync)
     except Exception as e:
         return server_error_response(e)
 
@@ -199,24 +215,26 @@ def switch():
 @manager.route('/rm', methods=['POST'])  # noqa: F821
 @login_required
 @validate_request("chunk_ids", "doc_id")
-def rm():
-    from rag.utils.storage_factory import STORAGE_IMPL
-    req = request.json
+async def rm():
+    req = await get_request_json()
     try:
-        e, doc = DocumentService.get_by_id(req["doc_id"])
-        if not e:
-            return get_data_error_result(message="Document not found!")
-        if not settings.docStoreConn.delete({"id": req["chunk_ids"]},
-                                            search.index_name(DocumentService.get_tenant_id(req["doc_id"])),
-                                            doc.kb_id):
-            return get_data_error_result(message="Chunk deleting failure")
-        deleted_chunk_ids = req["chunk_ids"]
-        chunk_number = len(deleted_chunk_ids)
-        DocumentService.decrement_chunk_num(doc.id, doc.kb_id, 1, chunk_number, 0)
-        for cid in deleted_chunk_ids:
-            if STORAGE_IMPL.obj_exist(doc.kb_id, cid):
-                STORAGE_IMPL.rm(doc.kb_id, cid)
-        return get_json_result(data=True)
+        def _rm_sync():
+            e, doc = DocumentService.get_by_id(req["doc_id"])
+            if not e:
+                return get_data_error_result(message="Document not found!")
+            if not settings.docStoreConn.delete({"id": req["chunk_ids"]},
+                                                search.index_name(DocumentService.get_tenant_id(req["doc_id"])),
+                                                doc.kb_id):
+                return get_data_error_result(message="Chunk deleting failure")
+            deleted_chunk_ids = req["chunk_ids"]
+            chunk_number = len(deleted_chunk_ids)
+            DocumentService.decrement_chunk_num(doc.id, doc.kb_id, 1, chunk_number, 0)
+            for cid in deleted_chunk_ids:
+                if settings.STORAGE_IMPL.obj_exist(doc.kb_id, cid):
+                    settings.STORAGE_IMPL.rm(doc.kb_id, cid)
+            return get_json_result(data=True)
+
+        return await asyncio.to_thread(_rm_sync)
     except Exception as e:
         return server_error_response(e)
 
@@ -224,8 +242,8 @@ def rm():
 @manager.route('/create', methods=['POST'])  # noqa: F821
 @login_required
 @validate_request("doc_id", "content_with_weight")
-def create():
-    req = request.json
+async def create():
+    req = await get_request_json()
     chunck_id = xxhash.xxh64((req["content_with_weight"] + req["doc_id"]).encode("utf-8")).hexdigest()
     d = {"id": chunck_id, "content_ltks": rag_tokenizer.tokenize(req["content_with_weight"]),
          "content_with_weight": req["content_with_weight"]}
@@ -246,35 +264,38 @@ def create():
         d["tag_feas"] = req["tag_feas"]
 
     try:
-        e, doc = DocumentService.get_by_id(req["doc_id"])
-        if not e:
-            return get_data_error_result(message="Document not found!")
-        d["kb_id"] = [doc.kb_id]
-        d["docnm_kwd"] = doc.name
-        d["title_tks"] = rag_tokenizer.tokenize(doc.name)
-        d["doc_id"] = doc.id
+        def _create_sync():
+            e, doc = DocumentService.get_by_id(req["doc_id"])
+            if not e:
+                return get_data_error_result(message="Document not found!")
+            d["kb_id"] = [doc.kb_id]
+            d["docnm_kwd"] = doc.name
+            d["title_tks"] = rag_tokenizer.tokenize(doc.name)
+            d["doc_id"] = doc.id
 
-        tenant_id = DocumentService.get_tenant_id(req["doc_id"])
-        if not tenant_id:
-            return get_data_error_result(message="Tenant not found!")
+            tenant_id = DocumentService.get_tenant_id(req["doc_id"])
+            if not tenant_id:
+                return get_data_error_result(message="Tenant not found!")
 
-        e, kb = KnowledgebaseService.get_by_id(doc.kb_id)
-        if not e:
-            return get_data_error_result(message="Knowledgebase not found!")
-        if kb.pagerank:
-            d[PAGERANK_FLD] = kb.pagerank
+            e, kb = KnowledgebaseService.get_by_id(doc.kb_id)
+            if not e:
+                return get_data_error_result(message="Knowledgebase not found!")
+            if kb.pagerank:
+                d[PAGERANK_FLD] = kb.pagerank
 
-        embd_id = DocumentService.get_embd_id(req["doc_id"])
-        embd_mdl = LLMBundle(tenant_id, LLMType.EMBEDDING.value, embd_id)
+            embd_id = DocumentService.get_embd_id(req["doc_id"])
+            embd_mdl = LLMBundle(tenant_id, LLMType.EMBEDDING.value, embd_id)
 
-        v, c = embd_mdl.encode([doc.name, req["content_with_weight"] if not d["question_kwd"] else "\n".join(d["question_kwd"])])
-        v = 0.1 * v[0] + 0.9 * v[1]
-        d["q_%d_vec" % len(v)] = v.tolist()
-        settings.docStoreConn.insert([d], search.index_name(tenant_id), doc.kb_id)
+            v, c = embd_mdl.encode([doc.name, req["content_with_weight"] if not d["question_kwd"] else "\n".join(d["question_kwd"])])
+            v = 0.1 * v[0] + 0.9 * v[1]
+            d["q_%d_vec" % len(v)] = v.tolist()
+            settings.docStoreConn.insert([d], search.index_name(tenant_id), doc.kb_id)
 
-        DocumentService.increment_chunk_num(
-            doc.id, doc.kb_id, c, 1, 0)
-        return get_json_result(data={"chunk_id": chunck_id})
+            DocumentService.increment_chunk_num(
+                doc.id, doc.kb_id, c, 1, 0)
+            return get_json_result(data={"chunk_id": chunck_id})
+
+        return await asyncio.to_thread(_create_sync)
     except Exception as e:
         return server_error_response(e)
 
@@ -282,8 +303,8 @@ def create():
 @manager.route('/retrieval_test', methods=['POST'])  # noqa: F821
 @login_required
 @validate_request("kb_id", "question")
-def retrieval_test():
-    req = request.json
+async def retrieval_test():
+    req = await get_request_json()
     page = int(req.get("page", 1))
     size = int(req.get("size", 30))
     question = req["question"]
@@ -298,25 +319,29 @@ def retrieval_test():
     use_kg = req.get("use_kg", False)
     top = int(req.get("top_k", 1024))
     langs = req.get("cross_languages", [])
-    tenant_ids = []
+    user_id = current_user.id
 
-    if req.get("search_id", ""):
-        search_config = SearchService.get_detail(req.get("search_id", "")).get("search_config", {})
-        meta_data_filter = search_config.get("meta_data_filter", {})
-        metas = DocumentService.get_meta_by_kbs(kb_ids)
-        if meta_data_filter.get("method") == "auto":
-            chat_mdl = LLMBundle(current_user.id, LLMType.CHAT, llm_name=search_config.get("chat_id", ""))
-            filters = gen_meta_filter(chat_mdl, metas, question)
-            doc_ids.extend(meta_filter(metas, filters))
-            if not doc_ids:
-                doc_ids = None
-        elif meta_data_filter.get("method") == "manual":
-            doc_ids.extend(meta_filter(metas, meta_data_filter["manual"]))
-            if not doc_ids:
-                doc_ids = None
+    async def _retrieval():
+        local_doc_ids = list(doc_ids) if doc_ids else []
+        tenant_ids = []
 
-    try:
-        tenants = UserTenantService.query(user_id=current_user.id)
+        meta_data_filter = {}
+        chat_mdl = None
+        if req.get("search_id", ""):
+            search_config = SearchService.get_detail(req.get("search_id", "")).get("search_config", {})
+            meta_data_filter = search_config.get("meta_data_filter", {})
+            if meta_data_filter.get("method") in ["auto", "semi_auto"]:
+                chat_mdl = LLMBundle(user_id, LLMType.CHAT, llm_name=search_config.get("chat_id", ""))
+        else:
+            meta_data_filter = req.get("meta_data_filter") or {}
+            if meta_data_filter.get("method") in ["auto", "semi_auto"]:
+                chat_mdl = LLMBundle(user_id, LLMType.CHAT)
+
+        if meta_data_filter:
+            metas = DocumentService.get_meta_by_kbs(kb_ids)
+            local_doc_ids = await apply_meta_data_filter(meta_data_filter, metas, question, chat_mdl, local_doc_ids)
+
+        tenants = UserTenantService.query(user_id=user_id)
         for kb_id in kb_ids:
             for tenant in tenants:
                 if KnowledgebaseService.query(
@@ -325,15 +350,16 @@ def retrieval_test():
                     break
             else:
                 return get_json_result(
-                    data=False, message='Only owner of knowledgebase authorized for this operation.',
+                    data=False, message='Only owner of dataset authorized for this operation.',
                     code=RetCode.OPERATING_ERROR)
 
         e, kb = KnowledgebaseService.get_by_id(kb_ids[0])
         if not e:
             return get_data_error_result(message="Knowledgebase not found!")
 
+        _question = question
         if langs:
-            question = cross_languages(kb.tenant_id, None, question, langs)
+            _question = await cross_languages(kb.tenant_id, None, _question, langs)
 
         embd_mdl = LLMBundle(kb.tenant_id, LLMType.EMBEDDING.value, llm_name=kb.embd_id)
 
@@ -343,31 +369,35 @@ def retrieval_test():
 
         if req.get("keyword", False):
             chat_mdl = LLMBundle(kb.tenant_id, LLMType.CHAT)
-            question += keyword_extraction(chat_mdl, question)
+            _question += await keyword_extraction(chat_mdl, _question)
 
-        labels = label_question(question, [kb])
-        ranks = settings.retriever.retrieval(question, embd_mdl, tenant_ids, kb_ids, page, size,
+        labels = label_question(_question, [kb])
+        ranks = settings.retriever.retrieval(_question, embd_mdl, tenant_ids, kb_ids, page, size,
                                float(req.get("similarity_threshold", 0.0)),
                                float(req.get("vector_similarity_weight", 0.3)),
                                top,
-                               doc_ids, rerank_mdl=rerank_mdl,
+                               local_doc_ids, rerank_mdl=rerank_mdl,
                                              highlight=req.get("highlight", False),
                                rank_feature=labels
                                )
         if use_kg:
-            ck = settings.kg_retriever.retrieval(question,
+            ck = settings.kg_retriever.retrieval(_question,
                                                    tenant_ids,
                                                    kb_ids,
                                                    embd_mdl,
                                                    LLMBundle(kb.tenant_id, LLMType.CHAT))
             if ck["content_with_weight"]:
                 ranks["chunks"].insert(0, ck)
+        ranks["chunks"] = settings.retriever.retrieval_by_children(ranks["chunks"], tenant_ids)
 
         for c in ranks["chunks"]:
             c.pop("vector", None)
         ranks["labels"] = labels
 
         return get_json_result(data=ranks)
+
+    try:
+        return await _retrieval()
     except Exception as e:
         if str(e).find("not_found") > 0:
             return get_json_result(data=False, message='No chunk found! Check the chunk status please!',

api/apps/connector_app.py

@@ -13,21 +13,33 @@
 #  See the License for the specific language governing permissions and
 #  limitations under the License.
 #
+import asyncio
+import json
+import logging
 import time
+import uuid
+from html import escape
+from typing import Any
 
-from flask import request
-from flask_login import login_required, current_user
+from quart import request, make_response
+from google_auth_oauthlib.flow import Flow
 
 from api.db import InputType
-from api.db.services.connector_service import ConnectorService, Connector2KbService, SyncLogsService
-from api.utils.api_utils import get_json_result, validate_request, get_data_error_result
-from common.misc_utils import get_uuid
+from api.db.services.connector_service import ConnectorService, SyncLogsService
+from api.utils.api_utils import get_data_error_result, get_json_result, get_request_json, validate_request
 from common.constants import RetCode, TaskStatus
+from common.data_source.config import GOOGLE_DRIVE_WEB_OAUTH_REDIRECT_URI, GMAIL_WEB_OAUTH_REDIRECT_URI, BOX_WEB_OAUTH_REDIRECT_URI, DocumentSource
+from common.data_source.google_util.constant import WEB_OAUTH_POPUP_TEMPLATE, GOOGLE_SCOPES
+from common.misc_utils import get_uuid
+from rag.utils.redis_conn import REDIS_CONN
+from api.apps import login_required, current_user
+from box_sdk_gen import BoxOAuth, OAuthConfig, GetAuthorizeUrlOptions
+
 
 @manager.route("/set", methods=["POST"])  # noqa: F821
 @login_required
-def set_connector():
-    req = request.json
+async def set_connector():
+    req = await get_request_json()
     if req.get("id"):
         conn = {fld: req[fld] for fld in ["prune_freq", "refresh_freq", "config", "timeout_secs"] if fld in req}
         ConnectorService.update_by_id(req["id"], conn)
@@ -42,13 +54,12 @@ def set_connector():
             "config": req["config"],
             "refresh_freq": int(req.get("refresh_freq", 30)),
             "prune_freq": int(req.get("prune_freq", 720)),
-            "timeout_secs": int(req.get("timeout_secs", 60*29)),
-            "status": TaskStatus.SCHEDULE
+            "timeout_secs": int(req.get("timeout_secs", 60 * 29)),
+            "status": TaskStatus.SCHEDULE,
         }
-        conn["status"] = TaskStatus.SCHEDULE
+        ConnectorService.save(**conn)
 
-    ConnectorService.save(**conn)
-    time.sleep(1)
+    await asyncio.sleep(1)
     e, conn = ConnectorService.get_by_id(req["id"])
 
     return get_json_result(data=conn.to_dict())
@@ -73,13 +84,14 @@ def get_connector(connector_id):
 @login_required
 def list_logs(connector_id):
     req = request.args.to_dict(flat=True)
-    return get_json_result(data=SyncLogsService.list_sync_tasks(connector_id, int(req.get("page", 1)), int(req.get("page_size", 15))))
+    arr, total = SyncLogsService.list_sync_tasks(connector_id, int(req.get("page", 1)), int(req.get("page_size", 15)))
+    return get_json_result(data={"total": total, "logs": arr})
 
 
 @manager.route("/<connector_id>/resume", methods=["PUT"])  # noqa: F821
 @login_required
-def resume(connector_id):
-    req = request.json
+async def resume(connector_id):
+    req = await get_request_json()
     if req.get("resume"):
         ConnectorService.resume(connector_id, TaskStatus.SCHEDULE)
     else:
@@ -87,14 +99,14 @@ def resume(connector_id):
     return get_json_result(data=True)
 
 
-@manager.route("/<connector_id>/link", methods=["POST"])  # noqa: F821
-@validate_request("kb_ids")
+@manager.route("/<connector_id>/rebuild", methods=["PUT"])  # noqa: F821
 @login_required
-def link_kb(connector_id):
-    req = request.json
-    errors = Connector2KbService.link_kb(connector_id, req["kb_ids"], current_user.id)
-    if errors:
-        return get_json_result(data=False, message=errors, code=RetCode.SERVER_ERROR)
+@validate_request("kb_id")
+async def rebuild(connector_id):
+    req = await get_request_json()
+    err = ConnectorService.rebuild(req["kb_id"], connector_id, current_user.id)
+    if err:
+        return get_json_result(data=False, message=err, code=RetCode.SERVER_ERROR)
     return get_json_result(data=True)
 
 
@@ -103,4 +115,369 @@ def link_kb(connector_id):
 def rm_connector(connector_id):
     ConnectorService.resume(connector_id, TaskStatus.CANCEL)
     ConnectorService.delete_by_id(connector_id)
-    return get_json_result(data=True)
+    return get_json_result(data=True)
+
+
+WEB_FLOW_TTL_SECS = 15 * 60
+
+
+def _web_state_cache_key(flow_id: str, source_type: str | None = None) -> str:
+    """Return Redis key for web OAuth state.
+
+    The default prefix keeps backward compatibility for Google Drive.
+    When source_type == "gmail", a different prefix is used so that
+    Drive/Gmail flows don't clash in Redis.
+    """
+    prefix = f"{source_type}_web_flow_state"
+    return f"{prefix}:{flow_id}"
+
+
+def _web_result_cache_key(flow_id: str, source_type: str | None = None) -> str:
+    """Return Redis key for web OAuth result.
+
+    Mirrors _web_state_cache_key logic for result storage.
+    """
+    prefix = f"{source_type}_web_flow_result"
+    return f"{prefix}:{flow_id}"
+
+
+def _load_credentials(payload: str | dict[str, Any]) -> dict[str, Any]:
+    if isinstance(payload, dict):
+        return payload
+    try:
+        return json.loads(payload)
+    except json.JSONDecodeError as exc:  # pragma: no cover - defensive
+        raise ValueError("Invalid Google credentials JSON.") from exc
+
+
+def _get_web_client_config(credentials: dict[str, Any]) -> dict[str, Any]:
+    web_section = credentials.get("web")
+    if not isinstance(web_section, dict):
+        raise ValueError("Google OAuth JSON must include a 'web' client configuration to use browser-based authorization.")
+    return {"web": web_section}
+
+
+async def _render_web_oauth_popup(flow_id: str, success: bool, message: str, source="drive"):
+    status = "success" if success else "error"
+    auto_close = "window.close();" if success else ""
+    escaped_message = escape(message)
+    #   Drive: ragflow-google-drive-oauth
+    #   Gmail: ragflow-gmail-oauth
+    payload_type = f"ragflow-{source}-oauth"
+    payload_json = json.dumps(
+        {
+            "type": payload_type,
+            "status": status,
+            "flowId": flow_id or "",
+            "message": message,
+        }
+    )
+    # TODO(google-oauth): title/heading/message may need to reflect drive/gmail based on cached type
+    html = WEB_OAUTH_POPUP_TEMPLATE.format(
+        title=f"Google {source.capitalize()} Authorization",
+        heading="Authorization complete" if success else "Authorization failed",
+        message=escaped_message,
+        payload_json=payload_json,
+        auto_close=auto_close,
+    )
+    response = await make_response(html, 200)
+    response.headers["Content-Type"] = "text/html; charset=utf-8"
+    return response
+
+
+@manager.route("/google/oauth/web/start", methods=["POST"])  # noqa: F821
+@login_required
+@validate_request("credentials")
+async def start_google_web_oauth():
+    source = request.args.get("type", "google-drive")
+    if source not in ("google-drive", "gmail"):
+        return get_json_result(code=RetCode.ARGUMENT_ERROR, message="Invalid Google OAuth type.")
+
+    if source == "gmail":
+        redirect_uri = GMAIL_WEB_OAUTH_REDIRECT_URI
+        scopes = GOOGLE_SCOPES[DocumentSource.GMAIL]
+    else:
+        redirect_uri = GOOGLE_DRIVE_WEB_OAUTH_REDIRECT_URI
+        scopes = GOOGLE_SCOPES[DocumentSource.GOOGLE_DRIVE]
+
+    if not redirect_uri:
+        return get_json_result(
+            code=RetCode.SERVER_ERROR,
+            message="Google OAuth redirect URI is not configured on the server.",
+        )
+
+    req = await get_request_json()
+    raw_credentials = req.get("credentials", "")
+
+    try:
+        credentials = _load_credentials(raw_credentials)
+        print(credentials)
+    except ValueError as exc:
+        return get_json_result(code=RetCode.ARGUMENT_ERROR, message=str(exc))
+
+    if credentials.get("refresh_token"):
+        return get_json_result(
+            code=RetCode.ARGUMENT_ERROR,
+            message="Uploaded credentials already include a refresh token.",
+        )
+
+    try:
+        client_config = _get_web_client_config(credentials)
+    except ValueError as exc:
+        return get_json_result(code=RetCode.ARGUMENT_ERROR, message=str(exc))
+
+    flow_id = str(uuid.uuid4())
+    try:
+        flow = Flow.from_client_config(client_config, scopes=scopes)
+        flow.redirect_uri = redirect_uri
+        authorization_url, _ = flow.authorization_url(
+            access_type="offline",
+            include_granted_scopes="true",
+            prompt="consent",
+            state=flow_id,
+        )
+    except Exception as exc:  # pragma: no cover - defensive
+        logging.exception("Failed to create Google OAuth flow: %s", exc)
+        return get_json_result(
+            code=RetCode.SERVER_ERROR,
+            message="Failed to initialize Google OAuth flow. Please verify the uploaded client configuration.",
+        )
+
+    cache_payload = {
+        "user_id": current_user.id,
+        "client_config": client_config,
+        "created_at": int(time.time()),
+    }
+    REDIS_CONN.set_obj(_web_state_cache_key(flow_id, source), cache_payload, WEB_FLOW_TTL_SECS)
+
+    return get_json_result(
+        data={
+            "flow_id": flow_id,
+            "authorization_url": authorization_url,
+            "expires_in": WEB_FLOW_TTL_SECS,
+        }
+    )
+
+
+@manager.route("/gmail/oauth/web/callback", methods=["GET"])  # noqa: F821
+async def google_gmail_web_oauth_callback():
+    state_id = request.args.get("state")
+    error = request.args.get("error")
+    source = "gmail"
+
+    error_description = request.args.get("error_description") or error
+
+    if not state_id:
+        return await _render_web_oauth_popup("", False, "Missing OAuth state parameter.", source)
+
+    state_cache = REDIS_CONN.get(_web_state_cache_key(state_id, source))
+    if not state_cache:
+        return await _render_web_oauth_popup(state_id, False, "Authorization session expired. Please restart from the main window.", source)
+
+    state_obj = json.loads(state_cache)
+    client_config = state_obj.get("client_config")
+    if not client_config:
+        REDIS_CONN.delete(_web_state_cache_key(state_id, source))
+        return await _render_web_oauth_popup(state_id, False, "Authorization session was invalid. Please retry.", source)
+
+    if error:
+        REDIS_CONN.delete(_web_state_cache_key(state_id, source))
+        return await _render_web_oauth_popup(state_id, False, error_description or "Authorization was cancelled.", source)
+
+    code = request.args.get("code")
+    if not code:
+        return await _render_web_oauth_popup(state_id, False, "Missing authorization code from Google.", source)
+
+    try:
+        # TODO(google-oauth): branch scopes/redirect_uri based on source_type (drive vs gmail)
+        flow = Flow.from_client_config(client_config, scopes=GOOGLE_SCOPES[DocumentSource.GMAIL])
+        flow.redirect_uri = GMAIL_WEB_OAUTH_REDIRECT_URI
+        flow.fetch_token(code=code)
+    except Exception as exc:  # pragma: no cover - defensive
+        logging.exception("Failed to exchange Google OAuth code: %s", exc)
+        REDIS_CONN.delete(_web_state_cache_key(state_id, source))
+        return await _render_web_oauth_popup(state_id, False, "Failed to exchange tokens with Google. Please retry.", source)
+
+    creds_json = flow.credentials.to_json()
+    result_payload = {
+        "user_id": state_obj.get("user_id"),
+        "credentials": creds_json,
+    }
+    REDIS_CONN.set_obj(_web_result_cache_key(state_id, source), result_payload, WEB_FLOW_TTL_SECS)
+    REDIS_CONN.delete(_web_state_cache_key(state_id, source))
+
+    return await _render_web_oauth_popup(state_id, True, "Authorization completed successfully.", source)
+
+
+@manager.route("/google-drive/oauth/web/callback", methods=["GET"])  # noqa: F821
+async def google_drive_web_oauth_callback():
+    state_id = request.args.get("state")
+    error = request.args.get("error")
+    source = "google-drive"
+
+    error_description = request.args.get("error_description") or error
+
+    if not state_id:
+        return await _render_web_oauth_popup("", False, "Missing OAuth state parameter.", source)
+
+    state_cache = REDIS_CONN.get(_web_state_cache_key(state_id, source))
+    if not state_cache:
+        return await _render_web_oauth_popup(state_id, False, "Authorization session expired. Please restart from the main window.", source)
+
+    state_obj = json.loads(state_cache)
+    client_config = state_obj.get("client_config")
+    if not client_config:
+        REDIS_CONN.delete(_web_state_cache_key(state_id, source))
+        return await _render_web_oauth_popup(state_id, False, "Authorization session was invalid. Please retry.", source)
+
+    if error:
+        REDIS_CONN.delete(_web_state_cache_key(state_id, source))
+        return await _render_web_oauth_popup(state_id, False, error_description or "Authorization was cancelled.", source)
+
+    code = request.args.get("code")
+    if not code:
+        return await _render_web_oauth_popup(state_id, False, "Missing authorization code from Google.", source)
+
+    try:
+        # TODO(google-oauth): branch scopes/redirect_uri based on source_type (drive vs gmail)
+        flow = Flow.from_client_config(client_config, scopes=GOOGLE_SCOPES[DocumentSource.GOOGLE_DRIVE])
+        flow.redirect_uri = GOOGLE_DRIVE_WEB_OAUTH_REDIRECT_URI
+        flow.fetch_token(code=code)
+    except Exception as exc:  # pragma: no cover - defensive
+        logging.exception("Failed to exchange Google OAuth code: %s", exc)
+        REDIS_CONN.delete(_web_state_cache_key(state_id, source))
+        return await _render_web_oauth_popup(state_id, False, "Failed to exchange tokens with Google. Please retry.", source)
+
+    creds_json = flow.credentials.to_json()
+    result_payload = {
+        "user_id": state_obj.get("user_id"),
+        "credentials": creds_json,
+    }
+    REDIS_CONN.set_obj(_web_result_cache_key(state_id, source), result_payload, WEB_FLOW_TTL_SECS)
+    REDIS_CONN.delete(_web_state_cache_key(state_id, source))
+
+    return await _render_web_oauth_popup(state_id, True, "Authorization completed successfully.", source)
+
+@manager.route("/google/oauth/web/result", methods=["POST"])  # noqa: F821
+@login_required
+@validate_request("flow_id")
+async def poll_google_web_result():
+    req = await request.json or {}
+    source = request.args.get("type")
+    if source not in ("google-drive", "gmail"):
+        return get_json_result(code=RetCode.ARGUMENT_ERROR, message="Invalid Google OAuth type.")
+    flow_id = req.get("flow_id")
+    cache_raw = REDIS_CONN.get(_web_result_cache_key(flow_id, source))
+    if not cache_raw:
+        return get_json_result(code=RetCode.RUNNING, message="Authorization is still pending.")
+
+    result = json.loads(cache_raw)
+    if result.get("user_id") != current_user.id:
+        return get_json_result(code=RetCode.PERMISSION_ERROR, message="You are not allowed to access this authorization result.")
+
+    REDIS_CONN.delete(_web_result_cache_key(flow_id, source))
+    return get_json_result(data={"credentials": result.get("credentials")})
+
+@manager.route("/box/oauth/web/start", methods=["POST"])  # noqa: F821
+@login_required
+async def start_box_web_oauth():
+    req = await get_request_json()
+
+    client_id = req.get("client_id")
+    client_secret = req.get("client_secret")    
+    redirect_uri = req.get("redirect_uri", BOX_WEB_OAUTH_REDIRECT_URI)
+
+    if not client_id or not client_secret:
+        return get_json_result(code=RetCode.ARGUMENT_ERROR, message="Box client_id and client_secret are required.")
+
+    flow_id = str(uuid.uuid4())
+
+    box_auth = BoxOAuth(
+        OAuthConfig(
+            client_id=client_id,
+            client_secret=client_secret,
+        )
+    )
+
+    auth_url = box_auth.get_authorize_url(
+        options=GetAuthorizeUrlOptions(
+            redirect_uri=redirect_uri,
+            state=flow_id,
+        )
+    )
+
+    cache_payload = {
+        "user_id": current_user.id,
+        "auth_url": auth_url,
+        "client_id": client_id,
+        "client_secret": client_secret,
+        "created_at": int(time.time()),
+    }
+    REDIS_CONN.set_obj(_web_state_cache_key(flow_id, "box"), cache_payload, WEB_FLOW_TTL_SECS)
+    return get_json_result(
+        data = {
+            "flow_id": flow_id,
+            "authorization_url": auth_url,
+            "expires_in": WEB_FLOW_TTL_SECS,}
+    )
+
+@manager.route("/box/oauth/web/callback", methods=["GET"])  # noqa: F821
+async def box_web_oauth_callback():
+    flow_id = request.args.get("state")
+    if not flow_id:
+        return await _render_web_oauth_popup("", False, "Missing OAuth parameters.", "box")
+    
+    code = request.args.get("code")
+    if not code:
+        return await _render_web_oauth_popup(flow_id, False, "Missing authorization code from Box.", "box")
+
+    cache_payload = json.loads(REDIS_CONN.get(_web_state_cache_key(flow_id, "box")))
+    if not cache_payload:
+        return get_json_result(code=RetCode.ARGUMENT_ERROR, message="Box OAuth session expired or invalid.")
+
+    error = request.args.get("error")
+    error_description = request.args.get("error_description") or error
+    if error:
+        REDIS_CONN.delete(_web_state_cache_key(flow_id, "box"))
+        return await _render_web_oauth_popup(flow_id, False, error_description or "Authorization failed.", "box")
+    
+    auth = BoxOAuth(
+        OAuthConfig(
+            client_id=cache_payload.get("client_id"),
+            client_secret=cache_payload.get("client_secret"),
+        )
+    )
+
+    auth.get_tokens_authorization_code_grant(code)
+    token = auth.retrieve_token()
+    result_payload = {
+        "user_id": cache_payload.get("user_id"),
+        "client_id": cache_payload.get("client_id"),
+        "client_secret": cache_payload.get("client_secret"),
+        "access_token": token.access_token,
+        "refresh_token": token.refresh_token,
+    }
+
+    REDIS_CONN.set_obj(_web_result_cache_key(flow_id, "box"), result_payload, WEB_FLOW_TTL_SECS)
+    REDIS_CONN.delete(_web_state_cache_key(flow_id, "box"))
+
+    return await _render_web_oauth_popup(flow_id, True, "Authorization completed successfully.", "box")
+
+@manager.route("/box/oauth/web/result", methods=["POST"])  # noqa: F821
+@login_required
+@validate_request("flow_id")
+async def poll_box_web_result():
+    req = await get_request_json()
+    flow_id = req.get("flow_id")
+
+    cache_blob = REDIS_CONN.get(_web_result_cache_key(flow_id, "box"))
+    if not cache_blob:
+        return get_json_result(code=RetCode.RUNNING, message="Authorization is still pending.")
+
+    cache_raw = json.loads(cache_blob)
+    if cache_raw.get("user_id") != current_user.id:
+        return get_json_result(code=RetCode.PERMISSION_ERROR, message="You are not allowed to access this authorization result.")
+    
+    REDIS_CONN.delete(_web_result_cache_key(flow_id, "box"))
+
+    return get_json_result(data={"credentials": cache_raw})

api/apps/conversation_app.py

@@ -14,19 +14,21 @@
 #  limitations under the License.
 #
 import json
+import os
 import re
 import logging
 from copy import deepcopy
-from flask import Response, request
-from flask_login import current_user, login_required
+import tempfile
+from quart import Response, request
+from api.apps import current_user, login_required
 from api.db.db_models import APIToken
 from api.db.services.conversation_service import ConversationService, structure_answer
-from api.db.services.dialog_service import DialogService, ask, chat, gen_mindmap
+from api.db.services.dialog_service import DialogService, async_ask, async_chat, gen_mindmap
 from api.db.services.llm_service import LLMBundle
 from api.db.services.search_service import SearchService
 from api.db.services.tenant_llm_service import TenantLLMService
 from api.db.services.user_service import TenantService, UserTenantService
-from api.utils.api_utils import get_data_error_result, get_json_result, server_error_response, validate_request
+from api.utils.api_utils import get_data_error_result, get_json_result, get_request_json, server_error_response, validate_request
 from rag.prompts.template import load_prompt
 from rag.prompts.generator import chunks_format
 from common.constants import RetCode, LLMType
@@ -34,8 +36,8 @@ from common.constants import RetCode, LLMType
 
 @manager.route("/set", methods=["POST"])  # noqa: F821
 @login_required
-def set_conversation():
-    req = request.json
+async def set_conversation():
+    req = await get_request_json()
     conv_id = req.get("conversation_id")
     is_new = req.get("is_new")
     name = req.get("name", "New conversation")
@@ -78,14 +80,13 @@ def set_conversation():
 
 @manager.route("/get", methods=["GET"])  # noqa: F821
 @login_required
-def get():
+async def get():
     conv_id = request.args["conversation_id"]
     try:
         e, conv = ConversationService.get_by_id(conv_id)
         if not e:
             return get_data_error_result(message="Conversation not found!")
         tenants = UserTenantService.query(user_id=current_user.id)
-        avatar = None
         for tenant in tenants:
             dialog = DialogService.query(tenant_id=tenant.tenant_id, id=conv.dialog_id)
             if dialog and len(dialog) > 0:
@@ -129,8 +130,9 @@ def getsse(dialog_id):
 
 @manager.route("/rm", methods=["POST"])  # noqa: F821
 @login_required
-def rm():
-    conv_ids = request.json["conversation_ids"]
+async def rm():
+    req = await get_request_json()
+    conv_ids = req["conversation_ids"]
     try:
         for cid in conv_ids:
             exist, conv = ConversationService.get_by_id(cid)
@@ -150,7 +152,7 @@ def rm():
 
 @manager.route("/list", methods=["GET"])  # noqa: F821
 @login_required
-def list_conversation():
+async def list_conversation():
     dialog_id = request.args["dialog_id"]
     try:
         if not DialogService.query(tenant_id=current_user.id, id=dialog_id):
@@ -166,8 +168,8 @@ def list_conversation():
 @manager.route("/completion", methods=["POST"])  # noqa: F821
 @login_required
 @validate_request("conversation_id", "messages")
-def completion():
-    req = request.json
+async def completion():
+    req = await get_request_json()
     msg = []
     for m in req["messages"]:
         if m["role"] == "system":
@@ -216,10 +218,10 @@ def completion():
             dia.llm_setting = chat_model_config
 
         is_embedded = bool(chat_model_id)
-        def stream():
+        async def stream():
             nonlocal dia, msg, req, conv
             try:
-                for ans in chat(dia, msg, True, **req):
+                async for ans in async_chat(dia, msg, True, **req):
                     ans = structure_answer(conv, ans, message_id, conv.id)
                     yield "data:" + json.dumps({"code": 0, "message": "", "data": ans}, ensure_ascii=False) + "\n\n"
                 if not is_embedded:
@@ -239,7 +241,7 @@ def completion():
 
         else:
             answer = None
-            for ans in chat(dia, msg, **req):
+            async for ans in async_chat(dia, msg, **req):
                 answer = structure_answer(conv, ans, message_id, conv.id)
                 if not is_embedded:
                     ConversationService.update_by_id(conv.id, conv.to_dict())
@@ -248,11 +250,69 @@ def completion():
     except Exception as e:
         return server_error_response(e)
 
+@manager.route("/sequence2txt", methods=["POST"])  # noqa: F821
+@login_required
+async def sequence2txt():
+    req = await request.form
+    stream_mode = req.get("stream", "false").lower() == "true"
+    files = await request.files
+    if "file" not in files:
+        return get_data_error_result(message="Missing 'file' in multipart form-data")
+
+    uploaded = files["file"]
+
+    ALLOWED_EXTS = {
+        ".wav", ".mp3", ".m4a", ".aac",
+        ".flac", ".ogg", ".webm",
+        ".opus", ".wma"
+    }
+
+    filename = uploaded.filename or ""
+    suffix = os.path.splitext(filename)[-1].lower()
+    if suffix not in ALLOWED_EXTS:
+        return get_data_error_result(message=
+            f"Unsupported audio format: {suffix}. "
+            f"Allowed: {', '.join(sorted(ALLOWED_EXTS))}"
+        )
+    fd, temp_audio_path = tempfile.mkstemp(suffix=suffix)
+    os.close(fd)
+    await uploaded.save(temp_audio_path)
+
+    tenants = TenantService.get_info_by(current_user.id)
+    if not tenants:
+        return get_data_error_result(message="Tenant not found!")
+
+    asr_id = tenants[0]["asr_id"]
+    if not asr_id:
+        return get_data_error_result(message="No default ASR model is set")
+
+    asr_mdl=LLMBundle(tenants[0]["tenant_id"], LLMType.SPEECH2TEXT, asr_id)
+    if not stream_mode:
+        text = asr_mdl.transcription(temp_audio_path)
+        try:
+            os.remove(temp_audio_path)
+        except Exception as e:
+            logging.error(f"Failed to remove temp audio file: {str(e)}")
+        return get_json_result(data={"text": text})
+    async def event_stream():
+        try:
+            for evt in asr_mdl.stream_transcription(temp_audio_path):
+                yield f"data: {json.dumps(evt, ensure_ascii=False)}\n\n"
+        except Exception as e:
+            err = {"event": "error", "text": str(e)}
+            yield f"data: {json.dumps(err, ensure_ascii=False)}\n\n"
+        finally:
+            try:
+                os.remove(temp_audio_path)
+            except Exception as e:
+                logging.error(f"Failed to remove temp audio file: {str(e)}")
+
+    return Response(event_stream(), content_type="text/event-stream")
 
 @manager.route("/tts", methods=["POST"])  # noqa: F821
 @login_required
-def tts():
-    req = request.json
+async def tts():
+    req = await get_request_json()
     text = req["text"]
 
     tenants = TenantService.get_info_by(current_user.id)
@@ -284,8 +344,8 @@ def tts():
 @manager.route("/delete_msg", methods=["POST"])  # noqa: F821
 @login_required
 @validate_request("conversation_id", "message_id")
-def delete_msg():
-    req = request.json
+async def delete_msg():
+    req = await get_request_json()
     e, conv = ConversationService.get_by_id(req["conversation_id"])
     if not e:
         return get_data_error_result(message="Conversation not found!")
@@ -307,8 +367,8 @@ def delete_msg():
 @manager.route("/thumbup", methods=["POST"])  # noqa: F821
 @login_required
 @validate_request("conversation_id", "message_id")
-def thumbup():
-    req = request.json
+async def thumbup():
+    req = await get_request_json()
     e, conv = ConversationService.get_by_id(req["conversation_id"])
     if not e:
         return get_data_error_result(message="Conversation not found!")
@@ -334,8 +394,8 @@ def thumbup():
 @manager.route("/ask", methods=["POST"])  # noqa: F821
 @login_required
 @validate_request("question", "kb_ids")
-def ask_about():
-    req = request.json
+async def ask_about():
+    req = await get_request_json()
     uid = current_user.id
 
     search_id = req.get("search_id", "")
@@ -346,10 +406,10 @@ def ask_about():
     if search_app:
         search_config = search_app.get("search_config", {})
 
-    def stream():
+    async def stream():
         nonlocal req, uid
         try:
-            for ans in ask(req["question"], req["kb_ids"], uid, search_config=search_config):
+            async for ans in async_ask(req["question"], req["kb_ids"], uid, search_config=search_config):
                 yield "data:" + json.dumps({"code": 0, "message": "", "data": ans}, ensure_ascii=False) + "\n\n"
         except Exception as e:
             yield "data:" + json.dumps({"code": 500, "message": str(e), "data": {"answer": "**ERROR**: " + str(e), "reference": []}}, ensure_ascii=False) + "\n\n"
@@ -366,8 +426,8 @@ def ask_about():
 @manager.route("/mindmap", methods=["POST"])  # noqa: F821
 @login_required
 @validate_request("question", "kb_ids")
-def mindmap():
-    req = request.json
+async def mindmap():
+    req = await get_request_json()
     search_id = req.get("search_id", "")
     search_app = SearchService.get_detail(search_id) if search_id else {}
     search_config = search_app.get("search_config", {}) if search_app else {}
@@ -375,7 +435,7 @@ def mindmap():
     kb_ids.extend(req["kb_ids"])
     kb_ids = list(set(kb_ids))
 
-    mind_map = gen_mindmap(req["question"], kb_ids, search_app.get("tenant_id", current_user.id), search_config)
+    mind_map = await gen_mindmap(req["question"], kb_ids, search_app.get("tenant_id", current_user.id), search_config)
     if "error" in mind_map:
         return server_error_response(Exception(mind_map["error"]))
     return get_json_result(data=mind_map)
@@ -384,8 +444,8 @@ def mindmap():
 @manager.route("/related_questions", methods=["POST"])  # noqa: F821
 @login_required
 @validate_request("question")
-def related_questions():
-    req = request.json
+async def related_questions():
+    req = await get_request_json()
 
     search_id = req.get("search_id", "")
     search_config = {}
@@ -402,7 +462,7 @@ def related_questions():
     if "parameter" in gen_conf:
         del gen_conf["parameter"]
     prompt = load_prompt("related_question")
-    ans = chat_mdl.chat(
+    ans = await chat_mdl.async_chat(
         prompt,
         [
             {

api/apps/dialog_app.py

@@ -14,25 +14,24 @@
 #  limitations under the License.
 #
 
-from flask import request
-from flask_login import login_required, current_user
+from quart import request
 from api.db.services import duplicate_name
 from api.db.services.dialog_service import DialogService
 from common.constants import StatusEnum
 from api.db.services.tenant_llm_service import TenantLLMService
 from api.db.services.knowledgebase_service import KnowledgebaseService
 from api.db.services.user_service import TenantService, UserTenantService
-from api.utils.api_utils import server_error_response, get_data_error_result, validate_request
+from api.utils.api_utils import get_data_error_result, get_json_result, get_request_json, server_error_response, validate_request
 from common.misc_utils import get_uuid
 from common.constants import RetCode
-from api.utils.api_utils import get_json_result
+from api.apps import login_required, current_user
 
 
 @manager.route('/set', methods=['POST'])  # noqa: F821
 @validate_request("prompt_config")
 @login_required
-def set_dialog():
-    req = request.json
+async def set_dialog():
+    req = await get_request_json()
     dialog_id = req.get("dialog_id", "")
     is_create = not dialog_id
     name = req.get("name", "New Dialog")
@@ -66,7 +65,7 @@ def set_dialog():
 
     if not is_create:
         if not req.get("kb_ids", []) and not prompt_config.get("tavily_api_key") and "{knowledge}" in prompt_config['system']:
-            return get_data_error_result(message="Please remove `{knowledge}` in system prompt since no knowledge base / Tavily used here.")
+            return get_data_error_result(message="Please remove `{knowledge}` in system prompt since no dataset / Tavily used here.")
 
         for p in prompt_config["parameters"]:
             if p["optional"]:
@@ -154,33 +153,34 @@ def get_kb_names(kb_ids):
 @login_required
 def list_dialogs():
     try:
-        diags = DialogService.query(
+        conversations = DialogService.query(
             tenant_id=current_user.id,
             status=StatusEnum.VALID.value,
             reverse=True,
             order_by=DialogService.model.create_time)
-        diags = [d.to_dict() for d in diags]
-        for d in diags:
-            d["kb_ids"], d["kb_names"] = get_kb_names(d["kb_ids"])
-        return get_json_result(data=diags)
+        conversations = [d.to_dict() for d in conversations]
+        for conversation in conversations:
+            conversation["kb_ids"], conversation["kb_names"] = get_kb_names(conversation["kb_ids"])
+        return get_json_result(data=conversations)
     except Exception as e:
         return server_error_response(e)
 
 
 @manager.route('/next', methods=['POST'])  # noqa: F821
 @login_required
-def list_dialogs_next():
-    keywords = request.args.get("keywords", "")
-    page_number = int(request.args.get("page", 0))
-    items_per_page = int(request.args.get("page_size", 0))
-    parser_id = request.args.get("parser_id")
-    orderby = request.args.get("orderby", "create_time")
-    if request.args.get("desc", "true").lower() == "false":
+async def list_dialogs_next():
+    args = request.args
+    keywords = args.get("keywords", "")
+    page_number = int(args.get("page", 0))
+    items_per_page = int(args.get("page_size", 0))
+    parser_id = args.get("parser_id")
+    orderby = args.get("orderby", "create_time")
+    if args.get("desc", "true").lower() == "false":
         desc = False
     else:
         desc = True
 
-    req = request.get_json()
+    req = await get_request_json()
     owner_ids = req.get("owner_ids", [])
     try:
         if not owner_ids:
@@ -207,8 +207,8 @@ def list_dialogs_next():
 @manager.route('/rm', methods=['POST'])  # noqa: F821
 @login_required
 @validate_request("dialog_ids")
-def rm():
-    req = request.json
+async def rm():
+    req = await get_request_json()
     dialog_list=[]
     tenants = UserTenantService.query(user_id=current_user.id)
     try:

api/apps/document_app.py

@@ -13,23 +13,21 @@
 #  See the License for the specific language governing permissions and
 #  limitations under the License
 #
+import asyncio
 import json
 import os.path
 import pathlib
 import re
 from pathlib import Path
-
-import flask
-from flask import request
-from flask_login import current_user, login_required
-
-from api import settings
+from quart import request, make_response
+from api.apps import current_user, login_required
 from api.common.check_team_permission import check_kb_team_permission
 from api.constants import FILE_NAME_LEN_LIMIT, IMG_BASE64_PREFIX
 from api.db import VALID_FILE_TYPES, FileType
 from api.db.db_models import Task
 from api.db.services import duplicate_name
 from api.db.services.document_service import DocumentService, doc_upload_and_parse
+from common.metadata_utils import meta_filter, convert_conditions
 from api.db.services.file2document_service import File2DocumentService
 from api.db.services.file_service import FileService
 from api.db.services.knowledgebase_service import KnowledgebaseService
@@ -40,7 +38,7 @@ from api.utils.api_utils import (
     get_data_error_result,
     get_json_result,
     server_error_response,
-    validate_request,
+    validate_request, get_request_json,
 )
 from api.utils.file_utils import filename_type, thumbnail
 from common.file_utils import get_project_base_directory
@@ -48,20 +46,22 @@ from common.constants import RetCode, VALID_TASK_STATUS, ParserType, TaskStatus
 from api.utils.web_utils import CONTENT_TYPE_MAP, html2pdf, is_valid_url
 from deepdoc.parser.html_parser import RAGFlowHtmlParser
 from rag.nlp import search, rag_tokenizer
-from rag.utils.storage_factory import STORAGE_IMPL
+from common import settings
 
 
 @manager.route("/upload", methods=["POST"])  # noqa: F821
 @login_required
 @validate_request("kb_id")
-def upload():
-    kb_id = request.form.get("kb_id")
+async def upload():
+    form = await request.form
+    kb_id = form.get("kb_id")
     if not kb_id:
         return get_json_result(data=False, message='Lack of "KB ID"', code=RetCode.ARGUMENT_ERROR)
-    if "file" not in request.files:
+    files = await request.files
+    if "file" not in files:
         return get_json_result(data=False, message="No file part!", code=RetCode.ARGUMENT_ERROR)
 
-    file_objs = request.files.getlist("file")
+    file_objs = files.getlist("file")
     for file_obj in file_objs:
         if file_obj.filename == "":
             return get_json_result(data=False, message="No file selected!", code=RetCode.ARGUMENT_ERROR)
@@ -70,11 +70,11 @@ def upload():
 
     e, kb = KnowledgebaseService.get_by_id(kb_id)
     if not e:
-        raise LookupError("Can't find this knowledgebase!")
+        raise LookupError("Can't find this dataset!")
     if not check_kb_team_permission(kb, current_user.id):
         return get_json_result(data=False, message="No authorization.", code=RetCode.AUTHENTICATION_ERROR)
 
-    err, files = FileService.upload_document(kb, file_objs, current_user.id)
+    err, files = await asyncio.to_thread(FileService.upload_document, kb, file_objs, current_user.id)
     if err:
         return get_json_result(data=files, message="\n".join(err), code=RetCode.SERVER_ERROR)
 
@@ -88,17 +88,18 @@ def upload():
 @manager.route("/web_crawl", methods=["POST"])  # noqa: F821
 @login_required
 @validate_request("kb_id", "name", "url")
-def web_crawl():
-    kb_id = request.form.get("kb_id")
+async def web_crawl():
+    form = await request.form
+    kb_id = form.get("kb_id")
     if not kb_id:
         return get_json_result(data=False, message='Lack of "KB ID"', code=RetCode.ARGUMENT_ERROR)
-    name = request.form.get("name")
-    url = request.form.get("url")
+    name = form.get("name")
+    url = form.get("url")
     if not is_valid_url(url):
         return get_json_result(data=False, message="The URL format is invalid", code=RetCode.ARGUMENT_ERROR)
     e, kb = KnowledgebaseService.get_by_id(kb_id)
     if not e:
-        raise LookupError("Can't find this knowledgebase!")
+        raise LookupError("Can't find this dataset!")
     if check_kb_team_permission(kb, current_user.id):
         return get_json_result(data=False, message="No authorization.", code=RetCode.AUTHENTICATION_ERROR)
 
@@ -119,9 +120,9 @@ def web_crawl():
             raise RuntimeError("This type of file has not been supported yet!")
 
         location = filename
-        while STORAGE_IMPL.obj_exist(kb_id, location):
+        while settings.STORAGE_IMPL.obj_exist(kb_id, location):
             location += "_"
-        STORAGE_IMPL.put(kb_id, location, blob)
+        settings.STORAGE_IMPL.put(kb_id, location, blob)
         doc = {
             "id": get_uuid(),
             "kb_id": kb.id,
@@ -153,8 +154,8 @@ def web_crawl():
 @manager.route("/create", methods=["POST"])  # noqa: F821
 @login_required
 @validate_request("name", "kb_id")
-def create():
-    req = request.json
+async def create():
+    req = await get_request_json()
     kb_id = req["kb_id"]
     if not kb_id:
         return get_json_result(data=False, message='Lack of "KB ID"', code=RetCode.ARGUMENT_ERROR)
@@ -168,10 +169,10 @@ def create():
     try:
         e, kb = KnowledgebaseService.get_by_id(kb_id)
         if not e:
-            return get_data_error_result(message="Can't find this knowledgebase!")
+            return get_data_error_result(message="Can't find this dataset!")
 
         if DocumentService.query(name=req["name"], kb_id=kb_id):
-            return get_data_error_result(message="Duplicated document name in the same knowledgebase.")
+            return get_data_error_result(message="Duplicated document name in the same dataset.")
 
         kb_root_folder = FileService.get_kb_folder(kb.tenant_id)
         if not kb_root_folder:
@@ -209,7 +210,7 @@ def create():
 
 @manager.route("/list", methods=["POST"])  # noqa: F821
 @login_required
-def list_docs():
+async def list_docs():
     kb_id = request.args.get("kb_id")
     if not kb_id:
         return get_json_result(data=False, message='Lack of "KB ID"', code=RetCode.ARGUMENT_ERROR)
@@ -218,7 +219,7 @@ def list_docs():
         if KnowledgebaseService.query(tenant_id=tenant.tenant_id, id=kb_id):
             break
     else:
-        return get_json_result(data=False, message="Only owner of knowledgebase authorized for this operation.", code=RetCode.OPERATING_ERROR)
+        return get_json_result(data=False, message="Only owner of dataset authorized for this operation.", code=RetCode.OPERATING_ERROR)
     keywords = request.args.get("keywords", "")
 
     page_number = int(request.args.get("page", 0))
@@ -231,7 +232,11 @@ def list_docs():
     create_time_from = int(request.args.get("create_time_from", 0))
     create_time_to = int(request.args.get("create_time_to", 0))
 
-    req = request.get_json()
+    req = await get_request_json()
+
+    return_empty_metadata = req.get("return_empty_metadata", False)
+    if isinstance(return_empty_metadata, str):
+        return_empty_metadata = return_empty_metadata.lower() == "true"
 
     run_status = req.get("run_status", [])
     if run_status:
@@ -246,9 +251,74 @@ def list_docs():
             return get_data_error_result(message=f"Invalid filter conditions: {', '.join(invalid_types)} type{'s' if len(invalid_types) > 1 else ''}")
 
     suffix = req.get("suffix", [])
+    metadata_condition = req.get("metadata_condition", {}) or {}
+    metadata = req.get("metadata", {}) or {}
+    if isinstance(metadata, dict) and metadata.get("empty_metadata"):
+        return_empty_metadata = True
+        metadata = {k: v for k, v in metadata.items() if k != "empty_metadata"}
+    if return_empty_metadata:
+        metadata_condition = {}
+        metadata = {}
+    else:
+        if metadata_condition and not isinstance(metadata_condition, dict):
+            return get_data_error_result(message="metadata_condition must be an object.")
+        if metadata and not isinstance(metadata, dict):
+            return get_data_error_result(message="metadata must be an object.")
+
+    doc_ids_filter = None
+    metas = None
+    if metadata_condition or metadata:
+        metas = DocumentService.get_flatted_meta_by_kbs([kb_id])
+
+    if metadata_condition:
+        doc_ids_filter = set(meta_filter(metas, convert_conditions(metadata_condition), metadata_condition.get("logic", "and")))
+        if metadata_condition.get("conditions") and not doc_ids_filter:
+            return get_json_result(data={"total": 0, "docs": []})
+
+    if metadata:
+        metadata_doc_ids = None
+        for key, values in metadata.items():
+            if not values:
+                continue
+            if not isinstance(values, list):
+                values = [values]
+            values = [str(v) for v in values if v is not None and str(v).strip()]
+            if not values:
+                continue
+            key_doc_ids = set()
+            for value in values:
+                key_doc_ids.update(metas.get(key, {}).get(value, []))
+            if metadata_doc_ids is None:
+                metadata_doc_ids = key_doc_ids
+            else:
+                metadata_doc_ids &= key_doc_ids
+            if not metadata_doc_ids:
+                return get_json_result(data={"total": 0, "docs": []})
+        if metadata_doc_ids is not None:
+            if doc_ids_filter is None:
+                doc_ids_filter = metadata_doc_ids
+            else:
+                doc_ids_filter &= metadata_doc_ids
+            if not doc_ids_filter:
+                return get_json_result(data={"total": 0, "docs": []})
+
+    if doc_ids_filter is not None:
+        doc_ids_filter = list(doc_ids_filter)
 
     try:
-        docs, tol = DocumentService.get_by_kb_id(kb_id, page_number, items_per_page, orderby, desc, keywords, run_status, types, suffix)
+        docs, tol = DocumentService.get_by_kb_id(
+            kb_id,
+            page_number,
+            items_per_page,
+            orderby,
+            desc,
+            keywords,
+            run_status,
+            types,
+            suffix,
+            doc_ids_filter,
+            return_empty_metadata=return_empty_metadata,
+        )
 
         if create_time_from or create_time_to:
             filtered_docs = []
@@ -261,6 +331,8 @@ def list_docs():
         for doc_item in docs:
             if doc_item["thumbnail"] and not doc_item["thumbnail"].startswith(IMG_BASE64_PREFIX):
                 doc_item["thumbnail"] = f"/v1/document/image/{kb_id}-{doc_item['thumbnail']}"
+            if doc_item.get("source_type"):
+                doc_item["source_type"] = doc_item["source_type"].split("/")[0]
 
         return get_json_result(data={"total": tol, "docs": docs})
     except Exception as e:
@@ -269,8 +341,8 @@ def list_docs():
 
 @manager.route("/filter", methods=["POST"])  # noqa: F821
 @login_required
-def get_filter():
-    req = request.get_json()
+async def get_filter():
+    req = await get_request_json()
 
     kb_id = req.get("kb_id")
     if not kb_id:
@@ -280,7 +352,7 @@ def get_filter():
         if KnowledgebaseService.query(tenant_id=tenant.tenant_id, id=kb_id):
             break
     else:
-        return get_json_result(data=False, message="Only owner of knowledgebase authorized for this operation.", code=RetCode.OPERATING_ERROR)
+        return get_json_result(data=False, message="Only owner of dataset authorized for this operation.", code=RetCode.OPERATING_ERROR)
 
     keywords = req.get("keywords", "")
 
@@ -307,8 +379,8 @@ def get_filter():
 
 @manager.route("/infos", methods=["POST"])  # noqa: F821
 @login_required
-def docinfos():
-    req = request.json
+async def doc_infos():
+    req = await get_request_json()
     doc_ids = req["doc_ids"]
     for doc_id in doc_ids:
         if not DocumentService.accessible(doc_id, current_user.id):
@@ -317,6 +389,107 @@ def docinfos():
     return get_json_result(data=list(docs.dicts()))
 
 
+@manager.route("/metadata/summary", methods=["POST"])  # noqa: F821
+@login_required
+async def metadata_summary():
+    req = await get_request_json()
+    kb_id = req.get("kb_id")
+    if not kb_id:
+        return get_json_result(data=False, message='Lack of "KB ID"', code=RetCode.ARGUMENT_ERROR)
+
+    tenants = UserTenantService.query(user_id=current_user.id)
+    for tenant in tenants:
+        if KnowledgebaseService.query(tenant_id=tenant.tenant_id, id=kb_id):
+            break
+    else:
+        return get_json_result(data=False, message="Only owner of dataset authorized for this operation.", code=RetCode.OPERATING_ERROR)
+
+    try:
+        summary = DocumentService.get_metadata_summary(kb_id)
+        return get_json_result(data={"summary": summary})
+    except Exception as e:
+        return server_error_response(e)
+
+
+@manager.route("/metadata/update", methods=["POST"])  # noqa: F821
+@login_required
+async def metadata_update():
+    req = await get_request_json()
+    kb_id = req.get("kb_id")
+    if not kb_id:
+        return get_json_result(data=False, message='Lack of "KB ID"', code=RetCode.ARGUMENT_ERROR)
+
+    tenants = UserTenantService.query(user_id=current_user.id)
+    for tenant in tenants:
+        if KnowledgebaseService.query(tenant_id=tenant.tenant_id, id=kb_id):
+            break
+    else:
+        return get_json_result(data=False, message="Only owner of dataset authorized for this operation.", code=RetCode.OPERATING_ERROR)
+
+    selector = req.get("selector", {}) or {}
+    updates = req.get("updates", []) or []
+    deletes = req.get("deletes", []) or []
+
+    if not isinstance(selector, dict):
+        return get_json_result(data=False, message="selector must be an object.", code=RetCode.ARGUMENT_ERROR)
+    if not isinstance(updates, list) or not isinstance(deletes, list):
+        return get_json_result(data=False, message="updates and deletes must be lists.", code=RetCode.ARGUMENT_ERROR)
+
+    metadata_condition = selector.get("metadata_condition", {}) or {}
+    if metadata_condition and not isinstance(metadata_condition, dict):
+        return get_json_result(data=False, message="metadata_condition must be an object.", code=RetCode.ARGUMENT_ERROR)
+
+    document_ids = selector.get("document_ids", []) or []
+    if document_ids and not isinstance(document_ids, list):
+        return get_json_result(data=False, message="document_ids must be a list.", code=RetCode.ARGUMENT_ERROR)
+
+    for upd in updates:
+        if not isinstance(upd, dict) or not upd.get("key") or "value" not in upd:
+            return get_json_result(data=False, message="Each update requires key and value.", code=RetCode.ARGUMENT_ERROR)
+    for d in deletes:
+        if not isinstance(d, dict) or not d.get("key"):
+            return get_json_result(data=False, message="Each delete requires key.", code=RetCode.ARGUMENT_ERROR)
+
+    kb_doc_ids = KnowledgebaseService.list_documents_by_ids([kb_id])
+    target_doc_ids = set(kb_doc_ids)
+    if document_ids:
+        invalid_ids = set(document_ids) - set(kb_doc_ids)
+        if invalid_ids:
+            return get_json_result(data=False, message=f"These documents do not belong to dataset {kb_id}: {', '.join(invalid_ids)}", code=RetCode.ARGUMENT_ERROR)
+        target_doc_ids = set(document_ids)
+
+    if metadata_condition:
+        metas = DocumentService.get_flatted_meta_by_kbs([kb_id])
+        filtered_ids = set(meta_filter(metas, convert_conditions(metadata_condition), metadata_condition.get("logic", "and")))
+        target_doc_ids = target_doc_ids & filtered_ids
+        if metadata_condition.get("conditions") and not target_doc_ids:
+            return get_json_result(data={"updated": 0, "matched_docs": 0})
+
+    target_doc_ids = list(target_doc_ids)
+    updated = DocumentService.batch_update_metadata(kb_id, target_doc_ids, updates, deletes)
+    return get_json_result(data={"updated": updated, "matched_docs": len(target_doc_ids)})
+
+
+@manager.route("/update_metadata_setting", methods=["POST"])  # noqa: F821
+@login_required
+@validate_request("doc_id", "metadata")
+async def update_metadata_setting():
+    req = await get_request_json()
+    if not DocumentService.accessible(req["doc_id"], current_user.id):
+        return get_json_result(data=False, message="No authorization.", code=RetCode.AUTHENTICATION_ERROR)
+
+    e, doc = DocumentService.get_by_id(req["doc_id"])
+    if not e:
+        return get_data_error_result(message="Document not found!")
+
+    DocumentService.update_parser_config(doc.id, {"metadata": req["metadata"]})
+    e, doc = DocumentService.get_by_id(doc.id)
+    if not e:
+        return get_data_error_result(message="Document not found!")
+
+    return get_json_result(data=doc.to_dict())
+
+
 @manager.route("/thumbnails", methods=["GET"])  # noqa: F821
 # @login_required
 def thumbnails():
@@ -339,8 +512,8 @@ def thumbnails():
 @manager.route("/change_status", methods=["POST"])  # noqa: F821
 @login_required
 @validate_request("doc_ids", "status")
-def change_status():
-    req = request.get_json()
+async def change_status():
+    req = await get_request_json()
     doc_ids = req.get("doc_ids", [])
     status = str(req.get("status", ""))
 
@@ -360,7 +533,7 @@ def change_status():
                 continue
             e, kb = KnowledgebaseService.get_by_id(doc.kb_id)
             if not e:
-                result[doc_id] = {"error": "Can't find this knowledgebase!"}
+                result[doc_id] = {"error": "Can't find this dataset!"}
                 continue
             if not DocumentService.update_by_id(doc_id, {"status": str(status)}):
                 result[doc_id] = {"error": "Database error (Document update)!"}
@@ -379,8 +552,8 @@ def change_status():
 @manager.route("/rm", methods=["POST"])  # noqa: F821
 @login_required
 @validate_request("doc_id")
-def rm():
-    req = request.json
+async def rm():
+    req = await get_request_json()
     doc_ids = req["doc_id"]
     if isinstance(doc_ids, str):
         doc_ids = [doc_ids]
@@ -389,7 +562,7 @@ def rm():
         if not DocumentService.accessible4deletion(doc_id, current_user.id):
             return get_json_result(data=False, message="No authorization.", code=RetCode.AUTHENTICATION_ERROR)
 
-    errors = FileService.delete_docs(doc_ids, current_user.id)
+    errors = await asyncio.to_thread(FileService.delete_docs, doc_ids, current_user.id)
 
     if errors:
         return get_json_result(data=False, message=errors, code=RetCode.SERVER_ERROR)
@@ -400,46 +573,57 @@ def rm():
 @manager.route("/run", methods=["POST"])  # noqa: F821
 @login_required
 @validate_request("doc_ids", "run")
-def run():
-    req = request.json
-    for doc_id in req["doc_ids"]:
-        if not DocumentService.accessible(doc_id, current_user.id):
-            return get_json_result(data=False, message="No authorization.", code=RetCode.AUTHENTICATION_ERROR)
+async def run():
+    req = await get_request_json()
     try:
-        kb_table_num_map = {}
-        for id in req["doc_ids"]:
-            info = {"run": str(req["run"]), "progress": 0}
-            if str(req["run"]) == TaskStatus.RUNNING.value and req.get("delete", False):
-                info["progress_msg"] = ""
-                info["chunk_num"] = 0
-                info["token_num"] = 0
+        def _run_sync():
+            for doc_id in req["doc_ids"]:
+                if not DocumentService.accessible(doc_id, current_user.id):
+                    return get_json_result(data=False, message="No authorization.", code=RetCode.AUTHENTICATION_ERROR)
 
-            tenant_id = DocumentService.get_tenant_id(id)
-            if not tenant_id:
-                return get_data_error_result(message="Tenant not found!")
-            e, doc = DocumentService.get_by_id(id)
-            if not e:
-                return get_data_error_result(message="Document not found!")
+            kb_table_num_map = {}
+            for id in req["doc_ids"]:
+                info = {"run": str(req["run"]), "progress": 0}
+                if str(req["run"]) == TaskStatus.RUNNING.value and req.get("delete", False):
+                    info["progress_msg"] = ""
+                    info["chunk_num"] = 0
+                    info["token_num"] = 0
 
-            if str(req["run"]) == TaskStatus.CANCEL.value:
-                if str(doc.run) == TaskStatus.RUNNING.value:
-                    cancel_all_task_of(id)
-                else:
-                    return get_data_error_result(message="Cannot cancel a task that is not in RUNNING status")
-            if all([("delete" not in req or req["delete"]), str(req["run"]) == TaskStatus.RUNNING.value, str(doc.run) == TaskStatus.DONE.value]):
-                DocumentService.clear_chunk_num_when_rerun(doc.id)
+                tenant_id = DocumentService.get_tenant_id(id)
+                if not tenant_id:
+                    return get_data_error_result(message="Tenant not found!")
+                e, doc = DocumentService.get_by_id(id)
+                if not e:
+                    return get_data_error_result(message="Document not found!")
 
-            DocumentService.update_by_id(id, info)
-            if req.get("delete", False):
-                TaskService.filter_delete([Task.doc_id == id])
-                if settings.docStoreConn.indexExist(search.index_name(tenant_id), doc.kb_id):
-                    settings.docStoreConn.delete({"doc_id": id}, search.index_name(tenant_id), doc.kb_id)
+                if str(req["run"]) == TaskStatus.CANCEL.value:
+                    if str(doc.run) == TaskStatus.RUNNING.value:
+                        cancel_all_task_of(id)
+                    else:
+                        return get_data_error_result(message="Cannot cancel a task that is not in RUNNING status")
+                if all([("delete" not in req or req["delete"]), str(req["run"]) == TaskStatus.RUNNING.value, str(doc.run) == TaskStatus.DONE.value]):
+                    DocumentService.clear_chunk_num_when_rerun(doc.id)
 
-            if str(req["run"]) == TaskStatus.RUNNING.value:
-                doc = doc.to_dict()
-                DocumentService.run(tenant_id, doc, kb_table_num_map)
+                DocumentService.update_by_id(id, info)
+                if req.get("delete", False):
+                    TaskService.filter_delete([Task.doc_id == id])
+                    if settings.docStoreConn.index_exist(search.index_name(tenant_id), doc.kb_id):
+                        settings.docStoreConn.delete({"doc_id": id}, search.index_name(tenant_id), doc.kb_id)
 
-        return get_json_result(data=True)
+                if str(req["run"]) == TaskStatus.RUNNING.value:
+                    if req.get("apply_kb"):
+                        e, kb = KnowledgebaseService.get_by_id(doc.kb_id)
+                        if not e:
+                            raise LookupError("Can't find this dataset!")
+                        doc.parser_config["enable_metadata"] = kb.parser_config.get("enable_metadata", False)
+                        doc.parser_config["metadata"] = kb.parser_config.get("metadata", {})
+                        DocumentService.update_parser_config(doc.id, doc.parser_config)
+                    doc_dict = doc.to_dict()
+                    DocumentService.run(tenant_id, doc_dict, kb_table_num_map)
+
+            return get_json_result(data=True)
+
+        return await asyncio.to_thread(_run_sync)
     except Exception as e:
         return server_error_response(e)
 
@@ -447,66 +631,72 @@ def run():
 @manager.route("/rename", methods=["POST"])  # noqa: F821
 @login_required
 @validate_request("doc_id", "name")
-def rename():
-    req = request.json
-    if not DocumentService.accessible(req["doc_id"], current_user.id):
-        return get_json_result(data=False, message="No authorization.", code=RetCode.AUTHENTICATION_ERROR)
+async def rename():
+    req = await get_request_json()
     try:
-        e, doc = DocumentService.get_by_id(req["doc_id"])
-        if not e:
-            return get_data_error_result(message="Document not found!")
-        if pathlib.Path(req["name"].lower()).suffix != pathlib.Path(doc.name.lower()).suffix:
-            return get_json_result(data=False, message="The extension of file can't be changed", code=RetCode.ARGUMENT_ERROR)
-        if len(req["name"].encode("utf-8")) > FILE_NAME_LEN_LIMIT:
-            return get_json_result(data=False, message=f"File name must be {FILE_NAME_LEN_LIMIT} bytes or less.", code=RetCode.ARGUMENT_ERROR)
+        def _rename_sync():
+            if not DocumentService.accessible(req["doc_id"], current_user.id):
+                return get_json_result(data=False, message="No authorization.", code=RetCode.AUTHENTICATION_ERROR)
 
-        for d in DocumentService.query(name=req["name"], kb_id=doc.kb_id):
-            if d.name == req["name"]:
-                return get_data_error_result(message="Duplicated document name in the same knowledgebase.")
+            e, doc = DocumentService.get_by_id(req["doc_id"])
+            if not e:
+                return get_data_error_result(message="Document not found!")
+            if pathlib.Path(req["name"].lower()).suffix != pathlib.Path(doc.name.lower()).suffix:
+                return get_json_result(data=False, message="The extension of file can't be changed", code=RetCode.ARGUMENT_ERROR)
+            if len(req["name"].encode("utf-8")) > FILE_NAME_LEN_LIMIT:
+                return get_json_result(data=False, message=f"File name must be {FILE_NAME_LEN_LIMIT} bytes or less.", code=RetCode.ARGUMENT_ERROR)
 
-        if not DocumentService.update_by_id(req["doc_id"], {"name": req["name"]}):
-            return get_data_error_result(message="Database error (Document rename)!")
+            for d in DocumentService.query(name=req["name"], kb_id=doc.kb_id):
+                if d.name == req["name"]:
+                    return get_data_error_result(message="Duplicated document name in the same dataset.")
 
-        informs = File2DocumentService.get_by_document_id(req["doc_id"])
-        if informs:
-            e, file = FileService.get_by_id(informs[0].file_id)
-            FileService.update_by_id(file.id, {"name": req["name"]})
+            if not DocumentService.update_by_id(req["doc_id"], {"name": req["name"]}):
+                return get_data_error_result(message="Database error (Document rename)!")
 
-        tenant_id = DocumentService.get_tenant_id(req["doc_id"])
-        title_tks = rag_tokenizer.tokenize(req["name"])
-        es_body = {
-            "docnm_kwd": req["name"],
-            "title_tks": title_tks,
-            "title_sm_tks": rag_tokenizer.fine_grained_tokenize(title_tks),
-        }
-        if settings.docStoreConn.indexExist(search.index_name(tenant_id), doc.kb_id):
-            settings.docStoreConn.update(
-                {"doc_id": req["doc_id"]},
-                es_body,
-                search.index_name(tenant_id),
-                doc.kb_id,
-            )
+            informs = File2DocumentService.get_by_document_id(req["doc_id"])
+            if informs:
+                e, file = FileService.get_by_id(informs[0].file_id)
+                FileService.update_by_id(file.id, {"name": req["name"]})
+
+            tenant_id = DocumentService.get_tenant_id(req["doc_id"])
+            title_tks = rag_tokenizer.tokenize(req["name"])
+            es_body = {
+                "docnm_kwd": req["name"],
+                "title_tks": title_tks,
+                "title_sm_tks": rag_tokenizer.fine_grained_tokenize(title_tks),
+            }
+            if settings.docStoreConn.index_exist(search.index_name(tenant_id), doc.kb_id):
+                settings.docStoreConn.update(
+                    {"doc_id": req["doc_id"]},
+                    es_body,
+                    search.index_name(tenant_id),
+                    doc.kb_id,
+                )
+            return get_json_result(data=True)
+
+        return await asyncio.to_thread(_rename_sync)
 
-        return get_json_result(data=True)
     except Exception as e:
         return server_error_response(e)
 
 
 @manager.route("/get/<doc_id>", methods=["GET"])  # noqa: F821
 # @login_required
-def get(doc_id):
+async def get(doc_id):
     try:
         e, doc = DocumentService.get_by_id(doc_id)
         if not e:
             return get_data_error_result(message="Document not found!")
 
         b, n = File2DocumentService.get_storage_address(doc_id=doc_id)
-        response = flask.make_response(STORAGE_IMPL.get(b, n))
+        data = await asyncio.to_thread(settings.STORAGE_IMPL.get, b, n)
+        response = await make_response(data)
 
         ext = re.search(r"\.([^.]+)$", doc.name.lower())
         ext = ext.group(1) if ext else None
         if ext:
             if doc.type == FileType.VISUAL.value:
+
                 content_type = CONTENT_TYPE_MAP.get(ext, f"image/{ext}")
             else:
                 content_type = CONTENT_TYPE_MAP.get(ext, f"application/{ext}")
@@ -516,12 +706,27 @@ def get(doc_id):
         return server_error_response(e)
 
 
+@manager.route("/download/<attachment_id>", methods=["GET"])  # noqa: F821
+@login_required
+async def download_attachment(attachment_id):
+    try:
+        ext = request.args.get("ext", "markdown")
+        data = await asyncio.to_thread(settings.STORAGE_IMPL.get, current_user.id, attachment_id)
+        response = await make_response(data)
+        response.headers.set("Content-Type", CONTENT_TYPE_MAP.get(ext, f"application/{ext}"))
+
+        return response
+
+    except Exception as e:
+        return server_error_response(e)
+
+
 @manager.route("/change_parser", methods=["POST"])  # noqa: F821
 @login_required
 @validate_request("doc_id")
-def change_parser():
+async def change_parser():
 
-    req = request.json
+    req = await get_request_json()
     if not DocumentService.accessible(req["doc_id"], current_user.id):
         return get_json_result(data=False, message="No authorization.", code=RetCode.AUTHENTICATION_ERROR)
 
@@ -541,8 +746,10 @@ def change_parser():
             tenant_id = DocumentService.get_tenant_id(req["doc_id"])
             if not tenant_id:
                 return get_data_error_result(message="Tenant not found!")
-            if settings.docStoreConn.indexExist(search.index_name(tenant_id), doc.kb_id):
+            DocumentService.delete_chunk_images(doc, tenant_id)
+            if settings.docStoreConn.index_exist(search.index_name(tenant_id), doc.kb_id):
                 settings.docStoreConn.delete({"doc_id": doc.id}, search.index_name(tenant_id), doc.kb_id)
+        return None
 
     try:
         if "pipeline_id" in req and req["pipeline_id"] != "":
@@ -571,13 +778,14 @@ def change_parser():
 
 @manager.route("/image/<image_id>", methods=["GET"])  # noqa: F821
 # @login_required
-def get_image(image_id):
+async def get_image(image_id):
     try:
         arr = image_id.split("-")
         if len(arr) != 2:
             return get_data_error_result(message="Image not found.")
         bkt, nm = image_id.split("-")
-        response = flask.make_response(STORAGE_IMPL.get(bkt, nm))
+        data = await asyncio.to_thread(settings.STORAGE_IMPL.get, bkt, nm)
+        response = await make_response(data)
         response.headers.set("Content-Type", "image/JPEG")
         return response
     except Exception as e:
@@ -587,24 +795,26 @@ def get_image(image_id):
 @manager.route("/upload_and_parse", methods=["POST"])  # noqa: F821
 @login_required
 @validate_request("conversation_id")
-def upload_and_parse():
-    if "file" not in request.files:
+async def upload_and_parse():
+    files = await request.files
+    if "file" not in files:
         return get_json_result(data=False, message="No file part!", code=RetCode.ARGUMENT_ERROR)
 
-    file_objs = request.files.getlist("file")
+    file_objs = files.getlist("file")
     for file_obj in file_objs:
         if file_obj.filename == "":
             return get_json_result(data=False, message="No file selected!", code=RetCode.ARGUMENT_ERROR)
 
-    doc_ids = doc_upload_and_parse(request.form.get("conversation_id"), file_objs, current_user.id)
-
+    form = await request.form
+    doc_ids = doc_upload_and_parse(form.get("conversation_id"), file_objs, current_user.id)
     return get_json_result(data=doc_ids)
 
 
 @manager.route("/parse", methods=["POST"])  # noqa: F821
 @login_required
-def parse():
-    url = request.json.get("url") if request.json else ""
+async def parse():
+    req = await get_request_json()
+    url = req.get("url", "")
     if url:
         if not is_valid_url(url):
             return get_json_result(data=False, message="The URL format is invalid", code=RetCode.ARGUMENT_ERROR)
@@ -645,10 +855,11 @@ def parse():
         txt = FileService.parse_docs([f], current_user.id)
         return get_json_result(data=txt)
 
-    if "file" not in request.files:
+    files = await request.files
+    if "file" not in files:
         return get_json_result(data=False, message="No file part!", code=RetCode.ARGUMENT_ERROR)
 
-    file_objs = request.files.getlist("file")
+    file_objs = files.getlist("file")
     txt = FileService.parse_docs(file_objs, current_user.id)
 
     return get_json_result(data=txt)
@@ -657,8 +868,8 @@ def parse():
 @manager.route("/set_meta", methods=["POST"])  # noqa: F821
 @login_required
 @validate_request("doc_id", "meta")
-def set_meta():
-    req = request.json
+async def set_meta():
+    req = await get_request_json()
     if not DocumentService.accessible(req["doc_id"], current_user.id):
         return get_json_result(data=False, message="No authorization.", code=RetCode.AUTHENTICATION_ERROR)
     try:
@@ -666,7 +877,10 @@ def set_meta():
         if not isinstance(meta, dict):
             return get_json_result(data=False, message="Only dictionary type supported.", code=RetCode.ARGUMENT_ERROR)
         for k, v in meta.items():
-            if not isinstance(v, str) and not isinstance(v, int) and not isinstance(v, float):
+            if isinstance(v, list):
+                if not all(isinstance(i, (str, int, float)) for i in v):
+                    return get_json_result(data=False, message=f"The type is not supported in list: {v}", code=RetCode.ARGUMENT_ERROR)
+            elif not isinstance(v, (str, int, float)):
                 return get_json_result(data=False, message=f"The type is not supported: {v}", code=RetCode.ARGUMENT_ERROR)
     except Exception as e:
         return get_json_result(data=False, message=f"Json syntax error: {e}", code=RetCode.ARGUMENT_ERROR)
@@ -684,3 +898,13 @@ def set_meta():
         return get_json_result(data=True)
     except Exception as e:
         return server_error_response(e)
+
+
+@manager.route("/upload_info", methods=["POST"])  # noqa: F821
+async def upload_info():
+    files = await request.files
+    file = files['file'] if files and files.get("file") else None
+    try:
+        return get_json_result(data=FileService.upload_info(current_user.id, file, request.args.get("url")))
+    except Exception as e:
+        return  server_error_response(e)

api/apps/evaluation_app.py

@@ -0,0 +1,479 @@
+#
+#  Copyright 2025 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+#
+
+"""
+RAG Evaluation API Endpoints
+
+Provides REST API for RAG evaluation functionality including:
+- Dataset management
+- Test case management
+- Evaluation execution
+- Results retrieval
+- Configuration recommendations
+"""
+
+from quart import request
+from api.apps import login_required, current_user
+from api.db.services.evaluation_service import EvaluationService
+from api.utils.api_utils import (
+    get_data_error_result,
+    get_json_result,
+    get_request_json,
+    server_error_response,
+    validate_request
+)
+from common.constants import RetCode
+
+
+# ==================== Dataset Management ====================
+
+@manager.route('/dataset/create', methods=['POST'])  # noqa: F821
+@login_required
+@validate_request("name", "kb_ids")
+async def create_dataset():
+    """
+    Create a new evaluation dataset.
+    
+    Request body:
+    {
+        "name": "Dataset name",
+        "description": "Optional description",
+        "kb_ids": ["kb_id1", "kb_id2"]
+    }
+    """
+    try:
+        req = await get_request_json()
+        name = req.get("name", "").strip()
+        description = req.get("description", "")
+        kb_ids = req.get("kb_ids", [])
+        
+        if not name:
+            return get_data_error_result(message="Dataset name cannot be empty")
+        
+        if not kb_ids or not isinstance(kb_ids, list):
+            return get_data_error_result(message="kb_ids must be a non-empty list")
+        
+        success, result = EvaluationService.create_dataset(
+            name=name,
+            description=description,
+            kb_ids=kb_ids,
+            tenant_id=current_user.id,
+            user_id=current_user.id
+        )
+        
+        if not success:
+            return get_data_error_result(message=result)
+        
+        return get_json_result(data={"dataset_id": result})
+    except Exception as e:
+        return server_error_response(e)
+
+
+@manager.route('/dataset/list', methods=['GET'])  # noqa: F821
+@login_required
+async def list_datasets():
+    """
+    List evaluation datasets for current tenant.
+    
+    Query params:
+    - page: Page number (default: 1)
+    - page_size: Items per page (default: 20)
+    """
+    try:
+        page = int(request.args.get("page", 1))
+        page_size = int(request.args.get("page_size", 20))
+        
+        result = EvaluationService.list_datasets(
+            tenant_id=current_user.id,
+            user_id=current_user.id,
+            page=page,
+            page_size=page_size
+        )
+        
+        return get_json_result(data=result)
+    except Exception as e:
+        return server_error_response(e)
+
+
+@manager.route('/dataset/<dataset_id>', methods=['GET'])  # noqa: F821
+@login_required
+async def get_dataset(dataset_id):
+    """Get dataset details by ID"""
+    try:
+        dataset = EvaluationService.get_dataset(dataset_id)
+        if not dataset:
+            return get_data_error_result(
+                message="Dataset not found",
+                code=RetCode.DATA_ERROR
+            )
+        
+        return get_json_result(data=dataset)
+    except Exception as e:
+        return server_error_response(e)
+
+
+@manager.route('/dataset/<dataset_id>', methods=['PUT'])  # noqa: F821
+@login_required
+async def update_dataset(dataset_id):
+    """
+    Update dataset.
+    
+    Request body:
+    {
+        "name": "New name",
+        "description": "New description",
+        "kb_ids": ["kb_id1", "kb_id2"]
+    }
+    """
+    try:
+        req = await get_request_json()
+        
+        # Remove fields that shouldn't be updated
+        req.pop("id", None)
+        req.pop("tenant_id", None)
+        req.pop("created_by", None)
+        req.pop("create_time", None)
+        
+        success = EvaluationService.update_dataset(dataset_id, **req)
+        
+        if not success:
+            return get_data_error_result(message="Failed to update dataset")
+        
+        return get_json_result(data={"dataset_id": dataset_id})
+    except Exception as e:
+        return server_error_response(e)
+
+
+@manager.route('/dataset/<dataset_id>', methods=['DELETE'])  # noqa: F821
+@login_required
+async def delete_dataset(dataset_id):
+    """Delete dataset (soft delete)"""
+    try:
+        success = EvaluationService.delete_dataset(dataset_id)
+        
+        if not success:
+            return get_data_error_result(message="Failed to delete dataset")
+        
+        return get_json_result(data={"dataset_id": dataset_id})
+    except Exception as e:
+        return server_error_response(e)
+
+
+# ==================== Test Case Management ====================
+
+@manager.route('/dataset/<dataset_id>/case/add', methods=['POST'])  # noqa: F821
+@login_required
+@validate_request("question")
+async def add_test_case(dataset_id):
+    """
+    Add a test case to a dataset.
+    
+    Request body:
+    {
+        "question": "Test question",
+        "reference_answer": "Optional ground truth answer",
+        "relevant_doc_ids": ["doc_id1", "doc_id2"],
+        "relevant_chunk_ids": ["chunk_id1", "chunk_id2"],
+        "metadata": {"key": "value"}
+    }
+    """
+    try:
+        req = await get_request_json()
+        question = req.get("question", "").strip()
+        
+        if not question:
+            return get_data_error_result(message="Question cannot be empty")
+        
+        success, result = EvaluationService.add_test_case(
+            dataset_id=dataset_id,
+            question=question,
+            reference_answer=req.get("reference_answer"),
+            relevant_doc_ids=req.get("relevant_doc_ids"),
+            relevant_chunk_ids=req.get("relevant_chunk_ids"),
+            metadata=req.get("metadata")
+        )
+        
+        if not success:
+            return get_data_error_result(message=result)
+        
+        return get_json_result(data={"case_id": result})
+    except Exception as e:
+        return server_error_response(e)
+
+
+@manager.route('/dataset/<dataset_id>/case/import', methods=['POST'])  # noqa: F821
+@login_required
+@validate_request("cases")
+async def import_test_cases(dataset_id):
+    """
+    Bulk import test cases.
+    
+    Request body:
+    {
+        "cases": [
+            {
+                "question": "Question 1",
+                "reference_answer": "Answer 1",
+                ...
+            },
+            {
+                "question": "Question 2",
+                ...
+            }
+        ]
+    }
+    """
+    try:
+        req = await get_request_json()
+        cases = req.get("cases", [])
+        
+        if not cases or not isinstance(cases, list):
+            return get_data_error_result(message="cases must be a non-empty list")
+        
+        success_count, failure_count = EvaluationService.import_test_cases(
+            dataset_id=dataset_id,
+            cases=cases
+        )
+        
+        return get_json_result(data={
+            "success_count": success_count,
+            "failure_count": failure_count,
+            "total": len(cases)
+        })
+    except Exception as e:
+        return server_error_response(e)
+
+
+@manager.route('/dataset/<dataset_id>/cases', methods=['GET'])  # noqa: F821
+@login_required
+async def get_test_cases(dataset_id):
+    """Get all test cases for a dataset"""
+    try:
+        cases = EvaluationService.get_test_cases(dataset_id)
+        return get_json_result(data={"cases": cases, "total": len(cases)})
+    except Exception as e:
+        return server_error_response(e)
+
+
+@manager.route('/case/<case_id>', methods=['DELETE'])  # noqa: F821
+@login_required
+async def delete_test_case(case_id):
+    """Delete a test case"""
+    try:
+        success = EvaluationService.delete_test_case(case_id)
+        
+        if not success:
+            return get_data_error_result(message="Failed to delete test case")
+        
+        return get_json_result(data={"case_id": case_id})
+    except Exception as e:
+        return server_error_response(e)
+
+
+# ==================== Evaluation Execution ====================
+
+@manager.route('/run/start', methods=['POST'])  # noqa: F821
+@login_required
+@validate_request("dataset_id", "dialog_id")
+async def start_evaluation():
+    """
+    Start an evaluation run.
+    
+    Request body:
+    {
+        "dataset_id": "dataset_id",
+        "dialog_id": "dialog_id",
+        "name": "Optional run name"
+    }
+    """
+    try:
+        req = await get_request_json()
+        dataset_id = req.get("dataset_id")
+        dialog_id = req.get("dialog_id")
+        name = req.get("name")
+        
+        success, result = EvaluationService.start_evaluation(
+            dataset_id=dataset_id,
+            dialog_id=dialog_id,
+            user_id=current_user.id,
+            name=name
+        )
+        
+        if not success:
+            return get_data_error_result(message=result)
+        
+        return get_json_result(data={"run_id": result})
+    except Exception as e:
+        return server_error_response(e)
+
+
+@manager.route('/run/<run_id>', methods=['GET'])  # noqa: F821
+@login_required
+async def get_evaluation_run(run_id):
+    """Get evaluation run details"""
+    try:
+        result = EvaluationService.get_run_results(run_id)
+        
+        if not result:
+            return get_data_error_result(
+                message="Evaluation run not found",
+                code=RetCode.DATA_ERROR
+            )
+        
+        return get_json_result(data=result)
+    except Exception as e:
+        return server_error_response(e)
+
+
+@manager.route('/run/<run_id>/results', methods=['GET'])  # noqa: F821
+@login_required
+async def get_run_results(run_id):
+    """Get detailed results for an evaluation run"""
+    try:
+        result = EvaluationService.get_run_results(run_id)
+        
+        if not result:
+            return get_data_error_result(
+                message="Evaluation run not found",
+                code=RetCode.DATA_ERROR
+            )
+        
+        return get_json_result(data=result)
+    except Exception as e:
+        return server_error_response(e)
+
+
+@manager.route('/run/list', methods=['GET'])  # noqa: F821
+@login_required
+async def list_evaluation_runs():
+    """
+    List evaluation runs.
+    
+    Query params:
+    - dataset_id: Filter by dataset (optional)
+    - dialog_id: Filter by dialog (optional)
+    - page: Page number (default: 1)
+    - page_size: Items per page (default: 20)
+    """
+    try:
+        # TODO: Implement list_runs in EvaluationService
+        return get_json_result(data={"runs": [], "total": 0})
+    except Exception as e:
+        return server_error_response(e)
+
+
+@manager.route('/run/<run_id>', methods=['DELETE'])  # noqa: F821
+@login_required
+async def delete_evaluation_run(run_id):
+    """Delete an evaluation run"""
+    try:
+        # TODO: Implement delete_run in EvaluationService
+        return get_json_result(data={"run_id": run_id})
+    except Exception as e:
+        return server_error_response(e)
+
+
+# ==================== Analysis & Recommendations ====================
+
+@manager.route('/run/<run_id>/recommendations', methods=['GET'])  # noqa: F821
+@login_required
+async def get_recommendations(run_id):
+    """Get configuration recommendations based on evaluation results"""
+    try:
+        recommendations = EvaluationService.get_recommendations(run_id)
+        return get_json_result(data={"recommendations": recommendations})
+    except Exception as e:
+        return server_error_response(e)
+
+
+@manager.route('/compare', methods=['POST'])  # noqa: F821
+@login_required
+@validate_request("run_ids")
+async def compare_runs():
+    """
+    Compare multiple evaluation runs.
+    
+    Request body:
+    {
+        "run_ids": ["run_id1", "run_id2", "run_id3"]
+    }
+    """
+    try:
+        req = await get_request_json()
+        run_ids = req.get("run_ids", [])
+        
+        if not run_ids or not isinstance(run_ids, list) or len(run_ids) < 2:
+            return get_data_error_result(
+                message="run_ids must be a list with at least 2 run IDs"
+            )
+        
+        # TODO: Implement compare_runs in EvaluationService
+        return get_json_result(data={"comparison": {}})
+    except Exception as e:
+        return server_error_response(e)
+
+
+@manager.route('/run/<run_id>/export', methods=['GET'])  # noqa: F821
+@login_required
+async def export_results(run_id):
+    """Export evaluation results as JSON/CSV"""
+    try:
+        # format_type = request.args.get("format", "json")  # TODO: Use for CSV export
+        
+        result = EvaluationService.get_run_results(run_id)
+        
+        if not result:
+            return get_data_error_result(
+                message="Evaluation run not found",
+                code=RetCode.DATA_ERROR
+            )
+        
+        # TODO: Implement CSV export
+        return get_json_result(data=result)
+    except Exception as e:
+        return server_error_response(e)
+
+
+# ==================== Real-time Evaluation ====================
+
+@manager.route('/evaluate_single', methods=['POST'])  # noqa: F821
+@login_required
+@validate_request("question", "dialog_id")
+async def evaluate_single():
+    """
+    Evaluate a single question-answer pair in real-time.
+    
+    Request body:
+    {
+        "question": "Test question",
+        "dialog_id": "dialog_id",
+        "reference_answer": "Optional ground truth",
+        "relevant_chunk_ids": ["chunk_id1", "chunk_id2"]
+    }
+    """
+    try:
+        # req = await get_request_json()  # TODO: Use for single evaluation implementation
+        
+        # TODO: Implement single evaluation
+        # This would execute the RAG pipeline and return metrics immediately
+        
+        return get_json_result(data={
+            "answer": "",
+            "metrics": {},
+            "retrieved_chunks": []
+        })
+    except Exception as e:
+        return server_error_response(e)

api/apps/file2document_app.py

@@ -19,22 +19,20 @@ from pathlib import Path
 from api.db.services.file2document_service import File2DocumentService
 from api.db.services.file_service import FileService
 
-from flask import request
-from flask_login import login_required, current_user
+from api.apps import login_required, current_user
 from api.db.services.knowledgebase_service import KnowledgebaseService
-from api.utils.api_utils import server_error_response, get_data_error_result, validate_request
+from api.utils.api_utils import get_data_error_result, get_json_result, get_request_json, server_error_response, validate_request
 from common.misc_utils import get_uuid
 from common.constants import RetCode
 from api.db import FileType
 from api.db.services.document_service import DocumentService
-from api.utils.api_utils import get_json_result
 
 
 @manager.route('/convert', methods=['POST'])  # noqa: F821
 @login_required
 @validate_request("file_ids", "kb_ids")
-def convert():
-    req = request.json
+async def convert():
+    req = await get_request_json()
     kb_ids = req["kb_ids"]
     file_ids = req["file_ids"]
     file2documents = []
@@ -70,7 +68,7 @@ def convert():
                     e, kb = KnowledgebaseService.get_by_id(kb_id)
                     if not e:
                         return get_data_error_result(
-                            message="Can't find this knowledgebase!")
+                            message="Can't find this dataset!")
                     e, file = FileService.get_by_id(id)
                     if not e:
                         return get_data_error_result(
@@ -79,7 +77,8 @@ def convert():
                     doc = DocumentService.insert({
                         "id": get_uuid(),
                         "kb_id": kb.id,
-                        "parser_id": FileService.get_parser(file.type, file.name, kb.parser_id),
+                        "parser_id": kb.parser_id,
+                        "pipeline_id": kb.pipeline_id,
                         "parser_config": kb.parser_config,
                         "created_by": current_user.id,
                         "type": file.type,
@@ -103,8 +102,8 @@ def convert():
 @manager.route('/rm', methods=['POST'])  # noqa: F821
 @login_required
 @validate_request("file_ids")
-def rm():
-    req = request.json
+async def rm():
+    req = await get_request_json()
     file_ids = req["file_ids"]
     if not file_ids:
         return get_json_result(

api/apps/file_app.py

@@ -14,13 +14,12 @@
 #  limitations under the License
 #
 import logging
+import asyncio
 import os
 import pathlib
 import re
-
-import flask
-from flask import request
-from flask_login import login_required, current_user
+from quart import request, make_response
+from api.apps import login_required, current_user
 
 from api.common.check_team_permission import check_file_team_permission
 from api.db.services.document_service import DocumentService
@@ -31,26 +30,28 @@ from common.constants import RetCode, FileSource
 from api.db import FileType
 from api.db.services import duplicate_name
 from api.db.services.file_service import FileService
-from api.utils.api_utils import get_json_result
+from api.utils.api_utils import get_json_result, get_request_json
 from api.utils.file_utils import filename_type
 from api.utils.web_utils import CONTENT_TYPE_MAP
-from rag.utils.storage_factory import STORAGE_IMPL
+from common import settings
 
 
 @manager.route('/upload', methods=['POST'])  # noqa: F821
 @login_required
 # @validate_request("parent_id")
-def upload():
-    pf_id = request.form.get("parent_id")
+async def upload():
+    form = await request.form
+    pf_id = form.get("parent_id")
 
     if not pf_id:
         root_folder = FileService.get_root_folder(current_user.id)
         pf_id = root_folder["id"]
 
-    if 'file' not in request.files:
+    files = await request.files
+    if 'file' not in files:
         return get_json_result(
             data=False, message='No file part!', code=RetCode.ARGUMENT_ERROR)
-    file_objs = request.files.getlist('file')
+    file_objs = files.getlist('file')
 
     for file_obj in file_objs:
         if file_obj.filename == '':
@@ -61,9 +62,10 @@ def upload():
         e, pf_folder = FileService.get_by_id(pf_id)
         if not e:
             return get_data_error_result( message="Can't find this folder!")
-        for file_obj in file_objs:
+
+        async def _handle_single_file(file_obj):
             MAX_FILE_NUM_PER_USER: int = int(os.environ.get('MAX_FILE_NUM_PER_USER', 0))
-            if 0 < MAX_FILE_NUM_PER_USER <= DocumentService.get_doc_count(current_user.id):
+            if 0 < MAX_FILE_NUM_PER_USER <= await asyncio.to_thread(DocumentService.get_doc_count, current_user.id):
                 return get_data_error_result( message="Exceed the maximum file number of a free user!")
 
             # split file name path
@@ -75,35 +77,36 @@ def upload():
             file_len = len(file_obj_names)
 
             # get folder
-            file_id_list = FileService.get_id_list_by_id(pf_id, file_obj_names, 1, [pf_id])
+            file_id_list = await asyncio.to_thread(FileService.get_id_list_by_id, pf_id, file_obj_names, 1, [pf_id])
             len_id_list = len(file_id_list)
 
             # create folder
             if file_len != len_id_list:
-                e, file = FileService.get_by_id(file_id_list[len_id_list - 1])
+                e, file = await asyncio.to_thread(FileService.get_by_id, file_id_list[len_id_list - 1])
                 if not e:
                     return get_data_error_result(message="Folder not found!")
-                last_folder = FileService.create_folder(file, file_id_list[len_id_list - 1], file_obj_names,
+                last_folder = await asyncio.to_thread(FileService.create_folder, file, file_id_list[len_id_list - 1], file_obj_names,
                                                         len_id_list)
             else:
-                e, file = FileService.get_by_id(file_id_list[len_id_list - 2])
+                e, file = await asyncio.to_thread(FileService.get_by_id, file_id_list[len_id_list - 2])
                 if not e:
                     return get_data_error_result(message="Folder not found!")
-                last_folder = FileService.create_folder(file, file_id_list[len_id_list - 2], file_obj_names,
+                last_folder = await asyncio.to_thread(FileService.create_folder, file, file_id_list[len_id_list - 2], file_obj_names,
                                                         len_id_list)
 
             # file type
             filetype = filename_type(file_obj_names[file_len - 1])
             location = file_obj_names[file_len - 1]
-            while STORAGE_IMPL.obj_exist(last_folder.id, location):
+            while await asyncio.to_thread(settings.STORAGE_IMPL.obj_exist, last_folder.id, location):
                 location += "_"
-            blob = file_obj.read()
-            filename = duplicate_name(
+            blob = await asyncio.to_thread(file_obj.read)
+            filename = await asyncio.to_thread(
+                duplicate_name,
                 FileService.query,
                 name=file_obj_names[file_len - 1],
                 parent_id=last_folder.id)
-            STORAGE_IMPL.put(last_folder.id, location, blob)
-            file = {
+            await asyncio.to_thread(settings.STORAGE_IMPL.put, last_folder.id, location, blob)
+            file_data = {
                 "id": get_uuid(),
                 "parent_id": last_folder.id,
                 "tenant_id": current_user.id,
@@ -113,8 +116,13 @@ def upload():
                 "location": location,
                 "size": len(blob),
             }
-            file = FileService.insert(file)
-            file_res.append(file.to_json())
+            inserted = await asyncio.to_thread(FileService.insert, file_data)
+            return inserted.to_json()
+
+        for file_obj in file_objs:
+            res = await _handle_single_file(file_obj)
+            file_res.append(res)
+
         return get_json_result(data=file_res)
     except Exception as e:
         return server_error_response(e)
@@ -123,10 +131,10 @@ def upload():
 @manager.route('/create', methods=['POST'])  # noqa: F821
 @login_required
 @validate_request("name")
-def create():
-    req = request.json
-    pf_id = request.json.get("parent_id")
-    input_file_type = request.json.get("type")
+async def create():
+    req = await get_request_json()
+    pf_id = req.get("parent_id")
+    input_file_type = req.get("type")
     if not pf_id:
         root_folder = FileService.get_root_folder(current_user.id)
         pf_id = root_folder["id"]
@@ -238,59 +246,62 @@ def get_all_parent_folders():
 @manager.route("/rm", methods=["POST"])  # noqa: F821
 @login_required
 @validate_request("file_ids")
-def rm():
-    req = request.json
+async def rm():
+    req = await get_request_json()
     file_ids = req["file_ids"]
 
-    def _delete_single_file(file):
-        try:
-            if file.location:
-                STORAGE_IMPL.rm(file.parent_id, file.location)
-        except Exception:
-            logging.exception(f"Fail to remove object: {file.parent_id}/{file.location}")
-
-        informs = File2DocumentService.get_by_file_id(file.id)
-        for inform in informs:
-            doc_id = inform.document_id
-            e, doc = DocumentService.get_by_id(doc_id)
-            if e and doc:
-                tenant_id = DocumentService.get_tenant_id(doc_id)
-                if tenant_id:
-                    DocumentService.remove_document(doc, tenant_id)
-            File2DocumentService.delete_by_file_id(file.id)
-
-        FileService.delete(file)
-
-    def _delete_folder_recursive(folder, tenant_id):
-        sub_files = FileService.list_all_files_by_parent_id(folder.id)
-        for sub_file in sub_files:
-            if sub_file.type == FileType.FOLDER.value:
-                _delete_folder_recursive(sub_file, tenant_id)
-            else:
-                _delete_single_file(sub_file)
-
-        FileService.delete(folder)
-
     try:
-        for file_id in file_ids:
-            e, file = FileService.get_by_id(file_id)
-            if not e or not file:
-                return get_data_error_result(message="File or Folder not found!")
-            if not file.tenant_id:
-                return get_data_error_result(message="Tenant not found!")
-            if not check_file_team_permission(file, current_user.id):
-                return get_json_result(data=False, message="No authorization.", code=RetCode.AUTHENTICATION_ERROR)
+        def _delete_single_file(file):
+            try:
+                if file.location:
+                    settings.STORAGE_IMPL.rm(file.parent_id, file.location)
+            except Exception as e:
+                logging.exception(f"Fail to remove object: {file.parent_id}/{file.location}, error: {e}")
 
-            if file.source_type == FileSource.KNOWLEDGEBASE:
-                continue
+            informs = File2DocumentService.get_by_file_id(file.id)
+            for inform in informs:
+                doc_id = inform.document_id
+                e, doc = DocumentService.get_by_id(doc_id)
+                if e and doc:
+                    tenant_id = DocumentService.get_tenant_id(doc_id)
+                    if tenant_id:
+                        DocumentService.remove_document(doc, tenant_id)
+                File2DocumentService.delete_by_file_id(file.id)
 
-            if file.type == FileType.FOLDER.value:
-                _delete_folder_recursive(file, current_user.id)
-                continue
+            FileService.delete(file)
 
-            _delete_single_file(file)
+        def _delete_folder_recursive(folder, tenant_id):
+            sub_files = FileService.list_all_files_by_parent_id(folder.id)
+            for sub_file in sub_files:
+                if sub_file.type == FileType.FOLDER.value:
+                    _delete_folder_recursive(sub_file, tenant_id)
+                else:
+                    _delete_single_file(sub_file)
 
-        return get_json_result(data=True)
+            FileService.delete(folder)
+
+        def _rm_sync():
+            for file_id in file_ids:
+                e, file = FileService.get_by_id(file_id)
+                if not e or not file:
+                    return get_data_error_result(message="File or Folder not found!")
+                if not file.tenant_id:
+                    return get_data_error_result(message="Tenant not found!")
+                if not check_file_team_permission(file, current_user.id):
+                    return get_json_result(data=False, message="No authorization.", code=RetCode.AUTHENTICATION_ERROR)
+
+                if file.source_type == FileSource.KNOWLEDGEBASE:
+                    continue
+
+                if file.type == FileType.FOLDER.value:
+                    _delete_folder_recursive(file, current_user.id)
+                    continue
+
+                _delete_single_file(file)
+
+            return get_json_result(data=True)
+
+        return await asyncio.to_thread(_rm_sync)
 
     except Exception as e:
         return server_error_response(e)
@@ -299,8 +310,8 @@ def rm():
 @manager.route('/rename', methods=['POST'])  # noqa: F821
 @login_required
 @validate_request("file_id", "name")
-def rename():
-    req = request.json
+async def rename():
+    req = await get_request_json()
     try:
         e, file = FileService.get_by_id(req["file_id"])
         if not e:
@@ -338,7 +349,7 @@ def rename():
 
 @manager.route('/get/<file_id>', methods=['GET'])  # noqa: F821
 @login_required
-def get(file_id):
+async def get(file_id):
     try:
         e, file = FileService.get_by_id(file_id)
         if not e:
@@ -346,12 +357,12 @@ def get(file_id):
         if not check_file_team_permission(file, current_user.id):
             return get_json_result(data=False, message='No authorization.', code=RetCode.AUTHENTICATION_ERROR)
 
-        blob = STORAGE_IMPL.get(file.parent_id, file.location)
+        blob = await asyncio.to_thread(settings.STORAGE_IMPL.get, file.parent_id, file.location)
         if not blob:
             b, n = File2DocumentService.get_storage_address(file_id=file_id)
-            blob = STORAGE_IMPL.get(b, n)
+            blob = await asyncio.to_thread(settings.STORAGE_IMPL.get, b, n)
 
-        response = flask.make_response(blob)
+        response = await make_response(blob)
         ext = re.search(r"\.([^.]+)$", file.name.lower())
         ext = ext.group(1) if ext else None
         if ext:
@@ -368,8 +379,8 @@ def get(file_id):
 @manager.route("/mv", methods=["POST"])  # noqa: F821
 @login_required
 @validate_request("src_file_ids", "dest_file_id")
-def move():
-    req = request.json
+async def move():
+    req = await get_request_json()
     try:
         file_ids = req["src_file_ids"]
         dest_parent_id = req["dest_file_id"]
@@ -428,11 +439,11 @@ def move():
             filename = source_file_entry.name
 
             new_location = filename
-            while STORAGE_IMPL.obj_exist(dest_folder.id, new_location):
+            while settings.STORAGE_IMPL.obj_exist(dest_folder.id, new_location):
                 new_location += "_"
 
             try:
-                STORAGE_IMPL.move(old_parent_id, old_location, dest_folder.id, new_location)
+                settings.STORAGE_IMPL.move(old_parent_id, old_location, dest_folder.id, new_location)
             except Exception as storage_err:
                 raise RuntimeError(f"Move file failed at storage layer: {str(storage_err)}")
 
@@ -444,10 +455,12 @@ def move():
                 },
             )
 
-        for file in files:
-            _move_entry_recursive(file, dest_folder)
+        def _move_sync():
+            for file in files:
+                _move_entry_recursive(file, dest_folder)
+            return get_json_result(data=True)
 
-        return get_json_result(data=True)
+        return await asyncio.to_thread(_move_sync)
 
     except Exception as e:
         return server_error_response(e)

api/apps/kb_app.py

@@ -16,12 +16,12 @@
 import json
 import logging
 import random
+import re
+import asyncio
 
-from flask import request
-from flask_login import login_required, current_user
+from quart import request
 import numpy as np
 
-
 from api.db.services.connector_service import Connector2KbService
 from api.db.services.llm_service import LLMBundle
 from api.db.services.document_service import DocumentService, queue_raptor_o_graphrag_tasks
@@ -30,37 +30,40 @@ from api.db.services.file_service import FileService
 from api.db.services.pipeline_operation_log_service import PipelineOperationLogService
 from api.db.services.task_service import TaskService, GRAPH_RAPTOR_FAKE_DOC_ID
 from api.db.services.user_service import TenantService, UserTenantService
-from api.utils.api_utils import get_error_data_result, server_error_response, get_data_error_result, validate_request, not_allowed_parameters
+from api.utils.api_utils import get_error_data_result, server_error_response, get_data_error_result, validate_request, not_allowed_parameters, \
+    get_request_json
 from api.db import VALID_FILE_TYPES
 from api.db.services.knowledgebase_service import KnowledgebaseService
 from api.db.db_models import File
 from api.utils.api_utils import get_json_result
-from api import settings
 from rag.nlp import search
 from api.constants import DATASET_NAME_LIMIT
-from rag.settings import PAGERANK_FLD
 from rag.utils.redis_conn import REDIS_CONN
-from rag.utils.storage_factory import STORAGE_IMPL
-from rag.utils.doc_store_conn import OrderByExpr  
-from common.constants import RetCode, PipelineTaskType, StatusEnum, VALID_TASK_STATUS, FileSource, LLMType
+from common.constants import RetCode, PipelineTaskType, StatusEnum, VALID_TASK_STATUS, FileSource, LLMType, PAGERANK_FLD
+from common import settings
+from common.doc_store.doc_store_base import OrderByExpr
+from api.apps import login_required, current_user
 
 
 @manager.route('/create', methods=['post'])  # noqa: F821
 @login_required
 @validate_request("name")
-def create():
-    req = request.json
-    req = KnowledgebaseService.create_with_name(
+async def create():
+    req = await get_request_json()
+    e, res = KnowledgebaseService.create_with_name(
         name = req.pop("name", None),
         tenant_id = current_user.id,
         parser_id = req.pop("parser_id", None),
         **req
-    )        
+    )
+
+    if not e:
+        return res
 
     try:
-        if not KnowledgebaseService.save(**req):
+        if not KnowledgebaseService.save(**res):
             return get_data_error_result()
-        return get_json_result(data={"kb_id":req["id"]})
+        return get_json_result(data={"kb_id":res["id"]})
     except Exception as e:
         return server_error_response(e)
 
@@ -69,8 +72,8 @@ def create():
 @login_required
 @validate_request("kb_id", "name", "description", "parser_id")
 @not_allowed_parameters("id", "tenant_id", "created_by", "create_time", "update_time", "create_date", "update_date", "created_by")
-def update():
-    req = request.json
+async def update():
+    req = await get_request_json()
     if not isinstance(req["name"], str):
         return get_data_error_result(message="Dataset name must be string.")
     if req["name"].strip() == "":
@@ -90,45 +93,91 @@ def update():
         if not KnowledgebaseService.query(
                 created_by=current_user.id, id=req["kb_id"]):
             return get_json_result(
-                data=False, message='Only owner of knowledgebase authorized for this operation.',
+                data=False, message='Only owner of dataset authorized for this operation.',
                 code=RetCode.OPERATING_ERROR)
 
         e, kb = KnowledgebaseService.get_by_id(req["kb_id"])
+
+        # Rename folder in FileService
+        if e and req["name"].lower() != kb.name.lower():
+            FileService.filter_update(
+                [
+                    File.tenant_id == kb.tenant_id,
+                    File.source_type == FileSource.KNOWLEDGEBASE,
+                    File.type == "folder",
+                    File.name == kb.name,
+                ],
+                {"name": req["name"]},
+            )
+
         if not e:
             return get_data_error_result(
-                message="Can't find this knowledgebase!")
+                message="Can't find this dataset!")
 
         if req["name"].lower() != kb.name.lower() \
                 and len(
             KnowledgebaseService.query(name=req["name"], tenant_id=current_user.id, status=StatusEnum.VALID.value)) >= 1:
             return get_data_error_result(
-                message="Duplicated knowledgebase name.")
+                message="Duplicated dataset name.")
 
         del req["kb_id"]
+        connectors = []
+        if "connectors" in req:
+            connectors = req["connectors"]
+            del req["connectors"]
         if not KnowledgebaseService.update_by_id(kb.id, req):
             return get_data_error_result()
 
         if kb.pagerank != req.get("pagerank", 0):
             if req.get("pagerank", 0) > 0:
-                settings.docStoreConn.update({"kb_id": kb.id}, {PAGERANK_FLD: req["pagerank"]},
-                                         search.index_name(kb.tenant_id), kb.id)
+                await asyncio.to_thread(
+                    settings.docStoreConn.update,
+                    {"kb_id": kb.id},
+                    {PAGERANK_FLD: req["pagerank"]},
+                    search.index_name(kb.tenant_id),
+                    kb.id,
+                )
             else:
                 # Elasticsearch requires PAGERANK_FLD be non-zero!
-                settings.docStoreConn.update({"exists": PAGERANK_FLD}, {"remove": PAGERANK_FLD},
-                                         search.index_name(kb.tenant_id), kb.id)
+                await asyncio.to_thread(
+                    settings.docStoreConn.update,
+                    {"exists": PAGERANK_FLD},
+                    {"remove": PAGERANK_FLD},
+                    search.index_name(kb.tenant_id),
+                    kb.id,
+                )
 
         e, kb = KnowledgebaseService.get_by_id(kb.id)
         if not e:
             return get_data_error_result(
                 message="Database error (Knowledgebase rename)!")
+        errors = Connector2KbService.link_connectors(kb.id, [conn for conn in connectors], current_user.id)
+        if errors:
+            logging.error("Link KB errors: ", errors)
         kb = kb.to_dict()
         kb.update(req)
+        kb["connectors"] = connectors
 
         return get_json_result(data=kb)
     except Exception as e:
         return server_error_response(e)
 
 
+@manager.route('/update_metadata_setting', methods=['post'])  # noqa: F821
+@login_required
+@validate_request("kb_id", "metadata")
+async def update_metadata_setting():
+    req = await get_request_json()
+    e, kb = KnowledgebaseService.get_by_id(req["kb_id"])
+    if not e:
+        return get_data_error_result(
+            message="Database error (Knowledgebase rename)!")
+    kb = kb.to_dict()
+    kb["parser_config"]["metadata"] = req["metadata"]
+    KnowledgebaseService.update_by_id(kb["id"], kb)
+    return get_json_result(data=kb)
+
+
 @manager.route('/detail', methods=['GET'])  # noqa: F821
 @login_required
 def detail():
@@ -141,12 +190,12 @@ def detail():
                 break
         else:
             return get_json_result(
-                data=False, message='Only owner of knowledgebase authorized for this operation.',
+                data=False, message='Only owner of dataset authorized for this operation.',
                 code=RetCode.OPERATING_ERROR)
         kb = KnowledgebaseService.get_detail(kb_id)
         if not kb:
             return get_data_error_result(
-                message="Can't find this knowledgebase!")
+                message="Can't find this dataset!")
         kb["size"] = DocumentService.get_total_size_by_kb_id(kb_id=kb["id"],keywords="", run_status=[], types=[])
         kb["connectors"] = Connector2KbService.list_connectors(kb_id)
 
@@ -160,18 +209,19 @@ def detail():
 
 @manager.route('/list', methods=['POST'])  # noqa: F821
 @login_required
-def list_kbs():
-    keywords = request.args.get("keywords", "")
-    page_number = int(request.args.get("page", 0))
-    items_per_page = int(request.args.get("page_size", 0))
-    parser_id = request.args.get("parser_id")
-    orderby = request.args.get("orderby", "create_time")
-    if request.args.get("desc", "true").lower() == "false":
+async def list_kbs():
+    args = request.args
+    keywords = args.get("keywords", "")
+    page_number = int(args.get("page", 0))
+    items_per_page = int(args.get("page_size", 0))
+    parser_id = args.get("parser_id")
+    orderby = args.get("orderby", "create_time")
+    if args.get("desc", "true").lower() == "false":
         desc = False
     else:
         desc = True
 
-    req = request.get_json()
+    req = await get_request_json()
     owner_ids = req.get("owner_ids", [])
     try:
         if not owner_ids:
@@ -193,11 +243,12 @@ def list_kbs():
     except Exception as e:
         return server_error_response(e)
 
+
 @manager.route('/rm', methods=['post'])  # noqa: F821
 @login_required
 @validate_request("kb_id")
-def rm():
-    req = request.json
+async def rm():
+    req = await get_request_json()
     if not KnowledgebaseService.accessible4deletion(req["kb_id"], current_user.id):
         return get_json_result(
             data=False,
@@ -209,28 +260,37 @@ def rm():
             created_by=current_user.id, id=req["kb_id"])
         if not kbs:
             return get_json_result(
-                data=False, message='Only owner of knowledgebase authorized for this operation.',
+                data=False, message='Only owner of dataset authorized for this operation.',
                 code=RetCode.OPERATING_ERROR)
 
-        for doc in DocumentService.query(kb_id=req["kb_id"]):
-            if not DocumentService.remove_document(doc, kbs[0].tenant_id):
+        def _rm_sync():
+            for doc in DocumentService.query(kb_id=req["kb_id"]):
+                if not DocumentService.remove_document(doc, kbs[0].tenant_id):
+                    return get_data_error_result(
+                        message="Database error (Document removal)!")
+                f2d = File2DocumentService.get_by_document_id(doc.id)
+                if f2d:
+                    FileService.filter_delete([File.source_type == FileSource.KNOWLEDGEBASE, File.id == f2d[0].file_id])
+                File2DocumentService.delete_by_document_id(doc.id)
+            FileService.filter_delete(
+                [
+                    File.tenant_id == kbs[0].tenant_id,
+                    File.source_type == FileSource.KNOWLEDGEBASE,
+                    File.type == "folder",
+                    File.name == kbs[0].name,
+                ]
+            )
+            if not KnowledgebaseService.delete_by_id(req["kb_id"]):
                 return get_data_error_result(
-                    message="Database error (Document removal)!")
-            f2d = File2DocumentService.get_by_document_id(doc.id)
-            if f2d:
-                FileService.filter_delete([File.source_type == FileSource.KNOWLEDGEBASE, File.id == f2d[0].file_id])
-            File2DocumentService.delete_by_document_id(doc.id)
-        FileService.filter_delete(
-            [File.source_type == FileSource.KNOWLEDGEBASE, File.type == "folder", File.name == kbs[0].name])
-        if not KnowledgebaseService.delete_by_id(req["kb_id"]):
-            return get_data_error_result(
-                message="Database error (Knowledgebase removal)!")
-        for kb in kbs:
-            settings.docStoreConn.delete({"kb_id": kb.id}, search.index_name(kb.tenant_id), kb.id)
-            settings.docStoreConn.deleteIdx(search.index_name(kb.tenant_id), kb.id)
-            if hasattr(STORAGE_IMPL, 'remove_bucket'):
-                STORAGE_IMPL.remove_bucket(kb.id)
-        return get_json_result(data=True)
+                    message="Database error (Knowledgebase removal)!")
+            for kb in kbs:
+                settings.docStoreConn.delete({"kb_id": kb.id}, search.index_name(kb.tenant_id), kb.id)
+                settings.docStoreConn.delete_idx(search.index_name(kb.tenant_id), kb.id)
+                if hasattr(settings.STORAGE_IMPL, 'remove_bucket'):
+                    settings.STORAGE_IMPL.remove_bucket(kb.id)
+            return get_json_result(data=True)
+
+        return await asyncio.to_thread(_rm_sync)
     except Exception as e:
         return server_error_response(e)
 
@@ -273,8 +333,8 @@ def list_tags_from_kbs():
 
 @manager.route('/<kb_id>/rm_tags', methods=['POST'])  # noqa: F821
 @login_required
-def rm_tags(kb_id):
-    req = request.json
+async def rm_tags(kb_id):
+    req = await get_request_json()
     if not KnowledgebaseService.accessible(kb_id, current_user.id):
         return get_json_result(
             data=False,
@@ -293,8 +353,8 @@ def rm_tags(kb_id):
 
 @manager.route('/<kb_id>/rename_tag', methods=['POST'])  # noqa: F821
 @login_required
-def rename_tags(kb_id):
-    req = request.json
+async def rename_tags(kb_id):
+    req = await get_request_json()
     if not KnowledgebaseService.accessible(kb_id, current_user.id):
         return get_json_result(
             data=False,
@@ -326,7 +386,7 @@ def knowledge_graph(kb_id):
     }
 
     obj = {"graph": {}, "mind_map": {}}
-    if not settings.docStoreConn.indexExist(search.index_name(kb.tenant_id), kb_id):
+    if not settings.docStoreConn.index_exist(search.index_name(kb.tenant_id), kb_id):
         return get_json_result(data=obj)
     sres = settings.retriever.search(req, search.index_name(kb.tenant_id), [kb_id])
     if not len(sres.ids):
@@ -397,7 +457,7 @@ def get_basic_info():
 
 @manager.route("/list_pipeline_logs", methods=["POST"])  # noqa: F821
 @login_required
-def list_pipeline_logs():
+async def list_pipeline_logs():
     kb_id = request.args.get("kb_id")
     if not kb_id:
         return get_json_result(data=False, message='Lack of "KB ID"', code=RetCode.ARGUMENT_ERROR)
@@ -416,7 +476,7 @@ def list_pipeline_logs():
     if create_date_to > create_date_from:
         return get_data_error_result(message="Create data filter is abnormal.")
 
-    req = request.get_json()
+    req = await get_request_json()
 
     operation_status = req.get("operation_status", [])
     if operation_status:
@@ -441,7 +501,7 @@ def list_pipeline_logs():
 
 @manager.route("/list_pipeline_dataset_logs", methods=["POST"])  # noqa: F821
 @login_required
-def list_pipeline_dataset_logs():
+async def list_pipeline_dataset_logs():
     kb_id = request.args.get("kb_id")
     if not kb_id:
         return get_json_result(data=False, message='Lack of "KB ID"', code=RetCode.ARGUMENT_ERROR)
@@ -458,7 +518,7 @@ def list_pipeline_dataset_logs():
     if create_date_to > create_date_from:
         return get_data_error_result(message="Create data filter is abnormal.")
 
-    req = request.get_json()
+    req = await get_request_json()
 
     operation_status = req.get("operation_status", [])
     if operation_status:
@@ -475,12 +535,12 @@ def list_pipeline_dataset_logs():
 
 @manager.route("/delete_pipeline_logs", methods=["POST"])  # noqa: F821
 @login_required
-def delete_pipeline_logs():
+async def delete_pipeline_logs():
     kb_id = request.args.get("kb_id")
     if not kb_id:
         return get_json_result(data=False, message='Lack of "KB ID"', code=RetCode.ARGUMENT_ERROR)
 
-    req = request.get_json()
+    req = await get_request_json()
     log_ids = req.get("log_ids", [])
 
     PipelineOperationLogService.delete_by_ids(log_ids)
@@ -504,8 +564,8 @@ def pipeline_log_detail():
 
 @manager.route("/run_graphrag", methods=["POST"])  # noqa: F821
 @login_required
-def run_graphrag():
-    req = request.json
+async def run_graphrag():
+    req = await get_request_json()
 
     kb_id = req.get("kb_id", "")
     if not kb_id:
@@ -566,15 +626,15 @@ def trace_graphrag():
 
     ok, task = TaskService.get_by_id(task_id)
     if not ok:
-        return get_error_data_result(message="GraphRAG Task Not Found or Error Occurred")
+        return get_json_result(data={})
 
     return get_json_result(data=task.to_dict())
 
 
 @manager.route("/run_raptor", methods=["POST"])  # noqa: F821
 @login_required
-def run_raptor():
-    req = request.json
+async def run_raptor():
+    req = await get_request_json()
 
     kb_id = req.get("kb_id", "")
     if not kb_id:
@@ -642,8 +702,8 @@ def trace_raptor():
 
 @manager.route("/run_mindmap", methods=["POST"])  # noqa: F821
 @login_required
-def run_mindmap():
-    req = request.json
+async def run_mindmap():
+    req = await get_request_json()
 
     kb_id = req.get("kb_id", "")
     if not kb_id:
@@ -726,6 +786,8 @@ def delete_kb_task():
     def cancel_task(task_id):
         REDIS_CONN.set(f"{task_id}-cancel", "x")
 
+    kb_task_id_field: str = ""
+    kb_task_finish_at: str = ""
     match pipeline_task_type:
         case PipelineTaskType.GRAPH_RAG:
             kb_task_id_field = "graphrag_task_id"
@@ -756,7 +818,7 @@ def delete_kb_task():
 
 @manager.route("/check_embedding", methods=["post"])  # noqa: F821
 @login_required
-def check_embedding():
+async def check_embedding():
 
     def _guess_vec_field(src: dict) -> str | None:
         for k in src or {}:
@@ -775,14 +837,14 @@ def check_embedding():
 
     def _to_1d(x):
         a = np.asarray(x, dtype=np.float32)
-        return a.reshape(-1)  
+        return a.reshape(-1)
 
     def _cos_sim(a, b, eps=1e-12):
         a = _to_1d(a)
         b = _to_1d(b)
         na = np.linalg.norm(a)
         nb = np.linalg.norm(b)
-        if na < eps or nb < eps: 
+        if na < eps or nb < eps:
             return 0.0
         return float(np.dot(a, b) / (na * nb))
 
@@ -796,31 +858,31 @@ def check_embedding():
         index_nm = search.index_name(tenant_id)
 
         res0 = docStoreConn.search(
-            selectFields=[], highlightFields=[],
+            select_fields=[], highlight_fields=[],
             condition={"kb_id": kb_id, "available_int": 1},
-            matchExprs=[], orderBy=OrderByExpr(),
+            match_expressions=[], order_by=OrderByExpr(),
             offset=0, limit=1,
-            indexNames=index_nm, knowledgebaseIds=[kb_id]
+            index_names=index_nm, knowledgebase_ids=[kb_id]
         )
-        total = docStoreConn.getTotal(res0)
+        total = docStoreConn.get_total(res0)
         if total <= 0:
             return []
 
         n = min(n, total)
-        offsets = sorted(random.sample(range(total), n))
+        offsets = sorted(random.sample(range(min(total,1000)), n))
         out = []
 
         for off in offsets:
             res1 = docStoreConn.search(
-                selectFields=list(base_fields),
-                highlightFields=[],
+                select_fields=list(base_fields),
+                highlight_fields=[],
                 condition={"kb_id": kb_id, "available_int": 1},
-                matchExprs=[], orderBy=OrderByExpr(),
+                match_expressions=[], order_by=OrderByExpr(),
                 offset=off, limit=1,
-                indexNames=index_nm, knowledgebaseIds=[kb_id]
+                index_names=index_nm, knowledgebase_ids=[kb_id]
             )
-            ids = docStoreConn.getChunkIds(res1)
-            if not ids: 
+            ids = docStoreConn.get_doc_ids(res1)
+            if not ids:
                 continue
 
             cid = ids[0]
@@ -840,9 +902,14 @@ def check_embedding():
                 "position_int": full_doc.get("position_int"),
                 "top_int": full_doc.get("top_int"),
                 "content_with_weight": full_doc.get("content_with_weight") or "",
+                "question_kwd": full_doc.get("question_kwd") or []
             })
         return out
-    req = request.json
+
+    def _clean(s: str) -> str:
+        s = re.sub(r"</?(table|td|caption|tr|th)( [^<>]{0,12})?>", " ", s or "")
+        return s if s else "None"
+    req = await get_request_json()
     kb_id = req.get("kb_id", "")
     embd_id = req.get("embd_id", "")
     n = int(req.get("check_num", 5))
@@ -854,8 +921,10 @@ def check_embedding():
 
     results, eff_sims = [], []
     for ck in samples:
-        txt = (ck.get("content_with_weight") or "").strip()
-        if not txt:
+        title = ck.get("doc_name") or "Title"
+        txt_in = "\n".join(ck.get("question_kwd") or []) or ck.get("content_with_weight") or ""
+        txt_in = _clean(txt_in)
+        if not txt_in:
             results.append({"chunk_id": ck["chunk_id"], "reason": "no_text"})
             continue
 
@@ -864,10 +933,19 @@ def check_embedding():
             continue
 
         try:
-            qv, _ = emb_mdl.encode_queries(txt)  
-            sim = _cos_sim(qv, ck["vector"])
-        except Exception:
-            return get_error_data_result(message="embedding failure")
+            v, _ = emb_mdl.encode([title, txt_in])
+            assert len(v[1]) == len(ck["vector"]), f"The dimension ({len(v[1])}) of given embedding model is different from the original ({len(ck['vector'])})"
+            sim_content = _cos_sim(v[1], ck["vector"])
+            title_w = 0.1
+            qv_mix = title_w * v[0] + (1 - title_w) * v[1]
+            sim_mix = _cos_sim(qv_mix, ck["vector"])
+            sim = sim_content
+            mode = "content_only"
+            if sim_mix > sim:
+                sim = sim_mix
+                mode = "title+content"
+        except Exception as e:
+            return get_error_data_result(message=f"Embedding failure. {e}")
 
         eff_sims.append(sim)
         results.append({
@@ -887,19 +965,8 @@ def check_embedding():
         "avg_cos_sim": round(float(np.mean(eff_sims)) if eff_sims else 0.0, 6),
         "min_cos_sim": round(float(np.min(eff_sims)) if eff_sims else 0.0, 6),
         "max_cos_sim": round(float(np.max(eff_sims)) if eff_sims else 0.0, 6),
+        "match_mode": mode,
     }
-    if summary["avg_cos_sim"] > 0.99:
+    if summary["avg_cos_sim"] > 0.9:
         return get_json_result(data={"summary": summary, "results": results})
-    return get_json_result(code=RetCode.NOT_EFFECTIVE, message="failed", data={"summary": summary, "results": results})
-
-
-@manager.route("/<kb_id>/link", methods=["POST"])  # noqa: F821
-@validate_request("connector_ids")
-@login_required
-def link_connector(kb_id):
-    req = request.json
-    errors = Connector2KbService.link_connectors(kb_id, req["connector_ids"], current_user.id)
-    if errors:
-        return get_json_result(data=False, message=errors, code=RetCode.SERVER_ERROR)
-    return get_json_result(data=True)
-
+    return get_json_result(code=RetCode.NOT_EFFECTIVE, message="Embedding model switch failed: the average similarity between old and new vectors is below 0.9, indicating incompatible vector spaces.", data={"summary": summary, "results": results})

api/apps/langfuse_app.py

@@ -15,28 +15,28 @@
 #
 
 
-from flask import request
-from flask_login import current_user, login_required
+from api.apps import current_user, login_required
 from langfuse import Langfuse
 
 from api.db.db_models import DB
 from api.db.services.langfuse_service import TenantLangfuseService
-from api.utils.api_utils import get_error_data_result, get_json_result, server_error_response, validate_request
+from api.utils.api_utils import get_error_data_result, get_json_result, get_request_json, server_error_response, validate_request
 
 
 @manager.route("/api_key", methods=["POST", "PUT"])  # noqa: F821
 @login_required
 @validate_request("secret_key", "public_key", "host")
-def set_api_key():
-    req = request.get_json()
+async def set_api_key():
+    req = await get_request_json()
     secret_key = req.get("secret_key", "")
     public_key = req.get("public_key", "")
     host = req.get("host", "")
     if not all([secret_key, public_key, host]):
         return get_error_data_result(message="Missing required fields")
 
+    current_user_id = current_user.id
     langfuse_keys = dict(
-        tenant_id=current_user.id,
+        tenant_id=current_user_id,
         secret_key=secret_key,
         public_key=public_key,
         host=host,
@@ -46,23 +46,24 @@ def set_api_key():
     if not langfuse.auth_check():
         return get_error_data_result(message="Invalid Langfuse keys")
 
-    langfuse_entry = TenantLangfuseService.filter_by_tenant(tenant_id=current_user.id)
+    langfuse_entry = TenantLangfuseService.filter_by_tenant(tenant_id=current_user_id)
     with DB.atomic():
         try:
             if not langfuse_entry:
                 TenantLangfuseService.save(**langfuse_keys)
             else:
-                TenantLangfuseService.update_by_tenant(tenant_id=current_user.id, langfuse_keys=langfuse_keys)
+                TenantLangfuseService.update_by_tenant(tenant_id=current_user_id, langfuse_keys=langfuse_keys)
             return get_json_result(data=langfuse_keys)
         except Exception as e:
-            server_error_response(e)
+            return server_error_response(e)
 
 
 @manager.route("/api_key", methods=["GET"])  # noqa: F821
 @login_required
 @validate_request()
 def get_api_key():
-    langfuse_entry = TenantLangfuseService.filter_by_tenant_with_info(tenant_id=current_user.id)
+    current_user_id = current_user.id
+    langfuse_entry = TenantLangfuseService.filter_by_tenant_with_info(tenant_id=current_user_id)
     if not langfuse_entry:
         return get_json_result(message="Have not record any Langfuse keys.")
 
@@ -73,7 +74,7 @@ def get_api_key():
     except langfuse.api.core.api_error.ApiError as api_err:
         return get_json_result(message=f"Error from Langfuse: {api_err}")
     except Exception as e:
-        server_error_response(e)
+        return server_error_response(e)
 
     langfuse_entry["project_id"] = langfuse.api.projects.get().dict()["data"][0]["id"]
     langfuse_entry["project_name"] = langfuse.api.projects.get().dict()["data"][0]["name"]
@@ -85,7 +86,8 @@ def get_api_key():
 @login_required
 @validate_request()
 def delete_api_key():
-    langfuse_entry = TenantLangfuseService.filter_by_tenant(tenant_id=current_user.id)
+    current_user_id = current_user.id
+    langfuse_entry = TenantLangfuseService.filter_by_tenant(tenant_id=current_user_id)
     if not langfuse_entry:
         return get_json_result(message="Have not record any Langfuse keys.")
 
@@ -94,4 +96,4 @@ def delete_api_key():
             TenantLangfuseService.delete_model(langfuse_entry)
             return get_json_result(data=True)
         except Exception as e:
-            server_error_response(e)
+            return server_error_response(e)

api/apps/llm_app.py

@@ -16,16 +16,16 @@
 import logging
 import json
 import os
-from flask import request
-from flask_login import login_required, current_user
+from quart import request
+
+from api.apps import login_required, current_user
 from api.db.services.tenant_llm_service import LLMFactoriesService, TenantLLMService
 from api.db.services.llm_service import LLMService
-from api.utils.api_utils import server_error_response, get_data_error_result, validate_request
+from api.utils.api_utils import get_allowed_llm_factories, get_data_error_result, get_json_result, get_request_json, server_error_response, validate_request
 from common.constants import StatusEnum, LLMType
 from api.db.db_models import TenantLLM
-from api.utils.api_utils import get_json_result, get_allowed_llm_factories
-from common.base64_image import test_image
-from rag.llm import EmbeddingModel, ChatModel, RerankModel, CvModel, TTSModel
+from rag.utils.base64_image import test_image
+from rag.llm import EmbeddingModel, ChatModel, RerankModel, CvModel, TTSModel, OcrModel, Seq2txtModel
 
 
 @manager.route("/factories", methods=["GET"])  # noqa: F821
@@ -33,7 +33,7 @@ from rag.llm import EmbeddingModel, ChatModel, RerankModel, CvModel, TTSModel
 def factories():
     try:
         fac = get_allowed_llm_factories()
-        fac = [f.to_dict() for f in fac if f.name not in ["Youdao", "FastEmbed", "BAAI"]]
+        fac = [f.to_dict() for f in fac if f.name not in ["Youdao", "FastEmbed", "BAAI", "Builtin"]]
         llms = LLMService.get_all()
         mdl_types = {}
         for m in llms:
@@ -43,7 +43,13 @@ def factories():
                 mdl_types[m.fid] = set([])
             mdl_types[m.fid].add(m.model_type)
         for f in fac:
-            f["model_types"] = list(mdl_types.get(f["name"], [LLMType.CHAT, LLMType.EMBEDDING, LLMType.RERANK, LLMType.IMAGE2TEXT, LLMType.SPEECH2TEXT, LLMType.TTS]))
+            f["model_types"] = list(
+                mdl_types.get(
+                    f["name"],
+                    [LLMType.CHAT, LLMType.EMBEDDING, LLMType.RERANK, LLMType.IMAGE2TEXT, LLMType.SPEECH2TEXT, LLMType.TTS, LLMType.OCR],
+                )
+            )
+
         return get_json_result(data=fac)
     except Exception as e:
         return server_error_response(e)
@@ -52,8 +58,8 @@ def factories():
 @manager.route("/set_api_key", methods=["POST"])  # noqa: F821
 @login_required
 @validate_request("llm_factory", "api_key")
-def set_api_key():
-    req = request.json
+async def set_api_key():
+    req = await get_request_json()
     # test if api key works
     chat_passed, embd_passed, rerank_passed = False, False, False
     factory = req["llm_factory"]
@@ -74,7 +80,7 @@ def set_api_key():
             assert factory in ChatModel, f"Chat model from {factory} is not supported yet."
             mdl = ChatModel[factory](req["api_key"], llm.llm_name, base_url=req.get("base_url"), **extra)
             try:
-                m, tc = mdl.chat(None, [{"role": "user", "content": "Hello! How are you doing!"}], {"temperature": 0.9, "max_tokens": 50})
+                m, tc = await mdl.async_chat(None, [{"role": "user", "content": "Hello! How are you doing!"}], {"temperature": 0.9, "max_tokens": 50})
                 if m.find("**ERROR**") >= 0:
                     raise Exception(m)
                 chat_passed = True
@@ -122,13 +128,13 @@ def set_api_key():
 @manager.route("/add_llm", methods=["POST"])  # noqa: F821
 @login_required
 @validate_request("llm_factory")
-def add_llm():
-    req = request.json
+async def add_llm():
+    req = await get_request_json()
     factory = req["llm_factory"]
     api_key = req.get("api_key", "x")
     llm_name = req.get("llm_name")
 
-    if factory not in get_allowed_llm_factories():
+    if factory not in [f.name for f in get_allowed_llm_factories()]:
         return get_data_error_result(message=f"LLM factory {factory} is not allowed")
 
     def apikey_json(keys):
@@ -142,16 +148,16 @@ def add_llm():
 
     elif factory == "Tencent Hunyuan":
         req["api_key"] = apikey_json(["hunyuan_sid", "hunyuan_sk"])
-        return set_api_key()
+        return await set_api_key()
 
     elif factory == "Tencent Cloud":
         req["api_key"] = apikey_json(["tencent_cloud_sid", "tencent_cloud_sk"])
-        return set_api_key()
+        return await set_api_key()
 
     elif factory == "Bedrock":
         # For Bedrock, due to its special authentication method
         # Assemble bedrock_ak, bedrock_sk, bedrock_region
-        api_key = apikey_json(["bedrock_ak", "bedrock_sk", "bedrock_region"])
+        api_key = apikey_json(["auth_mode", "bedrock_ak", "bedrock_sk", "bedrock_region", "aws_role_arn"])
 
     elif factory == "LocalAI":
         llm_name += "___LocalAI"
@@ -186,6 +192,9 @@ def add_llm():
     elif factory == "OpenRouter":
         api_key = apikey_json(["api_key", "provider_order"])
 
+    elif factory == "MinerU":
+        api_key = apikey_json(["api_key", "provider_order"])
+
     llm = {
         "tenant_id": current_user.id,
         "llm_factory": factory,
@@ -199,61 +208,83 @@ def add_llm():
     msg = ""
     mdl_nm = llm["llm_name"].split("___")[0]
     extra = {"provider": factory}
-    if llm["model_type"] == LLMType.EMBEDDING.value:
-        assert factory in EmbeddingModel, f"Embedding model from {factory} is not supported yet."
-        mdl = EmbeddingModel[factory](key=llm["api_key"], model_name=mdl_nm, base_url=llm["api_base"])
-        try:
-            arr, tc = mdl.encode(["Test if the api key is available"])
-            if len(arr[0]) == 0:
-                raise Exception("Fail")
-        except Exception as e:
-            msg += f"\nFail to access embedding model({mdl_nm})." + str(e)
-    elif llm["model_type"] == LLMType.CHAT.value:
-        assert factory in ChatModel, f"Chat model from {factory} is not supported yet."
-        mdl = ChatModel[factory](
-            key=llm["api_key"],
-            model_name=mdl_nm,
-            base_url=llm["api_base"],
-            **extra,
-        )
-        try:
-            m, tc = mdl.chat(None, [{"role": "user", "content": "Hello! How are you doing!"}], {"temperature": 0.9})
-            if not tc and m.find("**ERROR**:") >= 0:
-                raise Exception(m)
-        except Exception as e:
-            msg += f"\nFail to access model({factory}/{mdl_nm})." + str(e)
-    elif llm["model_type"] == LLMType.RERANK:
-        assert factory in RerankModel, f"RE-rank model from {factory} is not supported yet."
-        try:
-            mdl = RerankModel[factory](key=llm["api_key"], model_name=mdl_nm, base_url=llm["api_base"])
-            arr, tc = mdl.similarity("Hello~ RAGFlower!", ["Hi, there!", "Ohh, my friend!"])
-            if len(arr) == 0:
-                raise Exception("Not known.")
-        except KeyError:
-            msg += f"{factory} dose not support this model({factory}/{mdl_nm})"
-        except Exception as e:
-            msg += f"\nFail to access model({factory}/{mdl_nm})." + str(e)
-    elif llm["model_type"] == LLMType.IMAGE2TEXT.value:
-        assert factory in CvModel, f"Image to text model from {factory} is not supported yet."
-        mdl = CvModel[factory](key=llm["api_key"], model_name=mdl_nm, base_url=llm["api_base"])
-        try:
-            image_data = test_image
-            m, tc = mdl.describe(image_data)
-            if not tc and m.find("**ERROR**:") >= 0:
-                raise Exception(m)
-        except Exception as e:
-            msg += f"\nFail to access model({factory}/{mdl_nm})." + str(e)
-    elif llm["model_type"] == LLMType.TTS:
-        assert factory in TTSModel, f"TTS model from {factory} is not supported yet."
-        mdl = TTSModel[factory](key=llm["api_key"], model_name=mdl_nm, base_url=llm["api_base"])
-        try:
-            for resp in mdl.tts("Hello~ RAGFlower!"):
-                pass
-        except RuntimeError as e:
-            msg += f"\nFail to access model({factory}/{mdl_nm})." + str(e)
-    else:
-        # TODO: check other type of models
-        pass
+    model_type = llm["model_type"]
+    model_api_key = llm["api_key"]
+    model_base_url = llm.get("api_base", "")
+    match model_type:
+        case LLMType.EMBEDDING.value:
+            assert factory in EmbeddingModel, f"Embedding model from {factory} is not supported yet."
+            mdl = EmbeddingModel[factory](key=model_api_key, model_name=mdl_nm, base_url=model_base_url)
+            try:
+                arr, tc = mdl.encode(["Test if the api key is available"])
+                if len(arr[0]) == 0:
+                    raise Exception("Fail")
+            except Exception as e:
+                msg += f"\nFail to access embedding model({mdl_nm})." + str(e)
+        case LLMType.CHAT.value:
+            assert factory in ChatModel, f"Chat model from {factory} is not supported yet."
+            mdl = ChatModel[factory](
+                key=model_api_key,
+                model_name=mdl_nm,
+                base_url=model_base_url,
+                **extra,
+            )
+            try:
+                m, tc = await mdl.async_chat(None, [{"role": "user", "content": "Hello! How are you doing!"}],
+                                             {"temperature": 0.9})
+                if not tc and m.find("**ERROR**:") >= 0:
+                    raise Exception(m)
+            except Exception as e:
+                msg += f"\nFail to access model({factory}/{mdl_nm})." + str(e)
+
+        case LLMType.RERANK.value:
+            assert factory in RerankModel, f"RE-rank model from {factory} is not supported yet."
+            try:
+                mdl = RerankModel[factory](key=model_api_key, model_name=mdl_nm, base_url=model_base_url)
+                arr, tc = mdl.similarity("Hello~ RAGFlower!", ["Hi, there!", "Ohh, my friend!"])
+                if len(arr) == 0:
+                    raise Exception("Not known.")
+            except KeyError:
+                msg += f"{factory} dose not support this model({factory}/{mdl_nm})"
+            except Exception as e:
+                msg += f"\nFail to access model({factory}/{mdl_nm})." + str(e)
+
+        case LLMType.IMAGE2TEXT.value:
+            assert factory in CvModel, f"Image to text model from {factory} is not supported yet."
+            mdl = CvModel[factory](key=model_api_key, model_name=mdl_nm, base_url=model_base_url)
+            try:
+                image_data = test_image
+                m, tc = mdl.describe(image_data)
+                if not tc and m.find("**ERROR**:") >= 0:
+                    raise Exception(m)
+            except Exception as e:
+                msg += f"\nFail to access model({factory}/{mdl_nm})." + str(e)
+        case LLMType.TTS.value:
+            assert factory in TTSModel, f"TTS model from {factory} is not supported yet."
+            mdl = TTSModel[factory](key=model_api_key, model_name=mdl_nm, base_url=model_base_url)
+            try:
+                for resp in mdl.tts("Hello~ RAGFlower!"):
+                    pass
+            except RuntimeError as e:
+                msg += f"\nFail to access model({factory}/{mdl_nm})." + str(e)
+        case LLMType.OCR.value:
+            assert factory in OcrModel, f"OCR model from {factory} is not supported yet."
+            try:
+                mdl = OcrModel[factory](key=model_api_key, model_name=mdl_nm, base_url=model_base_url)
+                ok, reason = mdl.check_available()
+                if not ok:
+                    raise RuntimeError(reason or "Model not available")
+            except Exception as e:
+                msg += f"\nFail to access model({factory}/{mdl_nm})." + str(e)
+        case LLMType.SPEECH2TEXT:
+            assert factory in Seq2txtModel, f"Speech model from {factory} is not supported yet."
+            try:
+                mdl = Seq2txtModel[factory](key=model_api_key, model_name=mdl_nm, base_url=model_base_url)
+                # TODO: check the availability
+            except Exception as e:
+                msg += f"\nFail to access model({factory}/{mdl_nm})." + str(e)
+        case _:
+            raise RuntimeError(f"Unknown model type: {model_type}")
 
     if msg:
         return get_data_error_result(message=msg)
@@ -267,8 +298,8 @@ def add_llm():
 @manager.route("/delete_llm", methods=["POST"])  # noqa: F821
 @login_required
 @validate_request("llm_factory", "llm_name")
-def delete_llm():
-    req = request.json
+async def delete_llm():
+    req = await get_request_json()
     TenantLLMService.filter_delete([TenantLLM.tenant_id == current_user.id, TenantLLM.llm_factory == req["llm_factory"], TenantLLM.llm_name == req["llm_name"]])
     return get_json_result(data=True)
 
@@ -276,8 +307,8 @@ def delete_llm():
 @manager.route("/enable_llm", methods=["POST"])  # noqa: F821
 @login_required
 @validate_request("llm_factory", "llm_name")
-def enable_llm():
-    req = request.json
+async def enable_llm():
+    req = await get_request_json()
     TenantLLMService.filter_update(
         [TenantLLM.tenant_id == current_user.id, TenantLLM.llm_factory == req["llm_factory"], TenantLLM.llm_name == req["llm_name"]], {"status": str(req.get("status", "1"))}
     )
@@ -287,8 +318,8 @@ def enable_llm():
 @manager.route("/delete_factory", methods=["POST"])  # noqa: F821
 @login_required
 @validate_request("llm_factory")
-def delete_factory():
-    req = request.json
+async def delete_factory():
+    req = await get_request_json()
     TenantLLMService.filter_delete([TenantLLM.tenant_id == current_user.id, TenantLLM.llm_factory == req["llm_factory"]])
     return get_json_result(data=True)
 
@@ -297,6 +328,7 @@ def delete_factory():
 @login_required
 def my_llms():
     try:
+        TenantLLMService.ensure_mineru_from_env(current_user.id)
         include_details = request.args.get("include_details", "false").lower() == "true"
 
         if include_details:
@@ -344,11 +376,12 @@ def list_app():
     weighted = []
     model_type = request.args.get("model_type")
     try:
+        TenantLLMService.ensure_mineru_from_env(current_user.id)
         objs = TenantLLMService.query(tenant_id=current_user.id)
         facts = set([o.to_dict()["llm_factory"] for o in objs if o.api_key and o.status == StatusEnum.VALID.value])
         status = {(o.llm_name + "@" + o.llm_factory) for o in objs if o.status == StatusEnum.VALID.value}
         llms = LLMService.get_all()
-        llms = [m.to_dict() for m in llms if m.status == StatusEnum.VALID.value and m.fid not in weighted and (m.llm_name + "@" + m.fid) in status]
+        llms = [m.to_dict() for m in llms if m.status == StatusEnum.VALID.value and m.fid not in weighted and (m.fid == 'Builtin' or (m.llm_name + "@" + m.fid) in status)]
         for m in llms:
             m["available"] = m["fid"] in facts or m["llm_name"].lower() == "flag-embedding" or m["fid"] in self_deployed
             if "tei-" in os.getenv("COMPOSE_PROFILES", "") and m["model_type"] == LLMType.EMBEDDING and m["fid"] == "Builtin" and m["llm_name"] == os.getenv("TEI_MODEL", ""):
@@ -358,7 +391,7 @@ def list_app():
         for o in objs:
             if o.llm_name + "@" + o.llm_factory in llm_set:
                 continue
-            llms.append({"llm_name": o.llm_name, "model_type": o.model_type, "fid": o.llm_factory, "available": True})
+            llms.append({"llm_name": o.llm_name, "model_type": o.model_type, "fid": o.llm_factory, "available": True, "status": StatusEnum.VALID.value})
 
         res = {}
         for m in llms: