Feat: Submit clean data operations form data to the backend. #10427 (#11030)

### What problem does this PR solve?

Feat: Submit clean data operations form data to the backend. #10427

### Type of change


- [x] New Feature (non-breaking change which adds functionality)

.github/workflows/release.yml

@@ -16,7 +16,7 @@ concurrency:
 
 jobs:
   release:
-    runs-on: [ "self-hosted", "overseas" ]
+    runs-on: [ "self-hosted", "ragflow-test" ]
     steps:
       - name: Ensure workspace ownership
         run: echo "chown -R $USER $GITHUB_WORKSPACE" && sudo chown -R $USER $GITHUB_WORKSPACE
@@ -25,7 +25,7 @@ jobs:
       - name: Check out code
         uses: actions/checkout@v4
         with:
-          token: ${{ secrets.MY_GITHUB_TOKEN }}  # Use the secret as an environment variable
+          token: ${{ secrets.GITHUB_TOKEN }}  # Use the secret as an environment variable
           fetch-depth: 0
           fetch-tags: true
 
@@ -69,50 +69,26 @@ jobs:
         # https://github.com/actions/upload-release-asset has been replaced by https://github.com/softprops/action-gh-release
         uses: softprops/action-gh-release@v2
         with:
-          token: ${{ secrets.MY_GITHUB_TOKEN }}  # Use the secret as an environment variable
+          token: ${{ secrets.GITHUB_TOKEN }}  # Use the secret as an environment variable
           prerelease: ${{ env.PRERELEASE }}
           tag_name: ${{ env.RELEASE_TAG }}
           # The body field does not support environment variable substitution directly.
           body_path: release_body.md
 
-      # https://github.com/marketplace/actions/docker-login
+      - name: Build and push ragflow-sdk
-      - name: Login to Docker Hub
-        uses: docker/login-action@v3
-        with:
-          username: infiniflow
-          password: ${{ secrets.DOCKERHUB_TOKEN }}
-
-      # https://github.com/marketplace/actions/build-and-push-docker-images
-      - name: Build and push full image
-        uses: docker/build-push-action@v6
-        with:
-          context: .
-          push: true
-          tags: infiniflow/ragflow:${{ env.RELEASE_TAG }}
-          file: Dockerfile
-          platforms: linux/amd64
-
-      # https://github.com/marketplace/actions/build-and-push-docker-images
-      - name: Build and push slim image
-        uses: docker/build-push-action@v6
-        with:
-          context: .
-          push: true
-          tags: infiniflow/ragflow:${{ env.RELEASE_TAG }}-slim
-          file: Dockerfile
-          build-args: LIGHTEN=1
-          platforms: linux/amd64
-
-      - name: Build ragflow-sdk
         if: startsWith(github.ref, 'refs/tags/v')
         run: |
-          cd sdk/python && \
+          cd sdk/python && uv build && uv publish --token ${{ secrets.PYPI_API_TOKEN }}
-          uv build
 
-      - name: Publish package distributions to PyPI
+      - name: Build and push ragflow-cli
         if: startsWith(github.ref, 'refs/tags/v')
-        uses: pypa/gh-action-pypi-publish@release/v1
+        run: |
-        with:
+          cd admin/client && uv build && uv publish --token ${{ secrets.PYPI_API_TOKEN }}
-          packages-dir: sdk/python/dist/
+
-          password: ${{ secrets.PYPI_API_TOKEN }}
+      - name: Build and push image
-          verbose: true
+        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

@@ -10,7 +10,7 @@ on:
       - '*.md'
       - '*.mdx'
   pull_request:
-    types: [ opened, synchronize, reopened, labeled ]
+    types: [ labeled, synchronize, reopened ]
     paths-ignore:
       - 'docs/**'
       - '*.md'
@@ -29,17 +29,15 @@ jobs:
     # 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') }}
-    runs-on: [ "self-hosted", "debug" ]
+    runs-on: [ "self-hosted", "ragflow-test" ]
     steps:
       # https://github.com/hmarr/debug-action
       #- uses: hmarr/debug-action@v2
 
-      - name: Show who triggered this workflow
+      - name: Ensure workspace ownership
         run: |
           echo "Workflow triggered by ${{ github.event_name }}"
-
+          echo "chown -R $USER $GITHUB_WORKSPACE" && sudo chown -R $USER $GITHUB_WORKSPACE
-      - name: Ensure workspace ownership
-        run: echo "chown -R $USER $GITHUB_WORKSPACE" && sudo chown -R $USER $GITHUB_WORKSPACE
 
       # https://github.com/actions/checkout/issues/1781
       - name: Check out code
@@ -48,6 +46,44 @@ jobs:
           fetch-depth: 0
           fetch-tags: true
 
+      - name: Check workflow duplication
+        if: ${{ !cancelled() && !failure() && (github.event_name != 'pull_request' || contains(github.event.pull_request.labels.*.name, 'ci')) }}
+        run: |
+          if [[ "$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 <<< "${{ secrets.GITHUB_TOKEN }}"
+            PR_NUMBER=$(gh pr list --search ${HEAD} --state merged --json number --jq .[0].number)
+            echo "HEAD=${HEAD}"
+            echo "PR_NUMBER=${PR_NUMBER}"
+            if [[ -n "${PR_NUMBER}" ]]; then
+              PR_SHA_FP=${RUNNER_WORKSPACE_PREFIX}/artifacts/${GITHUB_REPOSITORY}/PR_${PR_NUMBER}
+              if [[ -f "${PR_SHA_FP}" ]]; then
+                read -r PR_SHA PR_RUN_ID < "${PR_SHA_FP}"
+                # Calculate the hash of the current workspace content
+                HEAD_SHA=$(git rev-parse HEAD^{tree})
+                if [[ "${HEAD_SHA}" == "${PR_SHA}" ]]; then
+                  echo "Cancel myself since the workspace content hash is the same with PR #${PR_NUMBER} merged. See ${GITHUB_SERVER_URL}/${GITHUB_REPOSITORY}/actions/runs/${PR_RUN_ID} for details."
+                  gh run cancel ${GITHUB_RUN_ID}
+                  while true; do
+                    status=$(gh run view ${GITHUB_RUN_ID} --json status -q .status)
+                    [ "$status" = "completed" ] && break
+                    sleep 5
+                  done
+                  exit 1
+                fi
+              fi
+            fi
+          else
+            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
+            PR_SHA=$(git rev-parse HEAD^{tree})
+            echo "PR #${PR_NUMBER} workspace content hash: ${PR_SHA}"
+            mkdir -p ${RUNNER_WORKSPACE_PREFIX}/artifacts/${GITHUB_REPOSITORY}
+            echo "${PR_SHA} ${GITHUB_RUN_ID}" > ${PR_SHA_FP}
+          fi
+
       # https://github.com/astral-sh/ruff-action
       - name: Static check with Ruff
         uses: astral-sh/ruff-action@v3
@@ -55,122 +91,140 @@ jobs:
           version: ">=0.11.x"
           args: "check"
 
-      - name: Build ragflow:nightly-slim
-        run: |
-          RUNNER_WORKSPACE_PREFIX=${RUNNER_WORKSPACE_PREFIX:-$HOME}
-          sudo docker pull ubuntu:22.04
-          sudo docker build --progress=plain --build-arg LIGHTEN=1 --build-arg NEED_MIRROR=1 -f Dockerfile -t infiniflow/ragflow:nightly-slim .
-
       - name: Build ragflow:nightly
         run: |
-          sudo docker build --progress=plain --build-arg NEED_MIRROR=1 -f Dockerfile -t infiniflow/ragflow:nightly .
+          RUNNER_WORKSPACE_PREFIX=${RUNNER_WORKSPACE_PREFIX:-$HOME}
-
+          RAGFLOW_IMAGE=infiniflow/ragflow:${GITHUB_RUN_ID}
-      - name: Start ragflow:nightly-slim
+          echo "RAGFLOW_IMAGE=${RAGFLOW_IMAGE}" >> $GITHUB_ENV
-        run: |
+          sudo docker pull ubuntu:22.04
-          sudo docker compose -f docker/docker-compose.yml down --volumes --remove-orphans
+          sudo DOCKER_BUILDKIT=1 docker build --build-arg NEED_MIRROR=1 -f Dockerfile -t ${RAGFLOW_IMAGE} .
-          echo -e "\nRAGFLOW_IMAGE=infiniflow/ragflow:nightly-slim" >> docker/.env
+          if [[ "$GITHUB_EVENT_NAME" == "schedule" ]]; then
-          sudo docker compose -f docker/docker-compose.yml up -d
+            export HTTP_API_TEST_LEVEL=p3
-
+          else
-      - name: Stop ragflow:nightly-slim
+            export HTTP_API_TEST_LEVEL=p2
-        if: always()  # always run this step even if previous steps failed
+          fi
-        run: |
+          echo "HTTP_API_TEST_LEVEL=${HTTP_API_TEST_LEVEL}" >> $GITHUB_ENV
-          sudo docker compose -f docker/docker-compose.yml down -v
+          echo "RAGFLOW_CONTAINER=${GITHUB_RUN_ID}-ragflow-cpu-1" >> $GITHUB_ENV
 
       - name: Start ragflow:nightly
         run: |
-          echo -e "\nRAGFLOW_IMAGE=infiniflow/ragflow:nightly" >> docker/.env
+          # Determine runner number (default to 1 if not found)
-          sudo docker compose -f docker/docker-compose.yml up -d
+          RUNNER_NUM=$(sudo docker inspect $(hostname) --format '{{index .Config.Labels "com.docker.compose.container-number"}}' 2>/dev/null || true)
+          RUNNER_NUM=${RUNNER_NUM:-1}
+
+          # Compute port numbers using bash arithmetic
+          ES_PORT=$((1200 + RUNNER_NUM * 10))
+          OS_PORT=$((1201 + RUNNER_NUM * 10))
+          INFINITY_THRIFT_PORT=$((23817 + RUNNER_NUM * 10))
+          INFINITY_HTTP_PORT=$((23820 + RUNNER_NUM * 10))
+          INFINITY_PSQL_PORT=$((5432 + RUNNER_NUM * 10))
+          MYSQL_PORT=$((5455 + RUNNER_NUM * 10))
+          MINIO_PORT=$((9000 + RUNNER_NUM * 10))
+          MINIO_CONSOLE_PORT=$((9001 + RUNNER_NUM * 10))
+          REDIS_PORT=$((6379 + RUNNER_NUM * 10))
+          TEI_PORT=$((6380 + RUNNER_NUM * 10))
+          KIBANA_PORT=$((6601 + RUNNER_NUM * 10))
+          SVR_HTTP_PORT=$((9380 + RUNNER_NUM * 10))
+          ADMIN_SVR_HTTP_PORT=$((9381 + RUNNER_NUM * 10))
+          SVR_MCP_PORT=$((9382 + RUNNER_NUM * 10))
+          SANDBOX_EXECUTOR_MANAGER_PORT=$((9385 + RUNNER_NUM * 10))
+          SVR_WEB_HTTP_PORT=$((80 + RUNNER_NUM * 10))
+          SVR_WEB_HTTPS_PORT=$((443 + RUNNER_NUM * 10))
+
+          # Persist computed ports into docker/.env so docker-compose uses the correct host bindings
+          echo "" >> docker/.env
+          echo -e "ES_PORT=${ES_PORT}" >> docker/.env
+          echo -e "OS_PORT=${OS_PORT}" >> docker/.env
+          echo -e "INFINITY_THRIFT_PORT=${INFINITY_THRIFT_PORT}" >> docker/.env
+          echo -e "INFINITY_HTTP_PORT=${INFINITY_HTTP_PORT}" >> docker/.env
+          echo -e "INFINITY_PSQL_PORT=${INFINITY_PSQL_PORT}" >> docker/.env
+          echo -e "MYSQL_PORT=${MYSQL_PORT}" >> docker/.env
+          echo -e "MINIO_PORT=${MINIO_PORT}" >> docker/.env
+          echo -e "MINIO_CONSOLE_PORT=${MINIO_CONSOLE_PORT}" >> docker/.env
+          echo -e "REDIS_PORT=${REDIS_PORT}" >> docker/.env
+          echo -e "TEI_PORT=${TEI_PORT}" >> docker/.env
+          echo -e "KIBANA_PORT=${KIBANA_PORT}" >> docker/.env
+          echo -e "SVR_HTTP_PORT=${SVR_HTTP_PORT}" >> docker/.env
+          echo -e "ADMIN_SVR_HTTP_PORT=${ADMIN_SVR_HTTP_PORT}" >> docker/.env
+          echo -e "SVR_MCP_PORT=${SVR_MCP_PORT}" >> docker/.env
+          echo -e "SANDBOX_EXECUTOR_MANAGER_PORT=${SANDBOX_EXECUTOR_MANAGER_PORT}" >> docker/.env
+          echo -e "SVR_WEB_HTTP_PORT=${SVR_WEB_HTTP_PORT}" >> docker/.env
+          echo -e "SVR_WEB_HTTPS_PORT=${SVR_WEB_HTTPS_PORT}" >> docker/.env
+
+          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
+
+          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
 
       - 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=""
-          export HOST_ADDRESS=http://host.docker.internal:9380
+          until sudo docker exec ${RAGFLOW_CONTAINER} curl -s --connect-timeout 5 ${HOST_ADDRESS} > /dev/null; do
-          until sudo docker exec ragflow-server curl -s --connect-timeout 5 ${HOST_ADDRESS} > /dev/null; do
             echo "Waiting for service to be available..."
             sleep 5
           done
-          if [[ $GITHUB_EVENT_NAME == 'schedule' ]]; then
+          source .venv/bin/activate && pytest -s --tb=short --level=${HTTP_API_TEST_LEVEL} test/testcases/test_sdk_api
-            export HTTP_API_TEST_LEVEL=p3
-          else
-            export HTTP_API_TEST_LEVEL=p2
-          fi
-          UV_LINK_MODE=copy uv sync --python 3.10 --only-group test --no-default-groups --frozen && uv pip install sdk/python && uv run --only-group test --no-default-groups pytest -s --tb=short --level=${HTTP_API_TEST_LEVEL} test/testcases/test_sdk_api
 
       - 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=""
-          export HOST_ADDRESS=http://host.docker.internal:9380
+          until sudo docker exec ${RAGFLOW_CONTAINER} curl -s --connect-timeout 5 ${HOST_ADDRESS} > /dev/null; do
-          until sudo docker exec ragflow-server curl -s --connect-timeout 5 ${HOST_ADDRESS} > /dev/null; do
             echo "Waiting for service to be available..."
             sleep 5
           done
-          cd sdk/python && UV_LINK_MODE=copy uv sync --python 3.10 --group test --frozen && source .venv/bin/activate && cd test/test_frontend_api && pytest -s --tb=short get_email.py test_dataset.py
+          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
           
       - 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=""
-          export HOST_ADDRESS=http://host.docker.internal:9380
+          until sudo docker exec ${RAGFLOW_CONTAINER} curl -s --connect-timeout 5 ${HOST_ADDRESS} > /dev/null; do
-          until sudo docker exec ragflow-server curl -s --connect-timeout 5 ${HOST_ADDRESS} > /dev/null; do
             echo "Waiting for service to be available..."
             sleep 5
           done
-          if [[ $GITHUB_EVENT_NAME == 'schedule' ]]; then
+          source .venv/bin/activate && pytest -s --tb=short --level=${HTTP_API_TEST_LEVEL} test/testcases/test_http_api
-            export HTTP_API_TEST_LEVEL=p3
-          else
-            export HTTP_API_TEST_LEVEL=p2
-          fi
-          UV_LINK_MODE=copy uv sync --python 3.10 --only-group test --no-default-groups --frozen && uv run --only-group test --no-default-groups pytest -s --tb=short --level=${HTTP_API_TEST_LEVEL} test/testcases/test_http_api
 
       - name: Stop ragflow:nightly
         if: always()  # always run this step even if previous steps failed
         run: |
-          sudo docker compose -f docker/docker-compose.yml down -v
+          sudo docker compose -f docker/docker-compose.yml -p ${GITHUB_RUN_ID} down -v
 
       - name: Start ragflow:nightly
         run: |
-          sudo DOC_ENGINE=infinity docker compose -f docker/docker-compose.yml up -d
+          sed -i '1i DOC_ENGINE=infinity' docker/.env
+          sudo docker compose -f docker/docker-compose.yml -p ${GITHUB_RUN_ID} up -d
 
       - 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=""
-          export HOST_ADDRESS=http://host.docker.internal:9380
+          until sudo docker exec ${RAGFLOW_CONTAINER} curl -s --connect-timeout 5 ${HOST_ADDRESS} > /dev/null; do
-          until sudo docker exec ragflow-server curl -s --connect-timeout 5 ${HOST_ADDRESS} > /dev/null; do
             echo "Waiting for service to be available..."
             sleep 5
           done
-          if [[ $GITHUB_EVENT_NAME == 'schedule' ]]; then
+          source .venv/bin/activate && DOC_ENGINE=infinity pytest -s --tb=short --level=${HTTP_API_TEST_LEVEL} test/testcases/test_sdk_api
-            export HTTP_API_TEST_LEVEL=p3
-          else
-            export HTTP_API_TEST_LEVEL=p2
-          fi
-          UV_LINK_MODE=copy uv sync --python 3.10 --only-group test --no-default-groups --frozen && uv pip install sdk/python && DOC_ENGINE=infinity uv run --only-group test --no-default-groups pytest -s --tb=short --level=${HTTP_API_TEST_LEVEL} test/testcases/test_sdk_api
 
       - 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=""
-          export HOST_ADDRESS=http://host.docker.internal:9380
+          until sudo docker exec ${RAGFLOW_CONTAINER} curl -s --connect-timeout 5 ${HOST_ADDRESS} > /dev/null; do
-          until sudo docker exec ragflow-server curl -s --connect-timeout 5 ${HOST_ADDRESS} > /dev/null; do
             echo "Waiting for service to be available..."
             sleep 5
           done
-          cd sdk/python && UV_LINK_MODE=copy uv sync --python 3.10 --group test --frozen && source .venv/bin/activate && cd test/test_frontend_api && pytest -s --tb=short get_email.py test_dataset.py
+          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
 
       - 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=""
-          export HOST_ADDRESS=http://host.docker.internal:9380
+          until sudo docker exec ${RAGFLOW_CONTAINER} curl -s --connect-timeout 5 ${HOST_ADDRESS} > /dev/null; do
-          until sudo docker exec ragflow-server curl -s --connect-timeout 5 ${HOST_ADDRESS} > /dev/null; do
             echo "Waiting for service to be available..."
             sleep 5
           done
-          if [[ $GITHUB_EVENT_NAME == 'schedule' ]]; then
+          source .venv/bin/activate && DOC_ENGINE=infinity pytest -s --tb=short --level=${HTTP_API_TEST_LEVEL} test/testcases/test_http_api
-            export HTTP_API_TEST_LEVEL=p3
-          else
-            export HTTP_API_TEST_LEVEL=p2
-          fi
-          UV_LINK_MODE=copy uv sync --python 3.10 --only-group test --no-default-groups --frozen && DOC_ENGINE=infinity uv run --only-group test --no-default-groups pytest -s --tb=short --level=${HTTP_API_TEST_LEVEL} test/testcases/test_http_api
 
       - name: Stop ragflow:nightly
         if: always()  # always run this step even if previous steps failed
         run: |
-          sudo DOC_ENGINE=infinity docker compose -f docker/docker-compose.yml down -v
+          sudo docker compose -f docker/docker-compose.yml -p ${GITHUB_RUN_ID} down -v
+          sudo docker rmi -f ${RAGFLOW_IMAGE:-NO_IMAGE} || true

.gitignore

@@ -149,7 +149,7 @@ out
 # Nuxt.js build / generate output
 .nuxt
 dist
-
+ragflow_cli.egg-info
 # Gatsby files
 .cache/
 # Comment in the public line in if your project uses Gatsby and not Next.js

CLAUDE.md

@@ -0,0 +1,116 @@
+# CLAUDE.md
+
+This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.
+
+## Project Overview
+
+RAGFlow is an open-source RAG (Retrieval-Augmented Generation) engine based on deep document understanding. It's a full-stack application with:
+- Python backend (Flask-based API server)
+- React/TypeScript frontend (built with UmiJS)
+- Microservices architecture with Docker deployment
+- Multiple data stores (MySQL, Elasticsearch/Infinity, Redis, MinIO)
+
+## Architecture
+
+### Backend (`/api/`)
+- **Main Server**: `api/ragflow_server.py` - Flask application entry point
+- **Apps**: Modular Flask blueprints in `api/apps/` for different functionalities:
+  - `kb_app.py` - Knowledge base management
+  - `dialog_app.py` - Chat/conversation handling
+  - `document_app.py` - Document processing
+  - `canvas_app.py` - Agent workflow canvas
+  - `file_app.py` - File upload/management
+- **Services**: Business logic in `api/db/services/`
+- **Models**: Database models in `api/db/db_models.py`
+
+### Core Processing (`/rag/`)
+- **Document Processing**: `deepdoc/` - PDF parsing, OCR, layout analysis
+- **LLM Integration**: `rag/llm/` - Model abstractions for chat, embedding, reranking
+- **RAG Pipeline**: `rag/flow/` - Chunking, parsing, tokenization
+- **Graph RAG**: `graphrag/` - Knowledge graph construction and querying
+
+### Agent System (`/agent/`)
+- **Components**: Modular workflow components (LLM, retrieval, categorize, etc.)
+- **Templates**: Pre-built agent workflows in `agent/templates/`
+- **Tools**: External API integrations (Tavily, Wikipedia, SQL execution, etc.)
+
+### Frontend (`/web/`)
+- React/TypeScript with UmiJS framework
+- Ant Design + shadcn/ui components
+- State management with Zustand
+- Tailwind CSS for styling
+
+## Common Development Commands
+
+### Backend Development
+```bash
+# Install Python dependencies
+uv sync --python 3.10 --all-extras
+uv run download_deps.py
+pre-commit install
+
+# Start dependent services
+docker compose -f docker/docker-compose-base.yml up -d
+
+# Run backend (requires services to be running)
+source .venv/bin/activate
+export PYTHONPATH=$(pwd)
+bash docker/launch_backend_service.sh
+
+# Run tests
+uv run pytest
+
+# Linting
+ruff check
+ruff format
+```
+
+### Frontend Development
+```bash
+cd web
+npm install
+npm run dev        # Development server
+npm run build      # Production build
+npm run lint       # ESLint
+npm run test       # Jest tests
+```
+
+### Docker Development
+```bash
+# Full stack with Docker
+cd docker
+docker compose -f docker-compose.yml up -d
+
+# Check server status
+docker logs -f ragflow-server
+
+# Rebuild images
+docker build --platform linux/amd64 -f Dockerfile -t infiniflow/ragflow:nightly .
+```
+
+## Key Configuration Files
+
+- `docker/.env` - Environment variables for Docker deployment
+- `docker/service_conf.yaml.template` - Backend service configuration
+- `pyproject.toml` - Python dependencies and project configuration
+- `web/package.json` - Frontend dependencies and scripts
+
+## Testing
+
+- **Python**: pytest with markers (p1/p2/p3 priority levels)
+- **Frontend**: Jest with React Testing Library
+- **API Tests**: HTTP API and SDK tests in `test/` and `sdk/python/test/`
+
+## Database Engines
+
+RAGFlow supports switching between Elasticsearch (default) and Infinity:
+- Set `DOC_ENGINE=infinity` in `docker/.env` to use Infinity
+- Requires container restart: `docker compose down -v && docker compose up -d`
+
+## Development Environment Requirements
+
+- Python 3.10-3.12
+- Node.js >=18.20.4
+- Docker & Docker Compose
+- uv package manager
+- 16GB+ RAM, 50GB+ disk space

Dockerfile

@@ -4,8 +4,6 @@ USER root
 SHELL ["/bin/bash", "-c"]
 
 ARG NEED_MIRROR=0
-ARG LIGHTEN=0
-ENV LIGHTEN=${LIGHTEN}
 
 WORKDIR /ragflow
 
@@ -17,13 +15,6 @@ RUN --mount=type=bind,from=infiniflow/ragflow_deps:latest,source=/huggingface.co
         /huggingface.co/InfiniFlow/text_concat_xgb_v1.0 \
         /huggingface.co/InfiniFlow/deepdoc \
         | tar -xf - --strip-components=3 -C /ragflow/rag/res/deepdoc 
-RUN --mount=type=bind,from=infiniflow/ragflow_deps:latest,source=/huggingface.co,target=/huggingface.co \
-    if [ "$LIGHTEN" != "1" ]; then \
-        (tar -cf - \
-            /huggingface.co/BAAI/bge-large-zh-v1.5 \
-            /huggingface.co/maidalun1020/bce-embedding-base_v1 \
-            | tar -xf - --strip-components=2 -C /root/.ragflow) \
-    fi
 
 # 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.
@@ -63,11 +54,11 @@ RUN --mount=type=cache,id=ragflow_apt,target=/var/cache/apt,sharing=locked \
     apt install -y ghostscript
 
 RUN if [ "$NEED_MIRROR" == "1" ]; then \
-        pip3 config set global.index-url https://mirrors.aliyun.com/pypi/simple && \
+        pip3 config set global.index-url https://pypi.tuna.tsinghua.edu.cn/simple && \
-        pip3 config set global.trusted-host mirrors.aliyun.com; \
+        pip3 config set global.trusted-host pypi.tuna.tsinghua.edu.cn; \
         mkdir -p /etc/uv && \
         echo "[[index]]" > /etc/uv/uv.toml && \
-        echo 'url = "https://mirrors.aliyun.com/pypi/simple"' >> /etc/uv/uv.toml && \
+        echo 'url = "https://pypi.tuna.tsinghua.edu.cn/simple"' >> /etc/uv/uv.toml && \
         echo "default = true" >> /etc/uv/uv.toml; \
     fi; \
     pipx install uv
@@ -151,15 +142,11 @@ COPY pyproject.toml uv.lock ./
 # uv records index url into uv.lock but doesn't failover among multiple indexes
 RUN --mount=type=cache,id=ragflow_uv,target=/root/.cache/uv,sharing=locked \
     if [ "$NEED_MIRROR" == "1" ]; then \
-        sed -i 's|pypi.org|mirrors.aliyun.com/pypi|g' uv.lock; \
+        sed -i 's|pypi.org|pypi.tuna.tsinghua.edu.cn|g' uv.lock; \
     else \
-        sed -i 's|mirrors.aliyun.com/pypi|pypi.org|g' uv.lock; \
+        sed -i 's|pypi.tuna.tsinghua.edu.cn|pypi.org|g' uv.lock; \
     fi; \
-    if [ "$LIGHTEN" == "1" ]; then \
+    uv sync --python 3.10 --frozen
-        uv sync --python 3.10 --frozen; \
-    else \
-        uv sync --python 3.10 --frozen --all-extras; \
-    fi
 
 COPY web web
 COPY docs docs
@@ -169,11 +156,7 @@ RUN --mount=type=cache,id=ragflow_npm,target=/root/.npm,sharing=locked \
 COPY .git /ragflow/.git
 
 RUN version_info=$(git describe --tags --match=v* --first-parent --always); \
-    if [ "$LIGHTEN" == "1" ]; then \
+    version_info="$version_info"; \
-        version_info="$version_info slim"; \
-    else \
-        version_info="$version_info full"; \
-    fi; \
     echo "RAGFlow version: $version_info"; \
     echo $version_info > /ragflow/VERSION
 
@@ -191,6 +174,7 @@ ENV PATH="${VIRTUAL_ENV}/bin:${PATH}"
 ENV PYTHONPATH=/ragflow/
 
 COPY web web
+COPY admin admin
 COPY api api
 COPY conf conf
 COPY deepdoc deepdoc
@@ -201,6 +185,7 @@ COPY agentic_reasoning agentic_reasoning
 COPY pyproject.toml uv.lock ./
 COPY mcp mcp
 COPY plugin plugin
+COPY common common
 
 COPY docker/service_conf.yaml.template ./conf/service_conf.yaml.template
 COPY docker/entrypoint.sh ./

Dockerfile_tei

@@ -0,0 +1,14 @@
+FROM ghcr.io/huggingface/text-embeddings-inference:cpu-1.8
+
+# uv tool install huggingface_hub
+# hf download --local-dir tei_data/BAAI/bge-small-en-v1.5 BAAI/bge-small-en-v1.5
+# hf download --local-dir tei_data/BAAI/bge-m3 BAAI/bge-m3
+# hf download --local-dir tei_data/Qwen/Qwen3-Embedding-0.6B Qwen/Qwen3-Embedding-0.6B
+COPY tei_data /data
+
+# curl -X POST http://localhost:6380/embed -H "Content-Type: application/json" -d '{"inputs": "Hello, world! This is a test sentence."}'
+# curl -X POST http://tei:80/embed -H "Content-Type: application/json" -d '{"inputs": "Hello, world! This is a test sentence."}'
+# [[-0.058816575,0.019564206,0.026697718,...]]
+
+# curl -X POST http://localhost:6380/v1/embeddings -H "Content-Type: application/json" -d '{"input": "Hello, world! This is a test sentence."}'
+# {"object":"list","data":[{"object":"embedding","embedding":[-0.058816575,0.019564206,...],"index":0}],"model":"BAAI/bge-small-en-v1.5","usage":{"prompt_tokens":12,"total_tokens":12}}

README.md

@@ -1,6 +1,6 @@
 <div align="center">
 <a href="https://demo.ragflow.io/">
-<img src="web/src/assets/logo-with-text.png" width="520" alt="ragflow logo">
+<img src="web/src/assets/logo-with-text.svg" width="520" alt="ragflow logo">
 </a>
 </div>
 
@@ -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:v0.20.5">
+        <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">
     </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">
@@ -43,7 +43,9 @@
   <a href="https://demo.ragflow.io">Demo</a>
 </h4>
 
-#
+<div align="center" style="margin-top:20px;margin-bottom:20px;">
+<img src="https://raw.githubusercontent.com/infiniflow/ragflow-docs/refs/heads/image/image/ragflow-octoverse.png" width="1200"/>
+</div>
 
 <div align="center">
 <a href="https://trendshift.io/repositories/9064" target="_blank"><img src="https://trendshift.io/api/badge/repositories/9064" alt="infiniflow%2Fragflow | Trendshift" style="width: 250px; height: 55px;" width="250" height="55"/></a>
@@ -84,8 +86,9 @@ Try our demo at [https://demo.ragflow.io](https://demo.ragflow.io).
 
 ## 🔥 Latest Updates
 
+- 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.
-- 2025-08-04 Supports new models, including Kimi K2 and Grok 4.
 - 2025-08-01 Supports agentic workflow and MCP.
 - 2025-05-23 Adds a Python/JavaScript code executor component to Agent.
 - 2025-05-05 Supports cross-language query.
@@ -135,7 +138,7 @@ releases! 🌟
 ## 🔎 System Architecture
 
 <div align="center" style="margin-top:20px;margin-bottom:20px;">
-<img src="https://github.com/infiniflow/ragflow/assets/12318111/d6ac5664-c237-4200-a7c2-a4a00691b485" width="1000"/>
+<img src="https://github.com/user-attachments/assets/31b0dd6f-ca4f-445a-9457-70cb44a381b2" width="1000"/>
 </div>
 
 ## 🎬 Get Started
@@ -174,41 +177,42 @@ releases! 🌟
    > ```bash
    > vm.max_map_count=262144
    > ```
-
+   >
 2. Clone the repo:
 
    ```bash
    $ git clone https://github.com/infiniflow/ragflow.git
    ```
-
 3. Start up the server using the pre-built Docker images:
 
 > [!CAUTION]
 > 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 `v0.20.5-slim` edition of the RAGFlow Docker image. See the following table for descriptions of different RAGFlow editions. To download a RAGFlow edition different from `v0.20.5-slim`, update the `RAGFLOW_IMAGE` variable accordingly in **docker/.env** before using `docker compose` to start the server. For example: set `RAGFLOW_IMAGE=infiniflow/ragflow:v0.20.5` for the full edition `v0.20.5`.
+> 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.
 
-   ```bash
+```bash
    $ cd ragflow/docker
    # Use CPU for embedding and DeepDoc tasks:
    $ docker compose -f docker-compose.yml up -d
 
    # To use GPU to accelerate embedding and DeepDoc tasks:
-   # docker compose -f docker-compose-gpu.yml up -d
+   # sed -i '1i DEVICE=gpu' .env
-   ```
+   # docker compose -f docker-compose.yml up -d
+```
 
-   | RAGFlow image tag | Image size (GB) | Has embedding models? | Stable?                  |
+| RAGFlow image tag | Image size (GB) | Has embedding models? | Stable?                    |
-   |-------------------|-----------------|-----------------------|--------------------------|
+| ----------------- | --------------- | --------------------- | -------------------------- |
-   | v0.20.5           | &approx;9       | :heavy_check_mark:    | Stable release           |
+| v0.21.1           | &approx;9       | ✔️                  | Stable release             |
-   | v0.20.5-slim      | &approx;2       | ❌                   | Stable release            |
+| v0.21.1-slim      | &approx;2       | ❌                    | Stable release             |
-   | nightly           | &approx;9       | :heavy_check_mark:    | _Unstable_ nightly build |
+| nightly           | &approx;2       | ❌                    | _Unstable_ nightly build |
-   | nightly-slim      | &approx;2       | ❌                   | _Unstable_ nightly build  |
+
+> Note: 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:
 
    ```bash
-   $ docker logs -f ragflow-server
+   $ docker logs -f docker-ragflow-cpu-1
    ```
 
    _The following output confirms a successful launch of the system:_
@@ -226,14 +230,17 @@ releases! 🌟
 
    > If you skip this confirmation step and directly log in to RAGFlow, your browser may prompt a `network anormal`
    > 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.
+
    > With the default settings, you only need to enter `http://IP_OF_YOUR_MACHINE` (**sans** port number) as the default
    > HTTP serving port `80` can be omitted when using the default configurations.
+   >
 6. In [service_conf.yaml.template](./docker/service_conf.yaml.template), select the desired LLM factory in `user_default_llm` and update
    the `API_KEY` field with the corresponding API key.
 
    > See [llm_api_key_setup](https://ragflow.io/docs/dev/llm_api_key_setup) for more information.
+   >
 
    _The show is on!_
 
@@ -272,7 +279,6 @@ RAGFlow uses Elasticsearch by default for storing full text and vectors. To swit
 > `-v` will delete the docker container volumes, and the existing data will be cleared.
 
 2. Set `DOC_ENGINE` in **docker/.env** to `infinity`.
-
 3. Start the containers:
 
    ```bash
@@ -286,16 +292,6 @@ RAGFlow uses Elasticsearch by default for storing full text and vectors. To swit
 
 This image is approximately 2 GB in size and relies on external LLM and embedding services.
 
-```bash
-git clone https://github.com/infiniflow/ragflow.git
-cd ragflow/
-docker build --platform linux/amd64 --build-arg LIGHTEN=1 -f Dockerfile -t infiniflow/ragflow:nightly-slim .
-```
-
-## 🔧 Build a Docker image including embedding models
-
-This image is approximately 9 GB in size. As it includes embedding models, it relies on external LLM services only.
-
 ```bash
 git clone https://github.com/infiniflow/ragflow.git
 cd ragflow/
@@ -309,17 +305,15 @@ docker build --platform linux/amd64 -f Dockerfile -t infiniflow/ragflow:nightly
    ```bash
    pipx install uv pre-commit
    ```
-
 2. Clone the source code and install Python dependencies:
 
    ```bash
    git clone https://github.com/infiniflow/ragflow.git
    cd ragflow/
-   uv sync --python 3.10 --all-extras # install RAGFlow dependent python modules
+   uv sync --python 3.10 # install RAGFlow dependent python modules
    uv run download_deps.py
    pre-commit install
    ```
-
 3. Launch the dependent services (MinIO, Elasticsearch, Redis, and MySQL) using Docker Compose:
 
    ```bash
@@ -331,24 +325,23 @@ docker build --platform linux/amd64 -f Dockerfile -t infiniflow/ragflow:nightly
    ```
    127.0.0.1       es01 infinity mysql minio redis sandbox-executor-manager
    ```
-
 4. If you cannot access HuggingFace, set the `HF_ENDPOINT` environment variable to use a mirror site:
 
    ```bash
    export HF_ENDPOINT=https://hf-mirror.com
    ```
-
 5. If your operating system does not have jemalloc, please install it as follows:
 
    ```bash
-   # ubuntu
+   # Ubuntu
    sudo apt-get install libjemalloc-dev
-   # centos
+   # CentOS
    sudo yum install jemalloc
-   # mac
+   # OpenSUSE
+   sudo zypper install jemalloc
+   # macOS
    sudo brew install jemalloc
    ```
-
 6. Launch backend service:
 
    ```bash
@@ -356,14 +349,12 @@ docker build --platform linux/amd64 -f Dockerfile -t infiniflow/ragflow:nightly
    export PYTHONPATH=$(pwd)
    bash docker/launch_backend_service.sh
    ```
-
 7. Install frontend dependencies:
 
    ```bash
    cd web
    npm install
    ```
-
 8. Launch frontend service:
 
    ```bash
@@ -373,14 +364,12 @@ docker build --platform linux/amd64 -f Dockerfile -t infiniflow/ragflow:nightly
    _The following output confirms a successful launch of the system:_
 
    ![](https://github.com/user-attachments/assets/0daf462c-a24d-4496-a66f-92533534e187)
-
 9. Stop RAGFlow front-end and back-end service after development is complete:
 
    ```bash
    pkill -f "ragflow_server.py|task_executor.py"
    ```
 
-
 ## 📚 Documentation
 
 - [Quickstart](https://ragflow.io/docs/dev/)

README_id.md

@@ -1,6 +1,6 @@
 <div align="center">
 <a href="https://demo.ragflow.io/">
-<img src="web/src/assets/logo-with-text.png" width="520" alt="Logo ragflow">
+<img src="web/src/assets/logo-with-text.svg" width="520" alt="Logo ragflow">
 </a>
 </div>
 
@@ -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:v0.20.5">
+        <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">
     </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">
@@ -43,7 +43,13 @@
   <a href="https://demo.ragflow.io">Demo</a>
 </h4>
 
-#
+<div align="center" style="margin-top:20px;margin-bottom:20px;">
+<img src="https://raw.githubusercontent.com/infiniflow/ragflow-docs/refs/heads/image/image/ragflow-octoverse.png" width="1200"/>
+</div>
+
+<div align="center">
+<a href="https://trendshift.io/repositories/9064" target="_blank"><img src="https://trendshift.io/api/badge/repositories/9064" alt="infiniflow%2Fragflow | Trendshift" style="width: 250px; height: 55px;" width="250" height="55"/></a>
+</div>
 
 <details open>
 <summary><b>📕 Daftar Isi </b> </summary>
@@ -80,8 +86,9 @@ Coba demo kami di [https://demo.ragflow.io](https://demo.ragflow.io).
 
 ## 🔥 Pembaruan Terbaru
 
+- 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.
-- 2025-08-04 Mendukung model baru, termasuk Kimi K2 dan Grok 4.
 - 2025-08-01 Mendukung alur kerja agen dan MCP.
 - 2025-05-23 Menambahkan komponen pelaksana kode Python/JS ke Agen.
 - 2025-05-05 Mendukung kueri lintas bahasa.
@@ -129,7 +136,7 @@ Coba demo kami di [https://demo.ragflow.io](https://demo.ragflow.io).
 ## 🔎 Arsitektur Sistem
 
 <div align="center" style="margin-top:20px;margin-bottom:20px;">
-<img src="https://github.com/infiniflow/ragflow/assets/12318111/d6ac5664-c237-4200-a7c2-a4a00691b485" width="1000"/>
+<img src="https://github.com/user-attachments/assets/31b0dd6f-ca4f-445a-9457-70cb44a381b2" width="1000"/>
 </div>
 
 ## 🎬 Mulai
@@ -168,41 +175,42 @@ Coba demo kami di [https://demo.ragflow.io](https://demo.ragflow.io).
    > ```bash
    > vm.max_map_count=262144
    > ```
-
+   >
 2. Clone repositori:
 
    ```bash
    $ git clone https://github.com/infiniflow/ragflow.git
    ```
-
 3. Bangun image Docker pre-built dan jalankan server:
 
 > [!CAUTION]
 > 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 v0.20.5-slim dari gambar Docker RAGFlow. Silakan merujuk ke tabel berikut untuk deskripsi berbagai edisi RAGFlow. Untuk mengunduh edisi RAGFlow yang berbeda dari v0.20.5-slim, perbarui variabel RAGFLOW_IMAGE di docker/.env sebelum menggunakan docker compose untuk memulai server. Misalnya, atur RAGFLOW_IMAGE=infiniflow/ragflow:v0.20.5 untuk edisi lengkap v0.20.5.
+> 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.
 
 ```bash
-$ cd ragflow/docker
+   $ cd ragflow/docker
-# Use CPU for embedding and DeepDoc tasks:
+   # Use CPU for embedding and DeepDoc tasks:
-$ docker compose -f docker-compose.yml up -d
+   $ docker compose -f docker-compose.yml up -d
 
-# To use GPU to accelerate embedding and DeepDoc tasks:
+   # To use GPU to accelerate embedding and DeepDoc tasks:
-# docker compose -f docker-compose-gpu.yml up -d
+   # sed -i '1i DEVICE=gpu' .env
+   # docker compose -f docker-compose.yml up -d
 ```
 
-| RAGFlow image tag | Image size (GB) | Has embedding models? | Stable?                  |
+| RAGFlow image tag | Image size (GB) | Has embedding models? | Stable?                    |
-| ----------------- | --------------- | --------------------- | ------------------------ |
+| ----------------- | --------------- | --------------------- | -------------------------- |
-| v0.20.5           | &approx;9       | :heavy_check_mark:    | Stable release           |
+| v0.21.1           | &approx;9       | ✔️                  | Stable release             |
-| v0.20.5-slim      | &approx;2       | ❌                    | Stable release           |
+| v0.21.1-slim      | &approx;2       | ❌                    | Stable release             |
-| nightly           | &approx;9       | :heavy_check_mark:    | _Unstable_ nightly build |
+| nightly           | &approx;2       | ❌                    | _Unstable_ nightly build |
-| nightly-slim      | &approx;2       | ❌                    | _Unstable_ nightly build |
+
+> Catatan: 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:
 
    ```bash
-   $ docker logs -f ragflow-server
+   $ docker logs -f docker-ragflow-cpu-1
    ```
 
    _Output berikut menandakan bahwa sistem berhasil diluncurkan:_
@@ -220,14 +228,17 @@ $ docker compose -f docker-compose.yml up -d
 
    > Jika Anda melewatkan langkah ini dan langsung login ke RAGFlow, browser Anda mungkin menampilkan error `network anormal`
    > karena RAGFlow mungkin belum sepenuhnya siap.
-
+   >
 2. Buka browser web Anda, masukkan alamat IP server Anda, dan login ke RAGFlow.
+
    > Dengan pengaturan default, Anda hanya perlu memasukkan `http://IP_DEVICE_ANDA` (**tanpa** nomor port) karena
    > port HTTP default `80` bisa dihilangkan saat menggunakan konfigurasi default.
+   >
 3. Dalam [service_conf.yaml.template](./docker/service_conf.yaml.template), pilih LLM factory yang diinginkan di `user_default_llm` dan perbarui
    bidang `API_KEY` dengan kunci API yang sesuai.
 
    > Lihat [llm_api_key_setup](https://ragflow.io/docs/dev/llm_api_key_setup) untuk informasi lebih lanjut.
+   >
 
    _Sistem telah siap digunakan!_
 
@@ -253,16 +264,6 @@ Pembaruan konfigurasi ini memerlukan reboot semua kontainer agar efektif:
 
 Image ini berukuran sekitar 2 GB dan bergantung pada aplikasi LLM eksternal dan embedding.
 
-```bash
-git clone https://github.com/infiniflow/ragflow.git
-cd ragflow/
-docker build --platform linux/amd64 --build-arg LIGHTEN=1 -f Dockerfile -t infiniflow/ragflow:nightly-slim .
-```
-
-## 🔧 Membangun Docker Image Termasuk Model Embedding
-
-Image ini berukuran sekitar 9 GB. Karena sudah termasuk model embedding, ia hanya bergantung pada aplikasi LLM eksternal.
-
 ```bash
 git clone https://github.com/infiniflow/ragflow.git
 cd ragflow/
@@ -276,17 +277,15 @@ docker build --platform linux/amd64 -f Dockerfile -t infiniflow/ragflow:nightly
    ```bash
    pipx install uv pre-commit
    ```
-
 2. Clone kode sumber dan instal dependensi Python:
 
    ```bash
    git clone https://github.com/infiniflow/ragflow.git
    cd ragflow/
-   uv sync --python 3.10 --all-extras # install RAGFlow dependent python modules
+   uv sync --python 3.10 # install RAGFlow dependent python modules
    uv run download_deps.py
    pre-commit install
    ```
-
 3. Jalankan aplikasi yang diperlukan (MinIO, Elasticsearch, Redis, dan MySQL) menggunakan Docker Compose:
 
    ```bash
@@ -298,13 +297,11 @@ docker build --platform linux/amd64 -f Dockerfile -t infiniflow/ragflow:nightly
    ```
    127.0.0.1       es01 infinity mysql minio redis sandbox-executor-manager
    ```
-
 4. Jika Anda tidak dapat mengakses HuggingFace, atur variabel lingkungan `HF_ENDPOINT` untuk menggunakan situs mirror:
 
    ```bash
    export HF_ENDPOINT=https://hf-mirror.com
    ```
-
 5. Jika sistem operasi Anda tidak memiliki jemalloc, instal sebagai berikut:
 
    ```bash
@@ -315,7 +312,6 @@ docker build --platform linux/amd64 -f Dockerfile -t infiniflow/ragflow:nightly
    # mac
    sudo brew install jemalloc
    ```
-
 6. Jalankan aplikasi backend:
 
    ```bash
@@ -323,14 +319,12 @@ docker build --platform linux/amd64 -f Dockerfile -t infiniflow/ragflow:nightly
    export PYTHONPATH=$(pwd)
    bash docker/launch_backend_service.sh
    ```
-
 7. Instal dependensi frontend:
 
    ```bash
    cd web
    npm install
    ```
-
 8. Jalankan aplikasi frontend:
 
    ```bash
@@ -340,15 +334,12 @@ docker build --platform linux/amd64 -f Dockerfile -t infiniflow/ragflow:nightly
    _Output berikut menandakan bahwa sistem berhasil diluncurkan:_
 
    ![](https://github.com/user-attachments/assets/0daf462c-a24d-4496-a66f-92533534e187)
-
-
 9. Hentikan layanan front-end dan back-end RAGFlow setelah pengembangan selesai:
 
    ```bash
    pkill -f "ragflow_server.py|task_executor.py"
    ```
 
-
 ## 📚 Dokumentasi
 
 - [Quickstart](https://ragflow.io/docs/dev/)

README_ja.md

@@ -1,6 +1,6 @@
 <div align="center">
 <a href="https://demo.ragflow.io/">
-<img src="web/src/assets/logo-with-text.png" width="350" alt="ragflow logo">
+<img src="web/src/assets/logo-with-text.svg" width="350" alt="ragflow logo">
 </a>
 </div>
 
@@ -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:v0.20.5">
+        <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">
     </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">
@@ -43,7 +43,13 @@
   <a href="https://demo.ragflow.io">Demo</a>
 </h4>
 
-#
+<div align="center" style="margin-top:20px;margin-bottom:20px;">
+<img src="https://raw.githubusercontent.com/infiniflow/ragflow-docs/refs/heads/image/image/ragflow-octoverse.png" width="1200"/>
+</div>
+
+<div align="center">
+<a href="https://trendshift.io/repositories/9064" target="_blank"><img src="https://trendshift.io/api/badge/repositories/9064" alt="infiniflow%2Fragflow | Trendshift" style="width: 250px; height: 55px;" width="250" height="55"/></a>
+</div>
 
 ## 💡 RAGFlow とは？
 
@@ -60,8 +66,9 @@
 
 ## 🔥 最新情報
 
+- 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。
-- 2025-08-04 新モデル、キミK2およびGrok 4をサポート。
 - 2025-08-01 エージェントワークフローとMCPをサポート。
 - 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。
@@ -109,7 +116,7 @@
 ## 🔎 システム構成
 
 <div align="center" style="margin-top:20px;margin-bottom:20px;">
-<img src="https://github.com/infiniflow/ragflow/assets/12318111/d6ac5664-c237-4200-a7c2-a4a00691b485" width="1000"/>
+<img src="https://github.com/user-attachments/assets/31b0dd6f-ca4f-445a-9457-70cb44a381b2" width="1000"/>
 </div>
 
 ## 🎬 初期設定
@@ -147,41 +154,42 @@
    > ```bash
    > vm.max_map_count=262144
    > ```
-
+   >
 2. リポジトリをクローンする:
 
    ```bash
    $ git clone https://github.com/infiniflow/ragflow.git
    ```
-
 3. ビルド済みの Docker イメージをビルドし、サーバーを起動する:
 
 > [!CAUTION]
 > 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 v0.20.5-slim 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。v0.20.5-slim 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。例えば、完全版 v0.20.5 をダウンロードするには、RAGFLOW_IMAGE=infiniflow/ragflow:v0.20.5 と設定します。
+> 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。
 
-   ```bash
+```bash
    $ cd ragflow/docker
    # Use CPU for embedding and DeepDoc tasks:
    $ docker compose -f docker-compose.yml up -d
 
    # To use GPU to accelerate embedding and DeepDoc tasks:
-   # docker compose -f docker-compose-gpu.yml up -d
+   # sed -i '1i DEVICE=gpu' .env
-   ```
+   # docker compose -f docker-compose.yml up -d
+```
 
-   | RAGFlow image tag | Image size (GB) | Has embedding models? | Stable?                  |
+| RAGFlow image tag | Image size (GB) | Has embedding models? | Stable?                    |
-   | ----------------- | --------------- | --------------------- | ------------------------ |
+| ----------------- | --------------- | --------------------- | -------------------------- |
-   | v0.20.5           | &approx;9       | :heavy_check_mark:    | Stable release           |
+| v0.21.1           | &approx;9       | ✔️                  | Stable release             |
-   | v0.20.5-slim      | &approx;2       | ❌                    | Stable release           |
+| v0.21.1-slim      | &approx;2       | ❌                    | Stable release             |
-   | nightly           | &approx;9       | :heavy_check_mark:    | _Unstable_ nightly build |
+| nightly           | &approx;2       | ❌                    | _Unstable_ nightly build |
-   | nightly-slim      | &approx;2       | ❌                     | _Unstable_ nightly build |
+
+> 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。
 
 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 ragflow-server
+   $ docker logs -f docker-ragflow-cpu-1
    ```
 
    _以下の出力は、システムが正常に起動したことを確認するものです:_
@@ -197,12 +205,15 @@
    ```
 
    > もし確認ステップをスキップして直接 RAGFlow にログインした場合、その時点で RAGFlow が完全に初期化されていない可能性があるため、ブラウザーがネットワーク異常エラーを表示するかもしれません。
-
+   >
 2. ウェブブラウザで、プロンプトに従ってサーバーの IP アドレスを入力し、RAGFlow にログインします。
+
    > デフォルトの設定を使用する場合、デフォルトの HTTP サービングポート `80` は省略できるので、与えられたシナリオでは、`http://IP_OF_YOUR_MACHINE`（ポート番号は省略）だけを入力すればよい。
+   >
 3. [service_conf.yaml.template](./docker/service_conf.yaml.template) で、`user_default_llm` で希望の LLM ファクトリを選択し、`API_KEY` フィールドを対応する API キーで更新する。
 
    > 詳しくは [llm_api_key_setup](https://ragflow.io/docs/dev/llm_api_key_setup) を参照してください。
+   >
 
    _これで初期設定完了！ショーの開幕です！_
 
@@ -231,33 +242,27 @@
 RAGFlow はデフォルトで Elasticsearch を使用して全文とベクトルを保存します。［Infinity］に切り替え（https://github.com/infiniflow/infinity/)、次の手順に従います。
 
 1. 実行中のすべてのコンテナを停止するには：
+
    ```bash
    $ docker compose -f docker/docker-compose.yml down -v
    ```
+
    Note: `-v` は docker コンテナのボリュームを削除し、既存のデータをクリアします。
 2. **docker/.env** の「DOC \_ ENGINE」を「infinity」に設定します。
-
 3. 起動コンテナ：
+
    ```bash
    $ docker compose -f docker-compose.yml up -d
    ```
+
    > [!WARNING]
    > 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）
 
 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。
 
-```bash
-git clone https://github.com/infiniflow/ragflow.git
-cd ragflow/
-docker build --platform linux/amd64 --build-arg LIGHTEN=1 -f Dockerfile -t infiniflow/ragflow:nightly-slim .
-```
-
-## 🔧 ソースコードをコンパイルした Docker イメージ（埋め込みモデルを含む）
-
-この Docker のサイズは約 9GB で、埋め込みモデルを含むため、外部の大モデルサービスのみが必要です。
-
 ```bash
 git clone https://github.com/infiniflow/ragflow.git
 cd ragflow/
@@ -271,17 +276,15 @@ docker build --platform linux/amd64 -f Dockerfile -t infiniflow/ragflow:nightly
    ```bash
    pipx install uv pre-commit
    ```
-
 2. ソースコードをクローンし、Python の依存関係をインストールする:
 
    ```bash
    git clone https://github.com/infiniflow/ragflow.git
    cd ragflow/
-   uv sync --python 3.10 --all-extras # install RAGFlow dependent python modules
+   uv sync --python 3.10 # install RAGFlow dependent python modules
    uv run download_deps.py
    pre-commit install
    ```
-
 3. Docker Compose を使用して依存サービス（MinIO、Elasticsearch、Redis、MySQL）を起動する:
 
    ```bash
@@ -293,13 +296,11 @@ docker build --platform linux/amd64 -f Dockerfile -t infiniflow/ragflow:nightly
    ```
    127.0.0.1       es01 infinity mysql minio redis sandbox-executor-manager
    ```
-
 4. HuggingFace にアクセスできない場合は、`HF_ENDPOINT` 環境変数を設定してミラーサイトを使用してください:
 
    ```bash
    export HF_ENDPOINT=https://hf-mirror.com
    ```
-
 5. オペレーティングシステムにjemallocがない場合は、次のようにインストールします:
 
    ```bash
@@ -310,7 +311,6 @@ docker build --platform linux/amd64 -f Dockerfile -t infiniflow/ragflow:nightly
    # mac
    sudo brew install jemalloc
    ```
-
 6. バックエンドサービスを起動する:
 
    ```bash
@@ -318,14 +318,12 @@ docker build --platform linux/amd64 -f Dockerfile -t infiniflow/ragflow:nightly
    export PYTHONPATH=$(pwd)
    bash docker/launch_backend_service.sh
    ```
-
 7. フロントエンドの依存関係をインストールする:
 
    ```bash
    cd web
    npm install
    ```
-
 8. フロントエンドサービスを起動する:
 
    ```bash
@@ -335,14 +333,12 @@ docker build --platform linux/amd64 -f Dockerfile -t infiniflow/ragflow:nightly
    _以下の画面で、システムが正常に起動したことを示します:_
 
    ![](https://github.com/user-attachments/assets/0daf462c-a24d-4496-a66f-92533534e187)
-
 9. 開発が完了したら、RAGFlow のフロントエンド サービスとバックエンド サービスを停止します:
 
    ```bash
    pkill -f "ragflow_server.py|task_executor.py"
    ```
 
-
 ## 📚 ドキュメンテーション
 
 - [Quickstart](https://ragflow.io/docs/dev/)

README_ko.md

@@ -1,6 +1,6 @@
 <div align="center">
 <a href="https://demo.ragflow.io/">
-<img src="web/src/assets/logo-with-text.png" width="520" alt="ragflow logo">
+<img src="web/src/assets/logo-with-text.svg" width="520" alt="ragflow logo">
 </a>
 </div>
 
@@ -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:v0.20.5">
+        <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">
     </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">
@@ -43,7 +43,14 @@
   <a href="https://demo.ragflow.io">Demo</a>
 </h4>
 
-#
+<div align="center" style="margin-top:20px;margin-bottom:20px;">
+<img src="https://raw.githubusercontent.com/infiniflow/ragflow-docs/refs/heads/image/image/ragflow-octoverse.png" width="1200"/>
+</div>
+
+<div align="center">
+<a href="https://trendshift.io/repositories/9064" target="_blank"><img src="https://trendshift.io/api/badge/repositories/9064" alt="infiniflow%2Fragflow | Trendshift" style="width: 250px; height: 55px;" width="250" height="55"/></a>
+</div>
+
 
 ## 💡 RAGFlow란?
 
@@ -60,8 +67,9 @@
 
 ## 🔥 업데이트
 
+- 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.
-- 2025-08-04 새로운 모델인 Kimi K2와 Grok 4를 포함하여 지원합니다.
 - 2025-08-01 에이전트 워크플로우와 MCP를 지원합니다.
 - 2025-05-23 Agent에 Python/JS 코드 실행기 구성 요소를 추가합니다.
 - 2025-05-05 언어 간 쿼리를 지원합니다.
@@ -109,7 +117,7 @@
 ## 🔎 시스템 아키텍처
 
 <div align="center" style="margin-top:20px;margin-bottom:20px;">
-<img src="https://github.com/infiniflow/ragflow/assets/12318111/d6ac5664-c237-4200-a7c2-a4a00691b485" width="1000"/>
+<img src="https://github.com/user-attachments/assets/31b0dd6f-ca4f-445a-9457-70cb44a381b2" width="1000"/>
 </div>
 
 ## 🎬 시작하기
@@ -160,7 +168,7 @@
 > 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 v0.20.5-slim 버전을 다운로드합니다. 다양한 RAGFlow 버전에 대한 설명은 다음 표를 참조하십시오. v0.20.5-slim과 다른 RAGFlow 버전을 다운로드하려면, docker/.env 파일에서 RAGFLOW_IMAGE 변수를 적절히 업데이트한 후 docker compose를 사용하여 서버를 시작하십시오. 예를 들어, 전체 버전인 v0.20.5을 다운로드하려면 RAGFLOW_IMAGE=infiniflow/ragflow:v0.20.5로 설정합니다.
+   > 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.
 
    ```bash
    $ cd ragflow/docker
@@ -168,20 +176,22 @@
    $ docker compose -f docker-compose.yml up -d
 
    # To use GPU to accelerate embedding and DeepDoc tasks:
-   # docker compose -f docker-compose-gpu.yml up -d
+   # sed -i '1i DEVICE=gpu' .env
+   # docker compose -f docker-compose.yml up -d
    ```
 
    | RAGFlow image tag | Image size (GB) | Has embedding models? | Stable?                  |
    | ----------------- | --------------- | --------------------- | ------------------------ |
-   | v0.20.5           | &approx;9       | :heavy_check_mark:    | Stable release           |
+   | v0.21.1           | &approx;9       | ✔️                    | Stable release           |
-   | v0.20.5-slim      | &approx;2       | ❌                    | Stable release           |
+   | v0.21.1-slim      | &approx;2       | ❌                    | Stable release           |
-   | nightly           | &approx;9       | :heavy_check_mark:    | _Unstable_ nightly build |
+   | nightly           | &approx;2       | ❌                    | _Unstable_ nightly build |
-   | nightly-slim      | &approx;2       | ❌                     | _Unstable_ nightly build |
+
+   > 참고: `v0.22.0`부터는 slim 에디션만 배포하며 이미지 태그에 **-slim** 접미사를 더 이상 붙이지 않습니다.
 
 1. 서버가 시작된 후 서버 상태를 확인하세요:
 
    ```bash
-   $ docker logs -f ragflow-server
+   $ docker logs -f docker-ragflow-cpu-1
    ```
 
    _다음 출력 결과로 시스템이 성공적으로 시작되었음을 확인합니다:_
@@ -247,16 +257,6 @@ RAGFlow 는 기본적으로 Elasticsearch 를 사용하여 전체 텍스트 및
 
 이 Docker 이미지의 크기는 약 1GB이며, 외부 대형 모델과 임베딩 서비스에 의존합니다.
 
-```bash
-git clone https://github.com/infiniflow/ragflow.git
-cd ragflow/
-docker build --platform linux/amd64 --build-arg LIGHTEN=1 -f Dockerfile -t infiniflow/ragflow:nightly-slim .
-```
-
-## 🔧 소스 코드로 Docker 이미지를 컴파일합니다(임베딩 모델 포함)
-
-이 Docker의 크기는 약 9GB이며, 이미 임베딩 모델을 포함하고 있으므로 외부 대형 모델 서비스에만 의존하면 됩니다.
-
 ```bash
 git clone https://github.com/infiniflow/ragflow.git
 cd ragflow/
@@ -276,7 +276,7 @@ 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 --all-extras # install RAGFlow dependent python modules
+   uv sync --python 3.10 # install RAGFlow dependent python modules
    uv run download_deps.py
    pre-commit install
    ```

README_pt_br.md

@@ -1,6 +1,6 @@
 <div align="center">
 <a href="https://demo.ragflow.io/">
-<img src="web/src/assets/logo-with-text.png" width="520" alt="ragflow logo">
+<img src="web/src/assets/logo-with-text.svg" width="520" alt="ragflow logo">
 </a>
 </div>
 
@@ -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:v0.20.5">
+        <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">
     </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">
@@ -43,7 +43,13 @@
   <a href="https://demo.ragflow.io">Demo</a>
 </h4>
 
-#
+<div align="center" style="margin-top:20px;margin-bottom:20px;">
+<img src="https://raw.githubusercontent.com/infiniflow/ragflow-docs/refs/heads/image/image/ragflow-octoverse.png" width="1200"/>
+</div>
+
+<div align="center">
+<a href="https://trendshift.io/repositories/9064" target="_blank"><img src="https://trendshift.io/api/badge/repositories/9064" alt="infiniflow%2Fragflow | Trendshift" style="width: 250px; height: 55px;" width="250" height="55"/></a>
+</div>
 
 <details open>
 <summary><b>📕 Índice</b></summary>
@@ -80,8 +86,9 @@ Experimente nossa demo em [https://demo.ragflow.io](https://demo.ragflow.io).
 
 ## 🔥 Últimas Atualizações
 
+- 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.
-- 04-08-2025 Suporta novos modelos, incluindo Kimi K2 e Grok 4.
 - 01-08-2025 Suporta fluxo de trabalho agente e MCP.
 - 23-05-2025 Adicione o componente executor de código Python/JS ao Agente.
 - 05-05-2025 Suporte a consultas entre idiomas.
@@ -129,7 +136,7 @@ Experimente nossa demo em [https://demo.ragflow.io](https://demo.ragflow.io).
 ## 🔎 Arquitetura do Sistema
 
 <div align="center" style="margin-top:20px;margin-bottom:20px;">
-<img src="https://github.com/infiniflow/ragflow/assets/12318111/d6ac5664-c237-4200-a7c2-a4a00691b485" width="1000"/>
+<img src="https://github.com/user-attachments/assets/31b0dd6f-ca4f-445a-9457-70cb44a381b2" width="1000"/>
 </div>
 
 ## 🎬 Primeiros Passos
@@ -147,84 +154,86 @@ Experimente nossa demo em [https://demo.ragflow.io](https://demo.ragflow.io).
 
 ### 🚀 Iniciar o servidor
 
-1.  Certifique-se de que `vm.max_map_count` >= 262144:
+1. Certifique-se de que `vm.max_map_count` >= 262144:
 
-    > Para verificar o valor de `vm.max_map_count`:
+   > Para verificar o valor de `vm.max_map_count`:
-    >
+   >
-    > ```bash
+   > ```bash
-    > $ sysctl vm.max_map_count
+   > $ sysctl vm.max_map_count
-    > ```
+   > ```
-    >
+   >
-    > Se necessário, redefina `vm.max_map_count` para um valor de pelo menos 262144:
+   > Se necessário, redefina `vm.max_map_count` para um valor de pelo menos 262144:
-    >
+   >
-    > ```bash
+   > ```bash
-    > # Neste caso, defina para 262144:
+   > # Neste caso, defina para 262144:
-    > $ sudo sysctl -w vm.max_map_count=262144
+   > $ sudo sysctl -w vm.max_map_count=262144
-    > ```
+   > ```
-    >
+   >
-    > Essa mudança será resetada após a reinicialização do sistema. Para garantir que a alteração permaneça permanente, adicione ou atualize o valor de `vm.max_map_count` em **/etc/sysctl.conf**:
+   > Essa mudança será resetada após a reinicialização do sistema. Para garantir que a alteração permaneça permanente, adicione ou atualize o valor de `vm.max_map_count` em **/etc/sysctl.conf**:
-    >
+   >
-    > ```bash
+   > ```bash
-    > vm.max_map_count=262144
+   > vm.max_map_count=262144
-    > ```
+   > ```
+   >
+2. Clone o repositório:
 
-2.  Clone o repositório:
+   ```bash
-
+   $ git clone https://github.com/infiniflow/ragflow.git
-    ```bash
+   ```
-    $ git clone https://github.com/infiniflow/ragflow.git
+3. Inicie o servidor usando as imagens Docker pré-compiladas:
-    ```
-
-3.  Inicie o servidor usando as imagens Docker pré-compiladas:
 
 > [!CAUTION]
 > 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 `v0.20.5-slim` 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 `v0.20.5-slim`, atualize a variável `RAGFLOW_IMAGE` conforme necessário no **docker/.env** antes de usar `docker compose` para iniciar o servidor. Por exemplo: defina `RAGFLOW_IMAGE=infiniflow/ragflow:v0.20.5` para a edição completa `v0.20.5`.
+    > 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.
 
-    ```bash
+```bash
-    $ cd ragflow/docker
+   $ cd ragflow/docker
-    # Use CPU for embedding and DeepDoc tasks:
+   # Use CPU for embedding and DeepDoc tasks:
-    $ docker compose -f docker-compose.yml up -d
+   $ docker compose -f docker-compose.yml up -d
 
-    # To use GPU to accelerate embedding and DeepDoc tasks:
+   # To use GPU to accelerate embedding and DeepDoc tasks:
-    # docker compose -f docker-compose-gpu.yml up -d
+   # 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?                 |
+| Tag da imagem RAGFlow | Tamanho da imagem (GB) | Possui modelos de incorporação? | Estável?                      |
-    | --------------------- | ---------------------- | ------------------------------- | ------------------------ |
+| --------------------- | ---------------------- | --------------------------------- | ------------------------------ |
-    | v0.20.5               | ~9                     | :heavy_check_mark:              | Lançamento estável       |
+| v0.21.1               | &approx;9              | ✔️                              | Lançamento estável           |
-    | v0.20.5-slim          | ~2                     | ❌                              | Lançamento estável       |
+| v0.21.1-slim          | &approx;2              | ❌                                | Lançamento estável           |
-    | nightly               | ~9                     | :heavy_check_mark:              | _Instável_ build noturno |
+| nightly               | &approx;2              | ❌                                | Construção noturna instável |
-    | nightly-slim          | ~2                     | ❌                               | _Instável_ build noturno |
 
-4.  Verifique o status do servidor após tê-lo iniciado:
+    > Observação: A partir da`v0.22.0`, distribuímos apenas a edição slim e não adicionamos mais o sufixo **-slim** às tags das imagens.
 
-    ```bash
+4. Verifique o status do servidor após tê-lo iniciado:
-    $ docker logs -f ragflow-server
-    ```
 
-    _O seguinte resultado confirma o lançamento bem-sucedido do sistema:_
+   ```bash
+   $ docker logs -f docker-ragflow-cpu-1
+   ```
 
-    ```bash
+   _O seguinte resultado confirma o lançamento bem-sucedido do sistema:_
-         ____   ___    ______ ______ __
-        / __ \ /   |  / ____// ____// /____  _      __
-       / /_/ // /| | / / __ / /_   / // __ \| | /| / /
-      / _, _// ___ |/ /_/ // __/  / // /_/ /| |/ |/ /
-     /_/ |_|/_/  |_|\____//_/    /_/ \____/ |__/|__/
 
-     * Rodando em todos os endereços (0.0.0.0)
+   ```bash
-    ```
+        ____   ___    ______ ______ __
+       / __ \ /   |  / ____// ____// /____  _      __
+      / /_/ // /| | / / __ / /_   / // __ \| | /| / /
+     / _, _// ___ |/ /_/ // __/  / // /_/ /| |/ |/ /
+    /_/ |_|/_/  |_|\____//_/    /_/ \____/ |__/|__/
 
-    > 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.
+    * Rodando em todos os endereços (0.0.0.0)
+   ```
 
-5.  No seu navegador, insira o endereço IP do seu servidor e faça login no RAGFlow.
+   > 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.
+   >
+5. No seu navegador, insira o endereço IP do seu servidor e faça login no RAGFlow.
 
-    > Com as configurações padrão, você só precisa digitar `http://IP_DO_SEU_MÁQUINA` (**sem** o número da porta), pois a porta HTTP padrão `80` pode ser omitida ao usar as configurações padrão.
+   > Com as configurações padrão, você só precisa digitar `http://IP_DO_SEU_MÁQUINA` (**sem** o número da porta), pois a porta HTTP padrão `80` pode ser omitida ao usar as configurações padrão.
+   >
+6. Em [service_conf.yaml.template](./docker/service_conf.yaml.template), selecione a fábrica LLM desejada em `user_default_llm` e atualize o campo `API_KEY` com a chave de API correspondente.
 
-6.  Em [service_conf.yaml.template](./docker/service_conf.yaml.template), selecione a fábrica LLM desejada em `user_default_llm` e atualize o campo `API_KEY` com a chave de API correspondente.
+   > Consulte [llm_api_key_setup](https://ragflow.io/docs/dev/llm_api_key_setup) para mais informações.
-
+   >
-    > Consulte [llm_api_key_setup](https://ragflow.io/docs/dev/llm_api_key_setup) para mais informações.
 
 _O show está no ar!_
 
@@ -255,9 +264,9 @@ O RAGFlow usa o Elasticsearch por padrão para armazenar texto completo e vetore
    ```bash
    $ docker compose -f docker/docker-compose.yml down -v
    ```
+
    Note: `-v` irá deletar os volumes do contêiner, e os dados existentes serão apagados.
 2. Defina `DOC_ENGINE` no **docker/.env** para `infinity`.
-
 3. Inicie os contêineres:
 
    ```bash
@@ -271,16 +280,6 @@ O RAGFlow usa o Elasticsearch por padrão para armazenar texto completo e vetore
 
 Esta imagem tem cerca de 2 GB de tamanho e depende de serviços externos de LLM e incorporação.
 
-```bash
-git clone https://github.com/infiniflow/ragflow.git
-cd ragflow/
-docker build --platform linux/amd64 --build-arg LIGHTEN=1 -f Dockerfile -t infiniflow/ragflow:nightly-slim .
-```
-
-## 🔧 Criar uma imagem Docker incluindo modelos de incorporação
-
-Esta imagem tem cerca de 9 GB de tamanho. Como inclui modelos de incorporação, depende apenas de serviços externos de LLM.
-
 ```bash
 git clone https://github.com/infiniflow/ragflow.git
 cd ragflow/
@@ -294,17 +293,15 @@ docker build --platform linux/amd64 -f Dockerfile -t infiniflow/ragflow:nightly
    ```bash
    pipx install uv pre-commit
    ```
-
 2. Clone o código-fonte e instale as dependências Python:
 
    ```bash
    git clone https://github.com/infiniflow/ragflow.git
    cd ragflow/
-   uv sync --python 3.10 --all-extras # instala os módulos Python dependentes do RAGFlow
+   uv sync --python 3.10 # instala os módulos Python dependentes do RAGFlow
    uv run download_deps.py
    pre-commit install
    ```
-
 3. Inicie os serviços dependentes (MinIO, Elasticsearch, Redis e MySQL) usando Docker Compose:
 
    ```bash
@@ -316,24 +313,21 @@ docker build --platform linux/amd64 -f Dockerfile -t infiniflow/ragflow:nightly
    ```
    127.0.0.1       es01 infinity mysql minio redis sandbox-executor-manager
    ```
-
 4. Se não conseguir acessar o HuggingFace, defina a variável de ambiente `HF_ENDPOINT` para usar um site espelho:
 
    ```bash
    export HF_ENDPOINT=https://hf-mirror.com
    ```
-
 5. Se o seu sistema operacional não tiver jemalloc, instale-o da seguinte maneira:
 
-    ```bash
+   ```bash
-    # ubuntu
+   # ubuntu
-    sudo apt-get install libjemalloc-dev
+   sudo apt-get install libjemalloc-dev
-    # centos
+   # centos
-    sudo yum instalar jemalloc
+   sudo yum instalar jemalloc
-    # mac
+   # mac
-    sudo brew install jemalloc
+   sudo brew install jemalloc
-    ```
+   ```
-
 6. Lance o serviço de back-end:
 
    ```bash
@@ -341,14 +335,12 @@ docker build --platform linux/amd64 -f Dockerfile -t infiniflow/ragflow:nightly
    export PYTHONPATH=$(pwd)
    bash docker/launch_backend_service.sh
    ```
-
 7. Instale as dependências do front-end:
 
    ```bash
    cd web
    npm install
    ```
-
 8. Lance o serviço de front-end:
 
    ```bash
@@ -358,13 +350,11 @@ docker build --platform linux/amd64 -f Dockerfile -t infiniflow/ragflow:nightly
    _O seguinte resultado confirma o lançamento bem-sucedido do sistema:_
 
    ![](https://github.com/user-attachments/assets/0daf462c-a24d-4496-a66f-92533534e187)
-
 9. Pare os serviços de front-end e back-end do RAGFlow após a conclusão do desenvolvimento:
 
-    ```bash
+   ```bash
-    pkill -f "ragflow_server.py|task_executor.py"
+   pkill -f "ragflow_server.py|task_executor.py"
-    ```
+   ```
-
 
 ## 📚 Documentação
 

README_tzh.md

@@ -1,6 +1,6 @@
 <div align="center">
 <a href="https://demo.ragflow.io/">
-<img src="web/src/assets/logo-with-text.png" width="350" alt="ragflow logo">
+<img src="web/src/assets/logo-with-text.svg" width="350" alt="ragflow logo">
 </a>
 </div>
 
@@ -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:v0.20.5">
+        <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">
     </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">
@@ -43,7 +43,9 @@
   <a href="https://demo.ragflow.io">Demo</a>
 </h4>
 
-#
+<div align="center" style="margin-top:20px;margin-bottom:20px;">
+<img src="https://raw.githubusercontent.com/infiniflow/ragflow-docs/refs/heads/image/image/ragflow-octoverse.png" width="1200"/>
+</div>
 
 <div align="center">
 <a href="https://trendshift.io/repositories/9064" target="_blank"><img src="https://trendshift.io/api/badge/repositories/9064" alt="infiniflow%2Fragflow | Trendshift" style="width: 250px; height: 55px;" width="250" height="55"/></a>
@@ -83,8 +85,9 @@
 
 ## 🔥 近期更新
 
+- 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。
-- 2025-08-04 支援 Kimi K2 和 Grok 4 等模型.
 - 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。
@@ -132,7 +135,7 @@
 ## 🔎 系統架構
 
 <div align="center" style="margin-top:20px;margin-bottom:20px;">
-<img src="https://github.com/infiniflow/ragflow/assets/12318111/d6ac5664-c237-4200-a7c2-a4a00691b485" width="1000"/>
+<img src="https://github.com/user-attachments/assets/31b0dd6f-ca4f-445a-9457-70cb44a381b2" width="1000"/>
 </div>
 
 ## 🎬 快速開始
@@ -170,47 +173,48 @@
    > ```bash
    > vm.max_map_count=262144
    > ```
-
+   >
 2. 克隆倉庫：
 
    ```bash
    $ git clone https://github.com/infiniflow/ragflow.git
    ```
-
 3. 進入 **docker** 資料夾，利用事先編譯好的 Docker 映像啟動伺服器：
 
 > [!CAUTION]
 > 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 `v0.20.5-slim`。請參考下表查看不同 Docker 發行版的說明。如需下載不同於 `v0.20.5-slim` 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。例如，你可以透過設定 `RAGFLOW_IMAGE=infiniflow/ragflow:v0.20.5` 來下載 RAGFlow 鏡像的 `v0.20.5` 完整發行版。
+> 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。
 
-   ```bash
+```bash
    $ cd ragflow/docker
    # Use CPU for embedding and DeepDoc tasks:
    $ docker compose -f docker-compose.yml up -d
 
    # To use GPU to accelerate embedding and DeepDoc tasks:
-   # docker compose -f docker-compose-gpu.yml up -d
+   # sed -i '1i DEVICE=gpu' .env
-   ```
+   # docker compose -f docker-compose.yml up -d
+```
 
-   | RAGFlow image tag | Image size (GB) | Has embedding models? | Stable?                  |
+| RAGFlow image tag | Image size (GB) | Has embedding models? | Stable?                    |
-   | ----------------- | --------------- | --------------------- | ------------------------ |
+| ----------------- | --------------- | --------------------- | -------------------------- |
-   | v0.20.5           | &approx;9       | :heavy_check_mark:    | Stable release           |
+| v0.21.1           | &approx;9       | ✔️                  | Stable release             |
-   | v0.20.5-slim      | &approx;2       | ❌                    | Stable release           |
+| v0.21.1-slim      | &approx;2       | ❌                    | Stable release             |
-   | nightly           | &approx;9       | :heavy_check_mark:    | _Unstable_ nightly build |
+| nightly           | &approx;2       | ❌                    | _Unstable_ nightly build |
-   | nightly-slim      | &approx;2       | ❌                     | _Unstable_ nightly build |
 
-   > [!TIP]
+> 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。
-   > 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。
+
-   >
+> [!TIP]
-   > - 華為雲鏡像名：`swr.cn-north-4.myhuaweicloud.com/infiniflow/ragflow`
+> 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。
-   > - 阿里雲鏡像名：`registry.cn-hangzhou.aliyuncs.com/infiniflow/ragflow`
+>
+> - 華為雲鏡像名：`swr.cn-north-4.myhuaweicloud.com/infiniflow/ragflow`
+> - 阿里雲鏡像名：`registry.cn-hangzhou.aliyuncs.com/infiniflow/ragflow`
 
 4. 伺服器啟動成功後再次確認伺服器狀態：
 
    ```bash
-   $ docker logs -f ragflow-server
+   $ docker logs -f docker-ragflow-cpu-1
    ```
 
    _出現以下介面提示說明伺服器啟動成功：_
@@ -226,12 +230,15 @@
    ```
 
    > 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。
-
+   >
 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。
+
    > 上面這個範例中，您只需輸入 http://IP_OF_YOUR_MACHINE 即可：未改動過設定則無需輸入連接埠（預設的 HTTP 服務連接埠 80）。
+   >
 6. 在 [service_conf.yaml.template](./docker/service_conf.yaml.template) 檔案的 `user_default_llm` 欄位設定 LLM factory，並在 `API_KEY` 欄填入和你選擇的大模型相對應的 API key。
 
    > 詳見 [llm_api_key_setup](https://ragflow.io/docs/dev/llm_api_key_setup)。
+   >
 
    _好戲開始，接著奏樂接著舞！ _
 
@@ -249,7 +256,7 @@
 
 > [./docker/README](./docker/README.md) 解釋了 [service_conf.yaml.template](./docker/service_conf.yaml.template) 用到的環境變數設定和服務配置。
 
-如需更新預設的 HTTP 服務連接埠(80), 可以在[docker-compose.yml](./docker/docker-compose.yml) 檔案中將配置`80:80` 改為`<YOUR_SERVING_PORT>:80` 。
+如需更新預設的 HTTP 服務連接埠(80), 可以在[docker-compose.yml](./docker/docker-compose.yml) 檔案中將配置 `80:80` 改為 `<YOUR_SERVING_PORT>:80` 。
 
 > 所有系統配置都需要透過系統重新啟動生效：
 >
@@ -266,10 +273,9 @@ RAGFlow 預設使用 Elasticsearch 儲存文字和向量資料. 如果要切換
    ```bash
    $ docker compose -f docker/docker-compose.yml down -v
    ```
+
    Note: `-v` 將會刪除 docker 容器的 volumes，已有的資料會被清空。
-
 2. 設定 **docker/.env** 目錄中的 `DOC_ENGINE` 為 `infinity`.
-
 3. 啟動容器:
 
    ```bash
@@ -286,17 +292,7 @@ RAGFlow 預設使用 Elasticsearch 儲存文字和向量資料. 如果要切換
 ```bash
 git clone https://github.com/infiniflow/ragflow.git
 cd ragflow/
-docker build --platform linux/amd64 --build-arg LIGHTEN=1 --build-arg NEED_MIRROR=1 -f Dockerfile -t infiniflow/ragflow:nightly-slim .
+docker build --platform linux/amd64 -f Dockerfile -t infiniflow/ragflow:nightly .
-```
-
-## 🔧 原始碼編譯 Docker 映像（包含 embedding 模型）
-
-本 Docker 大小約 9 GB 左右。由於已包含 embedding 模型，所以只需依賴外部的大模型服務即可。
-
-```bash
-git clone https://github.com/infiniflow/ragflow.git
-cd ragflow/
-docker build --platform linux/amd64 --build-arg NEED_MIRROR=1 -f Dockerfile -t infiniflow/ragflow:nightly .
 ```
 
 ## 🔨 以原始碼啟動服務
@@ -307,17 +303,15 @@ docker build --platform linux/amd64 --build-arg NEED_MIRROR=1 -f Dockerfile -t i
    pipx install uv pre-commit
    export UV_INDEX=https://mirrors.aliyun.com/pypi/simple
    ```
-
 2. 下載原始碼並安裝 Python 依賴：
 
    ```bash
    git clone https://github.com/infiniflow/ragflow.git
    cd ragflow/
-   uv sync --python 3.10 --all-extras # install RAGFlow dependent python modules
+   uv sync --python 3.10 # install RAGFlow dependent python modules
    uv run download_deps.py
    pre-commit install
    ```
-
 3. 透過 Docker Compose 啟動依賴的服務（MinIO, Elasticsearch, Redis, and MySQL）：
 
    ```bash
@@ -329,13 +323,11 @@ docker build --platform linux/amd64 --build-arg NEED_MIRROR=1 -f Dockerfile -t i
    ```
    127.0.0.1       es01 infinity mysql minio redis sandbox-executor-manager
    ```
-
 4. 如果無法存取 HuggingFace，可以把環境變數 `HF_ENDPOINT` 設為對應的鏡像網站：
 
    ```bash
    export HF_ENDPOINT=https://hf-mirror.com
    ```
-
 5. 如果你的操作系统没有 jemalloc，请按照如下方式安装：
 
    ```bash
@@ -346,7 +338,6 @@ docker build --platform linux/amd64 --build-arg NEED_MIRROR=1 -f Dockerfile -t i
    # mac
    sudo brew install jemalloc
    ```
-
 6. 啟動後端服務：
 
    ```bash
@@ -354,14 +345,12 @@ docker build --platform linux/amd64 --build-arg NEED_MIRROR=1 -f Dockerfile -t i
    export PYTHONPATH=$(pwd)
    bash docker/launch_backend_service.sh
    ```
-
 7. 安裝前端依賴：
 
    ```bash
    cd web
    npm install
    ```
-
 8. 啟動前端服務：
 
    ```bash
@@ -371,15 +360,16 @@ docker build --platform linux/amd64 --build-arg NEED_MIRROR=1 -f Dockerfile -t i
    以下界面說明系統已成功啟動：_
 
    ![](https://github.com/user-attachments/assets/0daf462c-a24d-4496-a66f-92533534e187)
+
    ```
 
+   ```
 9. 開發完成後停止 RAGFlow 前端和後端服務：
 
    ```bash
    pkill -f "ragflow_server.py|task_executor.py"
    ```
 
-
 ## 📚 技術文檔
 
 - [Quickstart](https://ragflow.io/docs/dev/)

README_zh.md

@@ -1,6 +1,6 @@
 <div align="center">
 <a href="https://demo.ragflow.io/">
-<img src="web/src/assets/logo-with-text.png" width="350" alt="ragflow logo">
+<img src="web/src/assets/logo-with-text.svg" width="350" alt="ragflow logo">
 </a>
 </div>
 
@@ -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:v0.20.5">
+        <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">
     </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">
@@ -43,7 +43,9 @@
   <a href="https://demo.ragflow.io">Demo</a>
 </h4>
 
-#
+<div align="center" style="margin-top:20px;margin-bottom:20px;">
+<img src="https://raw.githubusercontent.com/infiniflow/ragflow-docs/refs/heads/image/image/ragflow-octoverse.png" width="1200"/>
+</div>
 
 <div align="center">
 <a href="https://trendshift.io/repositories/9064" target="_blank"><img src="https://trendshift.io/api/badge/repositories/9064" alt="infiniflow%2Fragflow | Trendshift" style="width: 250px; height: 55px;" width="250" height="55"/></a>
@@ -83,8 +85,9 @@
 
 ## 🔥 近期更新
 
-- 2025-08-08 支持 OpenAI 最新的 GPT-5 系列模型.
+- 2025-10-23 支持 MinerU 和 Docling 作为文档解析方法。
-- 2025-08-04 新增对 Kimi K2 和 Grok 4 等模型的支持.
+- 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。
@@ -132,7 +135,7 @@
 ## 🔎 系统架构
 
 <div align="center" style="margin-top:20px;margin-bottom:20px;">
-<img src="https://github.com/infiniflow/ragflow/assets/12318111/d6ac5664-c237-4200-a7c2-a4a00691b485" width="1000"/>
+<img src="https://github.com/user-attachments/assets/31b0dd6f-ca4f-445a-9457-70cb44a381b2" width="1000"/>
 </div>
 
 ## 🎬 快速开始
@@ -183,7 +186,7 @@
 > 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 `v0.20.5-slim`。请参考下表查看不同 Docker 发行版的描述。如需下载不同于 `v0.20.5-slim` 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。比如，你可以通过设置 `RAGFLOW_IMAGE=infiniflow/ragflow:v0.20.5` 来下载 RAGFlow 镜像的 `v0.20.5` 完整发行版。
+   > 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。
 
    ```bash
    $ cd ragflow/docker
@@ -191,15 +194,17 @@
    $ docker compose -f docker-compose.yml up -d
 
    # To use GPU to accelerate embedding and DeepDoc tasks:
-   # docker compose -f docker-compose-gpu.yml up -d
+   # sed -i '1i DEVICE=gpu' .env
+   # docker compose -f docker-compose.yml up -d
    ```
 
    | RAGFlow image tag | Image size (GB) | Has embedding models? | Stable?                  |
    | ----------------- | --------------- | --------------------- | ------------------------ |
-   | v0.20.5           | &approx;9       | :heavy_check_mark:    | Stable release           |
+   | v0.21.1           | &approx;9       | ✔️                    | Stable release           |
-   | v0.20.5-slim      | &approx;2       | ❌                    | Stable release           |
+   | v0.21.1-slim      | &approx;2       | ❌                    | Stable release           |
-   | nightly           | &approx;9       | :heavy_check_mark:    | _Unstable_ nightly build |
+   | nightly           | &approx;2       | ❌                    | _Unstable_ nightly build |
-   | nightly-slim      | &approx;2       | ❌                     | _Unstable_ nightly build |
+
+   > 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。
 
    > [!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。
@@ -210,7 +215,7 @@
 4. 服务器启动成功后再次确认服务器状态：
 
    ```bash
-   $ docker logs -f ragflow-server
+   $ docker logs -f docker-ragflow-cpu-1
    ```
 
    _出现以下界面提示说明服务器启动成功：_
@@ -286,17 +291,7 @@ RAGFlow 默认使用 Elasticsearch 存储文本和向量数据. 如果要切换
 ```bash
 git clone https://github.com/infiniflow/ragflow.git
 cd ragflow/
-docker build --platform linux/amd64 --build-arg LIGHTEN=1 --build-arg NEED_MIRROR=1 -f Dockerfile -t infiniflow/ragflow:nightly-slim .
+docker build --platform linux/amd64 -f Dockerfile -t infiniflow/ragflow:nightly .
-```
-
-## 🔧 源码编译 Docker 镜像（包含 embedding 模型）
-
-本 Docker 大小约 9 GB 左右。由于已包含 embedding 模型，所以只需依赖外部的大模型服务即可。
-
-```bash
-git clone https://github.com/infiniflow/ragflow.git
-cd ragflow/
-docker build --platform linux/amd64 --build-arg NEED_MIRROR=1 -f Dockerfile -t infiniflow/ragflow:nightly .
 ```
 
 ## 🔨 以源代码启动服务
@@ -313,7 +308,7 @@ docker build --platform linux/amd64 --build-arg NEED_MIRROR=1 -f Dockerfile -t i
    ```bash
    git clone https://github.com/infiniflow/ragflow.git
    cd ragflow/
-   uv sync --python 3.10 --all-extras # install RAGFlow dependent python modules
+   uv sync --python 3.10 # install RAGFlow dependent python modules
    uv run download_deps.py
    pre-commit install
    ```

admin/build_cli_release.sh

@@ -0,0 +1,47 @@
+#!/bin/bash
+
+set -e
+
+echo "🚀 Start building..."
+echo "================================"
+
+PROJECT_NAME="ragflow-cli"
+
+RELEASE_DIR="release"
+BUILD_DIR="dist"
+SOURCE_DIR="src"
+PACKAGE_DIR="ragflow_cli"
+
+echo "🧹 Clean old build folder..."
+rm -rf release/
+
+echo "📁 Prepare source code..."
+mkdir release/$PROJECT_NAME/$SOURCE_DIR -p
+cp pyproject.toml release/$PROJECT_NAME/pyproject.toml
+cp README.md release/$PROJECT_NAME/README.md
+
+mkdir release/$PROJECT_NAME/$SOURCE_DIR/$PACKAGE_DIR -p
+cp admin_client.py release/$PROJECT_NAME/$SOURCE_DIR/$PACKAGE_DIR/admin_client.py
+
+if [ -d "release/$PROJECT_NAME/$SOURCE_DIR" ]; then
+    echo "✅ source dir: release/$PROJECT_NAME/$SOURCE_DIR"
+else
+    echo "❌ source dir not exist: release/$PROJECT_NAME/$SOURCE_DIR"
+    exit 1
+fi
+
+echo "🔨 Make build file..."
+cd release/$PROJECT_NAME
+export PYTHONPATH=$(pwd)
+python -m build
+
+echo "✅ check build result..."
+if [ -d "$BUILD_DIR" ]; then
+    echo "📦 Package generated:"
+    ls -la $BUILD_DIR/
+else
+    echo "❌ Build Failed: $BUILD_DIR not exist."
+    exit 1
+fi
+
+echo "🎉 Build finished successfully!"

admin/client/README.md

@@ -0,0 +1,136 @@
+# RAGFlow Admin Service & CLI
+
+### Introduction
+
+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.
+
+For user and system management, it supports listing, creating, modifying, and deleting users and their associated resources like knowledge bases and Agents.
+
+Built with scalability and reliability in mind, the Admin Service ensures smooth system operation and simplifies maintenance workflows.
+
+It consists of a server-side Service and a command-line client (CLI), both implemented in Python. User commands are parsed using the Lark parsing toolkit.
+
+- **Admin Service**: A backend service that interfaces with the RAGFlow system to execute administrative operations and monitor its status.
+- **Admin CLI**: A command-line interface that allows users to connect to the Admin Service and issue commands for system management.
+
+
+
+### Starting the Admin Service
+
+#### Launching from source code
+
+1. Before start Admin Service, please make sure RAGFlow system is already started.
+
+2. Launch from source code:
+
+   ```bash
+   python admin/server/admin_server.py
+   ```
+   The service will start and listen for incoming connections from the CLI on the configured port. 
+
+#### Using docker image
+
+1. Before startup, please configure the `docker_compose.yml`  file to enable admin server:
+
+   ```bash
+   command:
+     - --enable-adminserver
+   ```
+
+2. Start the containers, the service will start and listen for incoming connections from the CLI on the configured port.
+
+
+
+### Using the Admin CLI
+
+1.  Ensure the Admin Service is running.
+2.  Install ragflow-cli.
+    ```bash
+    pip install ragflow-cli==0.21.1
+    ```
+3.  Launch the CLI client:
+    ```bash
+    ragflow-cli -h 127.0.0.1 -p 9381
+    ```
+    You will be prompted to enter the superuser's password to log in.
+    The default password is admin.
+
+    **Parameters:**
+    
+    - -h: RAGFlow admin server host address
+    
+    - -p: RAGFlow admin server port
+
+
+
+## Supported Commands
+
+Commands are case-insensitive and must be terminated with a semicolon (`;`).
+
+### Service Management Commands
+
+-   `LIST SERVICES;`
+    -   Lists all available services within the RAGFlow system.
+-   `SHOW SERVICE <id>;`
+    -   Shows detailed status information for the service identified by `<id>`.
+
+
+### User Management Commands
+
+-   `LIST USERS;`
+    -   Lists all users known to the system.
+-   `SHOW USER '<username>';`
+    -   Shows details and permissions for the specified user. The username must be enclosed in single or double quotes.
+
+- `CREATE USER <username> <password>;`
+  - Create user by username and password. The username and password must be enclosed in single or double quotes.
+
+-   `DROP USER '<username>';`
+    -   Removes the specified user from the system. Use with caution.
+-   `ALTER USER PASSWORD '<username>' '<new_password>';`
+    -   Changes the password for the specified user.
+-   `ALTER USER ACTIVE <username> <on/off>;`
+    -   Changes the user to active or inactive.
+
+
+### Data and Agent Commands
+
+-   `LIST DATASETS OF '<username>';`
+    -   Lists the datasets associated with the specified user.
+-   `LIST AGENTS OF '<username>';`
+    -   Lists the agents associated with the specified user.
+
+### Meta-Commands
+
+Meta-commands are prefixed with a backslash (`\`).
+
+-   `\?` or `\help`
+    -   Shows help information for the available commands.
+-   `\q` or `\quit`
+    -   Exits the CLI application.
+
+## Examples
+
+```commandline
+admin> list users;
++-------------------------------+------------------------+-----------+-------------+
+| create_date                   | email                  | is_active | nickname    |
++-------------------------------+------------------------+-----------+-------------+
+| Fri, 22 Nov 2024 16:03:41 GMT | jeffery@infiniflow.org | 1         | Jeffery     |
+| Fri, 22 Nov 2024 16:10:55 GMT | aya@infiniflow.org     | 1         | Waterdancer |
++-------------------------------+------------------------+-----------+-------------+
+
+admin> list services;
++-------------------------------------------------------------------------------------------+-----------+----+---------------+-------+----------------+
+| extra                                                                                     | host      | id | name          | port  | service_type   |
++-------------------------------------------------------------------------------------------+-----------+----+---------------+-------+----------------+
+| {}                                                                                        | 0.0.0.0   | 0  | ragflow_0     | 9380  | ragflow_server |
+| {'meta_type': 'mysql', 'password': 'infini_rag_flow', 'username': 'root'}                 | localhost | 1  | mysql         | 5455  | meta_data      |
+| {'password': 'infini_rag_flow', 'store_type': 'minio', 'user': 'rag_flow'}                | localhost | 2  | minio         | 9000  | file_store     |
+| {'password': 'infini_rag_flow', 'retrieval_type': 'elasticsearch', 'username': 'elastic'} | localhost | 3  | elasticsearch | 1200  | retrieval      |
+| {'db_name': 'default_db', 'retrieval_type': 'infinity'}                                   | localhost | 4  | infinity      | 23817 | retrieval      |
+| {'database': 1, 'mq_type': 'redis', 'password': 'infini_rag_flow'}                        | localhost | 5  | redis         | 6379  | message_queue  |
++-------------------------------------------------------------------------------------------+-----------+----+---------------+-------+----------------+
+```

admin/client/admin_client.py

@@ -0,0 +1,931 @@
+#
+#  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.
+#
+
+import argparse
+import base64
+from cmd import Cmd
+
+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
+
+GRAMMAR = r"""
+start: command
+
+command: sql_command | meta_command
+
+sql_command: list_services
+           | show_service
+           | startup_service
+           | shutdown_service
+           | restart_service
+           | list_users
+           | show_user
+           | drop_user
+           | alter_user
+           | create_user
+           | activate_user
+           | list_datasets
+           | list_agents
+           | create_role
+           | drop_role
+           | alter_role
+           | list_roles
+           | show_role
+           | grant_permission
+           | revoke_permission
+           | alter_user_role
+           | show_user_permission
+
+// meta command definition
+meta_command: "\\" meta_command_name [meta_args]
+
+meta_command_name: /[a-zA-Z?]+/
+meta_args: (meta_arg)+
+
+meta_arg: /[^\\s"']+/ | quoted_string
+
+// command definition
+
+LIST: "LIST"i
+SERVICES: "SERVICES"i
+SHOW: "SHOW"i
+CREATE: "CREATE"i
+SERVICE: "SERVICE"i
+SHUTDOWN: "SHUTDOWN"i
+STARTUP: "STARTUP"i
+RESTART: "RESTART"i
+USERS: "USERS"i
+DROP: "DROP"i
+USER: "USER"i
+ALTER: "ALTER"i
+ACTIVE: "ACTIVE"i
+PASSWORD: "PASSWORD"i
+DATASETS: "DATASETS"i
+OF: "OF"i
+AGENTS: "AGENTS"i
+ROLE: "ROLE"i
+ROLES: "ROLES"i
+DESCRIPTION: "DESCRIPTION"i
+GRANT: "GRANT"i
+REVOKE: "REVOKE"i
+ALL: "ALL"i
+PERMISSION: "PERMISSION"i
+TO: "TO"i
+FROM: "FROM"i
+FOR: "FOR"i
+RESOURCES: "RESOURCES"i
+ON: "ON"i
+SET: "SET"i
+
+list_services: LIST SERVICES ";"
+show_service: SHOW SERVICE NUMBER ";"
+startup_service: STARTUP SERVICE NUMBER ";"
+shutdown_service: SHUTDOWN SERVICE NUMBER ";"
+restart_service: RESTART SERVICE NUMBER ";"
+
+list_users: LIST USERS ";"
+drop_user: DROP USER quoted_string ";"
+alter_user: ALTER USER PASSWORD quoted_string quoted_string ";"
+show_user: SHOW USER quoted_string ";"
+create_user: CREATE USER quoted_string quoted_string ";"
+activate_user: ALTER USER ACTIVE quoted_string status ";"
+
+list_datasets: LIST DATASETS OF quoted_string ";"
+list_agents: LIST AGENTS OF quoted_string ";"
+
+create_role: CREATE ROLE identifier [DESCRIPTION quoted_string] ";"
+drop_role: DROP ROLE identifier ";"
+alter_role: ALTER ROLE identifier SET DESCRIPTION quoted_string ";"
+list_roles: LIST ROLES ";"
+show_role: SHOW ROLE identifier ";"
+
+grant_permission: GRANT action_list ON identifier TO ROLE identifier ";"
+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 ";"
+
+action_list: identifier ("," identifier)*
+
+identifier: WORD
+quoted_string: QUOTED_STRING
+status: WORD
+
+QUOTED_STRING: /'[^']+'/ | /"[^"]+"/
+WORD: /[a-zA-Z0-9_\-\.]+/
+NUMBER: /[0-9]+/
+
+%import common.WS
+%ignore WS
+"""
+
+
+class AdminTransformer(Transformer):
+
+    def start(self, items):
+        return items[0]
+
+    def command(self, items):
+        return items[0]
+
+    def list_services(self, items):
+        result = {'type': 'list_services'}
+        return result
+
+    def show_service(self, items):
+        service_id = int(items[2])
+        return {"type": "show_service", "number": service_id}
+
+    def startup_service(self, items):
+        service_id = int(items[2])
+        return {"type": "startup_service", "number": service_id}
+
+    def shutdown_service(self, items):
+        service_id = int(items[2])
+        return {"type": "shutdown_service", "number": service_id}
+
+    def restart_service(self, items):
+        service_id = int(items[2])
+        return {"type": "restart_service", "number": service_id}
+
+    def list_users(self, items):
+        return {"type": "list_users"}
+
+    def show_user(self, items):
+        user_name = items[2]
+        return {"type": "show_user", "user_name": user_name}
+
+    def drop_user(self, items):
+        user_name = items[2]
+        return {"type": "drop_user", "user_name": user_name}
+
+    def alter_user(self, items):
+        user_name = items[3]
+        new_password = items[4]
+        return {"type": "alter_user", "user_name": user_name, "password": new_password}
+
+    def create_user(self, items):
+        user_name = items[2]
+        password = items[3]
+        return {"type": "create_user", "user_name": user_name, "password": password, "role": "user"}
+
+    def activate_user(self, items):
+        user_name = items[3]
+        activate_status = items[4]
+        return {"type": "activate_user", "activate_status": activate_status, "user_name": user_name}
+
+    def list_datasets(self, items):
+        user_name = items[3]
+        return {"type": "list_datasets", "user_name": user_name}
+
+    def list_agents(self, items):
+        user_name = items[3]
+        return {"type": "list_agents", "user_name": user_name}
+
+    def create_role(self, items):
+        role_name = items[2]
+        if len(items) > 4:
+            description = items[4]
+            return {"type": "create_role", "role_name": role_name, "description": description}
+        else:
+            return {"type": "create_role", "role_name": role_name}
+
+    def drop_role(self, items):
+        role_name = items[2]
+        return {"type": "drop_role", "role_name": role_name}
+
+    def alter_role(self, items):
+        role_name = items[2]
+        description = items[5]
+        return {"type": "alter_role", "role_name": role_name, "description": description}
+
+    def list_roles(self, items):
+        return {"type": "list_roles"}
+
+    def show_role(self, items):
+        role_name = items[2]
+        return {"type": "show_role", "role_name": role_name}
+
+    def grant_permission(self, items):
+        action_list = items[1]
+        resource = items[3]
+        role_name = items[6]
+        return {"type": "grant_permission", "role_name": role_name, "resource": resource, "actions": action_list}
+
+    def revoke_permission(self, items):
+        action_list = items[1]
+        resource = items[3]
+        role_name = items[6]
+        return {
+            "type": "revoke_permission",
+            "role_name": role_name,
+            "resource": resource, "actions": action_list
+        }
+
+    def alter_user_role(self, items):
+        user_name = items[2]
+        role_name = items[5]
+        return {"type": "alter_user_role", "user_name": user_name, "role_name": role_name}
+
+    def show_user_permission(self, items):
+        user_name = items[3]
+        return {"type": "show_user_permission", "user_name": user_name}
+
+    def action_list(self, items):
+        return items
+
+    def meta_command(self, items):
+        command_name = str(items[0]).lower()
+        args = items[1:] if len(items) > 1 else []
+
+        # handle quoted parameter
+        parsed_args = []
+        for arg in args:
+            if hasattr(arg, 'value'):
+                parsed_args.append(arg.value)
+            else:
+                parsed_args.append(str(arg))
+
+        return {'type': 'meta', 'command': command_name, 'args': parsed_args}
+
+    def meta_command_name(self, items):
+        return items[0]
+
+    def meta_args(self, items):
+        return items
+
+
+def encrypt(input_string):
+    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')))
+    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')
+
+
+class AdminCLI(Cmd):
+    def __init__(self):
+        super().__init__()
+        self.parser = Lark(GRAMMAR, start='start', parser='lalr', transformer=AdminTransformer())
+        self.command_history = []
+        self.is_interactive = False
+        self.admin_account = "admin@ragflow.io"
+        self.admin_password: str = "admin"
+        self.session = requests.Session()
+        self.access_token: str = ""
+        self.host: str = ""
+        self.port: int = 0
+
+    intro = r"""Type "\h" for help."""
+    prompt = "admin> "
+
+    def onecmd(self, command: str) -> bool:
+        try:
+            result = self.parse_command(command)
+
+            if isinstance(result, dict):
+                if 'type' in result and result.get('type') == 'empty':
+                    return False
+
+            self.execute_command(result)
+
+            if isinstance(result, Tree):
+                return False
+
+            if result.get('type') == 'meta' and result.get('command') in ['q', 'quit', 'exit']:
+                return True
+
+        except KeyboardInterrupt:
+            print("\nUse '\\q' to quit")
+        except EOFError:
+            print("\nGoodbye!")
+            return True
+        return False
+
+    def emptyline(self) -> bool:
+        return False
+
+    def default(self, line: str) -> bool:
+        return self.onecmd(line)
+
+    def parse_command(self, command_str: str) -> dict[str, str]:
+        if not command_str.strip():
+            return {'type': 'empty'}
+
+        self.command_history.append(command_str)
+
+        try:
+            result = self.parser.parse(command_str)
+            return result
+        except Exception as 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}")
+        url = f"http://{self.host}:{self.port}/api/v1/admin/login"
+
+        attempt_count = 3
+        if single_command:
+            attempt_count = 1
+
+        try_count = 0
+        while True:
+            try_count += 1
+            if try_count > attempt_count:
+                return False
+
+            if single_command:
+                admin_passwd = arguments['password']
+            else:
+                admin_passwd = input(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})
+                if response.status_code == 200:
+                    res_json = response.json()
+                    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'
+                        })
+                        print("Authentication successful.")
+                        return True
+                    else:
+                        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}")
+
+    def _print_table_simple(self, data):
+        if not data:
+            print("No data to print")
+            return
+        if isinstance(data, dict):
+            # handle single row data
+            data = [data]
+
+        columns = list(data[0].keys())
+        col_widths = {}
+
+        def get_string_width(text):
+            half_width_chars = (
+                " !\"#$%&'()*+,-./0123456789:;<=>?@"
+                "ABCDEFGHIJKLMNOPQRSTUVWXYZ[\\]^_`"
+                "abcdefghijklmnopqrstuvwxyz{|}~"
+                "\t\n\r"
+            )
+            width = 0
+            for char in text:
+                if char in half_width_chars:
+                    width += 1
+                else:
+                    width += 2
+            return width
+
+        for col in columns:
+            max_width = get_string_width(str(col))
+            for item in data:
+                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)
+
+        # Generate delimiter
+        separator = "+" + "+".join(["-" * (col_widths[col] + 2) for col in columns]) + "+"
+
+        # Print header
+        print(separator)
+        header = "|" + "|".join([f" {col:<{col_widths[col]}} " for col in columns]) + "|"
+        print(header)
+        print(separator)
+
+        # Print data
+        for item in data:
+            row = "|"
+            for col in columns:
+                value = str(item.get(col, ''))
+                if get_string_width(value) > col_widths[col]:
+                    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")
+
+        while True:
+            try:
+                command = input("admin> ").strip()
+                if not command:
+                    continue
+
+                print(f"command: {command}")
+                result = self.parse_command(command)
+                self.execute_command(result)
+
+                if isinstance(result, Tree):
+                    continue
+
+                if result.get('type') == 'meta' and result.get('command') in ['q', 'quit', 'exit']:
+                    break
+
+            except KeyboardInterrupt:
+                print("\nUse '\\q' to quit")
+            except EOFError:
+                print("\nGoodbye!")
+                break
+
+    def run_single_command(self, command: str):
+        result = self.parse_command(command)
+        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')
+        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
+                }
+            else:
+                return {
+                    'host': parsed_args.host,
+                    'port': parsed_args.port,
+                }
+        except SystemExit:
+            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':
+                print(f"Error: {parsed_command['message']}")
+                return
+            else:
+                command_dict = parsed_command
+
+        # print(f"Parsed command: {command_dict}")
+
+        command_type = command_dict['type']
+
+        match command_type:
+            case 'list_services':
+                self._handle_list_services(command_dict)
+            case 'show_service':
+                self._handle_show_service(command_dict)
+            case 'restart_service':
+                self._handle_restart_service(command_dict)
+            case 'shutdown_service':
+                self._handle_shutdown_service(command_dict)
+            case 'startup_service':
+                self._handle_startup_service(command_dict)
+            case 'list_users':
+                self._handle_list_users(command_dict)
+            case 'show_user':
+                self._handle_show_user(command_dict)
+            case 'drop_user':
+                self._handle_drop_user(command_dict)
+            case 'alter_user':
+                self._handle_alter_user(command_dict)
+            case 'create_user':
+                self._handle_create_user(command_dict)
+            case 'activate_user':
+                self._handle_activate_user(command_dict)
+            case 'list_datasets':
+                self._handle_list_datasets(command_dict)
+            case 'list_agents':
+                self._handle_list_agents(command_dict)
+            case 'create_role':
+                self._create_role(command_dict)
+            case 'drop_role':
+                self._drop_role(command_dict)
+            case 'alter_role':
+                self._alter_role(command_dict)
+            case 'list_roles':
+                self._list_roles(command_dict)
+            case 'show_role':
+                self._show_role(command_dict)
+            case 'grant_permission':
+                self._grant_permission(command_dict)
+            case 'revoke_permission':
+                self._revoke_permission(command_dict)
+            case 'alter_user_role':
+                self._alter_user_role(command_dict)
+            case 'show_user_permission':
+                self._show_user_permission(command_dict)
+            case 'meta':
+                self._handle_meta_command(command_dict)
+            case _:
+                print(f"Command '{command_type}' would be executed with API")
+
+    def _handle_list_services(self, command):
+        print("Listing all 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'])
+        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']
+        print(f"Showing service: {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':
+                print(f"Service {res_data['service_name']} is alive, ")
+                if isinstance(res_data['message'], str):
+                    print(res_data['message'])
+                else:
+                    self._print_table_simple(res_data['message'])
+            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']
+        print(f"Restart service {service_id}")
+
+    def _handle_shutdown_service(self, command):
+        service_id: int = command['number']
+        print(f"Shutdown service {service_id}")
+
+    def _handle_startup_service(self, command):
+        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'
+        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 get all users, code: {res_json['code']}, message: {res_json['message']}")
+
+    def _handle_show_user(self, command):
+        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}'
+        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 get user {user_name}, code: {res_json['code']}, message: {res_json['message']}")
+
+    def _handle_drop_user(self, command):
+        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}'
+        response = self.session.delete(url)
+        res_json = response.json()
+        if response.status_code == 200:
+            print(res_json["message"])
+        else:
+            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: str = user_name_tree.children[0].strip("'\"")
+        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)})
+        res_json = response.json()
+        if response.status_code == 200:
+            print(res_json["message"])
+        else:
+            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: str = user_name_tree.children[0].strip("'\"")
+        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}
+        )
+        res_json = response.json()
+        if response.status_code == 200:
+            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: str = user_name_tree.children[0].strip("'\"")
+        activate_tree: Tree = command['activate_status']
+        activate_status: str = activate_tree.children[0].strip("'\"")
+        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})
+            res_json = response.json()
+            if response.status_code == 200:
+                print(res_json["message"])
+            else:
+                print(f"Fail to alter activate status, code: {res_json['code']}, message: {res_json['message']}")
+        else:
+            print(f"Unknown activate status: {activate_status}.")
+
+    def _handle_list_datasets(self, command):
+        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'
+        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 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']
+        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'
+        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 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: str = role_name_tree.children[0].strip("'\"")
+        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}
+        )
+        res_json = response.json()
+        if response.status_code == 200:
+            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: 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}'
+        response = self.session.delete(url)
+        res_json = response.json()
+        if response.status_code == 200:
+            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: str = role_name_tree.children[0].strip("'\"")
+        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}
+        )
+        res_json = response.json()
+        if response.status_code == 200:
+            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']}")
+
+    def _list_roles(self, command):
+        print("Listing all 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'])
+        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: 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'
+        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 list roles, code: {res_json['code']}, message: {res_json['message']}")
+
+    def _grant_permission(self, command):
+        role_name_tree: Tree = command['role_name']
+        role_name_str: str = role_name_tree.children[0].strip("'\"")
+        resource_tree: Tree = command['resource']
+        resource_str: str = resource_tree.children[0].strip("'\"")
+        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}
+        )
+        res_json = response.json()
+        if response.status_code == 200:
+            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']}")
+
+    def _revoke_permission(self, command):
+        role_name_tree: Tree = command['role_name']
+        role_name_str: str = role_name_tree.children[0].strip("'\"")
+        resource_tree: Tree = command['resource']
+        resource_str: str = resource_tree.children[0].strip("'\"")
+        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}
+        )
+        res_json = response.json()
+        if response.status_code == 200:
+            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']}")
+
+    def _alter_user_role(self, command):
+        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_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}
+        )
+        res_json = response.json()
+        if response.status_code == 200:
+            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']}")
+
+    def _show_user_permission(self, command):
+        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'
+        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 user: {user_name_str} permission, code: {res_json['code']}, message: {res_json['message']}")
+
+    def _handle_meta_command(self, command):
+        meta_command = command['command']
+        args = command.get('args', [])
+
+        if meta_command in ['?', 'h', 'help']:
+            self.show_help()
+        elif meta_command in ['q', 'quit', 'exit']:
+            print("Goodbye!")
+        else:
+            print(f"Meta command '{meta_command}' with args {args}")
+
+    def show_help(self):
+        """Help info"""
+        help_text = """
+Commands:
+  LIST SERVICES
+  SHOW SERVICE <service>
+  STARTUP SERVICE <service>
+  SHUTDOWN SERVICE <service>
+  RESTART SERVICE <service>
+  LIST USERS
+  SHOW USER <user>
+  DROP USER <user>
+  CREATE USER <user> <password>
+  ALTER USER PASSWORD <user> <new_password>
+  ALTER USER ACTIVE <user> <on/off>
+  LIST DATASETS OF <user>
+  LIST AGENTS OF <user>
+
+Meta Commands:
+  \\?, \\h, \\help     Show this help
+  \\q, \\quit, \\exit   Quit the CLI
+        """
+        print(help_text)
+
+
+def main():
+    import sys
+
+    cli = AdminCLI()
+
+    args = cli.parse_connection_args(sys.argv)
+    if 'error' in args:
+        print(f"Error: {args['error']}")
+        return
+
+    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}")
+            cli.run_single_command(command)
+    else:
+        if cli.verify_admin(args, single_command=False):
+            print(r"""
+                ____  ___   ______________                 ___       __          _     
+               / __ \/   | / ____/ ____/ /___ _      __   /   | ____/ /___ ___  (_)___ 
+              / /_/ / /| |/ / __/ /_  / / __ \ | /| / /  / /| |/ __  / __ `__ \/ / __ \
+             / _, _/ ___ / /_/ / __/ / / /_/ / |/ |/ /  / ___ / /_/ / / / / / / / / / /
+            /_/ |_/_/  |_\____/_/   /_/\____/|__/|__/  /_/  |_\__,_/_/ /_/ /_/_/_/ /_/ 
+            """)
+            cli.cmdloop()
+
+
+if __name__ == '__main__':
+    main()

admin/client/pyproject.toml

@@ -0,0 +1,24 @@
+[project]
+name = "ragflow-cli"
+version = "0.21.1"
+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"
+dependencies = [
+    "requests>=2.30.0,<3.0.0",
+    "beartype>=0.18.5,<0.19.0",
+    "pycryptodomex>=3.10.0",
+    "lark>=1.1.0",
+]
+
+[dependency-groups]
+test = [
+    "pytest>=8.3.5",
+    "requests>=2.32.3",
+    "requests-toolbelt>=1.0.0",
+]
+
+[project.scripts]
+ragflow-cli = "admin_client:main"

admin/server/admin_server.py

@@ -0,0 +1,77 @@
+#
+#  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.
+#
+
+import os
+import signal
+import logging
+import time
+import threading
+import traceback
+from werkzeug.serving import run_simple
+from flask import Flask
+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 config import load_configurations, SERVICE_CONFIGS
+from auth import init_default_admin, setup_auth
+from flask_session import Session
+from flask_login import LoginManager
+
+stop_event = threading.Event()
+
+if __name__ == '__main__':
+    init_root_logger("admin_service")
+    logging.info(r"""
+        ____  ___   ______________                 ___       __          _     
+       / __ \/   | / ____/ ____/ /___ _      __   /   | ____/ /___ ___  (_)___ 
+      / /_/ / /| |/ / __/ /_  / / __ \ | /| / /  / /| |/ __  / __ `__ \/ / __ \
+     / _, _/ ___ / /_/ / __/ / / /_/ / |/ |/ /  / ___ / /_/ / / / / / / / / / /
+    /_/ |_/_/  |_\____/_/   /_/\____/|__/|__/  /_/  |_\__,_/_/ /_/ /_/_/_/ /_/ 
+    """)
+
+    app = Flask(__name__)
+    app.register_blueprint(admin_bp)
+    app.config["SESSION_PERMANENT"] = False
+    app.config["SESSION_TYPE"] = "filesystem"
+    app.config["MAX_CONTENT_LENGTH"] = int(
+        os.environ.get("MAX_CONTENT_LENGTH", 1024 * 1024 * 1024)
+    )
+    Session(app)
+    show_configs()
+    login_manager = LoginManager()
+    login_manager.init_app(app)
+    settings.init_settings()
+    setup_auth(login_manager)
+    init_default_admin()
+    SERVICE_CONFIGS.configs = load_configurations(SERVICE_CONF)
+
+    try:
+        logging.info("RAGFlow Admin service start...")
+        run_simple(
+            hostname="0.0.0.0",
+            port=9381,
+            application=app,
+            threaded=True,
+            use_reloader=True,
+            use_debugger=True,
+        )
+    except Exception:
+        traceback.print_exc()
+        stop_event.set()
+        time.sleep(1)
+        os.kill(os.getpid(), signal.SIGKILL)

admin/server/auth.py

@@ -0,0 +1,187 @@
+#
+#  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.
+#
+
+
+import logging
+import uuid
+from functools import wraps
+from datetime import datetime
+from flask import request, jsonify
+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.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
+
+
+def setup_auth(login_manager):
+    @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
+
+
+def init_default_admin():
+    # Verify that at least one active admin user exists. If not, create a default one.
+    users = UserService.query(is_superuser=True)
+    if not users:
+        default_admin = {
+            "id": uuid.uuid1().hex,
+            "password": encode_to_base64("admin"),
+            "nickname": "admin",
+            "is_superuser": True,
+            "email": "admin@ragflow.io",
+            "creator": "system",
+            "status": "1",
+        }
+        if not UserService.save(**default_admin):
+            raise AdminException("Can't init admin.", 500)
+    elif not any([u.is_active == ActiveEnum.ACTIVE.value for u in users]):
+        raise AdminException("No active admin. Please update 'is_active' in db manually.", 500)
+
+
+def check_admin_auth(func):
+    @wraps(func)
+    def wrapper(*args, **kwargs):
+        user = UserService.filter_by_id(current_user.id)
+        if not user:
+            raise UserNotFoundError(current_user.email)
+        if not user.is_superuser:
+            raise AdminException("Not admin", 403)
+        if user.is_active == ActiveEnum.INACTIVE.value:
+            raise AdminException(f"User {current_user.email} inactive", 403)
+
+        return func(*args, **kwargs)
+
+    return wrapper
+
+
+def login_admin(email: str, password: str):
+    """
+    :param email: admin email
+    :param password: string before decrypt
+    """
+    users = UserService.query(email=email)
+    if not users:
+        raise UserNotFoundError(email)
+    psw = decrypt(password)
+    user = UserService.query_user(email, psw)
+    if not user:
+        raise AdminException("Email and password do not match!")
+    if not user.is_superuser:
+        raise AdminException("Not admin", 403)
+    if user.is_active == ActiveEnum.INACTIVE.value:
+        raise AdminException(f"User {email} inactive", 403)
+
+    resp = user.to_json()
+    user.access_token = get_uuid()
+    login_user(user)
+    user.update_time = (current_timestamp(),)
+    user.update_date = (datetime_format(datetime.now()),)
+    user.last_login_time = get_format_time()
+    user.save()
+    msg = "Welcome back!"
+    return construct_response(data=resp, auth=user.get_id(), message=msg)
+
+
+def check_admin(username: str, password: str):
+    users = UserService.query(email=username)
+    if not users:
+        logging.info(f"Username: {username} is not registered!")
+        user_info = {
+            "id": uuid.uuid1().hex,
+            "password": encode_to_base64("admin"),
+            "nickname": "admin",
+            "is_superuser": True,
+            "email": "admin@ragflow.io",
+            "creator": "system",
+            "status": "1",
+        }
+        if not UserService.save(**user_info):
+            raise AdminException("Can't init admin.", 500)
+
+    user = UserService.query_user(username, password)
+    if user:
+        return True
+    else:
+        return False
+
+
+def login_verify(f):
+    @wraps(f)
+    def decorated(*args, **kwargs):
+        auth = request.authorization
+        if not auth or 'username' not in auth.parameters or 'password' not in auth.parameters:
+            return jsonify({
+                "code": 401,
+                "message": "Authentication required",
+                "data": None
+            }), 200
+
+        username = auth.parameters['username']
+        password = auth.parameters['password']
+        try:
+            if check_admin(username, password) is False:
+                return jsonify({
+                    "code": 500,
+                    "message": "Access denied",
+                    "data": None
+                }), 200
+        except Exception as e:
+            error_msg = str(e)
+            return jsonify({
+                "code": 500,
+                "message": error_msg
+            }), 200
+
+        return f(*args, **kwargs)
+
+    return decorated

admin/server/config.py

@@ -0,0 +1,306 @@
+#
+#  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.
+#
+
+
+import logging
+import threading
+from enum import Enum
+
+from pydantic import BaseModel
+from typing import Any
+from common.config_utils import read_config
+from urllib.parse import urlparse
+
+
+class ServiceConfigs:
+    configs = dict
+
+    def __init__(self):
+        self.configs = []
+        self.lock = threading.Lock()
+
+
+SERVICE_CONFIGS = ServiceConfigs
+
+
+class ServiceType(Enum):
+    METADATA = "metadata"
+    RETRIEVAL = "retrieval"
+    MESSAGE_QUEUE = "message_queue"
+    RAGFLOW_SERVER = "ragflow_server"
+    TASK_EXECUTOR = "task_executor"
+    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
+
+    def to_dict(self) -> dict[str, Any]:
+        result = super().to_dict()
+        if 'extra' not in result:
+            result['extra'] = dict()
+        extra_dict = result['extra'].copy()
+        extra_dict['meta_type'] = self.meta_type
+        result['extra'] = extra_dict
+        return result
+
+
+class MySQLConfig(MetaConfig):
+    username: str
+    password: str
+
+    def to_dict(self) -> dict[str, Any]:
+        result = super().to_dict()
+        if 'extra' not in result:
+            result['extra'] = dict()
+        extra_dict = result['extra'].copy()
+        extra_dict['username'] = self.username
+        extra_dict['password'] = self.password
+        result['extra'] = extra_dict
+        return result
+
+
+class PostgresConfig(MetaConfig):
+
+    def to_dict(self) -> dict[str, Any]:
+        result = super().to_dict()
+        if 'extra' not in result:
+            result['extra'] = dict()
+        return result
+
+
+class RetrievalConfig(BaseConfig):
+    retrieval_type: str
+
+    def to_dict(self) -> dict[str, Any]:
+        result = super().to_dict()
+        if 'extra' not in result:
+            result['extra'] = dict()
+        extra_dict = result['extra'].copy()
+        extra_dict['retrieval_type'] = self.retrieval_type
+        result['extra'] = extra_dict
+        return result
+
+
+class InfinityConfig(RetrievalConfig):
+    db_name: str
+
+    def to_dict(self) -> dict[str, Any]:
+        result = super().to_dict()
+        if 'extra' not in result:
+            result['extra'] = dict()
+        extra_dict = result['extra'].copy()
+        extra_dict['db_name'] = self.db_name
+        result['extra'] = extra_dict
+        return result
+
+
+class ElasticsearchConfig(RetrievalConfig):
+    username: str
+    password: str
+
+    def to_dict(self) -> dict[str, Any]:
+        result = super().to_dict()
+        if 'extra' not in result:
+            result['extra'] = dict()
+        extra_dict = result['extra'].copy()
+        extra_dict['username'] = self.username
+        extra_dict['password'] = self.password
+        result['extra'] = extra_dict
+        return result
+
+
+class MessageQueueConfig(BaseConfig):
+    mq_type: str
+
+    def to_dict(self) -> dict[str, Any]:
+        result = super().to_dict()
+        if 'extra' not in result:
+            result['extra'] = dict()
+        extra_dict = result['extra'].copy()
+        extra_dict['mq_type'] = self.mq_type
+        result['extra'] = extra_dict
+        return result
+
+
+class RedisConfig(MessageQueueConfig):
+    database: int
+    password: str
+
+    def to_dict(self) -> dict[str, Any]:
+        result = super().to_dict()
+        if 'extra' not in result:
+            result['extra'] = dict()
+        extra_dict = result['extra'].copy()
+        extra_dict['database'] = self.database
+        extra_dict['password'] = self.password
+        result['extra'] = extra_dict
+        return result
+
+
+class RabbitMQConfig(MessageQueueConfig):
+
+    def to_dict(self) -> dict[str, Any]:
+        result = super().to_dict()
+        if 'extra' not in result:
+            result['extra'] = dict()
+        return result
+
+
+class RAGFlowServerConfig(BaseConfig):
+
+    def to_dict(self) -> dict[str, Any]:
+        result = super().to_dict()
+        if 'extra' not in result:
+            result['extra'] = dict()
+        return result
+
+
+class TaskExecutorConfig(BaseConfig):
+
+    def to_dict(self) -> dict[str, Any]:
+        result = super().to_dict()
+        if 'extra' not in result:
+            result['extra'] = dict()
+        return result
+
+
+class FileStoreConfig(BaseConfig):
+    store_type: str
+
+    def to_dict(self) -> dict[str, Any]:
+        result = super().to_dict()
+        if 'extra' not in result:
+            result['extra'] = dict()
+        extra_dict = result['extra'].copy()
+        extra_dict['store_type'] = self.store_type
+        result['extra'] = extra_dict
+        return result
+
+
+class MinioConfig(FileStoreConfig):
+    user: str
+    password: str
+
+    def to_dict(self) -> dict[str, Any]:
+        result = super().to_dict()
+        if 'extra' not in result:
+            result['extra'] = dict()
+        extra_dict = result['extra'].copy()
+        extra_dict['user'] = self.user
+        extra_dict['password'] = self.password
+        result['extra'] = extra_dict
+        return result
+
+
+def load_configurations(config_path: str) -> list[BaseConfig]:
+    raw_configs = read_config(config_path)
+    configurations = []
+    ragflow_count = 0
+    id_count = 0
+    for k, v in raw_configs.items():
+        match (k):
+            case "ragflow":
+                name: str = f'ragflow_{ragflow_count}'
+                host: str = v['host']
+                http_port: int = v['http_port']
+                config = RAGFlowServerConfig(id=id_count, name=name, host=host, port=http_port,
+                                             service_type="ragflow_server",
+                                             detail_func_name="check_ragflow_server_alive")
+                configurations.append(config)
+                id_count += 1
+            case "es":
+                name: str = 'elasticsearch'
+                url = v['hosts']
+                parsed = urlparse(url)
+                host: str = parsed.hostname
+                port: int = parsed.port
+                username: str = v.get('username')
+                password: str = v.get('password')
+                config = ElasticsearchConfig(id=id_count, name=name, host=host, port=port, service_type="retrieval",
+                                             retrieval_type="elasticsearch",
+                                             username=username, password=password,
+                                             detail_func_name="get_es_cluster_stats")
+                configurations.append(config)
+                id_count += 1
+
+            case "infinity":
+                name: str = 'infinity'
+                url = v['uri']
+                parts = url.split(':', 1)
+                host = parts[0]
+                port = int(parts[1])
+                database: str = v.get('db_name', 'default_db')
+                config = InfinityConfig(id=id_count, name=name, host=host, port=port, service_type="retrieval",
+                                        retrieval_type="infinity",
+                                        db_name=database, detail_func_name="get_infinity_status")
+                configurations.append(config)
+                id_count += 1
+            case "minio":
+                name: str = 'minio'
+                url = v['host']
+                parts = url.split(':', 1)
+                host = parts[0]
+                port = int(parts[1])
+                user = v.get('user')
+                password = v.get('password')
+                config = MinioConfig(id=id_count, name=name, host=host, port=port, user=user, password=password,
+                                     service_type="file_store",
+                                     store_type="minio", detail_func_name="check_minio_alive")
+                configurations.append(config)
+                id_count += 1
+            case "redis":
+                name: str = 'redis'
+                url = v['host']
+                parts = url.split(':', 1)
+                host = parts[0]
+                port = int(parts[1])
+                password = v.get('password')
+                db: int = v.get('db')
+                config = RedisConfig(id=id_count, name=name, host=host, port=port, password=password, database=db,
+                                     service_type="message_queue", mq_type="redis", detail_func_name="get_redis_info")
+                configurations.append(config)
+                id_count += 1
+            case "mysql":
+                name: str = 'mysql'
+                host: str = v.get('host')
+                port: int = v.get('port')
+                username = v.get('user')
+                password = v.get('password')
+                config = MySQLConfig(id=id_count, name=name, host=host, port=port, username=username, password=password,
+                                     service_type="meta_data", meta_type="mysql", detail_func_name="get_mysql_status")
+                configurations.append(config)
+                id_count += 1
+            case "admin":
+                pass
+            case _:
+                logging.warning(f"Unknown configuration key: {k}")
+                continue
+
+    return configurations

admin/server/exceptions.py

@@ -0,0 +1,17 @@
+class AdminException(Exception):
+    def __init__(self, message, code=400):
+        super().__init__(message)
+        self.code = code
+        self.message = message
+
+class UserNotFoundError(AdminException):
+    def __init__(self, username):
+        super().__init__(f"User '{username}' not found", 404)
+
+class UserAlreadyExistsError(AdminException):
+    def __init__(self, username):
+        super().__init__(f"User '{username}' already exists", 409)
+
+class CannotDeleteAdminError(AdminException):
+    def __init__(self):
+        super().__init__("Cannot delete admin account", 403)

admin/server/responses.py

@@ -0,0 +1,34 @@
+#
+#  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.
+#
+
+
+from flask import jsonify
+
+
+def success_response(data=None, message="Success", code=0):
+    return jsonify({
+        "code": code,
+        "message": message,
+        "data": data
+    }), 200
+
+
+def error_response(message="Error", code=-1, data=None):
+    return jsonify({
+        "code": code,
+        "message": message,
+        "data": data
+    }), 400

admin/server/roles.py

@@ -0,0 +1,76 @@
+#
+#  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.
+#
+import logging
+
+from typing import Dict, Any
+
+from api.common.exceptions import AdminException
+
+
+class RoleMgr:
+    @staticmethod
+    def create_role(role_name: str, description: str):
+        error_msg = f"not implement: create role: {role_name}, description: {description}"
+        logging.error(error_msg)
+        raise AdminException(error_msg)
+
+    @staticmethod
+    def update_role_description(role_name: str, description: str) -> Dict[str, Any]:
+        error_msg = f"not implement: update role: {role_name} with description: {description}"
+        logging.error(error_msg)
+        raise AdminException(error_msg)
+
+    @staticmethod
+    def delete_role(role_name: str) -> Dict[str, Any]:
+        error_msg = f"not implement: drop role: {role_name}"
+        logging.error(error_msg)
+        raise AdminException(error_msg)
+
+    @staticmethod
+    def list_roles() -> Dict[str, Any]:
+        error_msg = "not implement: list roles"
+        logging.error(error_msg)
+        raise AdminException(error_msg)
+
+    @staticmethod
+    def get_role_permission(role_name: str) -> Dict[str, Any]:
+        error_msg = f"not implement: show role {role_name}"
+        logging.error(error_msg)
+        raise AdminException(error_msg)
+
+    @staticmethod
+    def grant_role_permission(role_name: str, actions: list, resource: str) -> Dict[str, Any]:
+        error_msg = f"not implement: grant role {role_name} actions: {actions} on {resource}"
+        logging.error(error_msg)
+        raise AdminException(error_msg)
+
+    @staticmethod
+    def revoke_role_permission(role_name: str, actions: list, resource: str) -> Dict[str, Any]:
+        error_msg = f"not implement: revoke role {role_name} actions: {actions} on {resource}"
+        logging.error(error_msg)
+        raise AdminException(error_msg)
+
+    @staticmethod
+    def update_user_role(user_name: str, role_name: str) -> Dict[str, Any]:
+        error_msg = f"not implement: update user role: {user_name} to role {role_name}"
+        logging.error(error_msg)
+        raise AdminException(error_msg)
+
+    @staticmethod
+    def get_user_permission(user_name: str) -> Dict[str, Any]:
+        error_msg = f"not implement: get user permission: {user_name}"
+        logging.error(error_msg)
+        raise AdminException(error_msg)

admin/server/routes.py

@@ -0,0 +1,371 @@
+#
+#  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.
+#
+
+import secrets
+
+from flask import Blueprint, request
+from flask_login import current_user, logout_user, login_required
+
+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
+
+admin_bp = Blueprint('admin', __name__, url_prefix='/api/v1/admin')
+
+
+@admin_bp.route('/login', methods=['POST'])
+def login():
+    if not request.json:
+        return error_response('Authorize admin failed.' ,400)
+    try:
+        email = request.json.get("email", "")
+        password = request.json.get("password", "")
+        return login_admin(email, password)
+    except Exception as e:
+        return error_response(str(e), 500)
+
+
+@admin_bp.route('/logout', methods=['GET'])
+@login_required
+def logout():
+    try:
+        current_user.access_token = f"INVALID_{secrets.token_hex(16)}"
+        current_user.save()
+        logout_user()
+        return success_response(True)
+    except Exception as e:
+        return error_response(str(e), 500)
+
+
+@admin_bp.route('/auth', methods=['GET'])
+@login_verify
+def auth_admin():
+    try:
+        return success_response(None, "Admin is authorized", 0)
+    except Exception as e:
+        return error_response(str(e), 500)
+
+
+@admin_bp.route('/users', methods=['GET'])
+@login_required
+@check_admin_auth
+def list_users():
+    try:
+        users = UserMgr.get_all_users()
+        return success_response(users, "Get all users", 0)
+    except Exception as e:
+        return error_response(str(e), 500)
+
+
+@admin_bp.route('/users', methods=['POST'])
+@login_required
+@check_admin_auth
+def create_user():
+    try:
+        data = request.get_json()
+        if not data or 'username' not in data or 'password' not in data:
+            return error_response("Username and password are required", 400)
+
+        username = data['username']
+        password = data['password']
+        role = data.get('role', 'user')
+
+        res = UserMgr.create_user(username, password, role)
+        if res["success"]:
+            user_info = res["user_info"]
+            user_info.pop("password")  # do not return password
+            return success_response(user_info, "User created successfully")
+        else:
+            return error_response("create user failed")
+
+    except AdminException as e:
+        return error_response(e.message, e.code)
+    except Exception as e:
+        return error_response(str(e))
+
+
+@admin_bp.route('/users/<username>', methods=['DELETE'])
+@login_required
+@check_admin_auth
+def delete_user(username):
+    try:
+        res = UserMgr.delete_user(username)
+        if res["success"]:
+            return success_response(None, res["message"])
+        else:
+            return error_response(res["message"])
+
+    except AdminException as e:
+        return error_response(e.message, e.code)
+    except Exception as e:
+        return error_response(str(e), 500)
+
+
+@admin_bp.route('/users/<username>/password', methods=['PUT'])
+@login_required
+@check_admin_auth
+def change_password(username):
+    try:
+        data = request.get_json()
+        if not data or 'new_password' not in data:
+            return error_response("New password is required", 400)
+
+        new_password = data['new_password']
+        msg = UserMgr.update_user_password(username, new_password)
+        return success_response(None, msg)
+
+    except AdminException as e:
+        return error_response(e.message, e.code)
+    except Exception as e:
+        return error_response(str(e), 500)
+
+
+@admin_bp.route('/users/<username>/activate', methods=['PUT'])
+@login_required
+@check_admin_auth
+def alter_user_activate_status(username):
+    try:
+        data = request.get_json()
+        if not data or 'activate_status' not in data:
+            return error_response("Activation status is required", 400)
+        activate_status = data['activate_status']
+        msg = UserMgr.update_user_activate_status(username, activate_status)
+        return success_response(None, msg)
+    except AdminException as e:
+        return error_response(e.message, e.code)
+    except Exception as e:
+        return error_response(str(e), 500)
+
+
+@admin_bp.route('/users/<username>', methods=['GET'])
+@login_required
+@check_admin_auth
+def get_user_details(username):
+    try:
+        user_details = UserMgr.get_user_details(username)
+        return success_response(user_details)
+
+    except AdminException as e:
+        return error_response(e.message, e.code)
+    except Exception as e:
+        return error_response(str(e), 500)
+
+
+@admin_bp.route('/users/<username>/datasets', methods=['GET'])
+@login_required
+@check_admin_auth
+def get_user_datasets(username):
+    try:
+        datasets_list = UserServiceMgr.get_user_datasets(username)
+        return success_response(datasets_list)
+
+    except AdminException as e:
+        return error_response(e.message, e.code)
+    except Exception as e:
+        return error_response(str(e), 500)
+
+
+@admin_bp.route('/users/<username>/agents', methods=['GET'])
+@login_required
+@check_admin_auth
+def get_user_agents(username):
+    try:
+        agents_list = UserServiceMgr.get_user_agents(username)
+        return success_response(agents_list)
+
+    except AdminException as e:
+        return error_response(e.message, e.code)
+    except Exception as e:
+        return error_response(str(e), 500)
+
+
+@admin_bp.route('/services', methods=['GET'])
+@login_required
+@check_admin_auth
+def get_services():
+    try:
+        services = ServiceMgr.get_all_services()
+        return success_response(services, "Get all services", 0)
+    except Exception as e:
+        return error_response(str(e), 500)
+
+
+@admin_bp.route('/service_types/<service_type>', methods=['GET'])
+@login_required
+@check_admin_auth
+def get_services_by_type(service_type_str):
+    try:
+        services = ServiceMgr.get_services_by_type(service_type_str)
+        return success_response(services)
+    except Exception as e:
+        return error_response(str(e), 500)
+
+
+@admin_bp.route('/services/<service_id>', methods=['GET'])
+@login_required
+@check_admin_auth
+def get_service(service_id):
+    try:
+        services = ServiceMgr.get_service_details(service_id)
+        return success_response(services)
+    except Exception as e:
+        return error_response(str(e), 500)
+
+
+@admin_bp.route('/services/<service_id>', methods=['DELETE'])
+@login_required
+@check_admin_auth
+def shutdown_service(service_id):
+    try:
+        services = ServiceMgr.shutdown_service(service_id)
+        return success_response(services)
+    except Exception as e:
+        return error_response(str(e), 500)
+
+
+@admin_bp.route('/services/<service_id>', methods=['PUT'])
+@login_required
+@check_admin_auth
+def restart_service(service_id):
+    try:
+        services = ServiceMgr.restart_service(service_id)
+        return success_response(services)
+    except Exception as e:
+        return error_response(str(e), 500)
+
+
+@admin_bp.route('/roles', methods=['POST'])
+@login_required
+@check_admin_auth
+def create_role():
+    try:
+        data = request.get_json()
+        if not data or 'role_name' not in data:
+            return error_response("Role name is required", 400)
+        role_name: str = data['role_name']
+        description: str = data['description']
+        res = RoleMgr.create_role(role_name, description)
+        return success_response(res)
+    except Exception as e:
+        return error_response(str(e), 500)
+
+
+@admin_bp.route('/roles/<role_name>', methods=['PUT'])
+@login_required
+@check_admin_auth
+def update_role(role_name: str):
+    try:
+        data = request.get_json()
+        if not data or 'description' not in data:
+            return error_response("Role description is required", 400)
+        description: str = data['description']
+        res = RoleMgr.update_role_description(role_name, description)
+        return success_response(res)
+    except Exception as e:
+        return error_response(str(e), 500)
+
+
+@admin_bp.route('/roles/<role_name>', methods=['DELETE'])
+@login_required
+@check_admin_auth
+def delete_role(role_name: str):
+    try:
+        res = RoleMgr.delete_role(role_name)
+        return success_response(res)
+    except Exception as e:
+        return error_response(str(e), 500)
+
+
+@admin_bp.route('/roles', methods=['GET'])
+@login_required
+@check_admin_auth
+def list_roles():
+    try:
+        res = RoleMgr.list_roles()
+        return success_response(res)
+    except Exception as e:
+        return error_response(str(e), 500)
+
+
+@admin_bp.route('/roles/<role_name>/permission', methods=['GET'])
+@login_required
+@check_admin_auth
+def get_role_permission(role_name: str):
+    try:
+        res = RoleMgr.get_role_permission(role_name)
+        return success_response(res)
+    except Exception as e:
+        return error_response(str(e), 500)
+
+
+@admin_bp.route('/roles/<role_name>/permission', methods=['POST'])
+@login_required
+@check_admin_auth
+def grant_role_permission(role_name: str):
+    try:
+        data = request.get_json()
+        if not data or 'actions' not in data or 'resource' not in data:
+            return error_response("Permission is required", 400)
+        actions: list = data['actions']
+        resource: str = data['resource']
+        res = RoleMgr.grant_role_permission(role_name, actions, resource)
+        return success_response(res)
+    except Exception as e:
+        return error_response(str(e), 500)
+
+
+@admin_bp.route('/roles/<role_name>/permission', methods=['DELETE'])
+@login_required
+@check_admin_auth
+def revoke_role_permission(role_name: str):
+    try:
+        data = request.get_json()
+        if not data or 'actions' not in data or 'resource' not in data:
+            return error_response("Permission is required", 400)
+        actions: list = data['actions']
+        resource: str = data['resource']
+        res = RoleMgr.revoke_role_permission(role_name, actions, resource)
+        return success_response(res)
+    except Exception as e:
+        return error_response(str(e), 500)
+
+
+@admin_bp.route('/users/<user_name>/role', methods=['PUT'])
+@login_required
+@check_admin_auth
+def update_user_role(user_name: str):
+    try:
+        data = request.get_json()
+        if not data or 'role_name' not in data:
+            return error_response("Role name is required", 400)
+        role_name: str = data['role_name']
+        res = RoleMgr.update_user_role(user_name, role_name)
+        return success_response(res)
+    except Exception as e:
+        return error_response(str(e), 500)
+
+
+@admin_bp.route('/users/<user_name>/permission', methods=['GET'])
+@login_required
+@check_admin_auth
+def get_user_permission(user_name: str):
+    try:
+        res = RoleMgr.get_user_permission(user_name)
+        return success_response(res)
+    except Exception as e:
+        return error_response(str(e), 500)

admin/server/services.py

@@ -0,0 +1,225 @@
+#
+#  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.
+#
+
+
+import re
+from werkzeug.security import check_password_hash
+from common.constants import ActiveEnum
+from api.db.services import UserService
+from api.db.joint_services.user_account_service import create_new_user, delete_user_data
+from api.db.services.canvas_service import UserCanvasService
+from api.db.services.user_service import TenantService
+from api.db.services.knowledgebase_service import KnowledgebaseService
+from api.utils.crypt import decrypt
+from api.utils import health_utils
+
+from api.common.exceptions import AdminException, UserAlreadyExistsError, UserNotFoundError
+from config import SERVICE_CONFIGS
+
+
+class UserMgr:
+    @staticmethod
+    def get_all_users():
+        users = UserService.get_all_users()
+        result = []
+        for user in users:
+            result.append({
+                'email': user.email,
+                'nickname': user.nickname,
+                'create_date': user.create_date,
+                'is_active': user.is_active,
+                'is_superuser': user.is_superuser,
+            })
+        return result
+
+    @staticmethod
+    def get_user_details(username):
+        # use email to query
+        users = UserService.query_user_by_email(username)
+        result = []
+        for user in users:
+            result.append({
+                'email': user.email,
+                'language': user.language,
+                'last_login_time': user.last_login_time,
+                'is_active': user.is_active,
+                'is_anonymous': user.is_anonymous,
+                'login_channel': user.login_channel,
+                'status': user.status,
+                'is_superuser': user.is_superuser,
+                'create_date': user.create_date,
+                'update_date': user.update_date
+            })
+        return result
+
+    @staticmethod
+    def create_user(username, password, role="user") -> dict:
+        # Validate the email address
+        if not re.match(r"^[\w\._-]+@([\w_-]+\.)+[\w-]{2,}$", username):
+            raise AdminException(f"Invalid email address: {username}!")
+        # Check if the email address is already used
+        if UserService.query(email=username):
+            raise UserAlreadyExistsError(username)
+        # Construct user info data
+        user_info_dict = {
+            "email": username,
+            "nickname": "",  # ask user to edit it manually in settings.
+            "password": decrypt(password),
+            "login_channel": "password",
+            "is_superuser": role == "admin",
+        }
+        return create_new_user(user_info_dict)
+
+    @staticmethod
+    def delete_user(username):
+        # use email to delete
+        user_list = UserService.query_user_by_email(username)
+        if not user_list:
+            raise UserNotFoundError(username)
+        if len(user_list) > 1:
+            raise AdminException(f"Exist more than 1 user: {username}!")
+        usr = user_list[0]
+        return delete_user_data(usr.id)
+
+    @staticmethod
+    def update_user_password(username, new_password) -> str:
+        # use email to find user. check exist and unique.
+        user_list = UserService.query_user_by_email(username)
+        if not user_list:
+            raise UserNotFoundError(username)
+        elif len(user_list) > 1:
+            raise AdminException(f"Exist more than 1 user: {username}!")
+        # check new_password different from old.
+        usr = user_list[0]
+        psw = decrypt(new_password)
+        if check_password_hash(usr.password, psw):
+            return "Same password, no need to update!"
+        # update password
+        UserService.update_user_password(usr.id, psw)
+        return "Password updated successfully!"
+
+    @staticmethod
+    def update_user_activate_status(username, activate_status: str):
+        # use email to find user. check exist and unique.
+        user_list = UserService.query_user_by_email(username)
+        if not user_list:
+            raise UserNotFoundError(username)
+        elif len(user_list) > 1:
+            raise AdminException(f"Exist more than 1 user: {username}!")
+        # check activate status different from new
+        usr = user_list[0]
+        # format activate_status before handle
+        _activate_status = activate_status.lower()
+        target_status = {
+            'on': ActiveEnum.ACTIVE.value,
+            'off': ActiveEnum.INACTIVE.value,
+        }.get(_activate_status)
+        if not target_status:
+            raise AdminException(f"Invalid activate_status: {activate_status}")
+        if target_status == usr.is_active:
+            return f"User activate status is already {_activate_status}!"
+        # update is_active
+        UserService.update_user(usr.id, {"is_active": target_status})
+        return f"Turn {_activate_status} user activate status successfully!"
+
+
+class UserServiceMgr:
+
+    @staticmethod
+    def get_user_datasets(username):
+        # use email to find user.
+        user_list = UserService.query_user_by_email(username)
+        if not user_list:
+            raise UserNotFoundError(username)
+        elif len(user_list) > 1:
+            raise AdminException(f"Exist more than 1 user: {username}!")
+        # find tenants
+        usr = user_list[0]
+        tenants = TenantService.get_joined_tenants_by_user_id(usr.id)
+        tenant_ids = [m["tenant_id"] for m in tenants]
+        # filter permitted kb and owned kb
+        return KnowledgebaseService.get_all_kb_by_tenant_ids(tenant_ids, usr.id)
+
+    @staticmethod
+    def get_user_agents(username):
+        # use email to find user.
+        user_list = UserService.query_user_by_email(username)
+        if not user_list:
+            raise UserNotFoundError(username)
+        elif len(user_list) > 1:
+            raise AdminException(f"Exist more than 1 user: {username}!")
+        # find tenants
+        usr = user_list[0]
+        tenants = TenantService.get_joined_tenants_by_user_id(usr.id)
+        tenant_ids = [m["tenant_id"] for m in tenants]
+        # filter permitted agents and owned agents
+        res = UserCanvasService.get_all_agents_by_tenant_ids(tenant_ids, usr.id)
+        return [{
+            'title': r['title'],
+            'permission': r['permission'],
+            'canvas_category': r['canvas_category'].split('_')[0]
+        } for r in res]
+
+
+class ServiceMgr:
+
+    @staticmethod
+    def get_all_services():
+        result = []
+        configs = SERVICE_CONFIGS.configs
+        for service_id, config in enumerate(configs):
+            config_dict = config.to_dict()
+            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:
+                config_dict['status'] = 'timeout'
+            result.append(config_dict)
+        return result
+
+    @staticmethod
+    def get_services_by_type(service_type_str: str):
+        raise AdminException("get_services_by_type: not implemented")
+
+    @staticmethod
+    def get_service_details(service_id: int):
+        service_id = 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}")
+
+        detail_func = getattr(health_utils, service_info.get('detail_func_name'))
+        res = detail_func()
+        res.update({'service_name': service_info.get('name')})
+        return res
+
+    @staticmethod
+    def shutdown_service(service_id: int):
+        raise AdminException("shutdown_service: not implemented")
+
+    @staticmethod
+    def restart_service(service_id: int):
+        raise AdminException("restart_service: not implemented")

agent/canvas.py

@@ -26,8 +26,8 @@ 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.utils import get_uuid, hash_str2int
+from common.misc_utils import get_uuid, hash_str2int
-from rag.prompts.prompts import chunks_format
+from rag.prompts.generator import chunks_format
 from rag.utils.redis_conn import REDIS_CONN
 
 class Graph:
@@ -153,6 +153,41 @@ class Graph:
     def get_tenant_id(self):
         return self._tenant_id
 
+    def get_variable_value(self, exp: str) -> Any:
+        exp = exp.strip("{").strip("}").strip(" ").strip("{").strip("}")
+        if exp.find("@") < 0:
+            return self.globals[exp]
+        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 ""
+        root_val = cpn["obj"].output(root_key)
+
+        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:
+            return cur
+        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)
+        return cur
+
 
 class Canvas(Graph):
 
@@ -193,7 +228,6 @@ class Canvas(Graph):
             self.history = []
             self.retrieval = []
             self.memory = []
-
         for k in self.globals.keys():
             if isinstance(self.globals[k], str):
                 self.globals[k] = ""
@@ -247,12 +281,21 @@ class Canvas(Graph):
         def _run_batch(f, t):
             with ThreadPoolExecutor(max_workers=5) as executor:
                 thr = []
-                for i in range(f, t):
+                i = f
+                while i < 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", {})))
+                        i += 1
                     else:
-                        thr.append(executor.submit(cpn.invoke, **cpn.get_input()))
+                        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]:
+                                self.path.pop(i)
+                                t -= 1
+                                break
+                        else:
+                            thr.append(executor.submit(cpn.invoke, **cpn.get_input()))
+                            i += 1
                 for t in thr:
                     t.result()
 
@@ -282,7 +325,7 @@ class Canvas(Graph):
                     "thoughts": self.get_component_thoughts(self.path[i])
                 })
             _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])
@@ -383,7 +426,6 @@ class Canvas(Graph):
                 self.path = path
                 yield decorate("user_inputs", {"inputs": another_inputs, "tips": tips})
                 return
-
         self.path = self.path[:idx]
         if not self.error:
             yield decorate("workflow_finished",
@@ -406,16 +448,6 @@ class Canvas(Graph):
             return False
         return True
 
-    def get_variable_value(self, exp: str) -> Any:
-        exp = exp.strip("{").strip("}").strip(" ").strip("{").strip("}")
-        if exp.find("@") < 0:
-            return self.globals[exp]
-        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}'")
-        return cpn["obj"].output(var_nm)
-
     def get_history(self, window_size):
         convs = []
         if window_size <= 0:
@@ -490,7 +522,8 @@ class Canvas(Graph):
 
         r = self.retrieval[-1]
         for ck in chunks_format({"chunks": chunks}):
-            cid = hash_str2int(ck["id"], 100)
+            cid = hash_str2int(ck["id"], 500)
+            # cid = uuid.uuid5(uuid.NAMESPACE_DNS, ck["id"])
             if cid not in r:
                 r["chunks"][cid] = ck
 

agent/component/agent_with_tools.py

@@ -27,10 +27,9 @@ from agent.tools.base import LLMToolPluginCallSession, ToolParamBase, ToolBase,
 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 api.utils.api_utils import timeout
+from common.connection_utils import timeout
-from rag.prompts import message_fit_in
+from rag.prompts.generator import next_step, COMPLETE_TASK, analyze_task, \
-from rag.prompts.prompts import next_step, COMPLETE_TASK, analyze_task, \
+    citation_prompt, reflect, rank_memories, kb_prompt, citation_plus, full_question, message_fit_in
-    citation_prompt, reflect, rank_memories, kb_prompt, citation_plus, full_question
 from rag.utils.mcp_tool_call_conn import MCPToolCallSession, mcp_tool_metadata_to_openai_tool
 from agent.component.llm import LLMParam, LLM
 
@@ -138,7 +137,7 @@ class Agent(LLM, ToolBase):
             res.update(cpn.get_input_form())
         return res
 
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 20*60))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 20*60)))
     def _invoke(self, **kwargs):
         if kwargs.get("user_prompt"):
             usr_pmt = ""
@@ -159,7 +158,12 @@ class Agent(LLM, ToolBase):
 
         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 self._param.output_structure and not (ex and ex["goto"]):
+        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))
             return
 
@@ -347,3 +351,11 @@ Respond immediately with your final comprehensive answer.
 
         return "Error occurred."
 
+    def reset(self, temp=False):
+        """
+        Reset all tools if they have a reset method. This avoids errors for tools like MCPToolCallSession.
+        """
+        for k, cpn in self.tools.items():
+            if hasattr(cpn, "reset") and callable(cpn.reset):
+                cpn.reset()
+

agent/component/base.py

@@ -25,7 +25,7 @@ from typing import Any, List, Union
 import pandas as pd
 import trio
 from agent import settings
-from api.utils.api_utils import timeout
+from common.connection_utils import timeout
 
 
 _FEEDED_DEPRECATED_PARAMS = "_feeded_deprecated_params"
@@ -244,7 +244,7 @@ class ComponentParamBase(ABC):
 
                 if not value_legal:
                     raise ValueError(
-                        "Plase check runtime conf, {} = {} does not match user-parameter restriction".format(
+                        "Please check runtime conf, {} = {} does not match user-parameter restriction".format(
                             variable, value
                         )
                     )
@@ -431,7 +431,7 @@ class ComponentBase(ABC):
         self.set_output("_elapsed_time", time.perf_counter() - self.output("_created_time"))
         return self.output()
 
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60)))
     def _invoke(self, **kwargs):
         raise NotImplementedError()
 

agent/component/categorize.py

@@ -18,17 +18,17 @@ import os
 import re
 from abc import ABC
 
-from api.db import LLMType
+from common.constants import LLMType
 from api.db.services.llm_service import LLMBundle
 from agent.component.llm import LLMParam, LLM
-from api.utils.api_utils import timeout
+from common.connection_utils import timeout
 from rag.llm.chat_model import ERROR_PREFIX
 
 
 class CategorizeParam(LLMParam):
 
     """
-    Define the Categorize component parameters.
+    Define the categorize component parameters.
     """
     def __init__(self):
         super().__init__()
@@ -80,7 +80,7 @@ Here's description of each category:
  - Prioritize the most specific applicable category
  - Return only the category name without explanations
  - Use "Other" only when no other category fits
- 
+
  """.format(
             "\n - ".join(list(self.category_description.keys())),
             "\n".join(descriptions)
@@ -96,7 +96,7 @@ Here's description of each category:
 class Categorize(LLM, ABC):
     component_name = "Categorize"
 
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60)))
     def _invoke(self, **kwargs):
         msg = self._canvas.get_history(self._param.message_history_window_size)
         if not msg:
@@ -112,7 +112,7 @@ class Categorize(LLM, ABC):
 
         user_prompt = """
 ---- 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())
         logging.info(f"input: {user_prompt}, answer: {str(ans)}")
@@ -134,4 +134,4 @@ class Categorize(LLM, ABC):
         self.set_output("_next", cpn_ids)
 
     def thoughts(self) -> str:
-        return "Which should it falls into {}? ...".format(",".join([f"`{c}`" for c, _ in self._param.category_description.items()]))
+        return "Which should it falls into {}? ...".format(",".join([f"`{c}`" for c, _ in self._param.category_description.items()]))

agent/component/data_operations.py

@@ -0,0 +1,201 @@
+from abc import ABC
+import ast
+import os
+from agent.component.base import ComponentBase, ComponentParamBase
+from api.utils.api_utils import timeout
+
+class DataOperationsParam(ComponentParamBase):
+    """
+    Define the Data Operations component parameters.
+    """
+    def __init__(self):
+        super().__init__()
+        self.inputs = []
+        self.operations = "literal_eval"
+        self.select_keys = []
+        self.filter_values=[]
+        self.updates=[]
+        self.remove_keys=[]
+        self.rename_keys=[]
+        self.outputs = {
+            "result": {
+                "value": [],
+                "type": "Array of Object"
+            }
+        }
+    
+    def check(self):
+        self.check_valid_value(self.operations, "Support operations", ["select_keys", "literal_eval","combine","filter_values","append_or_update","remove_keys","rename_keys"])
+    
+    
+
+class DataOperations(ComponentBase,ABC):
+    component_name = "DataOperations"
+
+    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 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)
+        if not isinstance(inputs, (list, tuple)):
+            inputs = [inputs]
+        for input_ref in self._param.inputs:
+            input_object=self._canvas.get_variable_value(input_ref)
+            if input_object is None:
+                continue
+            if isinstance(input_object,dict):
+                self.input_objects.append(input_object)
+            elif isinstance(input_object,list):
+                self.input_objects.extend(x for x in input_object if isinstance(x, dict))
+            else:
+                continue
+        if self._param.operations == "select_keys":
+            self._select_keys()
+        elif self._param.operations == "literal_eval":
+            self._literal_eval()
+        elif self._param.operations == "combine":
+            self._combine()
+        elif self._param.operations == "filter_values":
+            self._filter_values()
+        elif self._param.operations == "append_or_update":
+            self._append_or_update()
+        elif self._param.operations == "remove_keys":
+            self._remove_keys()
+        else:
+            self._rename_keys()
+    
+    def _select_keys(self):
+        filter_criteria: list[str] = self._param.select_keys
+        results = [{key: value for key, value in data_dict.items() if key in filter_criteria} for data_dict in self.input_objects]
+        self.set_output("result", results)
+
+
+    def _recursive_eval(self, data):
+        if isinstance(data, dict):
+            return {k: self.recursive_eval(v) for k, v in data.items()}
+        if isinstance(data, list):
+            return [self.recursive_eval(item) for item in data]
+        if isinstance(data, str):
+            try:
+                if (
+                    data.strip().startswith(("{", "[", "(", "'", '"'))
+                    or data.strip().lower() in ("true", "false", "none")
+                    or data.strip().replace(".", "").isdigit()
+                ):
+                    return ast.literal_eval(data)
+            except (ValueError, SyntaxError, TypeError, MemoryError):
+                return data
+            else:
+                return data
+        return data
+    
+    def _literal_eval(self):
+        self.set_output("result", self._recursive_eval(self.input_objects))
+
+    def _combine(self):
+        result={}
+        for obj in self.input_objects():
+            for key, value in obj.items():
+                if key not in result:
+                    result[key] = value
+                elif isinstance(result[key], list):
+                    if isinstance(value, list):
+                        result[key].extend(value)
+                    else:
+                        result[key].append(value)
+                else:
+                    result[key] = (
+                        [result[key], value] if not isinstance(value, list) else [result[key], *value]
+                    )
+        self.set_output("result", result)
+    
+    def norm(self,v):
+        s = "" if v is None else str(v)
+        return s
+    
+    def match_rule(self, obj, rule):
+        key = rule.get("key")
+        op = (rule.get("operator") or "equals").lower()
+        target = self.norm(rule.get("value"))
+        if key not in obj:
+            return False
+        val = obj.get(key, None)
+        v = self.norm(val)
+        if op == "=":
+            return v == target
+        if op == "≠":
+            return v != target
+        if op == "contains":
+            return target in v
+        if op == "start with":
+            return v.startswith(target)
+        if op == "end with":
+            return v.endswith(target)
+        return False
+        
+    def _filter_values(self):
+        results=[]
+        rules = (getattr(self._param, "filter_values", None) or [])
+        for obj in self.input_objects():
+            if not rules:
+                results.append(obj)
+                continue
+            if all(self.match_rule(obj, r) for r in rules):
+                results.append(obj)
+        self.set_output("result", results)
+            
+                
+    def _append_or_update(self):
+        results=[]
+        updates = getattr(self._param, "updates", []) or [] 
+        for obj in self.input_objects():
+            new_obj = dict(obj)
+            for item in updates:
+                if not isinstance(item, dict):
+                    continue
+                k = (item.get("key") or "").strip()
+                if not k:
+                    continue
+                new_obj[k] = item.get("value")
+            results.append(new_obj)
+        self.set_output("result", results)
+
+    def _remove_keys(self):
+        results = []
+        remove_keys = getattr(self._param, "remove_keys", []) or []
+
+        for obj in (self.input_objects or []):
+            new_obj = dict(obj)
+            for k in remove_keys:
+                if not isinstance(k, str):
+                    continue
+                new_obj.pop(k, None)
+            results.append(new_obj)
+        self.set_output("result", results)
+
+    def _rename_keys(self):
+        results = []
+        rename_pairs = getattr(self._param, "rename_keys", []) or []
+
+        for obj in (self.input_objects or []):
+            new_obj = dict(obj)
+            for pair in rename_pairs:
+                if not isinstance(pair, dict):
+                    continue
+                old = (pair.get("old_key") or "").strip()
+                new = (pair.get("new_key") or "").strip()
+                if not old or not new or old == new:
+                    continue
+                if old in new_obj:
+                    new_obj[new] = new_obj.pop(old)
+            results.append(new_obj)
+        self.set_output("result", results)
+
+    def thoughts(self) -> str:
+        return "DataOperation in progress"

agent/component/invoke.py

@@ -19,11 +19,12 @@ import os
 import re
 import time
 from abc import ABC
+
 import requests
 
-from api.utils.api_utils import timeout
-from deepdoc.parser import HtmlParser
 from agent.component.base import ComponentBase, ComponentParamBase
+from common.connection_utils import timeout
+from deepdoc.parser import HtmlParser
 
 
 class InvokeParam(ComponentParamBase):
@@ -43,17 +44,17 @@ class InvokeParam(ComponentParamBase):
         self.datatype = "json"  # New parameter to determine data posting type
 
     def check(self):
-        self.check_valid_value(self.method.lower(), "Type of content from the crawler", ['get', 'post', 'put'])
+        self.check_valid_value(self.method.lower(), "Type of content from the crawler", ["get", "post", "put"])
         self.check_empty(self.url, "End point URL")
         self.check_positive_integer(self.timeout, "Timeout time in second")
         self.check_boolean(self.clean_html, "Clean HTML")
-        self.check_valid_value(self.datatype.lower(), "Data post type", ['json', 'formdata'])  # Check for valid datapost value
+        self.check_valid_value(self.datatype.lower(), "Data post type", ["json", "formdata"])  # Check for valid datapost value
 
 
 class Invoke(ComponentBase, ABC):
     component_name = "Invoke"
 
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 3))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 3)))
     def _invoke(self, **kwargs):
         args = {}
         for para in self._param.variables:
@@ -63,6 +64,18 @@ class Invoke(ComponentBase, ABC):
                 args[para["key"]] = self._canvas.get_variable_value(para["ref"])
 
         url = self._param.url.strip()
+
+        def replace_variable(match):
+            var_name = match.group(1)
+            try:
+                value = self._canvas.get_variable_value(var_name)
+                return str(value or "")
+            except Exception:
+                return ""
+
+        # {base_url} or {component_id@variable_name}
+        url = re.sub(r"\{([a-zA-Z_][a-zA-Z0-9_.@-]*)\}", replace_variable, url)
+
         if url.find("http") != 0:
             url = "http://" + url
 
@@ -75,52 +88,32 @@ class Invoke(ComponentBase, ABC):
             proxies = {"http": self._param.proxy, "https": self._param.proxy}
 
         last_e = ""
-        for _ in range(self._param.max_retries+1):
+        for _ in range(self._param.max_retries + 1):
             try:
-                if method == 'get':
+                if method == "get":
-                    response = requests.get(url=url,
+                    response = requests.get(url=url, params=args, headers=headers, proxies=proxies, timeout=self._param.timeout)
-                                            params=args,
-                                            headers=headers,
-                                            proxies=proxies,
-                                            timeout=self._param.timeout)
                     if self._param.clean_html:
                         sections = HtmlParser()(None, response.content)
                         self.set_output("result", "\n".join(sections))
                     else:
                         self.set_output("result", response.text)
 
-                if method == 'put':
+                if method == "put":
-                    if self._param.datatype.lower() == 'json':
+                    if self._param.datatype.lower() == "json":
-                        response = requests.put(url=url,
+                        response = requests.put(url=url, json=args, headers=headers, proxies=proxies, timeout=self._param.timeout)
-                                                json=args,
-                                                headers=headers,
-                                                proxies=proxies,
-                                                timeout=self._param.timeout)
                     else:
-                        response = requests.put(url=url,
+                        response = requests.put(url=url, data=args, headers=headers, proxies=proxies, timeout=self._param.timeout)
-                                                data=args,
-                                                headers=headers,
-                                                proxies=proxies,
-                                                timeout=self._param.timeout)
                     if self._param.clean_html:
                         sections = HtmlParser()(None, response.content)
                         self.set_output("result", "\n".join(sections))
                     else:
                         self.set_output("result", response.text)
 
-                if method == 'post':
+                if method == "post":
-                    if self._param.datatype.lower() == 'json':
+                    if self._param.datatype.lower() == "json":
-                        response = requests.post(url=url,
+                        response = requests.post(url=url, json=args, headers=headers, proxies=proxies, timeout=self._param.timeout)
-                                                 json=args,
-                                                 headers=headers,
-                                                 proxies=proxies,
-                                                 timeout=self._param.timeout)
                     else:
-                        response = requests.post(url=url,
+                        response = requests.post(url=url, data=args, headers=headers, proxies=proxies, timeout=self._param.timeout)
-                                                 data=args,
-                                                 headers=headers,
-                                                 proxies=proxies,
-                                                 timeout=self._param.timeout)
                     if self._param.clean_html:
                         self.set_output("result", "\n".join(sections))
                     else:

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):
     """

agent/component/llm.py

@@ -21,13 +21,12 @@ from copy import deepcopy
 from typing import Any, Generator
 import json_repair
 from functools import partial
-from api.db import LLMType
+from common.constants import LLMType
 from api.db.services.llm_service import LLMBundle
 from api.db.services.tenant_llm_service import TenantLLMService
 from agent.component.base import ComponentBase, ComponentParamBase
-from api.utils.api_utils import timeout
+from common.connection_utils import timeout
-from rag.prompts import message_fit_in, citation_prompt
+from rag.prompts.generator import tool_call_summary, message_fit_in, citation_prompt, structured_output_prompt
-from rag.prompts.prompts import tool_call_summary
 
 
 class LLMParam(ComponentParamBase):
@@ -82,9 +81,9 @@ class LLMParam(ComponentParamBase):
 
 class LLM(ComponentBase):
     component_name = "LLM"
-    
+
-    def __init__(self, canvas, id, param: ComponentParamBase):
+    def __init__(self, canvas, component_id, param: ComponentParamBase):
-        super().__init__(canvas, id, param)
+        super().__init__(canvas, component_id, param)
         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,
                                   retry_interval=self._param.delay_after_error
@@ -102,6 +101,8 @@ class LLM(ComponentBase):
 
     def get_input_elements(self) -> dict[str, Any]:
         res = self.get_input_elements_from_text(self._param.sys_prompt)
+        if isinstance(self._param.prompts, str):
+            self._param.prompts = [{"role": "user", "content": self._param.prompts}]
         for prompt in self._param.prompts:
             d = self.get_input_elements_from_text(prompt["content"])
             res.update(d)
@@ -113,6 +114,17 @@ class LLM(ComponentBase):
     def add2system_prompt(self, txt):
         self._param.sys_prompt += txt
 
+    def _sys_prompt_and_msg(self, msg, args):
+        if isinstance(self._param.prompts, str):
+            self._param.prompts = [{"role": "user", "content": self._param.prompts}]
+        for p in self._param.prompts:
+            if msg and msg[-1]["role"] == p["role"]:
+                continue
+            p = deepcopy(p)
+            p["content"] = self.string_format(p["content"], args)
+            msg.append(p)
+        return msg, self.string_format(self._param.sys_prompt, args)
+
     def _prepare_prompt_variables(self):
         if self._param.visual_files_var:
             self.imgs = self._canvas.get_variable_value(self._param.visual_files_var)
@@ -128,7 +140,6 @@ class LLM(ComponentBase):
 
         args = {}
         vars = self.get_input_elements() if not self._param.debug_inputs else self._param.debug_inputs
-        sys_prompt = self._param.sys_prompt
         for k, o in vars.items():
             args[k] = o["value"]
             if not isinstance(args[k], str):
@@ -138,16 +149,8 @@ class LLM(ComponentBase):
                     args[k] = str(args[k])
             self.set_input_value(k, args[k])
 
-        msg = self._canvas.get_history(self._param.message_history_window_size)[:-1]
+        msg, sys_prompt = self._sys_prompt_and_msg(self._canvas.get_history(self._param.message_history_window_size)[:-1], args)
-        for p in self._param.prompts:
-            if msg and msg[-1]["role"] == p["role"]:
-                continue
-            msg.append(deepcopy(p))
-
-        sys_prompt = self.string_format(sys_prompt, args)
         user_defined_prompt, sys_prompt = self._extract_prompts(sys_prompt)
-        for m in msg:
-            m["content"] = self.string_format(m["content"], args)
         if self._param.cite and self._canvas.get_reference()["chunks"]:
             sys_prompt += citation_prompt(user_defined_prompt)
 
@@ -202,7 +205,7 @@ class LLM(ComponentBase):
             for txt in self.chat_mdl.chat_streamly(msg[0]["content"], msg[1:], self._param.gen_conf(), images=self.imgs, **kwargs):
                 yield delta(txt)
 
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60)))
     def _invoke(self, **kwargs):
         def clean_formated_answer(ans: str) -> str:
             ans = re.sub(r"^.*</think>", "", ans, flags=re.DOTALL)
@@ -210,11 +213,15 @@ class LLM(ComponentBase):
             return re.sub(r"```\n*$", "", ans, flags=re.DOTALL)
 
         prompt, msg, _ = self._prepare_prompt_variables()
-        error = ""
+        error: str = ""
-
+        output_structure=None
-        if self._param.output_structure:
+        try:
-            prompt += "\nThe output MUST follow this JSON format:\n"+json.dumps(self._param.output_structure, ensure_ascii=False, indent=2)
+            output_structure = None#self._param.outputs['structured']
-            prompt += "\nRedundant information is FORBIDDEN."
+        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))
                 error = ""
@@ -225,7 +232,7 @@ class LLM(ComponentBase):
                     error = ans
                     continue
                 try:
-                    self.set_output("structured_content", json_repair.loads(clean_formated_answer(ans)))
+                    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"})
@@ -236,7 +243,7 @@ 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 self._param.output_structure and not (ex and ex["goto"]):
+        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))
             return
 

agent/component/message.py

@@ -23,7 +23,7 @@ from typing import Any
 from agent.component.base import ComponentBase, ComponentParamBase
 from jinja2 import Template as Jinja2Template
 
-from api.utils.api_utils import timeout
+from common.connection_utils import timeout
 
 
 class MessageParam(ComponentParamBase):
@@ -49,7 +49,10 @@ class MessageParam(ComponentParamBase):
 class Message(ComponentBase):
     component_name = "Message"
 
-    def get_kwargs(self, script:str, kwargs:dict = {}, delimeter:str=None) -> tuple[str, dict[str, str | list | Any]]:
+    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:
                 continue
@@ -60,8 +63,8 @@ class Message(ComponentBase):
             if isinstance(v, partial):
                 for t in v():
                     ans += t
-            elif isinstance(v, list) and delimeter:
+            elif isinstance(v, list) and delimiter:
-                ans = delimeter.join([str(vv) for vv in v])
+                ans = delimiter.join([str(vv) for vv in v])
             elif not isinstance(v, str):
                 try:
                     ans = json.dumps(v, ensure_ascii=False)
@@ -127,7 +130,7 @@ class Message(ComponentBase):
         ]
         return any([re.search(p, content) for p in patt])
 
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60)))
     def _invoke(self, **kwargs):
         rand_cnt = random.choice(self._param.content)
         if self._param.stream and not self._is_jinjia2(rand_cnt):

agent/component/string_transform.py

@@ -16,9 +16,11 @@
 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 api.utils.api_utils import timeout
+from common.connection_utils import timeout
 from .message import Message
 
 
@@ -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 {
@@ -56,7 +61,7 @@ class StringTransform(Message, ABC):
             "type": "line"
         } for k, o in self.get_input_elements_from_text(self._param.script).items()}
 
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60)))
     def _invoke(self, **kwargs):
         if self._param.method == "split":
             self._split(kwargs.get("line"))
@@ -90,7 +95,7 @@ class StringTransform(Message, ABC):
         for k,v in kwargs.items():
             if not v:
                 v = ""
-            script = re.sub(k, v, script)
+            script = re.sub(k, lambda match: v, script)
 
         self.set_output("result", script)
 

agent/component/switch.py

@@ -19,7 +19,7 @@ from abc import ABC
 from typing import Any
 
 from agent.component.base import ComponentBase, ComponentParamBase
-from api.utils.api_utils import timeout
+from common.connection_utils import timeout
 
 
 class SwitchParam(ComponentParamBase):
@@ -61,7 +61,7 @@ class SwitchParam(ComponentParamBase):
 class Switch(ComponentBase, ABC):
     component_name = "Switch"
 
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 3))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 3)))
     def _invoke(self, **kwargs):
         for cond in self._param.conditions:
             res = []

agent/templates/generate_SEO_blog.json

@@ -2,9 +2,11 @@
     "id": 8,
     "title": {
         "en": "Generate SEO Blog",
+        "de": "SEO Blog generieren",
         "zh": "生成SEO博客"},
     "description": {
         "en": "This is a multi-agent version of the SEO blog generation workflow. It simulates a small team of AI “writers”, where each agent plays a specialized role — just like a real editorial team.",
+        "de": "Dies ist eine Multi-Agenten-Version des Workflows zur Erstellung von SEO-Blogs. Sie simuliert ein kleines Team von KI-„Autoren“, in dem jeder Agent eine spezielle Rolle übernimmt – genau wie in einem echten Redaktionsteam.",
         "zh": "多智能体架构可根据简单的用户输入自动生成完整的SEO博客文章。模拟小型“作家”团队，其中每个智能体扮演一个专业角色——就像真正的编辑团队。"},
     "canvas_type": "Agent",
     "dsl": {

agent/templates/knowledge_base_report.json

@@ -2,9 +2,11 @@
     "id": 20,
     "title": {
         "en": "Report Agent Using Knowledge Base",
+        "de": "Berichtsagent mit Wissensdatenbank",
         "zh": "知识库检索智能体"},
     "description": {
         "en": "A report generation assistant using local knowledge base, with advanced capabilities in task planning, reasoning, and reflective analysis. Recommended for academic research paper Q&A",
+        "de": "Ein Berichtsgenerierungsassistent, der eine lokale Wissensdatenbank nutzt, mit erweiterten Fähigkeiten in Aufgabenplanung, Schlussfolgerung und reflektierender Analyse. Empfohlen für akademische Forschungspapier-Fragen und -Antworten.",
         "zh": "一个使用本地知识库的报告生成助手，具备高级能力，包括任务规划、推理和反思性分析。推荐用于学术研究论文问答。"},
     "canvas_type": "Agent",
     "dsl": {

agent/templates/knowledge_base_report_r.json

@@ -1,10 +1,12 @@
 {
     "id": 21,
     "title": {
-        "en": "Report Agent Using Knowledge Base", 
+        "en": "Report Agent Using Knowledge Base",
+        "de": "Berichtsagent mit Wissensdatenbank",
         "zh": "知识库检索智能体"},
     "description": {
         "en": "A report generation assistant using local knowledge base, with advanced capabilities in task planning, reasoning, and reflective analysis. Recommended for academic research paper Q&A",
+        "de": "Ein Berichtsgenerierungsassistent, der eine lokale Wissensdatenbank nutzt, mit erweiterten Fähigkeiten in Aufgabenplanung, Schlussfolgerung und reflektierender Analyse. Empfohlen für akademische Forschungspapier-Fragen und -Antworten.",
         "zh": "一个使用本地知识库的报告生成助手，具备高级能力，包括任务规划、推理和反思性分析。推荐用于学术研究论文问答。"},
     "canvas_type": "Recommended",
     "dsl": {

agent/templates/market_generate_seo_blog.json

@@ -2,9 +2,11 @@
     "id": 12,
     "title": {
         "en": "Generate SEO Blog",
+        "de": "SEO Blog generieren",
         "zh": "生成SEO博客"},
     "description": {
-        "en": "This workflow automatically generates a complete SEO-optimized blog article based on a simple user input. You don’t need any writing experience. Just provide a topic or short request — the system will handle the rest.",
+        "en": "This workflow automatically generates a complete SEO-optimized blog article based on a simple user input. You don't need any writing experience. Just provide a topic or short request — the system will handle the rest.",
+        "de": "Dieser Workflow generiert automatisch einen vollständigen SEO-optimierten Blogartikel basierend auf einer einfachen Benutzereingabe. Sie benötigen keine Schreiberfahrung. Geben Sie einfach ein Thema oder eine kurze Anfrage ein – das System übernimmt den Rest.",
         "zh": "此工作流根据简单的用户输入自动生成完整的SEO博客文章。你无需任何写作经验，只需提供一个主题或简短请求，系统将处理其余部分。"},
     "canvas_type": "Marketing",
     "dsl": {
@@ -916,4 +918,4 @@
             "retrieval": []
         },
     "avatar": "data:image/jpeg;base64,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"
-}
+}

agent/templates/seo_blog.json

@@ -2,9 +2,11 @@
     "id": 4,
     "title": {
         "en": "Generate SEO Blog",
+        "de": "SEO Blog generieren",
         "zh": "生成SEO博客"},
     "description": {
-        "en": "This workflow automatically generates a complete SEO-optimized blog article based on a simple user input. You don’t need any writing experience. Just provide a topic or short request — the system will handle the rest.",
+        "en": "This workflow automatically generates a complete SEO-optimized blog article based on a simple user input. You don't need any writing experience. Just provide a topic or short request — the system will handle the rest.",
+        "de": "Dieser Workflow generiert automatisch einen vollständigen SEO-optimierten Blogartikel basierend auf einer einfachen Benutzereingabe. Sie benötigen keine Schreiberfahrung. Geben Sie einfach ein Thema oder eine kurze Anfrage ein – das System übernimmt den Rest.",
         "zh": "此工作流根据简单的用户输入自动生成完整的SEO博客文章。你无需任何写作经验，只需提供一个主题或简短请求，系统将处理其余部分。"},
     "canvas_type": "Recommended",
     "dsl": {
@@ -916,4 +918,4 @@
             "retrieval": []
         },
     "avatar": "data:image/jpeg;base64,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"
-}
+}

agent/templates/sql_assistant.json

@@ -2,10 +2,12 @@
     "id": 17,
     "title": {
         "en": "SQL Assistant",
+        "de": "SQL Assistent",
         "zh": "SQL助理"},
     "description": {
-        "en": "SQL Assistant is an AI-powered tool that lets business users turn plain-English questions into fully formed SQL queries. Simply type your question (e.g., “Show me last quarter’s top 10 products by revenue”) and SQL Assistant generates the exact SQL, runs it against your database, and returns the results in seconds. ",
+        "en": "SQL Assistant is an AI-powered tool that lets business users turn plain-English questions into fully formed SQL queries. Simply type your question (e.g., 'Show me last quarter's top 10 products by revenue') and SQL Assistant generates the exact SQL, runs it against your database, and returns the results in seconds. ",
-        "zh": "用户能够将简单文本问题转化为完整的SQL查询并输出结果。只需输入您的问题（例如，“展示上个季度前十名按收入排序的产品”），SQL助理就会生成精确的SQL语句，对其运行您的数据库，并几秒钟内返回结果。"},
+        "de": "SQL-Assistent ist ein KI-gestütztes Tool, mit dem Geschäftsanwender einfache englische Fragen in vollständige SQL-Abfragen umwandeln können. Geben Sie einfach Ihre Frage ein (z.B. 'Zeige mir die Top 10 Produkte des letzten Quartals nach Umsatz') und der SQL-Assistent generiert das exakte SQL, führt es gegen Ihre Datenbank aus und liefert die Ergebnisse in Sekunden.",
+        "zh": "用户能够将简单文本问题转化为完整的SQL查询并输出结果。只需输入您的问题（例如，展示上个季度前十名按收入排序的产品），SQL助理就会生成精确的SQL语句，对其运行您的数据库，并几秒钟内返回结果。"},
     "canvas_type": "Marketing",
     "dsl": {
             "components": {
@@ -83,7 +85,7 @@
                             },
                             "password": "20010812Yy!",
                             "port": 3306,
-                            "sql": "Agent:WickedGoatsDivide@content",
+                            "sql": "{Agent:WickedGoatsDivide@content}",
                             "username": "13637682833@163.com"
                         }
                     },
@@ -114,9 +116,7 @@
                         "params": {
                             "cross_languages": [],
                             "empty_response": "",
-                            "kb_ids": [
+                            "kb_ids": [],
-                                "ed31364c727211f0bdb2bafe6e7908e6"
-                            ],
                             "keywords_similarity_weight": 0.7,
                             "outputs": {
                                 "formalized_content": {
@@ -124,7 +124,7 @@
                                     "value": ""
                                 }
                             },
-                            "query": "sys.query",
+                            "query": "{sys.query}",
                             "rerank_id": "",
                             "similarity_threshold": 0.2,
                             "top_k": 1024,
@@ -145,9 +145,7 @@
                         "params": {
                             "cross_languages": [],
                             "empty_response": "",
-                            "kb_ids": [
+                            "kb_ids": [],
-                                "0f968106727311f08357bafe6e7908e6"
-                            ],
                             "keywords_similarity_weight": 0.7,
                             "outputs": {
                                 "formalized_content": {
@@ -155,7 +153,7 @@
                                     "value": ""
                                 }
                             },
-                            "query": "sys.query",
+                            "query": "{sys.query}",
                             "rerank_id": "",
                             "similarity_threshold": 0.2,
                             "top_k": 1024,
@@ -176,9 +174,7 @@
                         "params": {
                             "cross_languages": [],
                             "empty_response": "",
-                            "kb_ids": [
+                            "kb_ids": [],
-                                "4ad1f9d0727311f0827dbafe6e7908e6"
-                            ],
                             "keywords_similarity_weight": 0.7,
                             "outputs": {
                                 "formalized_content": {
@@ -186,7 +182,7 @@
                                     "value": ""
                                 }
                             },
-                            "query": "sys.query",
+                            "query": "{sys.query}",
                             "rerank_id": "",
                             "similarity_threshold": 0.2,
                             "top_k": 1024,
@@ -347,9 +343,7 @@
                             "form": {
                                 "cross_languages": [],
                                 "empty_response": "",
-                                "kb_ids": [
+                                "kb_ids": [],
-                                    "ed31364c727211f0bdb2bafe6e7908e6"
-                                ],
                                 "keywords_similarity_weight": 0.7,
                                 "outputs": {
                                     "formalized_content": {
@@ -357,7 +351,7 @@
                                         "value": ""
                                     }
                                 },
-                                "query": "sys.query",
+                                "query": "{sys.query}",
                                 "rerank_id": "",
                                 "similarity_threshold": 0.2,
                                 "top_k": 1024,
@@ -387,9 +381,7 @@
                             "form": {
                                 "cross_languages": [],
                                 "empty_response": "",
-                                "kb_ids": [
+                                "kb_ids": [],
-                                    "0f968106727311f08357bafe6e7908e6"
-                                ],
                                 "keywords_similarity_weight": 0.7,
                                 "outputs": {
                                     "formalized_content": {
@@ -397,7 +389,7 @@
                                         "value": ""
                                     }
                                 },
-                                "query": "sys.query",
+                                "query": "{sys.query}",
                                 "rerank_id": "",
                                 "similarity_threshold": 0.2,
                                 "top_k": 1024,
@@ -427,9 +419,7 @@
                             "form": {
                                 "cross_languages": [],
                                 "empty_response": "",
-                                "kb_ids": [
+                                "kb_ids": [],
-                                    "4ad1f9d0727311f0827dbafe6e7908e6"
-                                ],
                                 "keywords_similarity_weight": 0.7,
                                 "outputs": {
                                     "formalized_content": {
@@ -437,7 +427,7 @@
                                         "value": ""
                                     }
                                 },
-                                "query": "sys.query",
+                                "query": "{sys.query}",
                                 "rerank_id": "",
                                 "similarity_threshold": 0.2,
                                 "top_k": 1024,
@@ -539,7 +529,7 @@
                                 },
                                 "password": "20010812Yy!",
                                 "port": 3306,
-                                "sql": "Agent:WickedGoatsDivide@content",
+                                "sql": "{Agent:WickedGoatsDivide@content}",
                                 "username": "13637682833@163.com"
                             },
                             "label": "ExeSQL",
@@ -725,4 +715,4 @@
             "retrieval": []
         },
     "avatar": "data:image/jpeg;base64,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"
-}
+}

agent/tools/arxiv.py

@@ -19,7 +19,7 @@ import time
 from abc import ABC
 import arxiv
 from agent.tools.base import ToolParamBase, ToolMeta, ToolBase
-from api.utils.api_utils import timeout
+from common.connection_utils import timeout
 
 
 class ArXivParam(ToolParamBase):
@@ -61,7 +61,7 @@ class ArXivParam(ToolParamBase):
 class ArXiv(ToolBase, ABC):
     component_name = "ArXiv"
 
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12)))
     def _invoke(self, **kwargs):
         if not kwargs.get("query"):
             self.set_output("formalized_content", "")
@@ -97,6 +97,6 @@ class ArXiv(ToolBase, ABC):
 
     def thoughts(self) -> str:
         return """
-Keywords: {} 
+Keywords: {}
 Looking for the most relevant articles.
-                """.format(self.get_input().get("query", "-_-!"))
+                """.format(self.get_input().get("query", "-_-!"))

agent/tools/base.py

@@ -20,9 +20,9 @@ from copy import deepcopy
 from functools import partial
 from typing import TypedDict, List, Any
 from agent.component.base import ComponentParamBase, ComponentBase
-from api.utils import hash_str2int
+from common.misc_utils import hash_str2int
 from rag.llm.chat_model import ToolCallSession
-from rag.prompts.prompts import kb_prompt
+from rag.prompts.generator import kb_prompt
 from rag.utils.mcp_tool_call_conn import MCPToolCallSession
 from timeit import default_timer as timer
 

agent/tools/code_exec.py

@@ -22,7 +22,7 @@ from typing import Optional
 from pydantic import BaseModel, Field, field_validator
 from agent.tools.base import ToolParamBase, ToolBase, ToolMeta
 from api import settings
-from api.utils.api_utils import timeout
+from common.connection_utils import timeout
 
 
 class Language(StrEnum):
@@ -129,7 +129,7 @@ module.exports = { main };
 class CodeExec(ToolBase, ABC):
     component_name = "CodeExec"
 
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60)))
     def _invoke(self, **kwargs):
         lang = kwargs.get("lang", self._param.lang)
         script = kwargs.get("script", self._param.script)
@@ -156,8 +156,8 @@ class CodeExec(ToolBase, ABC):
             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=os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60))
+            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, resp.status_code)
+            logging.info(f"http://{settings.SANDBOX_HOST}:9385/run,  code_req: {code_req}, resp.status_code {resp.status_code}:")
             if resp.status_code != 200:
                 resp.raise_for_status()
             body = resp.json()

agent/tools/duckduckgo.py

@@ -19,7 +19,7 @@ import time
 from abc import ABC
 from duckduckgo_search import DDGS
 from agent.tools.base import ToolMeta, ToolParamBase, ToolBase
-from api.utils.api_utils import timeout
+from common.connection_utils import timeout
 
 
 class DuckDuckGoParam(ToolParamBase):
@@ -73,7 +73,7 @@ class DuckDuckGoParam(ToolParamBase):
 class DuckDuckGo(ToolBase, ABC):
     component_name = "DuckDuckGo"
 
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12)))
     def _invoke(self, **kwargs):
         if not kwargs.get("query"):
             self.set_output("formalized_content", "")
@@ -115,6 +115,6 @@ class DuckDuckGo(ToolBase, ABC):
 
     def thoughts(self) -> str:
         return """
-Keywords: {} 
+Keywords: {}
 Looking for the most relevant articles.
-                """.format(self.get_input().get("query", "-_-!"))
+                """.format(self.get_input().get("query", "-_-!"))

agent/tools/email.py

@@ -25,7 +25,7 @@ from email.header import Header
 from email.utils import formataddr
 
 from agent.tools.base import ToolParamBase, ToolBase, ToolMeta
-from api.utils.api_utils import timeout
+from common.connection_utils import timeout
 
 
 class EmailParam(ToolParamBase):
@@ -98,8 +98,8 @@ class EmailParam(ToolParamBase):
 
 class Email(ToolBase, ABC):
     component_name = "Email"
-    
+
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 60))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 60)))
     def _invoke(self, **kwargs):
         if not kwargs.get("to_email"):
             self.set_output("success", False)
@@ -212,4 +212,4 @@ class Email(ToolBase, ABC):
 To: {}
 Subject: {}
 Your email is on its way—sit tight!
-""".format(inputs.get("to_email", "-_-!"), inputs.get("subject", "-_-!"))
+""".format(inputs.get("to_email", "-_-!"), inputs.get("subject", "-_-!"))

agent/tools/exesql.py

@@ -22,7 +22,7 @@ import pymysql
 import psycopg2
 import pyodbc
 from agent.tools.base import ToolParamBase, ToolBase, ToolMeta
-from api.utils.api_utils import timeout
+from common.connection_utils import timeout
 
 
 class ExeSQLParam(ToolParamBase):
@@ -53,12 +53,13 @@ class ExeSQLParam(ToolParamBase):
         self.max_records = 1024
 
     def check(self):
-        self.check_valid_value(self.db_type, "Choose DB type", ['mysql', 'postgresql', 'mariadb', 'mssql'])
+        self.check_valid_value(self.db_type, "Choose DB type", ['mysql', 'postgres', 'mariadb', 'mssql', 'IBM DB2', 'trino'])
         self.check_empty(self.database, "Database name")
         self.check_empty(self.username, "database username")
         self.check_empty(self.host, "IP Address")
         self.check_positive_integer(self.port, "IP Port")
-        self.check_empty(self.password, "Database password")
+        if self.db_type != "trino":
+            self.check_empty(self.password, "Database password")
         self.check_positive_integer(self.max_records, "Maximum number of records")
         if self.database == "rag_flow":
             if self.host == "ragflow-mysql":
@@ -78,7 +79,7 @@ class ExeSQLParam(ToolParamBase):
 class ExeSQL(ToolBase, ABC):
     component_name = "ExeSQL"
 
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 60))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 60)))
     def _invoke(self, **kwargs):
 
         def convert_decimals(obj):
@@ -111,7 +112,7 @@ class ExeSQL(ToolBase, ABC):
         if self._param.db_type in ["mysql", "mariadb"]:
             db = pymysql.connect(db=self._param.database, user=self._param.username, host=self._param.host,
                                  port=self._param.port, password=self._param.password)
-        elif self._param.db_type == 'postgresql':
+        elif self._param.db_type == 'postgres':
             db = psycopg2.connect(dbname=self._param.database, user=self._param.username, host=self._param.host,
                                   port=self._param.port, password=self._param.password)
         elif self._param.db_type == 'mssql':
@@ -123,6 +124,94 @@ class ExeSQL(ToolBase, ABC):
                     r'PWD=' + self._param.password
             )
             db = pyodbc.connect(conn_str)
+        elif self._param.db_type == 'trino':
+            try:
+                import trino
+                from trino.auth import BasicAuthentication
+            except Exception:
+                raise Exception("Missing dependency 'trino'. Please install: pip install trino")
+
+            def _parse_catalog_schema(db: str):
+                if not db:
+                    return None, None
+                if "." in db:
+                    c, s = db.split(".", 1)
+                elif "/" in db:
+                    c, s = db.split("/", 1)
+                else:
+                    c, s = db, "default"
+                return c, s
+
+            catalog, schema = _parse_catalog_schema(self._param.database)
+            if not catalog:
+                raise Exception("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 self._param.password:
+                auth = BasicAuthentication(self._param.username, self._param.password)
+
+            try:
+                db = trino.dbapi.connect(
+                    host=self._param.host,
+                    port=int(self._param.port or 8080),
+                    user=self._param.username or "ragflow",
+                    catalog=catalog,
+                    schema=schema or "default",
+                    http_scheme=http_scheme,
+                    auth=auth
+                )
+            except Exception as e:
+                raise Exception("Database Connection Failed! \n" + str(e))
+        elif self._param.db_type == 'IBM DB2':
+            import ibm_db
+            conn_str = (
+                f"DATABASE={self._param.database};"
+                f"HOSTNAME={self._param.host};"
+                f"PORT={self._param.port};"
+                f"PROTOCOL=TCPIP;"
+                f"UID={self._param.username};"
+                f"PWD={self._param.password};"
+            )
+            try:
+                conn = ibm_db.connect(conn_str, "", "")
+            except Exception as e:
+                raise Exception("Database Connection Failed! \n" + str(e))
+
+            sql_res = []
+            formalized_content = []
+            for single_sql in sqls:
+                single_sql = single_sql.replace("```", "").strip()
+                if not single_sql:
+                    continue
+                single_sql = re.sub(r"\[ID:[0-9]+\]", "", single_sql)
+
+                stmt = ibm_db.exec_immediate(conn, single_sql)
+                rows = []
+                row = ibm_db.fetch_assoc(stmt)
+                while row and len(rows) < self._param.max_records:
+                    rows.append(row)
+                    row = ibm_db.fetch_assoc(stmt)
+
+                if not rows:
+                    sql_res.append({"content": "No record in the database!"})
+                    continue
+
+                df = pd.DataFrame(rows)
+                for col in df.columns:
+                    if pd.api.types.is_datetime64_any_dtype(df[col]):
+                        df[col] = df[col].dt.strftime("%Y-%m-%d")
+
+                df = df.where(pd.notnull(df), None)
+
+                sql_res.append(convert_decimals(df.to_dict(orient="records")))
+                formalized_content.append(df.to_markdown(index=False, floatfmt=".6f"))
+
+            ibm_db.close(conn)
+
+            self.set_output("json", sql_res)
+            self.set_output("formalized_content", "\n\n".join(formalized_content))
+            return self.output("formalized_content")
         try:
             cursor = db.cursor()
         except Exception as e:
@@ -150,6 +239,8 @@ class ExeSQL(ToolBase, ABC):
                 if pd.api.types.is_datetime64_any_dtype(single_res[col]):
                     single_res[col] = single_res[col].dt.strftime('%Y-%m-%d')
 
+            single_res = single_res.where(pd.notnull(single_res), None)
+
             sql_res.append(convert_decimals(single_res.to_dict(orient='records')))
             formalized_content.append(single_res.to_markdown(index=False, floatfmt=".6f"))
 

agent/tools/github.py

@@ -19,7 +19,7 @@ import time
 from abc import ABC
 import requests
 from agent.tools.base import ToolParamBase, ToolMeta, ToolBase
-from api.utils.api_utils import timeout
+from common.connection_utils import timeout
 
 
 class GitHubParam(ToolParamBase):
@@ -57,7 +57,7 @@ class GitHubParam(ToolParamBase):
 class GitHub(ToolBase, ABC):
     component_name = "GitHub"
 
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12)))
     def _invoke(self, **kwargs):
         if not kwargs.get("query"):
             self.set_output("formalized_content", "")
@@ -88,4 +88,4 @@ class GitHub(ToolBase, ABC):
         assert False, self.output()
 
     def thoughts(self) -> str:
-        return "Scanning GitHub repos related to `{}`.".format(self.get_input().get("query", "-_-!"))
+        return "Scanning GitHub repos related to `{}`.".format(self.get_input().get("query", "-_-!"))

agent/tools/google.py

@@ -19,7 +19,7 @@ import time
 from abc import ABC
 from serpapi import GoogleSearch
 from agent.tools.base import ToolParamBase, ToolMeta, ToolBase
-from api.utils.api_utils import timeout
+from common.connection_utils import timeout
 
 
 class GoogleParam(ToolParamBase):
@@ -116,7 +116,7 @@ class GoogleParam(ToolParamBase):
 class Google(ToolBase, ABC):
     component_name = "Google"
 
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12)))
     def _invoke(self, **kwargs):
         if not kwargs.get("q"):
             self.set_output("formalized_content", "")
@@ -154,6 +154,6 @@ class Google(ToolBase, ABC):
 
     def thoughts(self) -> str:
         return """
-Keywords: {} 
+Keywords: {}
 Looking for the most relevant articles.
-        """.format(self.get_input().get("query", "-_-!"))
+        """.format(self.get_input().get("query", "-_-!"))

agent/tools/googlescholar.py

@@ -19,7 +19,7 @@ import time
 from abc import ABC
 from scholarly import scholarly
 from agent.tools.base import ToolMeta, ToolParamBase, ToolBase
-from api.utils.api_utils import timeout
+from common.connection_utils import timeout
 
 
 class GoogleScholarParam(ToolParamBase):
@@ -63,7 +63,7 @@ class GoogleScholarParam(ToolParamBase):
 class GoogleScholar(ToolBase, ABC):
     component_name = "GoogleScholar"
 
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12)))
     def _invoke(self, **kwargs):
         if not kwargs.get("query"):
             self.set_output("formalized_content", "")
@@ -93,4 +93,4 @@ class GoogleScholar(ToolBase, ABC):
         assert False, self.output()
 
     def thoughts(self) -> str:
-        return "Looking for scholarly papers on `{}`,” prioritising reputable sources.".format(self.get_input().get("query", "-_-!"))
+        return "Looking for scholarly papers on `{}`,” prioritising reputable sources.".format(self.get_input().get("query", "-_-!"))

agent/tools/pubmed.py

@@ -21,7 +21,7 @@ from Bio import Entrez
 import re
 import xml.etree.ElementTree as ET
 from agent.tools.base import ToolParamBase, ToolMeta, ToolBase
-from api.utils.api_utils import timeout
+from common.connection_utils import timeout
 
 
 class PubMedParam(ToolParamBase):
@@ -33,7 +33,7 @@ class PubMedParam(ToolParamBase):
         self.meta:ToolMeta = {
             "name": "pubmed_search",
             "description": """
-PubMed is an openly accessible, free database which includes primarily the MEDLINE database of references and abstracts on life sciences and biomedical topics. 
+PubMed is an openly accessible, free database which includes primarily the MEDLINE database of references and abstracts on life sciences and biomedical topics.
 In addition to MEDLINE, PubMed provides access to:
  - older references from the print version of Index Medicus, back to 1951 and earlier
  - references to some journals before they were indexed in Index Medicus and MEDLINE, for instance Science, BMJ, and Annals of Surgery
@@ -69,7 +69,7 @@ In addition to MEDLINE, PubMed provides access to:
 class PubMed(ToolBase, ABC):
     component_name = "PubMed"
 
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12)))
     def _invoke(self, **kwargs):
         if not kwargs.get("query"):
             self.set_output("formalized_content", "")
@@ -85,13 +85,7 @@ class PubMed(ToolBase, ABC):
                 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: child.find("MedlineCitation") \
+                                      get_content=lambda child: self._format_pubmed_content(child),)
-                                                                    .find("Article") \
-                                                                    .find("Abstract") \
-                                                                    .find("AbstractText").text \
-                                                                    if child.find("MedlineCitation")\
-                                                                            .find("Article").find("Abstract")  \
-                                                                    else "No abstract available")
                 return self.output("formalized_content")
             except Exception as e:
                 last_e = e
@@ -104,5 +98,50 @@ class PubMed(ToolBase, ABC):
 
         assert False, self.output()
 
+    def _format_pubmed_content(self, child):
+        """Extract structured reference info from PubMed XML"""
+        def safe_find(path):
+            node = child
+            for p in path.split("/"):
+                if node is None:
+                    return None
+                node = node.find(p)
+            return node.text if node is not None and node.text else None
+
+        title = safe_find("MedlineCitation/Article/ArticleTitle") or "No title"
+        abstract = safe_find("MedlineCitation/Article/Abstract/AbstractText") or "No abstract available"
+        journal = safe_find("MedlineCitation/Article/Journal/Title") or "Unknown Journal"
+        volume = safe_find("MedlineCitation/Article/Journal/JournalIssue/Volume") or "-"
+        issue = safe_find("MedlineCitation/Article/Journal/JournalIssue/Issue") or "-"
+        pages = safe_find("MedlineCitation/Article/Pagination/MedlinePgn") or "-"
+
+        # Authors
+        authors = []
+        for author in child.findall(".//AuthorList/Author"):
+            lastname = safe_find("LastName") or ""
+            forename = safe_find("ForeName") or ""
+            fullname = f"{forename} {lastname}".strip()
+            if fullname:
+                authors.append(fullname)
+        authors_str = ", ".join(authors) if authors else "Unknown Authors"
+
+        # DOI
+        doi = None
+        for eid in child.findall(".//ArticleId"):
+            if eid.attrib.get("IdType") == "doi":
+                doi = eid.text
+                break
+
+        return (
+            f"Title: {title}\n"
+            f"Authors: {authors_str}\n"
+            f"Journal: {journal}\n"
+            f"Volume: {volume}\n"
+            f"Issue: {issue}\n"
+            f"Pages: {pages}\n"
+            f"DOI: {doi or '-'}\n"
+            f"Abstract: {abstract.strip()}"
+        )
+
     def thoughts(self) -> str:
-        return "Looking for scholarly papers on `{}`,” prioritising reputable sources.".format(self.get_input().get("query", "-_-!"))
+        return "Looking for scholarly papers on `{}`,” prioritising reputable sources.".format(self.get_input().get("query", "-_-!"))

agent/tools/retrieval.py

@@ -13,18 +13,22 @@
 #  See the License for the specific language governing permissions and
 #  limitations under the License.
 #
+from functools import partial
+import json
 import os
 import re
 from abc import ABC
 from agent.tools.base import ToolParamBase, ToolBase, ToolMeta
-from api.db import LLMType
+from common.constants import LLMType
+from api.db.services.document_service import DocumentService
+from api.db.services.dialog_service import meta_filter
 from api.db.services.knowledgebase_service import KnowledgebaseService
 from api.db.services.llm_service import LLMBundle
 from api import settings
-from api.utils.api_utils import timeout
+from common import globals
+from common.connection_utils import timeout
 from rag.app.tag import label_question
-from rag.prompts import kb_prompt
+from rag.prompts.generator import cross_languages, kb_prompt, gen_meta_filter
-from rag.prompts.prompts import cross_languages
 
 
 class RetrievalParam(ToolParamBase):
@@ -58,6 +62,8 @@ class RetrievalParam(ToolParamBase):
         self.empty_response = ""
         self.use_kg = False
         self.cross_languages = []
+        self.toc_enhance = False
+        self.meta_data_filter={}
 
     def check(self):
         self.check_decimal_float(self.similarity_threshold, "[Retrieval] Similarity threshold")
@@ -75,7 +81,7 @@ class RetrievalParam(ToolParamBase):
 class Retrieval(ToolBase, ABC):
     component_name = "Retrieval"
 
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12))
+    @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)
@@ -117,12 +123,55 @@ class Retrieval(ToolBase, ABC):
         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!={}:
+            metas = DocumentService.get_meta_by_kbs(kb_ids)
+            if self._param.meta_data_filter.get("method") == "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
+
         if self._param.cross_languages:
             query = cross_languages(kbs[0].tenant_id, None, query, self._param.cross_languages)
 
         if kbs:
             query = re.sub(r"^user[:：\s]*", "", query, flags=re.IGNORECASE)
-            kbinfos = settings.retrievaler.retrieval(
+            kbinfos = globals.retriever.retrieval(
                 query,
                 embd_mdl,
                 [kb.tenant_id for kb in kbs],
@@ -131,12 +180,18 @@ class Retrieval(ToolBase, ABC):
                 self._param.top_n,
                 self._param.similarity_threshold,
                 1 - self._param.keywords_similarity_weight,
+                doc_ids=doc_ids,
                 aggs=False,
                 rerank_mdl=rerank_mdl,
                 rank_feature=label_question(query, kbs),
             )
+            if self._param.toc_enhance:
+                chat_mdl = LLMBundle(self._canvas._tenant_id, LLMType.CHAT)
+                cks = globals.retriever.retrieval_by_toc(query, kbinfos["chunks"], [kb.tenant_id for kb in kbs], chat_mdl, self._param.top_n)
+                if cks:
+                    kbinfos["chunks"] = cks
             if self._param.use_kg:
-                ck = settings.kg_retrievaler.retrieval(query,
+                ck = settings.kg_retriever.retrieval(query,
                                                        [kb.tenant_id for kb in kbs],
                                                        kb_ids,
                                                        embd_mdl,
@@ -147,7 +202,7 @@ class Retrieval(ToolBase, ABC):
             kbinfos = {"chunks": [], "doc_aggs": []}
 
         if self._param.use_kg and kbs:
-            ck = settings.kg_retrievaler.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 ck["content_with_weight"]:
                 ck["content"] = ck["content_with_weight"]
                 del ck["content_with_weight"]
@@ -163,13 +218,20 @@ class Retrieval(ToolBase, ABC):
             self.set_output("formalized_content", self._param.empty_response)
             return
 
+        # Format the chunks for JSON output (similar to how other tools do it)
+        json_output = kbinfos["chunks"].copy()
+
         self._canvas.add_reference(kbinfos["chunks"], kbinfos["doc_aggs"])
         form_cnt = "\n".join(kb_prompt(kbinfos, 200000, True))
+
+        # Set both formalized content and JSON output
         self.set_output("formalized_content", form_cnt)
+        self.set_output("json", json_output)
+
         return form_cnt
 
     def thoughts(self) -> str:
         return """
-Keywords: {} 
+Keywords: {}
 Looking for the most relevant articles.
-        """.format(self.get_input().get("query", "-_-!"))
+        """.format(self.get_input().get("query", "-_-!"))

agent/tools/searxng.py

@@ -19,7 +19,7 @@ import time
 from abc import ABC
 import requests
 from agent.tools.base import ToolMeta, ToolParamBase, ToolBase
-from api.utils.api_utils import timeout
+from common.connection_utils import timeout
 
 
 class SearXNGParam(ToolParamBase):
@@ -77,7 +77,7 @@ class SearXNGParam(ToolParamBase):
 class SearXNG(ToolBase, ABC):
     component_name = "SearXNG"
 
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12)))
     def _invoke(self, **kwargs):
         # Gracefully handle try-run without inputs
         query = kwargs.get("query")
@@ -85,7 +85,7 @@ class SearXNG(ToolBase, ABC):
             self.set_output("formalized_content", "")
             return ""
 
-        searxng_url = (kwargs.get("searxng_url") or getattr(self._param, "searxng_url", "") or "").strip()
+        searxng_url = (getattr(self._param, "searxng_url", "") or kwargs.get("searxng_url") or "").strip()
         # In try-run, if no URL configured, just return empty instead of raising
         if not searxng_url:
             self.set_output("formalized_content", "")
@@ -94,7 +94,6 @@ class SearXNG(ToolBase, ABC):
         last_e = ""
         for _ in range(self._param.max_retries+1):
             try:
-                # 构建搜索参数
                 search_params = {
                     'q': query,
                     'format': 'json',
@@ -104,33 +103,29 @@ class SearXNG(ToolBase, ABC):
                     'pageno': 1
                 }
 
-                # 发送搜索请求
                 response = requests.get(
                     f"{searxng_url}/search",
                     params=search_params,
                     timeout=10
                 )
                 response.raise_for_status()
-                
+
                 data = response.json()
-                
+
-                # 验证响应数据
                 if not data or not isinstance(data, dict):
                     raise ValueError("Invalid response from SearXNG")
-                
+
                 results = data.get("results", [])
                 if not isinstance(results, list):
                     raise ValueError("Invalid results format from SearXNG")
-                
+
-                # 限制结果数量
                 results = results[:self._param.top_n]
-                
+
-                # 处理搜索结果
                 self._retrieve_chunks(results,
                                       get_title=lambda r: r.get("title", ""),
                                       get_url=lambda r: r.get("url", ""),
                                       get_content=lambda r: r.get("content", ""))
-                
+
                 self.set_output("json", results)
                 return self.output("formalized_content")
 
@@ -151,6 +146,6 @@ class SearXNG(ToolBase, ABC):
 
     def thoughts(self) -> str:
         return """
-Keywords: {} 
+Keywords: {}
 Searching with SearXNG for relevant results...
                 """.format(self.get_input().get("query", "-_-!"))

agent/tools/tavily.py

@@ -19,7 +19,7 @@ import time
 from abc import ABC
 from tavily import TavilyClient
 from agent.tools.base import ToolParamBase, ToolBase, ToolMeta
-from api.utils.api_utils import timeout
+from common.connection_utils import timeout
 
 
 class TavilySearchParam(ToolParamBase):
@@ -31,7 +31,7 @@ class TavilySearchParam(ToolParamBase):
         self.meta:ToolMeta = {
             "name": "tavily_search",
             "description": """
-Tavily is a search engine optimized for LLMs, aimed at efficient, quick and persistent search results. 
+Tavily is a search engine optimized for LLMs, aimed at efficient, quick and persistent search results.
 When searching:
    - Start with specific query which should focus on just a single aspect.
    - Number of keywords in query should be less than 5.
@@ -101,7 +101,7 @@ When searching:
 class TavilySearch(ToolBase, ABC):
     component_name = "TavilySearch"
 
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12)))
     def _invoke(self, **kwargs):
         if not kwargs.get("query"):
             self.set_output("formalized_content", "")
@@ -136,7 +136,7 @@ class TavilySearch(ToolBase, ABC):
 
     def thoughts(self) -> str:
         return """
-Keywords: {} 
+Keywords: {}
 Looking for the most relevant articles.
                 """.format(self.get_input().get("query", "-_-!"))
 
@@ -199,7 +199,7 @@ class TavilyExtractParam(ToolParamBase):
 class TavilyExtract(ToolBase, ABC):
     component_name = "TavilyExtract"
 
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60)))
     def _invoke(self, **kwargs):
         self.tavily_client = TavilyClient(api_key=self._param.api_key)
         last_e = None
@@ -224,4 +224,4 @@ class TavilyExtract(ToolBase, ABC):
         assert False, self.output()
 
     def thoughts(self) -> str:
-        return "Opened {}—pulling out the main text…".format(self.get_input().get("urls", "-_-!"))
+        return "Opened {}—pulling out the main text…".format(self.get_input().get("urls", "-_-!"))

agent/tools/wencai.py

@@ -21,7 +21,7 @@ import pandas as pd
 import pywencai
 
 from agent.tools.base import ToolParamBase, ToolMeta, ToolBase
-from api.utils.api_utils import timeout
+from common.connection_utils import timeout
 
 
 class WenCaiParam(ToolParamBase):
@@ -68,7 +68,7 @@ fund selection platform: through AI technology, is committed to providing excell
 class WenCai(ToolBase, ABC):
     component_name = "WenCai"
 
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12)))
     def _invoke(self, **kwargs):
         if not kwargs.get("query"):
             self.set_output("report", "")
@@ -111,4 +111,4 @@ class WenCai(ToolBase, ABC):
         assert False, self.output()
 
     def thoughts(self) -> str:
-        return "Pulling live financial data for `{}`.".format(self.get_input().get("query", "-_-!"))
+        return "Pulling live financial data for `{}`.".format(self.get_input().get("query", "-_-!"))

agent/tools/wikipedia.py

@@ -19,7 +19,7 @@ import time
 from abc import ABC
 import wikipedia
 from agent.tools.base import ToolMeta, ToolParamBase, ToolBase
-from api.utils.api_utils import timeout
+from common.connection_utils import timeout
 
 
 class WikipediaParam(ToolParamBase):
@@ -64,7 +64,7 @@ class WikipediaParam(ToolParamBase):
 class Wikipedia(ToolBase, ABC):
     component_name = "Wikipedia"
 
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 60))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 60)))
     def _invoke(self, **kwargs):
         if not kwargs.get("query"):
             self.set_output("formalized_content", "")
@@ -99,6 +99,6 @@ class Wikipedia(ToolBase, ABC):
 
     def thoughts(self) -> str:
         return """
-Keywords: {} 
+Keywords: {}
 Looking for the most relevant articles.
-        """.format(self.get_input().get("query", "-_-!"))
+        """.format(self.get_input().get("query", "-_-!"))

agent/tools/yahoofinance.py

@@ -20,7 +20,7 @@ from abc import ABC
 import pandas as pd
 import yfinance as yf
 from agent.tools.base import ToolMeta, ToolParamBase, ToolBase
-from api.utils.api_utils import timeout
+from common.connection_utils import timeout
 
 
 class YahooFinanceParam(ToolParamBase):
@@ -72,7 +72,7 @@ class YahooFinanceParam(ToolParamBase):
 class YahooFinance(ToolBase, ABC):
     component_name = "YahooFinance"
 
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 60))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 60)))
     def _invoke(self, **kwargs):
         if not kwargs.get("stock_code"):
             self.set_output("report", "")
@@ -111,4 +111,4 @@ class YahooFinance(ToolBase, ABC):
         assert False, self.output()
 
     def thoughts(self) -> str:
-        return "Pulling live financial data for `{}`.".format(self.get_input().get("stock_code", "-_-!"))
+        return "Pulling live financial data for `{}`.".format(self.get_input().get("stock_code", "-_-!"))

api/apps/__init__.py

@@ -24,10 +24,11 @@ from flask_cors import CORS
 from flasgger import Swagger
 from itsdangerous.url_safe import URLSafeTimedSerializer as Serializer
 
-from api.db import StatusEnum
+from common.constants import StatusEnum
 from api.db.db_models import close_connection
 from api.db.services import UserService
-from api.utils import CustomJSONEncoder, commands
+from api.utils.json_encode import CustomJSONEncoder
+from api.utils import commands
 
 from flask_mail import Mail
 from flask_session import Session

api/apps/api_app.py

@@ -21,7 +21,7 @@ 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, VALID_TASK_STATUS, FileType, LLMType, ParserType, FileSource
+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 api.db.services.api_service import APITokenService, API4ConversationService
@@ -32,20 +32,22 @@ 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 api.utils import get_uuid, current_timestamp, datetime_format
+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 import keyword_extraction
+from rag.prompts.generator import keyword_extraction
 from rag.utils.storage_factory import STORAGE_IMPL
+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 common import globals
 
 
 @manager.route('/new_token', methods=['POST'])  # noqa: F821
@@ -58,7 +60,7 @@ def new_token():
             return get_data_error_result(message="Tenant not found!")
 
         tenant_id = tenants[0].tenant_id
-        obj = {"tenant_id": tenant_id, "token": generate_confirmation_token(tenant_id),
+        obj = {"tenant_id": tenant_id, "token": generate_confirmation_token(),
                "create_time": current_timestamp(),
                "create_date": datetime_format(datetime.now()),
                "update_time": None,
@@ -144,7 +146,7 @@ def set_conversation():
     objs = APIToken.query(token=token)
     if not objs:
         return get_json_result(
-            data=False, message='Authentication error: API key is invalid!"', code=settings.RetCode.AUTHENTICATION_ERROR)
+            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)
@@ -185,7 +187,7 @@ def completion():
     objs = APIToken.query(token=token)
     if not objs:
         return get_json_result(
-            data=False, message='Authentication error: API key is invalid!"', code=settings.RetCode.AUTHENTICATION_ERROR)
+            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:
@@ -351,7 +353,7 @@ def get_conversation(conversation_id):
     objs = APIToken.query(token=token)
     if not objs:
         return get_json_result(
-            data=False, message='Authentication error: API key is invalid!"', code=settings.RetCode.AUTHENTICATION_ERROR)
+            data=False, message='Authentication error: API key is invalid!"', code=RetCode.AUTHENTICATION_ERROR)
 
     try:
         e, conv = API4ConversationService.get_by_id(conversation_id)
@@ -361,7 +363,7 @@ def get_conversation(conversation_id):
         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=settings.RetCode.AUTHENTICATION_ERROR)
+                                   code=RetCode.AUTHENTICATION_ERROR)
 
         for referenct_i in conv['reference']:
             if referenct_i is None or len(referenct_i) == 0:
@@ -382,7 +384,7 @@ def upload():
     objs = APIToken.query(token=token)
     if not objs:
         return get_json_result(
-            data=False, message='Authentication error: API key is invalid!"', code=settings.RetCode.AUTHENTICATION_ERROR)
+            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
@@ -398,12 +400,12 @@ def upload():
 
     if 'file' not in request.files:
         return get_json_result(
-            data=False, message='No file part!', code=settings.RetCode.ARGUMENT_ERROR)
+            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=settings.RetCode.ARGUMENT_ERROR)
+            data=False, message='No file selected!', code=RetCode.ARGUMENT_ERROR)
 
     root_folder = FileService.get_root_folder(tenant_id)
     pf_id = root_folder["id"]
@@ -495,17 +497,17 @@ def upload_parse():
     objs = APIToken.query(token=token)
     if not objs:
         return get_json_result(
-            data=False, message='Authentication error: API key is invalid!"', code=settings.RetCode.AUTHENTICATION_ERROR)
+            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=settings.RetCode.ARGUMENT_ERROR)
+            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=settings.RetCode.ARGUMENT_ERROR)
+                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)
@@ -518,7 +520,7 @@ def list_chunks():
     objs = APIToken.query(token=token)
     if not objs:
         return get_json_result(
-            data=False, message='Authentication error: API key is invalid!"', code=settings.RetCode.AUTHENTICATION_ERROR)
+            data=False, message='Authentication error: API key is invalid!"', code=RetCode.AUTHENTICATION_ERROR)
 
     req = request.json
 
@@ -536,7 +538,7 @@ def list_chunks():
             )
         kb_ids = KnowledgebaseService.get_kb_ids(tenant_id)
 
-        res = settings.retrievaler.chunk_list(doc_id, tenant_id, kb_ids)
+        res = globals.retriever.chunk_list(doc_id, tenant_id, kb_ids)
         res = [
             {
                 "content": res_item["content_with_weight"],
@@ -558,11 +560,11 @@ def get_chunk(chunk_id):
     objs = APIToken.query(token=token)
     if not objs:
         return get_json_result(
-            data=False, message='Authentication error: API key is invalid!"', code=settings.RetCode.AUTHENTICATION_ERROR)
+            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)
+        chunk = globals.docStoreConn.get(chunk_id, search.index_name(tenant_id), kb_ids)
         if chunk is None:
             return server_error_response(Exception("Chunk not found"))
         k = []
@@ -583,7 +585,7 @@ def list_kb_docs():
     objs = APIToken.query(token=token)
     if not objs:
         return get_json_result(
-            data=False, message='Authentication error: API key is invalid!"', code=settings.RetCode.AUTHENTICATION_ERROR)
+            data=False, message='Authentication error: API key is invalid!"', code=RetCode.AUTHENTICATION_ERROR)
 
     req = request.json
     tenant_id = objs[0].tenant_id
@@ -636,7 +638,7 @@ def docinfos():
     objs = APIToken.query(token=token)
     if not objs:
         return get_json_result(
-            data=False, message='Authentication error: API key is invalid!"', code=settings.RetCode.AUTHENTICATION_ERROR)
+            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)
@@ -650,7 +652,7 @@ def document_rm():
     objs = APIToken.query(token=token)
     if not objs:
         return get_json_result(
-            data=False, message='Authentication error: API key is invalid!"', code=settings.RetCode.AUTHENTICATION_ERROR)
+            data=False, message='Authentication error: API key is invalid!"', code=RetCode.AUTHENTICATION_ERROR)
 
     tenant_id = objs[0].tenant_id
     req = request.json
@@ -702,7 +704,7 @@ def document_rm():
             errors += str(e)
 
     if errors:
-        return get_json_result(data=False, message=errors, code=settings.RetCode.SERVER_ERROR)
+        return get_json_result(data=False, message=errors, code=RetCode.SERVER_ERROR)
 
     return get_json_result(data=True)
 
@@ -717,7 +719,7 @@ def completion_faq():
     objs = APIToken.query(token=token)
     if not objs:
         return get_json_result(
-            data=False, message='Authentication error: API key is invalid!"', code=settings.RetCode.AUTHENTICATION_ERROR)
+            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:
@@ -856,7 +858,7 @@ def retrieval():
     objs = APIToken.query(token=token)
     if not objs:
         return get_json_result(
-            data=False, message='Authentication error: API key is invalid!"', code=settings.RetCode.AUTHENTICATION_ERROR)
+            data=False, message='Authentication error: API key is invalid!"', code=RetCode.AUTHENTICATION_ERROR)
 
     req = request.json
     kb_ids = req.get("kb_id", [])
@@ -867,7 +869,7 @@ def retrieval():
     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)) 
+    highlight = bool(req.get("highlight", False))
 
     try:
         kbs = KnowledgebaseService.get_by_ids(kb_ids)
@@ -875,7 +877,7 @@ def retrieval():
         if len(embd_nms) != 1:
             return get_json_result(
                 data=False, message='Knowledge bases use different embedding models or does not exist."',
-                code=settings.RetCode.AUTHENTICATION_ERROR)
+                code=RetCode.AUTHENTICATION_ERROR)
 
         embd_mdl = LLMBundle(kbs[0].tenant_id, LLMType.EMBEDDING, llm_name=kbs[0].embd_id)
         rerank_mdl = None
@@ -884,7 +886,7 @@ def retrieval():
         if req.get("keyword", False):
             chat_mdl = LLMBundle(kbs[0].tenant_id, LLMType.CHAT)
             question += keyword_extraction(chat_mdl, question)
-        ranks = settings.retrievaler.retrieval(question, embd_mdl, kbs[0].tenant_id, kb_ids, page, size,
+        ranks = globals.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))
@@ -894,5 +896,5 @@ def 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!',
-                                   code=settings.RetCode.DATA_ERROR)
+                                   code=RetCode.DATA_ERROR)
         return server_error_response(e)

api/apps/canvas_app.py

@@ -19,6 +19,7 @@ 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
@@ -28,32 +29,29 @@ from api.db import CanvasCategory, FileType
 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
+from api.db.services.pipeline_operation_log_service import PipelineOperationLogService
+from api.db.services.task_service import queue_dataflow, CANVAS_DEBUG_DOC_ID, TaskService
 from api.db.services.user_service import TenantService
 from api.db.services.user_canvas_version import UserCanvasVersionService
-from api.settings import RetCode
+from common.constants import RetCode
-from api.utils import get_uuid
+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 agent.canvas import Canvas
 from peewee import MySQLDatabase, PostgresqlDatabase
-from api.db.db_models import APIToken
+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 globals
 
 
 @manager.route('/templates', methods=['GET'])  # noqa: F821
 @login_required
 def templates():
-    return get_json_result(data=[c.to_dict() for c in CanvasTemplateService.query(canvas_category=CanvasCategory.Agent)])
+    return get_json_result(data=[c.to_dict() for c in CanvasTemplateService.get_all()])
-
-
-@manager.route('/list', methods=['GET'])  # noqa: F821
-@login_required
-def canvas_list():
-    return get_json_result(data=sorted([c.to_dict() for c in \
-                                 UserCanvasService.query(user_id=current_user.id, canvas_category=CanvasCategory.Agent)], key=lambda x: x["update_time"]*-1)
-                           )
 
 
 @manager.route('/rm', methods=['POST'])  # noqa: F821
@@ -77,9 +75,10 @@ def save():
     if not isinstance(req["dsl"], str):
         req["dsl"] = json.dumps(req["dsl"], ensure_ascii=False)
     req["dsl"] = json.loads(req["dsl"])
+    cate = req.get("canvas_category", CanvasCategory.Agent)
     if "id" not in req:
         req["user_id"] = current_user.id
-        if UserCanvasService.query(user_id=current_user.id, title=req["title"].strip(), canvas_category=CanvasCategory.Agent):
+        if UserCanvasService.query(user_id=current_user.id, title=req["title"].strip(), canvas_category=cate):
             return get_data_error_result(message=f"{req['title'].strip()} already exists.")
         req["id"] = get_uuid()
         if not UserCanvasService.save(**req):
@@ -101,7 +100,7 @@ def save():
 def get(canvas_id):
     if not UserCanvasService.accessible(canvas_id, current_user.id):
         return get_data_error_result(message="canvas not found.")
-    e, c = UserCanvasService.get_by_tenant_id(canvas_id)
+    e, c = UserCanvasService.get_by_canvas_id(canvas_id)
     return get_json_result(data=c)
 
 
@@ -148,6 +147,14 @@ def run():
     if not isinstance(cvs.dsl, str):
         cvs.dsl = json.dumps(cvs.dsl, ensure_ascii=False)
 
+    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)
+        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"])
     except Exception as e:
@@ -173,6 +180,44 @@ def run():
     return resp
 
 
+@manager.route('/rerun', methods=['POST'])  # noqa: F821
+@validate_request("id", "dsl", "component_id")
+@login_required
+def rerun():
+    req = request.json
+    doc = PipelineOperationLogService.get_documents_info(req["id"])
+    if not doc:
+        return get_data_error_result(message="Document not found.")
+    doc = doc[0]
+    if 0 < doc["progress"] < 1:
+        return get_data_error_result(message=f"`{doc['name']}` is processing...")
+
+    if globals.docStoreConn.indexExist(search.index_name(current_user.id), doc["kb_id"]):
+        globals.docStoreConn.delete({"doc_id": doc["id"]}, search.index_name(current_user.id), doc["kb_id"])
+    doc["progress_msg"] = ""
+    doc["chunk_num"] = 0
+    doc["token_num"] = 0
+    DocumentService.clear_chunk_num_when_rerun(doc["id"])
+    DocumentService.update_by_id(id, doc)
+    TaskService.filter_delete([Task.doc_id == id])
+
+    dsl = req["dsl"]
+    dsl["path"] = [req["component_id"]]
+    PipelineOperationLogService.update_by_id(req["id"], {"dsl": dsl})
+    queue_dataflow(tenant_id=current_user.id, flow_id=req["id"], task_id=get_uuid(), doc_id=doc["id"], priority=0, rerun=True)
+    return get_json_result(data=True)
+
+
+@manager.route('/cancel/<task_id>', methods=['PUT'])  # noqa: F821
+@login_required
+def cancel(task_id):
+    try:
+        REDIS_CONN.set(f"{task_id}-cancel", "x")
+    except Exception as e:
+        logging.exception(e)
+    return get_json_result(data=True)
+
+
 @manager.route('/reset', methods=['POST'])  # noqa: F821
 @validate_request("id")
 @login_required
@@ -198,7 +243,7 @@ def reset():
 
 @manager.route("/upload/<canvas_id>", methods=["POST"])  # noqa: F821
 def upload(canvas_id):
-    e, cvs = UserCanvasService.get_by_tenant_id(canvas_id)
+    e, cvs = UserCanvasService.get_by_canvas_id(canvas_id)
     if not e:
         return get_data_error_result(message="canvas not found.")
 
@@ -332,7 +377,7 @@ def test_db_connect():
         if req["db_type"] in ["mysql", "mariadb"]:
             db = MySQLDatabase(req["database"], user=req["username"], host=req["host"], port=req["port"],
                                password=req["password"])
-        elif req["db_type"] == 'postgresql':
+        elif req["db_type"] == 'postgres':
             db = PostgresqlDatabase(req["database"], user=req["username"], host=req["host"], port=req["port"],
                                     password=req["password"])
         elif req["db_type"] == 'mssql':
@@ -348,6 +393,65 @@ def test_db_connect():
             cursor = db.cursor()
             cursor.execute("SELECT 1")
             cursor.close()
+        elif req["db_type"] == 'IBM DB2':
+            import ibm_db
+            conn_str = (
+                f"DATABASE={req['database']};"
+                f"HOSTNAME={req['host']};"
+                f"PORT={req['port']};"
+                f"PROTOCOL=TCPIP;"
+                f"UID={req['username']};"
+                f"PWD={req['password']};"
+            )
+            logging.info(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:
+                    return None, None
+                if "." in db:
+                    c, s = db.split(".", 1)
+                elif "/" in db:
+                    c, s = db.split("/", 1)
+                else:
+                    c, s = db, "default"
+                return c, s
+            try:
+                import trino
+                import os
+                from trino.auth import BasicAuthentication
+            except Exception:
+                return server_error_response("Missing dependency 'trino'. Please install: pip install trino")
+
+            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"])
+
+            conn = trino.dbapi.connect(
+                host=req["host"],
+                port=int(req["port"] or 8080),
+                user=req["username"] or "ragflow",
+                catalog=catalog,
+                schema=schema or "default",
+                http_scheme=http_scheme,
+                auth=auth
+            )
+            cur = conn.cursor()
+            cur.execute("SELECT 1")
+            cur.fetchall()
+            cur.close()
+            conn.close()
+            return get_json_result(data="Database Connection Successful!")
         else:
             return server_error_response("Unsupported database type.")
         if req["db_type"] != 'mssql':
@@ -383,22 +487,32 @@ def getversion( version_id):
         return get_json_result(data=f"Error getting history file: {e}")
 
 
-@manager.route('/listteam', methods=['GET'])  # noqa: F821
+@manager.route('/list', methods=['GET'])  # noqa: F821
 @login_required
 def list_canvas():
     keywords = request.args.get("keywords", "")
-    page_number = int(request.args.get("page", 1))
+    page_number = int(request.args.get("page", 0))
-    items_per_page = int(request.args.get("page_size", 150))
+    items_per_page = int(request.args.get("page_size", 0))
     orderby = request.args.get("orderby", "create_time")
-    desc = request.args.get("desc", True)
+    canvas_category = request.args.get("canvas_category")
-    try:
+    if request.args.get("desc", "true").lower() == "false":
+        desc = False
+    else:
+        desc = True
+    owner_ids = [id for id in request.args.get("owner_ids", "").strip().split(",") if id]
+    if not owner_ids:
         tenants = TenantService.get_joined_tenants_by_user_id(current_user.id)
+        tenants = [m["tenant_id"] for m in tenants]
+        tenants.append(current_user.id)
         canvas, total = UserCanvasService.get_by_tenant_ids(
-            [m["tenant_id"] for m in tenants], current_user.id, page_number,
+            tenants, current_user.id, page_number,
-            items_per_page, orderby, desc, keywords, canvas_category=CanvasCategory.Agent)
+            items_per_page, orderby, desc, keywords, canvas_category)
-        return get_json_result(data={"canvas": canvas, "total": total})
+    else:
-    except Exception as e:
+        tenants = owner_ids
-        return server_error_response(e)
+        canvas, total = UserCanvasService.get_by_tenant_ids(
+            tenants, current_user.id, 0,
+            0, orderby, desc, keywords, canvas_category)
+    return get_json_result(data={"canvas": canvas, "total": total})
 
 
 @manager.route('/setting', methods=['POST'])  # noqa: F821
@@ -474,7 +588,7 @@ def sessions(canvas_id):
 @manager.route('/prompts', methods=['GET'])  # noqa: F821
 @login_required
 def prompts():
-    from rag.prompts.prompts import ANALYZE_TASK_SYSTEM, ANALYZE_TASK_USER, NEXT_STEP, REFLECT, CITATION_PROMPT_TEMPLATE
+    from rag.prompts.generator import ANALYZE_TASK_SYSTEM, ANALYZE_TASK_USER, NEXT_STEP, REFLECT, CITATION_PROMPT_TEMPLATE
     return get_json_result(data={
         "task_analysis": ANALYZE_TASK_SYSTEM +"\n\n"+ ANALYZE_TASK_USER,
         "plan_generation": NEXT_STEP,
@@ -483,3 +597,11 @@ def prompts():
         #"context_ranking": RANK_MEMORY,
         "citation_guidelines": CITATION_PROMPT_TEMPLATE
     })
+
+
+@manager.route('/download', methods=['GET'])  # noqa: F821
+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)

api/apps/chunk_app.py

@@ -22,7 +22,6 @@ from flask import request
 from flask_login import current_user, login_required
 
 from api import settings
-from api.db import LLMType, ParserType
 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
@@ -33,10 +32,11 @@ from api.utils.api_utils import get_data_error_result, get_json_result, server_e
 from rag.app.qa import beAdoc, rmPrefix
 from rag.app.tag import label_question
 from rag.nlp import rag_tokenizer, search
-from rag.prompts import cross_languages, keyword_extraction
+from rag.prompts.generator import gen_meta_filter, cross_languages, keyword_extraction
-from rag.prompts.prompts import gen_meta_filter
 from rag.settings import PAGERANK_FLD
-from rag.utils import rmSpace
+from common.string_utils import remove_redundant_spaces
+from common.constants import RetCode, LLMType, ParserType
+from common import globals
 
 
 @manager.route('/list', methods=['POST'])  # noqa: F821
@@ -61,12 +61,12 @@ def list_chunk():
         }
         if "available_int" in req:
             query["available_int"] = int(req["available_int"])
-        sres = settings.retrievaler.search(query, search.index_name(tenant_id), kb_ids, highlight=True)
+        sres = globals.retriever.search(query, search.index_name(tenant_id), kb_ids, highlight=["content_ltks"])
         res = {"total": sres.total, "chunks": [], "doc": doc.to_dict()}
         for id in sres.ids:
             d = {
                 "chunk_id": id,
-                "content_with_weight": rmSpace(sres.highlight[id]) if question and id in sres.highlight else sres.field[
+                "content_with_weight": remove_redundant_spaces(sres.highlight[id]) if question and id in sres.highlight else sres.field[
                     id].get(
                     "content_with_weight", ""),
                 "doc_id": sres.field[id]["doc_id"],
@@ -84,7 +84,7 @@ def list_chunk():
     except Exception as e:
         if str(e).find("not_found") > 0:
             return get_json_result(data=False, message='No chunk found!',
-                                   code=settings.RetCode.DATA_ERROR)
+                                   code=RetCode.DATA_ERROR)
         return server_error_response(e)
 
 
@@ -99,7 +99,7 @@ def get():
             return get_data_error_result(message="Tenant not found!")
         for tenant in tenants:
             kb_ids = KnowledgebaseService.get_kb_ids(tenant.tenant_id)
-            chunk = settings.docStoreConn.get(chunk_id, search.index_name(tenant.tenant_id), kb_ids)
+            chunk = globals.docStoreConn.get(chunk_id, search.index_name(tenant.tenant_id), kb_ids)
             if chunk:
                 break
         if chunk is None:
@@ -116,7 +116,7 @@ def get():
     except Exception as e:
         if str(e).find("NotFoundError") >= 0:
             return get_json_result(data=False, message='Chunk not found!',
-                                   code=settings.RetCode.DATA_ERROR)
+                                   code=RetCode.DATA_ERROR)
         return server_error_response(e)
 
 
@@ -171,7 +171,7 @@ def set():
         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)
+        globals.docStoreConn.update({"id": req["chunk_id"]}, d, search.index_name(tenant_id), doc.kb_id)
         return get_json_result(data=True)
     except Exception as e:
         return server_error_response(e)
@@ -187,7 +187,7 @@ def switch():
         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},
+            if not globals.docStoreConn.update({"id": cid},
                                                 {"available_int": int(req["available_int"])},
                                                 search.index_name(DocumentService.get_tenant_id(req["doc_id"])),
                                                 doc.kb_id):
@@ -207,7 +207,7 @@ def rm():
         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"]},
+        if not globals.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")
@@ -271,7 +271,7 @@ def create():
         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)
+        globals.docStoreConn.insert([d], search.index_name(tenant_id), doc.kb_id)
 
         DocumentService.increment_chunk_num(
             doc.id, doc.kb_id, c, 1, 0)
@@ -293,7 +293,7 @@ def retrieval_test():
         kb_ids = [kb_ids]
     if not kb_ids:
         return get_json_result(data=False, message='Please specify dataset firstly.',
-                               code=settings.RetCode.DATA_ERROR)
+                               code=RetCode.DATA_ERROR)
 
     doc_ids = req.get("doc_ids", [])
     use_kg = req.get("use_kg", False)
@@ -327,7 +327,7 @@ def retrieval_test():
             else:
                 return get_json_result(
                     data=False, message='Only owner of knowledgebase authorized for this operation.',
-                    code=settings.RetCode.OPERATING_ERROR)
+                    code=RetCode.OPERATING_ERROR)
 
         e, kb = KnowledgebaseService.get_by_id(kb_ids[0])
         if not e:
@@ -347,15 +347,16 @@ def retrieval_test():
             question += keyword_extraction(chat_mdl, question)
 
         labels = label_question(question, [kb])
-        ranks = settings.retrievaler.retrieval(question, embd_mdl, tenant_ids, kb_ids, page, size,
+        ranks = globals.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, highlight=req.get("highlight"),
+                               doc_ids, rerank_mdl=rerank_mdl,
+                                             highlight=req.get("highlight", False),
                                rank_feature=labels
                                )
         if use_kg:
-            ck = settings.kg_retrievaler.retrieval(question,
+            ck = settings.kg_retriever.retrieval(question,
                                                    tenant_ids,
                                                    kb_ids,
                                                    embd_mdl,
@@ -371,7 +372,7 @@ def retrieval_test():
     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=settings.RetCode.DATA_ERROR)
+                                   code=RetCode.DATA_ERROR)
         return server_error_response(e)
 
 
@@ -385,7 +386,7 @@ def knowledge_graph():
         "doc_ids": [doc_id],
         "knowledge_graph_kwd": ["graph", "mind_map"]
     }
-    sres = settings.retrievaler.search(req, search.index_name(tenant_id), kb_ids)
+    sres = globals.retriever.search(req, search.index_name(tenant_id), kb_ids)
     obj = {"graph": {}, "mind_map": {}}
     for id in sres.ids[:2]:
         ty = sres.field[id]["knowledge_graph_kwd"]

api/apps/connector_app.py

@@ -0,0 +1,106 @@
+#
+#  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.
+#
+import time
+
+from flask import request
+from flask_login import login_required, current_user
+
+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 common.constants import RetCode, TaskStatus
+
+@manager.route("/set", methods=["POST"])  # noqa: F821
+@login_required
+def set_connector():
+    req = 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)
+    else:
+        req["id"] = get_uuid()
+        conn = {
+            "id": req["id"],
+            "tenant_id": current_user.id,
+            "name": req["name"],
+            "source": req["source"],
+            "input_type": InputType.POLL,
+            "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
+        }
+        conn["status"] = TaskStatus.SCHEDULE
+
+    ConnectorService.save(**conn)
+    time.sleep(1)
+    e, conn = ConnectorService.get_by_id(req["id"])
+
+    return get_json_result(data=conn.to_dict())
+
+
+@manager.route("/list", methods=["GET"])  # noqa: F821
+@login_required
+def list_connector():
+    return get_json_result(data=ConnectorService.list(current_user.id))
+
+
+@manager.route("/<connector_id>", methods=["GET"])  # noqa: F821
+@login_required
+def get_connector(connector_id):
+    e, conn = ConnectorService.get_by_id(connector_id)
+    if not e:
+        return get_data_error_result(message="Can't find this Connector!")
+    return get_json_result(data=conn.to_dict())
+
+
+@manager.route("/<connector_id>/logs", methods=["GET"])  # noqa: F821
+@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))))
+
+
+@manager.route("/<connector_id>/resume", methods=["PUT"])  # noqa: F821
+@login_required
+def resume(connector_id):
+    req = request.json
+    if req.get("resume"):
+        ConnectorService.resume(connector_id, TaskStatus.SCHEDULE)
+    else:
+        ConnectorService.resume(connector_id, TaskStatus.CANCEL)
+    return get_json_result(data=True)
+
+
+@manager.route("/<connector_id>/link", methods=["POST"])  # noqa: F821
+@validate_request("kb_ids")
+@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)
+    return get_json_result(data=True)
+
+
+@manager.route("/<connector_id>/rm", methods=["POST"])  # noqa: F821
+@login_required
+def rm_connector(connector_id):
+    ConnectorService.resume(connector_id, TaskStatus.CANCEL)
+    ConnectorService.delete_by_id(connector_id)
+    return get_json_result(data=True)

api/apps/conversation_app.py

@@ -15,12 +15,10 @@
 #
 import json
 import re
-import traceback
+import logging
 from copy import deepcopy
 from flask import Response, request
 from flask_login import current_user, login_required
-from api import settings
-from api.db import LLMType
 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
@@ -29,8 +27,9 @@ 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 rag.prompts.prompt_template import load_prompt
+from rag.prompts.template import load_prompt
-from rag.prompts.prompts import chunks_format
+from rag.prompts.generator import chunks_format
+from common.constants import RetCode, LLMType
 
 
 @manager.route("/set", methods=["POST"])  # noqa: F821
@@ -93,7 +92,7 @@ def get():
                 avatar = dialog[0].icon
                 break
         else:
-            return get_json_result(data=False, message="Only owner of conversation authorized for this operation.", code=settings.RetCode.OPERATING_ERROR)
+            return get_json_result(data=False, message="Only owner of conversation authorized for this operation.", code=RetCode.OPERATING_ERROR)
 
         for ref in conv.reference:
             if isinstance(ref, list):
@@ -142,7 +141,7 @@ def rm():
                 if DialogService.query(tenant_id=tenant.tenant_id, id=conv.dialog_id):
                     break
             else:
-                return get_json_result(data=False, message="Only owner of conversation authorized for this operation.", code=settings.RetCode.OPERATING_ERROR)
+                return get_json_result(data=False, message="Only owner of conversation authorized for this operation.", code=RetCode.OPERATING_ERROR)
             ConversationService.delete_by_id(cid)
         return get_json_result(data=True)
     except Exception as e:
@@ -155,7 +154,7 @@ def list_conversation():
     dialog_id = request.args["dialog_id"]
     try:
         if not DialogService.query(tenant_id=current_user.id, id=dialog_id):
-            return get_json_result(data=False, message="Only owner of dialog authorized for this operation.", code=settings.RetCode.OPERATING_ERROR)
+            return get_json_result(data=False, message="Only owner of dialog authorized for this operation.", code=RetCode.OPERATING_ERROR)
         convs = ConversationService.query(dialog_id=dialog_id, order_by=ConversationService.model.create_time, reverse=True)
 
         convs = [d.to_dict() for d in convs]
@@ -226,7 +225,7 @@ def completion():
                 if not is_embedded:
                     ConversationService.update_by_id(conv.id, conv.to_dict())
             except Exception as e:
-                traceback.print_exc()
+                logging.exception(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"
 

api/apps/dataflow_app.py

@@ -1,353 +0,0 @@
-#
-#  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.
-#
-import json
-import re
-import sys
-import time
-from functools import partial
-
-import trio
-from flask import request
-from flask_login import current_user, login_required
-
-from agent.canvas import Canvas
-from agent.component import LLM
-from api.db import CanvasCategory, FileType
-from api.db.services.canvas_service import CanvasTemplateService, UserCanvasService
-from api.db.services.document_service import DocumentService
-from api.db.services.file_service import FileService
-from api.db.services.task_service import queue_dataflow
-from api.db.services.user_canvas_version import UserCanvasVersionService
-from api.db.services.user_service import TenantService
-from api.settings import RetCode
-from api.utils import get_uuid
-from api.utils.api_utils import get_data_error_result, get_json_result, server_error_response, validate_request
-from api.utils.file_utils import filename_type, read_potential_broken_pdf
-from rag.flow.pipeline import Pipeline
-
-
-@manager.route("/templates", methods=["GET"])  # noqa: F821
-@login_required
-def templates():
-    return get_json_result(data=[c.to_dict() for c in CanvasTemplateService.query(canvas_category=CanvasCategory.DataFlow)])
-
-
-@manager.route("/list", methods=["GET"])  # noqa: F821
-@login_required
-def canvas_list():
-    return get_json_result(data=sorted([c.to_dict() for c in UserCanvasService.query(user_id=current_user.id, canvas_category=CanvasCategory.DataFlow)], key=lambda x: x["update_time"] * -1))
-
-
-@manager.route("/rm", methods=["POST"])  # noqa: F821
-@validate_request("canvas_ids")
-@login_required
-def rm():
-    for i in request.json["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.", code=RetCode.OPERATING_ERROR)
-        UserCanvasService.delete_by_id(i)
-    return get_json_result(data=True)
-
-
-@manager.route("/set", methods=["POST"])  # noqa: F821
-@validate_request("dsl", "title")
-@login_required
-def save():
-    req = request.json
-    if not isinstance(req["dsl"], str):
-        req["dsl"] = json.dumps(req["dsl"], ensure_ascii=False)
-    req["dsl"] = json.loads(req["dsl"])
-    req["canvas_category"] = CanvasCategory.DataFlow
-    if "id" not in req:
-        req["user_id"] = current_user.id
-        if UserCanvasService.query(user_id=current_user.id, title=req["title"].strip(), canvas_category=CanvasCategory.DataFlow):
-            return get_data_error_result(message=f"{req['title'].strip()} already exists.")
-        req["id"] = get_uuid()
-
-        if not UserCanvasService.save(**req):
-            return get_data_error_result(message="Fail to save canvas.")
-    else:
-        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)
-        UserCanvasService.update_by_id(req["id"], req)
-    # save version
-    UserCanvasVersionService.insert(user_canvas_id=req["id"], dsl=req["dsl"], title="{0}_{1}".format(req["title"], time.strftime("%Y_%m_%d_%H_%M_%S")))
-    UserCanvasVersionService.delete_all_versions(req["id"])
-    return get_json_result(data=req)
-
-
-@manager.route("/get/<canvas_id>", methods=["GET"])  # noqa: F821
-@login_required
-def get(canvas_id):
-    if not UserCanvasService.accessible(canvas_id, current_user.id):
-        return get_data_error_result(message="canvas not found.")
-    e, c = UserCanvasService.get_by_tenant_id(canvas_id)
-    return get_json_result(data=c)
-
-
-@manager.route("/run", methods=["POST"])  # noqa: F821
-@validate_request("id")
-@login_required
-def run():
-    req = request.json
-    flow_id = req.get("id", "")
-    doc_id = req.get("doc_id", "")
-    if not all([flow_id, doc_id]):
-        return get_data_error_result(message="id and doc_id are required.")
-
-    if not DocumentService.get_by_id(doc_id):
-        return get_data_error_result(message=f"Document for {doc_id} not found.")
-
-    user_id = req.get("user_id", current_user.id)
-    if not UserCanvasService.accessible(flow_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(flow_id)
-    if not e:
-        return get_data_error_result(message="canvas not found.")
-
-    if not isinstance(cvs.dsl, str):
-        cvs.dsl = json.dumps(cvs.dsl, ensure_ascii=False)
-
-    task_id = get_uuid()
-
-    ok, error_message = queue_dataflow(dsl=cvs.dsl, tenant_id=user_id, doc_id=doc_id, task_id=task_id, flow_id=flow_id, priority=0)
-    if not ok:
-        return server_error_response(error_message)
-
-    return get_json_result(data={"task_id": task_id, "flow_id": flow_id})
-
-
-@manager.route("/reset", methods=["POST"])  # noqa: F821
-@validate_request("id")
-@login_required
-def reset():
-    req = request.json
-    flow_id = req.get("id", "")
-    if not flow_id:
-        return get_data_error_result(message="id is required.")
-
-    if not UserCanvasService.accessible(flow_id, current_user.id):
-        return get_json_result(data=False, message="Only owner of canvas authorized for this operation.", code=RetCode.OPERATING_ERROR)
-
-    task_id = req.get("task_id", "")
-
-    try:
-        e, user_canvas = UserCanvasService.get_by_id(req["id"])
-        if not e:
-            return get_data_error_result(message="canvas not found.")
-
-        dataflow = Pipeline(dsl=json.dumps(user_canvas.dsl), tenant_id=current_user.id, flow_id=flow_id, task_id=task_id)
-        dataflow.reset()
-        req["dsl"] = json.loads(str(dataflow))
-        UserCanvasService.update_by_id(req["id"], {"dsl": req["dsl"]})
-        return get_json_result(data=req["dsl"])
-    except Exception as e:
-        return server_error_response(e)
-
-
-@manager.route("/upload/<canvas_id>", methods=["POST"])  # noqa: F821
-def upload(canvas_id):
-    e, cvs = UserCanvasService.get_by_tenant_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, CrawlResult, DefaultMarkdownGenerator, PruningContentFilter
-
-        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"]
-    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))
-    except Exception as e:
-        return server_error_response(e)
-
-
-@manager.route("/input_form", methods=["GET"])  # noqa: F821
-@login_required
-def input_form():
-    flow_id = request.args.get("id")
-    cpn_id = request.args.get("component_id")
-    try:
-        e, user_canvas = UserCanvasService.get_by_id(flow_id)
-        if not e:
-            return get_data_error_result(message="canvas not found.")
-        if not UserCanvasService.query(user_id=current_user.id, id=flow_id):
-            return get_json_result(data=False, message="Only owner of canvas authorized for this operation.", code=RetCode.OPERATING_ERROR)
-
-        dataflow = Pipeline(dsl=json.dumps(user_canvas.dsl), tenant_id=current_user.id, flow_id=flow_id, task_id="")
-
-        return get_json_result(data=dataflow.get_component_input_form(cpn_id))
-    except Exception as e:
-        return server_error_response(e)
-
-
-@manager.route("/debug", methods=["POST"])  # noqa: F821
-@validate_request("id", "component_id", "params")
-@login_required
-def debug():
-    req = 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.reset()
-        canvas.message_id = get_uuid()
-        component = canvas.get_component(req["component_id"])["obj"]
-        component.reset()
-
-        if isinstance(component, LLM):
-            component.set_debug_inputs(req["params"])
-        component.invoke(**{k: o["value"] for k, o in req["params"].items()})
-        outputs = component.output()
-        for k in outputs.keys():
-            if isinstance(outputs[k], partial):
-                txt = ""
-                for c in outputs[k]():
-                    txt += c
-                outputs[k] = txt
-        return get_json_result(data=outputs)
-    except Exception as e:
-        return server_error_response(e)
-
-
-# api get list version dsl of canvas
-@manager.route("/getlistversion/<canvas_id>", methods=["GET"])  # noqa: F821
-@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)
-    except Exception as e:
-        return get_data_error_result(message=f"Error getting history files: {e}")
-
-
-# api get version dsl of canvas
-@manager.route("/getversion/<version_id>", methods=["GET"])  # noqa: F821
-@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())
-    except Exception as e:
-        return get_json_result(data=f"Error getting history file: {e}")
-
-
-@manager.route("/listteam", methods=["GET"])  # noqa: F821
-@login_required
-def list_canvas():
-    keywords = request.args.get("keywords", "")
-    page_number = int(request.args.get("page", 1))
-    items_per_page = int(request.args.get("page_size", 150))
-    orderby = request.args.get("orderby", "create_time")
-    desc = request.args.get("desc", True)
-    try:
-        tenants = TenantService.get_joined_tenants_by_user_id(current_user.id)
-        canvas, total = UserCanvasService.get_by_tenant_ids(
-            [m["tenant_id"] for m in tenants], current_user.id, page_number, items_per_page, orderby, desc, keywords, canvas_category=CanvasCategory.DataFlow
-        )
-        return get_json_result(data={"canvas": canvas, "total": total})
-    except Exception as e:
-        return server_error_response(e)
-
-
-@manager.route("/setting", methods=["POST"])  # noqa: F821
-@validate_request("id", "title", "permission")
-@login_required
-def setting():
-    req = request.json
-    req["user_id"] = current_user.id
-
-    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)
-
-    e, flow = UserCanvasService.get_by_id(req["id"])
-    if not e:
-        return get_data_error_result(message="canvas not found.")
-    flow = flow.to_dict()
-    flow["title"] = req["title"]
-    for key in ("description", "permission", "avatar"):
-        if value := req.get(key):
-            flow[key] = value
-
-    num = UserCanvasService.update_by_id(req["id"], flow)
-    return get_json_result(data=num)
-
-
-@manager.route("/trace", methods=["GET"])  # noqa: F821
-def trace():
-    dataflow_id = request.args.get("dataflow_id")
-    task_id = request.args.get("task_id")
-    if not all([dataflow_id, task_id]):
-        return get_data_error_result(message="dataflow_id and task_id are required.")
-
-    e, dataflow_canvas = UserCanvasService.get_by_id(dataflow_id)
-    if not e:
-        return get_data_error_result(message="dataflow not found.")
-
-    dsl_str = json.dumps(dataflow_canvas.dsl, ensure_ascii=False)
-    dataflow = Pipeline(dsl=dsl_str, tenant_id=dataflow_canvas.user_id, flow_id=dataflow_id, task_id=task_id)
-    log = dataflow.fetch_logs()
-
-    return get_json_result(data=log)

api/apps/dialog_app.py

@@ -18,13 +18,13 @@ from flask import request
 from flask_login import login_required, current_user
 from api.db.services import duplicate_name
 from api.db.services.dialog_service import DialogService
-from api.db import StatusEnum
+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 import settings
 from api.utils.api_utils import server_error_response, get_data_error_result, validate_request
-from api.utils import get_uuid
+from common.misc_utils import get_uuid
+from common.constants import RetCode
 from api.utils.api_utils import get_json_result
 
 
@@ -219,7 +219,7 @@ def rm():
             else:
                 return get_json_result(
                     data=False, message='Only owner of dialog authorized for this operation.',
-                    code=settings.RetCode.OPERATING_ERROR)
+                    code=RetCode.OPERATING_ERROR)
             dialog_list.append({"id": id,"status":StatusEnum.INVALID.value})
         DialogService.update_many_by_id(dialog_list)
         return get_json_result(data=True)

api/apps/document_app.py

@@ -23,29 +23,32 @@ import flask
 from flask import request
 from flask_login import current_user, login_required
 
-from api import settings
+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, VALID_TASK_STATUS, FileSource, FileType, ParserType, TaskStatus
+from api.db import VALID_FILE_TYPES, FileType
-from api.db.db_models import File, Task
+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 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 TaskService, cancel_all_task_of, queue_tasks
+from api.db.services.task_service import TaskService, cancel_all_task_of
 from api.db.services.user_service import UserTenantService
-from api.utils import get_uuid
+from common.misc_utils import get_uuid
 from api.utils.api_utils import (
     get_data_error_result,
     get_json_result,
     server_error_response,
     validate_request,
 )
-from api.utils.file_utils import filename_type, get_project_base_directory, thumbnail
+from api.utils.file_utils import filename_type, thumbnail
+from common.file_utils import get_project_base_directory
+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
+from rag.nlp import search, rag_tokenizer
 from rag.utils.storage_factory import STORAGE_IMPL
+from common import globals
 
 
 @manager.route("/upload", methods=["POST"])  # noqa: F821
@@ -54,27 +57,29 @@ from rag.utils.storage_factory import STORAGE_IMPL
 def upload():
     kb_id = request.form.get("kb_id")
     if not kb_id:
-        return get_json_result(data=False, message='Lack of "KB ID"', code=settings.RetCode.ARGUMENT_ERROR)
+        return get_json_result(data=False, message='Lack of "KB ID"', code=RetCode.ARGUMENT_ERROR)
     if "file" not in request.files:
-        return get_json_result(data=False, message="No file part!", code=settings.RetCode.ARGUMENT_ERROR)
+        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=settings.RetCode.ARGUMENT_ERROR)
+            return get_json_result(data=False, message="No file selected!", code=RetCode.ARGUMENT_ERROR)
         if len(file_obj.filename.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=settings.RetCode.ARGUMENT_ERROR)
+            return get_json_result(data=False, message=f"File name must be {FILE_NAME_LEN_LIMIT} bytes or less.", code=RetCode.ARGUMENT_ERROR)
 
     e, kb = KnowledgebaseService.get_by_id(kb_id)
     if not e:
         raise LookupError("Can't find this knowledgebase!")
-    err, files = FileService.upload_document(kb, file_objs, current_user.id)
+    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)
     if err:
-        return get_json_result(data=files, message="\n".join(err), code=settings.RetCode.SERVER_ERROR)
+        return get_json_result(data=files, message="\n".join(err), code=RetCode.SERVER_ERROR)
 
     if not files:
-        return get_json_result(data=files, message="There seems to be an issue with your file format. Please verify it is correct and not corrupted.", code=settings.RetCode.DATA_ERROR)
+        return get_json_result(data=files, message="There seems to be an issue with your file format. Please verify it is correct and not corrupted.", code=RetCode.DATA_ERROR)
     files = [f[0] for f in files]  # remove the blob
 
     return get_json_result(data=files)
@@ -86,14 +91,16 @@ def upload():
 def web_crawl():
     kb_id = request.form.get("kb_id")
     if not kb_id:
-        return get_json_result(data=False, message='Lack of "KB ID"', code=settings.RetCode.ARGUMENT_ERROR)
+        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")
     if not is_valid_url(url):
-        return get_json_result(data=False, message="The URL format is invalid", code=settings.RetCode.ARGUMENT_ERROR)
+        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!")
+    if check_kb_team_permission(kb, current_user.id):
+        return get_json_result(data=False, message="No authorization.", code=RetCode.AUTHENTICATION_ERROR)
 
     blob = html2pdf(url)
     if not blob:
@@ -150,12 +157,12 @@ def create():
     req = request.json
     kb_id = req["kb_id"]
     if not kb_id:
-        return get_json_result(data=False, message='Lack of "KB ID"', code=settings.RetCode.ARGUMENT_ERROR)
+        return get_json_result(data=False, message='Lack of "KB ID"', 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=settings.RetCode.ARGUMENT_ERROR)
+        return get_json_result(data=False, message=f"File name must be {FILE_NAME_LEN_LIMIT} bytes or less.", code=RetCode.ARGUMENT_ERROR)
 
     if req["name"].strip() == "":
-        return get_json_result(data=False, message="File name can't be empty.", code=settings.RetCode.ARGUMENT_ERROR)
+        return get_json_result(data=False, message="File name can't be empty.", code=RetCode.ARGUMENT_ERROR)
     req["name"] = req["name"].strip()
 
     try:
@@ -182,6 +189,7 @@ def create():
                 "id": get_uuid(),
                 "kb_id": kb.id,
                 "parser_id": kb.parser_id,
+                "pipeline_id": kb.pipeline_id,
                 "parser_config": kb.parser_config,
                 "created_by": current_user.id,
                 "type": FileType.VIRTUAL,
@@ -204,13 +212,13 @@ def create():
 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=settings.RetCode.ARGUMENT_ERROR)
+        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 knowledgebase authorized for this operation.", code=settings.RetCode.OPERATING_ERROR)
+        return get_json_result(data=False, message="Only owner of knowledgebase authorized for this operation.", code=RetCode.OPERATING_ERROR)
     keywords = request.args.get("keywords", "")
 
     page_number = int(request.args.get("page", 0))
@@ -266,13 +274,13 @@ def get_filter():
 
     kb_id = req.get("kb_id")
     if not kb_id:
-        return get_json_result(data=False, message='Lack of "KB ID"', code=settings.RetCode.ARGUMENT_ERROR)
+        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 knowledgebase authorized for this operation.", code=settings.RetCode.OPERATING_ERROR)
+        return get_json_result(data=False, message="Only owner of knowledgebase authorized for this operation.", code=RetCode.OPERATING_ERROR)
 
     keywords = req.get("keywords", "")
 
@@ -304,7 +312,7 @@ def docinfos():
     doc_ids = req["doc_ids"]
     for doc_id in doc_ids:
         if not DocumentService.accessible(doc_id, current_user.id):
-            return get_json_result(data=False, message="No authorization.", code=settings.RetCode.AUTHENTICATION_ERROR)
+            return get_json_result(data=False, message="No authorization.", code=RetCode.AUTHENTICATION_ERROR)
     docs = DocumentService.get_by_ids(doc_ids)
     return get_json_result(data=list(docs.dicts()))
 
@@ -314,7 +322,7 @@ def docinfos():
 def thumbnails():
     doc_ids = request.args.getlist("doc_ids")
     if not doc_ids:
-        return get_json_result(data=False, message='Lack of "Document ID"', code=settings.RetCode.ARGUMENT_ERROR)
+        return get_json_result(data=False, message='Lack of "Document ID"', code=RetCode.ARGUMENT_ERROR)
 
     try:
         docs = DocumentService.get_thumbnails(doc_ids)
@@ -337,7 +345,7 @@ def change_status():
     status = str(req.get("status", ""))
 
     if status not in ["0", "1"]:
-        return get_json_result(data=False, message='"Status" must be either 0 or 1!', code=settings.RetCode.ARGUMENT_ERROR)
+        return get_json_result(data=False, message='"Status" must be either 0 or 1!', code=RetCode.ARGUMENT_ERROR)
 
     result = {}
     for doc_id in doc_ids:
@@ -359,7 +367,7 @@ def change_status():
                 continue
 
             status_int = int(status)
-            if not settings.docStoreConn.update({"doc_id": doc_id}, {"available_int": status_int}, search.index_name(kb.tenant_id), doc.kb_id):
+            if not globals.docStoreConn.update({"doc_id": doc_id}, {"available_int": status_int}, search.index_name(kb.tenant_id), doc.kb_id):
                 result[doc_id] = {"error": "Database error (docStore update)!"}
             result[doc_id] = {"status": status}
         except Exception as e:
@@ -379,50 +387,12 @@ def rm():
 
     for doc_id in doc_ids:
         if not DocumentService.accessible4deletion(doc_id, current_user.id):
-            return get_json_result(data=False, message="No authorization.", code=settings.RetCode.AUTHENTICATION_ERROR)
+            return get_json_result(data=False, message="No authorization.", code=RetCode.AUTHENTICATION_ERROR)
 
-    root_folder = FileService.get_root_folder(current_user.id)
+    errors = FileService.delete_docs(doc_ids, current_user.id)
-    pf_id = root_folder["id"]
-    FileService.init_knowledgebase_docs(pf_id, current_user.id)
-    errors = ""
-    kb_table_num_map = {}
-    for doc_id in doc_ids:
-        try:
-            e, doc = DocumentService.get_by_id(doc_id)
-            if not e:
-                return get_data_error_result(message="Document not found!")
-            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)
-
-            TaskService.filter_delete([Task.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)
-            deleted_file_count = 0
-            if f2d:
-                deleted_file_count = FileService.filter_delete([File.source_type == FileSource.KNOWLEDGEBASE, File.id == f2d[0].file_id])
-            File2DocumentService.delete_by_document_id(doc_id)
-            if deleted_file_count > 0:
-                STORAGE_IMPL.rm(b, n)
-
-            doc_parser = doc.parser_id
-            if doc_parser == ParserType.TABLE:
-                kb_id = doc.kb_id
-                if kb_id not in kb_table_num_map:
-                    counts = DocumentService.count_by_kb_id(kb_id=kb_id, keywords="", run_status=[TaskStatus.DONE], types=[])
-                    kb_table_num_map[kb_id] = counts
-                kb_table_num_map[kb_id] -= 1
-                if kb_table_num_map[kb_id] <= 0:
-                    KnowledgebaseService.delete_field_map(kb_id)
-        except Exception as e:
-            errors += str(e)
 
     if errors:
-        return get_json_result(data=False, message=errors, code=settings.RetCode.SERVER_ERROR)
+        return get_json_result(data=False, message=errors, code=RetCode.SERVER_ERROR)
 
     return get_json_result(data=True)
 
@@ -434,7 +404,7 @@ 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=settings.RetCode.AUTHENTICATION_ERROR)
+            return get_json_result(data=False, message="No authorization.", code=RetCode.AUTHENTICATION_ERROR)
     try:
         kb_table_num_map = {}
         for id in req["doc_ids"]:
@@ -462,25 +432,12 @@ def run():
             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):
+                if globals.docStoreConn.indexExist(search.index_name(tenant_id), doc.kb_id):
-                    settings.docStoreConn.delete({"doc_id": id}, search.index_name(tenant_id), doc.kb_id)
+                    globals.docStoreConn.delete({"doc_id": id}, search.index_name(tenant_id), doc.kb_id)
 
             if str(req["run"]) == TaskStatus.RUNNING.value:
                 doc = doc.to_dict()
-                doc["tenant_id"] = tenant_id
+                DocumentService.run(tenant_id, doc, kb_table_num_map)
-
-                doc_parser = doc.get("parser_id", ParserType.NAIVE)
-                if doc_parser == ParserType.TABLE:
-                    kb_id = doc.get("kb_id")
-                    if not kb_id:
-                        continue
-                    if kb_id not in kb_table_num_map:
-                        count = DocumentService.count_by_kb_id(kb_id=kb_id, keywords="", run_status=[TaskStatus.DONE], types=[])
-                        kb_table_num_map[kb_id] = count
-                        if kb_table_num_map[kb_id] <= 0:
-                            KnowledgebaseService.delete_field_map(kb_id)
-                bucket, name = File2DocumentService.get_storage_address(doc_id=doc["id"])
-                queue_tasks(doc, bucket, name, 0)
 
         return get_json_result(data=True)
     except Exception as e:
@@ -493,15 +450,15 @@ def run():
 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=settings.RetCode.AUTHENTICATION_ERROR)
+        return get_json_result(data=False, message="No authorization.", code=RetCode.AUTHENTICATION_ERROR)
     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=settings.RetCode.ARGUMENT_ERROR)
+            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=settings.RetCode.ARGUMENT_ERROR)
+            return get_json_result(data=False, message=f"File name must be {FILE_NAME_LEN_LIMIT} bytes or less.", code=RetCode.ARGUMENT_ERROR)
 
         for d in DocumentService.query(name=req["name"], kb_id=doc.kb_id):
             if d.name == req["name"]:
@@ -515,6 +472,21 @@ def rename():
             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 globals.docStoreConn.indexExist(search.index_name(tenant_id), doc.kb_id):
+            globals.docStoreConn.update(
+                {"doc_id": req["doc_id"]},
+                es_body,
+                search.index_name(tenant_id),
+                doc.kb_id,
+            )
+
         return get_json_result(data=True)
     except Exception as e:
         return server_error_response(e)
@@ -546,16 +518,40 @@ def get(doc_id):
 
 @manager.route("/change_parser", methods=["POST"])  # noqa: F821
 @login_required
-@validate_request("doc_id", "parser_id")
+@validate_request("doc_id")
 def change_parser():
-    req = request.json
 
+    req = request.json
     if not DocumentService.accessible(req["doc_id"], current_user.id):
-        return get_json_result(data=False, message="No authorization.", code=settings.RetCode.AUTHENTICATION_ERROR)
+        return get_json_result(data=False, message="No authorization.", code=RetCode.AUTHENTICATION_ERROR)
-    try:
+
-        e, doc = DocumentService.get_by_id(req["doc_id"])
+    e, doc = DocumentService.get_by_id(req["doc_id"])
+    if not e:
+        return get_data_error_result(message="Document not found!")
+
+    def reset_doc():
+        nonlocal doc
+        e = DocumentService.update_by_id(doc.id, {"pipeline_id": req["pipeline_id"], "parser_id": req["parser_id"], "progress": 0, "progress_msg": "", "run": TaskStatus.UNSTART.value})
         if not e:
             return get_data_error_result(message="Document not found!")
+        if doc.token_num > 0:
+            e = DocumentService.increment_chunk_num(doc.id, doc.kb_id, doc.token_num * -1, doc.chunk_num * -1, doc.process_duration * -1)
+            if not e:
+                return get_data_error_result(message="Document not found!")
+            tenant_id = DocumentService.get_tenant_id(req["doc_id"])
+            if not tenant_id:
+                return get_data_error_result(message="Tenant not found!")
+            if globals.docStoreConn.indexExist(search.index_name(tenant_id), doc.kb_id):
+                globals.docStoreConn.delete({"doc_id": doc.id}, search.index_name(tenant_id), doc.kb_id)
+
+    try:
+        if "pipeline_id" in req and req["pipeline_id"] != "":
+            if doc.pipeline_id == req["pipeline_id"]:
+                return get_json_result(data=True)
+            DocumentService.update_by_id(doc.id, {"pipeline_id": req["pipeline_id"]})
+            reset_doc()
+            return get_json_result(data=True)
+
         if doc.parser_id.lower() == req["parser_id"].lower():
             if "parser_config" in req:
                 if req["parser_config"] == doc.parser_config:
@@ -565,22 +561,9 @@ def change_parser():
 
         if (doc.type == FileType.VISUAL and req["parser_id"] != "picture") or (re.search(r"\.(ppt|pptx|pages)$", doc.name) and req["parser_id"] != "presentation"):
             return get_data_error_result(message="Not supported yet!")
-
-        e = DocumentService.update_by_id(doc.id, {"parser_id": req["parser_id"], "progress": 0, "progress_msg": "", "run": TaskStatus.UNSTART.value})
-        if not e:
-            return get_data_error_result(message="Document not found!")
         if "parser_config" in req:
             DocumentService.update_parser_config(doc.id, req["parser_config"])
-        if doc.token_num > 0:
+        reset_doc()
-            e = DocumentService.increment_chunk_num(doc.id, doc.kb_id, doc.token_num * -1, doc.chunk_num * -1, doc.process_duration * -1)
-            if not e:
-                return get_data_error_result(message="Document not found!")
-            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):
-                settings.docStoreConn.delete({"doc_id": doc.id}, search.index_name(tenant_id), doc.kb_id)
-
         return get_json_result(data=True)
     except Exception as e:
         return server_error_response(e)
@@ -606,12 +589,12 @@ def get_image(image_id):
 @validate_request("conversation_id")
 def upload_and_parse():
     if "file" not in request.files:
-        return get_json_result(data=False, message="No file part!", code=settings.RetCode.ARGUMENT_ERROR)
+        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=settings.RetCode.ARGUMENT_ERROR)
+            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)
 
@@ -624,7 +607,7 @@ def parse():
     url = request.json.get("url") if request.json else ""
     if url:
         if not is_valid_url(url):
-            return get_json_result(data=False, message="The URL format is invalid", code=settings.RetCode.ARGUMENT_ERROR)
+            return get_json_result(data=False, message="The URL format is invalid", code=RetCode.ARGUMENT_ERROR)
         download_path = os.path.join(get_project_base_directory(), "logs/downloads")
         os.makedirs(download_path, exist_ok=True)
         from seleniumwire.webdriver import Chrome, ChromeOptions
@@ -657,13 +640,13 @@ def parse():
 
         r = re.search(r"filename=\"([^\"]+)\"", str(res_headers))
         if not r or not r.group(1):
-            return get_json_result(data=False, message="Can't not identify downloaded file", code=settings.RetCode.ARGUMENT_ERROR)
+            return get_json_result(data=False, message="Can't not identify downloaded file", code=RetCode.ARGUMENT_ERROR)
         f = File(r.group(1), os.path.join(download_path, r.group(1)))
         txt = FileService.parse_docs([f], current_user.id)
         return get_json_result(data=txt)
 
     if "file" not in request.files:
-        return get_json_result(data=False, message="No file part!", code=settings.RetCode.ARGUMENT_ERROR)
+        return get_json_result(data=False, message="No file part!", code=RetCode.ARGUMENT_ERROR)
 
     file_objs = request.files.getlist("file")
     txt = FileService.parse_docs(file_objs, current_user.id)
@@ -677,18 +660,18 @@ def parse():
 def set_meta():
     req = request.json
     if not DocumentService.accessible(req["doc_id"], current_user.id):
-        return get_json_result(data=False, message="No authorization.", code=settings.RetCode.AUTHENTICATION_ERROR)
+        return get_json_result(data=False, message="No authorization.", code=RetCode.AUTHENTICATION_ERROR)
     try:
         meta = json.loads(req["meta"])
         if not isinstance(meta, dict):
-            return get_json_result(data=False, message="Only dictionary type supported.", code=settings.RetCode.ARGUMENT_ERROR)
+            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):
-                return get_json_result(data=False, message=f"The type is not supported: {v}", code=settings.RetCode.ARGUMENT_ERROR)
+                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=settings.RetCode.ARGUMENT_ERROR)
+        return get_json_result(data=False, message=f"Json syntax error: {e}", code=RetCode.ARGUMENT_ERROR)
     if not isinstance(meta, dict):
-        return get_json_result(data=False, message='Meta data should be in Json map format, like {"key": "value"}', code=settings.RetCode.ARGUMENT_ERROR)
+        return get_json_result(data=False, message='Meta data should be in Json map format, like {"key": "value"}', code=RetCode.ARGUMENT_ERROR)
 
     try:
         e, doc = DocumentService.get_by_id(req["doc_id"])

api/apps/file2document_app.py

@@ -23,10 +23,10 @@ from flask import request
 from flask_login 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 import get_uuid
+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 import settings
 from api.utils.api_utils import get_json_result
 
 
@@ -108,7 +108,7 @@ def rm():
     file_ids = req["file_ids"]
     if not file_ids:
         return get_json_result(
-            data=False, message='Lack of "Files ID"', code=settings.RetCode.ARGUMENT_ERROR)
+            data=False, message='Lack of "Files ID"', code=RetCode.ARGUMENT_ERROR)
     try:
         for file_id in file_ids:
             informs = File2DocumentService.get_by_file_id(file_id)

api/apps/file_app.py

@@ -13,6 +13,7 @@
 #  See the License for the specific language governing permissions and
 #  limitations under the License
 #
+import logging
 import os
 import pathlib
 import re
@@ -21,14 +22,15 @@ import flask
 from flask import request
 from flask_login import login_required, current_user
 
+from api.common.check_team_permission import check_file_team_permission
 from api.db.services.document_service import DocumentService
 from api.db.services.file2document_service import File2DocumentService
 from api.utils.api_utils import server_error_response, get_data_error_result, validate_request
-from api.utils import get_uuid
+from common.misc_utils import get_uuid
-from api.db import FileType, FileSource
+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 import settings
 from api.utils.api_utils import get_json_result
 from api.utils.file_utils import filename_type
 from api.utils.web_utils import CONTENT_TYPE_MAP
@@ -47,21 +49,21 @@ def upload():
 
     if 'file' not in request.files:
         return get_json_result(
-            data=False, message='No file part!', code=settings.RetCode.ARGUMENT_ERROR)
+            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=settings.RetCode.ARGUMENT_ERROR)
+                data=False, message='No file selected!', code=RetCode.ARGUMENT_ERROR)
     file_res = []
     try:
         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:
-            MAX_FILE_NUM_PER_USER = int(os.environ.get('MAX_FILE_NUM_PER_USER', 0))
+            MAX_FILE_NUM_PER_USER: int = int(os.environ.get('MAX_FILE_NUM_PER_USER', 0))
-            if MAX_FILE_NUM_PER_USER > 0 and DocumentService.get_doc_count(current_user.id) >= MAX_FILE_NUM_PER_USER:
+            if 0 < MAX_FILE_NUM_PER_USER <= 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
@@ -132,7 +134,7 @@ def create():
     try:
         if not FileService.is_parent_folder_exist(pf_id):
             return get_json_result(
-                data=False, message="Parent Folder Doesn't Exist!", code=settings.RetCode.OPERATING_ERROR)
+                data=False, message="Parent Folder Doesn't Exist!", code=RetCode.OPERATING_ERROR)
         if FileService.query(name=req["name"], parent_id=pf_id):
             return get_data_error_result(
                 message="Duplicated folder name in the same folder.")
@@ -233,52 +235,63 @@ def get_all_parent_folders():
         return server_error_response(e)
 
 
-@manager.route('/rm', methods=['POST'])  # noqa: F821
+@manager.route("/rm", methods=["POST"])  # noqa: F821
 @login_required
 @validate_request("file_ids")
 def rm():
     req = 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:
+            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:
-                file_id_list = FileService.get_all_innermost_file_ids(file_id, [])
+                _delete_folder_recursive(file, current_user.id)
-                for inner_file_id in file_id_list:
+                continue
-                    e, file = FileService.get_by_id(inner_file_id)
-                    if not e:
-                        return get_data_error_result(message="File not found!")
-                    STORAGE_IMPL.rm(file.parent_id, file.location)
-                FileService.delete_folder_by_pf_id(current_user.id, file_id)
-            else:
-                STORAGE_IMPL.rm(file.parent_id, file.location)
-                if not FileService.delete(file):
-                    return get_data_error_result(
-                        message="Database error (File removal)!")
 
-            # delete file2document
+            _delete_single_file(file)
-            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 not e:
-                    return get_data_error_result(message="Document not found!")
-                tenant_id = DocumentService.get_tenant_id(doc_id)
-                if not tenant_id:
-                    return get_data_error_result(message="Tenant not found!")
-                if not DocumentService.remove_document(doc, tenant_id):
-                    return get_data_error_result(
-                        message="Database error (Document removal)!")
-            File2DocumentService.delete_by_file_id(file_id)
 
         return get_json_result(data=True)
+
     except Exception as e:
         return server_error_response(e)
 
@@ -292,13 +305,15 @@ def rename():
         e, file = FileService.get_by_id(req["file_id"])
         if not e:
             return get_data_error_result(message="File 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.type != FileType.FOLDER.value \
             and pathlib.Path(req["name"].lower()).suffix != pathlib.Path(
                 file.name.lower()).suffix:
             return get_json_result(
                 data=False,
                 message="The extension of file can't be changed",
-                code=settings.RetCode.ARGUMENT_ERROR)
+                code=RetCode.ARGUMENT_ERROR)
         for file in FileService.query(name=req["name"], pf_id=file.parent_id):
             if file.name == req["name"]:
                 return get_data_error_result(
@@ -328,6 +343,8 @@ def get(file_id):
         e, file = FileService.get_by_id(file_id)
         if not e:
             return get_data_error_result(message="Document 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)
 
         blob = STORAGE_IMPL.get(file.parent_id, file.location)
         if not blob:
@@ -348,29 +365,89 @@ def get(file_id):
         return server_error_response(e)
 
 
-@manager.route('/mv', methods=['POST'])  # noqa: F821
+@manager.route("/mv", methods=["POST"])  # noqa: F821
 @login_required
 @validate_request("src_file_ids", "dest_file_id")
 def move():
     req = request.json
     try:
         file_ids = req["src_file_ids"]
-        parent_id = req["dest_file_id"]
+        dest_parent_id = req["dest_file_id"]
+
+        ok, dest_folder = FileService.get_by_id(dest_parent_id)
+        if not ok or not dest_folder:
+            return get_data_error_result(message="Parent folder not found!")
+
         files = FileService.get_by_ids(file_ids)
-        files_dict = {}
+        if not files:
-        for file in files:
+            return get_data_error_result(message="Source files not found!")
-            files_dict[file.id] = file
+
+        files_dict = {f.id: f for f in files}
 
         for file_id in file_ids:
-            file = files_dict[file_id]
+            file = files_dict.get(file_id)
             if not file:
-                return get_data_error_result(message="File or Folder not found!")
+                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!")
-        fe, _ = FileService.get_by_id(parent_id)
+            if not check_file_team_permission(file, current_user.id):
-        if not fe:
+                return get_json_result(
-            return get_data_error_result(message="Parent Folder not found!")
+                    data=False,
-        FileService.move_file(file_ids, parent_id)
+                    message="No authorization.",
+                    code=RetCode.AUTHENTICATION_ERROR,
+                )
+
+        def _move_entry_recursive(source_file_entry, dest_folder):
+            if source_file_entry.type == FileType.FOLDER.value:
+                existing_folder = FileService.query(name=source_file_entry.name, parent_id=dest_folder.id)
+                if existing_folder:
+                    new_folder = existing_folder[0]
+                else:
+                    new_folder = FileService.insert(
+                        {
+                            "id": get_uuid(),
+                            "parent_id": dest_folder.id,
+                            "tenant_id": source_file_entry.tenant_id,
+                            "created_by": current_user.id,
+                            "name": source_file_entry.name,
+                            "location": "",
+                            "size": 0,
+                            "type": FileType.FOLDER.value,
+                        }
+                    )
+
+                sub_files = FileService.list_all_files_by_parent_id(source_file_entry.id)
+                for sub_file in sub_files:
+                    _move_entry_recursive(sub_file, new_folder)
+
+                FileService.delete_by_id(source_file_entry.id)
+                return
+
+            old_parent_id = source_file_entry.parent_id
+            old_location = source_file_entry.location
+            filename = source_file_entry.name
+
+            new_location = filename
+            while 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)
+            except Exception as storage_err:
+                raise RuntimeError(f"Move file failed at storage layer: {str(storage_err)}")
+
+            FileService.update_by_id(
+                source_file_entry.id,
+                {
+                    "parent_id": dest_folder.id,
+                    "location": new_location,
+                },
+            )
+
+        for file in files:
+            _move_entry_recursive(file, dest_folder)
+
         return get_json_result(data=True)
+
     except Exception as e:
         return server_error_response(e)

api/apps/kb_app.py

@@ -14,60 +14,52 @@
 #  limitations under the License.
 #
 import json
+import logging
+import random
 
 from flask import request
 from flask_login import login_required, current_user
+import numpy as np
 
-from api.db.services import duplicate_name
+
-from api.db.services.document_service import DocumentService
+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
 from api.db.services.file2document_service import File2DocumentService
 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 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
-from api.utils import get_uuid
+from api.db import VALID_FILE_TYPES
-from api.db import StatusEnum, FileSource
 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 import globals
 
 @manager.route('/create', methods=['post'])  # noqa: F821
 @login_required
 @validate_request("name")
 def create():
     req = request.json
-    dataset_name = req["name"]
+    req = KnowledgebaseService.create_with_name(
-    if not isinstance(dataset_name, str):
+        name = req.pop("name", None),
-        return get_data_error_result(message="Dataset name must be string.")
+        tenant_id = current_user.id,
-    if dataset_name.strip() == "":
+        parser_id = req.pop("parser_id", None),
-        return get_data_error_result(message="Dataset name can't be empty.")
+        **req
-    if len(dataset_name.encode("utf-8")) > DATASET_NAME_LIMIT:
+    )        
-        return get_data_error_result(
-            message=f"Dataset name length is {len(dataset_name)} which is larger than {DATASET_NAME_LIMIT}")
 
-    dataset_name = dataset_name.strip()
-    dataset_name = duplicate_name(
-        KnowledgebaseService.query,
-        name=dataset_name,
-        tenant_id=current_user.id,
-        status=StatusEnum.VALID.value)
     try:
-        req["id"] = get_uuid()
-        req["name"] = dataset_name
-        req["tenant_id"] = current_user.id
-        req["created_by"] = current_user.id
-        e, t = TenantService.get_by_id(current_user.id)
-        if not e:
-            return get_data_error_result(message="Tenant not found.")
-        req["embd_id"] = t.embd_id
         if not KnowledgebaseService.save(**req):
             return get_data_error_result()
-        return get_json_result(data={"kb_id": req["id"]})
+        return get_json_result(data={"kb_id":req["id"]})
     except Exception as e:
         return server_error_response(e)
 
@@ -91,14 +83,14 @@ def update():
         return get_json_result(
             data=False,
             message='No authorization.',
-            code=settings.RetCode.AUTHENTICATION_ERROR
+            code=RetCode.AUTHENTICATION_ERROR
         )
     try:
         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.',
-                code=settings.RetCode.OPERATING_ERROR)
+                code=RetCode.OPERATING_ERROR)
 
         e, kb = KnowledgebaseService.get_by_id(req["kb_id"])
         if not e:
@@ -117,11 +109,11 @@ def update():
 
         if kb.pagerank != req.get("pagerank", 0):
             if req.get("pagerank", 0) > 0:
-                settings.docStoreConn.update({"kb_id": kb.id}, {PAGERANK_FLD: req["pagerank"]},
+                globals.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},
+                globals.docStoreConn.update({"exists": PAGERANK_FLD}, {"remove": PAGERANK_FLD},
                                          search.index_name(kb.tenant_id), kb.id)
 
         e, kb = KnowledgebaseService.get_by_id(kb.id)
@@ -149,12 +141,17 @@ def detail():
         else:
             return get_json_result(
                 data=False, message='Only owner of knowledgebase authorized for this operation.',
-                code=settings.RetCode.OPERATING_ERROR)
+                code=RetCode.OPERATING_ERROR)
         kb = KnowledgebaseService.get_detail(kb_id)
         if not kb:
             return get_data_error_result(
                 message="Can't find this knowledgebase!")
         kb["size"] = DocumentService.get_total_size_by_kb_id(kb_id=kb["id"],keywords="", run_status=[], types=[])
+        kb["connectors"] = Connector2KbService.list_connectors(kb_id)
+
+        for key in ["graphrag_task_finish_at", "raptor_task_finish_at", "mindmap_task_finish_at"]:
+            if finish_at := kb.get(key):
+                kb[key] = finish_at.strftime("%Y-%m-%d %H:%M:%S")
         return get_json_result(data=kb)
     except Exception as e:
         return server_error_response(e)
@@ -204,7 +201,7 @@ def rm():
         return get_json_result(
             data=False,
             message='No authorization.',
-            code=settings.RetCode.AUTHENTICATION_ERROR
+            code=RetCode.AUTHENTICATION_ERROR
         )
     try:
         kbs = KnowledgebaseService.query(
@@ -212,7 +209,7 @@ def rm():
         if not kbs:
             return get_json_result(
                 data=False, message='Only owner of knowledgebase authorized for this operation.',
-                code=settings.RetCode.OPERATING_ERROR)
+                code=RetCode.OPERATING_ERROR)
 
         for doc in DocumentService.query(kb_id=req["kb_id"]):
             if not DocumentService.remove_document(doc, kbs[0].tenant_id):
@@ -228,8 +225,8 @@ def rm():
             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)
+            globals.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)
+            globals.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)
@@ -244,13 +241,13 @@ def list_tags(kb_id):
         return get_json_result(
             data=False,
             message='No authorization.',
-            code=settings.RetCode.AUTHENTICATION_ERROR
+            code=RetCode.AUTHENTICATION_ERROR
         )
 
     tenants = UserTenantService.get_tenants_by_user_id(current_user.id)
     tags = []
     for tenant in tenants:
-        tags += settings.retrievaler.all_tags(tenant["tenant_id"], [kb_id])
+        tags += globals.retriever.all_tags(tenant["tenant_id"], [kb_id])
     return get_json_result(data=tags)
 
 
@@ -263,13 +260,13 @@ def list_tags_from_kbs():
             return get_json_result(
                 data=False,
                 message='No authorization.',
-                code=settings.RetCode.AUTHENTICATION_ERROR
+                code=RetCode.AUTHENTICATION_ERROR
             )
 
     tenants = UserTenantService.get_tenants_by_user_id(current_user.id)
     tags = []
     for tenant in tenants:
-        tags += settings.retrievaler.all_tags(tenant["tenant_id"], kb_ids)
+        tags += globals.retriever.all_tags(tenant["tenant_id"], kb_ids)
     return get_json_result(data=tags)
 
 
@@ -281,12 +278,12 @@ def rm_tags(kb_id):
         return get_json_result(
             data=False,
             message='No authorization.',
-            code=settings.RetCode.AUTHENTICATION_ERROR
+            code=RetCode.AUTHENTICATION_ERROR
         )
     e, kb = KnowledgebaseService.get_by_id(kb_id)
 
     for t in req["tags"]:
-        settings.docStoreConn.update({"tag_kwd": t, "kb_id": [kb_id]},
+        globals.docStoreConn.update({"tag_kwd": t, "kb_id": [kb_id]},
                                      {"remove": {"tag_kwd": t}},
                                      search.index_name(kb.tenant_id),
                                      kb_id)
@@ -301,11 +298,11 @@ def rename_tags(kb_id):
         return get_json_result(
             data=False,
             message='No authorization.',
-            code=settings.RetCode.AUTHENTICATION_ERROR
+            code=RetCode.AUTHENTICATION_ERROR
         )
     e, kb = KnowledgebaseService.get_by_id(kb_id)
 
-    settings.docStoreConn.update({"tag_kwd": req["from_tag"], "kb_id": [kb_id]},
+    globals.docStoreConn.update({"tag_kwd": req["from_tag"], "kb_id": [kb_id]},
                                      {"remove": {"tag_kwd": req["from_tag"].strip()}, "add": {"tag_kwd": req["to_tag"]}},
                                      search.index_name(kb.tenant_id),
                                      kb_id)
@@ -319,7 +316,7 @@ def knowledge_graph(kb_id):
         return get_json_result(
             data=False,
             message='No authorization.',
-            code=settings.RetCode.AUTHENTICATION_ERROR
+            code=RetCode.AUTHENTICATION_ERROR
         )
     _, kb = KnowledgebaseService.get_by_id(kb_id)
     req = {
@@ -328,9 +325,9 @@ def knowledge_graph(kb_id):
     }
 
     obj = {"graph": {}, "mind_map": {}}
-    if not settings.docStoreConn.indexExist(search.index_name(kb.tenant_id), kb_id):
+    if not globals.docStoreConn.indexExist(search.index_name(kb.tenant_id), kb_id):
         return get_json_result(data=obj)
-    sres = settings.retrievaler.search(req, search.index_name(kb.tenant_id), [kb_id])
+    sres = globals.retriever.search(req, search.index_name(kb.tenant_id), [kb_id])
     if not len(sres.ids):
         return get_json_result(data=obj)
 
@@ -359,10 +356,10 @@ def delete_knowledge_graph(kb_id):
         return get_json_result(
             data=False,
             message='No authorization.',
-            code=settings.RetCode.AUTHENTICATION_ERROR
+            code=RetCode.AUTHENTICATION_ERROR
         )
     _, kb = KnowledgebaseService.get_by_id(kb_id)
-    settings.docStoreConn.delete({"knowledge_graph_kwd": ["graph", "subgraph", "entity", "relation"]}, search.index_name(kb.tenant_id), kb_id)
+    globals.docStoreConn.delete({"knowledge_graph_kwd": ["graph", "subgraph", "entity", "relation"]}, search.index_name(kb.tenant_id), kb_id)
 
     return get_json_result(data=True)
 
@@ -376,6 +373,532 @@ def get_meta():
             return get_json_result(
                 data=False,
                 message='No authorization.',
-                code=settings.RetCode.AUTHENTICATION_ERROR
+                code=RetCode.AUTHENTICATION_ERROR
             )
     return get_json_result(data=DocumentService.get_meta_by_kbs(kb_ids))
+
+
+@manager.route("/basic_info", methods=["GET"])  # noqa: F821
+@login_required
+def get_basic_info():
+    kb_id = request.args.get("kb_id", "")
+    if not KnowledgebaseService.accessible(kb_id, current_user.id):
+        return get_json_result(
+            data=False,
+            message='No authorization.',
+            code=RetCode.AUTHENTICATION_ERROR
+        )
+
+    basic_info = DocumentService.knowledgebase_basic_info(kb_id)
+
+    return get_json_result(data=basic_info)
+
+
+@manager.route("/list_pipeline_logs", methods=["POST"])  # noqa: F821
+@login_required
+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)
+
+    keywords = request.args.get("keywords", "")
+
+    page_number = int(request.args.get("page", 0))
+    items_per_page = int(request.args.get("page_size", 0))
+    orderby = request.args.get("orderby", "create_time")
+    if request.args.get("desc", "true").lower() == "false":
+        desc = False
+    else:
+        desc = True
+    create_date_from = request.args.get("create_date_from", "")
+    create_date_to = request.args.get("create_date_to", "")
+    if create_date_to > create_date_from:
+        return get_data_error_result(message="Create data filter is abnormal.")
+
+    req = request.get_json()
+
+    operation_status = req.get("operation_status", [])
+    if operation_status:
+        invalid_status = {s for s in operation_status if s not in VALID_TASK_STATUS}
+        if invalid_status:
+            return get_data_error_result(message=f"Invalid filter operation_status 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 ''}")
+
+    suffix = req.get("suffix", [])
+
+    try:
+        logs, tol = PipelineOperationLogService.get_file_logs_by_kb_id(kb_id, page_number, items_per_page, orderby, desc, keywords, operation_status, types, suffix, create_date_from, create_date_to)
+        return get_json_result(data={"total": tol, "logs": logs})
+    except Exception as e:
+        return server_error_response(e)
+
+
+@manager.route("/list_pipeline_dataset_logs", methods=["POST"])  # noqa: F821
+@login_required
+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)
+
+    page_number = int(request.args.get("page", 0))
+    items_per_page = int(request.args.get("page_size", 0))
+    orderby = request.args.get("orderby", "create_time")
+    if request.args.get("desc", "true").lower() == "false":
+        desc = False
+    else:
+        desc = True
+    create_date_from = request.args.get("create_date_from", "")
+    create_date_to = request.args.get("create_date_to", "")
+    if create_date_to > create_date_from:
+        return get_data_error_result(message="Create data filter is abnormal.")
+
+    req = request.get_json()
+
+    operation_status = req.get("operation_status", [])
+    if operation_status:
+        invalid_status = {s for s in operation_status if s not in VALID_TASK_STATUS}
+        if invalid_status:
+            return get_data_error_result(message=f"Invalid filter operation_status status conditions: {', '.join(invalid_status)}")
+
+    try:
+        logs, tol = PipelineOperationLogService.get_dataset_logs_by_kb_id(kb_id, page_number, items_per_page, orderby, desc, operation_status, create_date_from, create_date_to)
+        return get_json_result(data={"total": tol, "logs": logs})
+    except Exception as e:
+        return server_error_response(e)
+
+
+@manager.route("/delete_pipeline_logs", methods=["POST"])  # noqa: F821
+@login_required
+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()
+    log_ids = req.get("log_ids", [])
+
+    PipelineOperationLogService.delete_by_ids(log_ids)
+
+    return get_json_result(data=True)
+
+
+@manager.route("/pipeline_log_detail", methods=["GET"])  # noqa: F821
+@login_required
+def pipeline_log_detail():
+    log_id = request.args.get("log_id")
+    if not log_id:
+        return get_json_result(data=False, message='Lack of "Pipeline log ID"', code=RetCode.ARGUMENT_ERROR)
+
+    ok, log = PipelineOperationLogService.get_by_id(log_id)
+    if not ok:
+        return get_data_error_result(message="Invalid pipeline log ID")
+
+    return get_json_result(data=log.to_dict())
+
+
+@manager.route("/run_graphrag", methods=["POST"])  # noqa: F821
+@login_required
+def run_graphrag():
+    req = request.json
+
+    kb_id = req.get("kb_id", "")
+    if not kb_id:
+        return get_error_data_result(message='Lack of "KB ID"')
+
+    ok, kb = KnowledgebaseService.get_by_id(kb_id)
+    if not ok:
+        return get_error_data_result(message="Invalid Knowledgebase ID")
+
+    task_id = kb.graphrag_task_id
+    if task_id:
+        ok, task = TaskService.get_by_id(task_id)
+        if not ok:
+            logging.warning(f"A valid GraphRAG task id is expected for kb {kb_id}")
+
+        if task and task.progress not in [-1, 1]:
+            return get_error_data_result(message=f"Task {task_id} in progress with status {task.progress}. A Graph Task is already running.")
+
+    documents, _ = DocumentService.get_by_kb_id(
+        kb_id=kb_id,
+        page_number=0,
+        items_per_page=0,
+        orderby="create_time",
+        desc=False,
+        keywords="",
+        run_status=[],
+        types=[],
+        suffix=[],
+    )
+    if not documents:
+        return get_error_data_result(message=f"No documents in Knowledgebase {kb_id}")
+
+    sample_document = documents[0]
+    document_ids = [document["id"] for document in documents]
+
+    task_id = queue_raptor_o_graphrag_tasks(sample_doc_id=sample_document, ty="graphrag", priority=0, fake_doc_id=GRAPH_RAPTOR_FAKE_DOC_ID, doc_ids=list(document_ids))
+
+    if not KnowledgebaseService.update_by_id(kb.id, {"graphrag_task_id": task_id}):
+        logging.warning(f"Cannot save graphrag_task_id for kb {kb_id}")
+
+    return get_json_result(data={"graphrag_task_id": task_id})
+
+
+@manager.route("/trace_graphrag", methods=["GET"])  # noqa: F821
+@login_required
+def trace_graphrag():
+    kb_id = request.args.get("kb_id", "")
+    if not kb_id:
+        return get_error_data_result(message='Lack of "KB ID"')
+
+    ok, kb = KnowledgebaseService.get_by_id(kb_id)
+    if not ok:
+        return get_error_data_result(message="Invalid Knowledgebase ID")
+
+    task_id = kb.graphrag_task_id
+    if not task_id:
+        return get_json_result(data={})
+
+    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=task.to_dict())
+
+
+@manager.route("/run_raptor", methods=["POST"])  # noqa: F821
+@login_required
+def run_raptor():
+    req = request.json
+
+    kb_id = req.get("kb_id", "")
+    if not kb_id:
+        return get_error_data_result(message='Lack of "KB ID"')
+
+    ok, kb = KnowledgebaseService.get_by_id(kb_id)
+    if not ok:
+        return get_error_data_result(message="Invalid Knowledgebase ID")
+
+    task_id = kb.raptor_task_id
+    if task_id:
+        ok, task = TaskService.get_by_id(task_id)
+        if not ok:
+            logging.warning(f"A valid RAPTOR task id is expected for kb {kb_id}")
+
+        if task and task.progress not in [-1, 1]:
+            return get_error_data_result(message=f"Task {task_id} in progress with status {task.progress}. A RAPTOR Task is already running.")
+
+    documents, _ = DocumentService.get_by_kb_id(
+        kb_id=kb_id,
+        page_number=0,
+        items_per_page=0,
+        orderby="create_time",
+        desc=False,
+        keywords="",
+        run_status=[],
+        types=[],
+        suffix=[],
+    )
+    if not documents:
+        return get_error_data_result(message=f"No documents in Knowledgebase {kb_id}")
+
+    sample_document = documents[0]
+    document_ids = [document["id"] for document in documents]
+
+    task_id = queue_raptor_o_graphrag_tasks(sample_doc_id=sample_document, ty="raptor", priority=0, fake_doc_id=GRAPH_RAPTOR_FAKE_DOC_ID, doc_ids=list(document_ids))
+
+    if not KnowledgebaseService.update_by_id(kb.id, {"raptor_task_id": task_id}):
+        logging.warning(f"Cannot save raptor_task_id for kb {kb_id}")
+
+    return get_json_result(data={"raptor_task_id": task_id})
+
+
+@manager.route("/trace_raptor", methods=["GET"])  # noqa: F821
+@login_required
+def trace_raptor():
+    kb_id = request.args.get("kb_id", "")
+    if not kb_id:
+        return get_error_data_result(message='Lack of "KB ID"')
+
+    ok, kb = KnowledgebaseService.get_by_id(kb_id)
+    if not ok:
+        return get_error_data_result(message="Invalid Knowledgebase ID")
+
+    task_id = kb.raptor_task_id
+    if not task_id:
+        return get_json_result(data={})
+
+    ok, task = TaskService.get_by_id(task_id)
+    if not ok:
+        return get_error_data_result(message="RAPTOR Task Not Found or Error Occurred")
+
+    return get_json_result(data=task.to_dict())
+
+
+@manager.route("/run_mindmap", methods=["POST"])  # noqa: F821
+@login_required
+def run_mindmap():
+    req = request.json
+
+    kb_id = req.get("kb_id", "")
+    if not kb_id:
+        return get_error_data_result(message='Lack of "KB ID"')
+
+    ok, kb = KnowledgebaseService.get_by_id(kb_id)
+    if not ok:
+        return get_error_data_result(message="Invalid Knowledgebase ID")
+
+    task_id = kb.mindmap_task_id
+    if task_id:
+        ok, task = TaskService.get_by_id(task_id)
+        if not ok:
+            logging.warning(f"A valid Mindmap task id is expected for kb {kb_id}")
+
+        if task and task.progress not in [-1, 1]:
+            return get_error_data_result(message=f"Task {task_id} in progress with status {task.progress}. A Mindmap Task is already running.")
+
+    documents, _ = DocumentService.get_by_kb_id(
+        kb_id=kb_id,
+        page_number=0,
+        items_per_page=0,
+        orderby="create_time",
+        desc=False,
+        keywords="",
+        run_status=[],
+        types=[],
+        suffix=[],
+    )
+    if not documents:
+        return get_error_data_result(message=f"No documents in Knowledgebase {kb_id}")
+
+    sample_document = documents[0]
+    document_ids = [document["id"] for document in documents]
+
+    task_id = queue_raptor_o_graphrag_tasks(sample_doc_id=sample_document, ty="mindmap", priority=0, fake_doc_id=GRAPH_RAPTOR_FAKE_DOC_ID, doc_ids=list(document_ids))
+
+    if not KnowledgebaseService.update_by_id(kb.id, {"mindmap_task_id": task_id}):
+        logging.warning(f"Cannot save mindmap_task_id for kb {kb_id}")
+
+    return get_json_result(data={"mindmap_task_id": task_id})
+
+
+@manager.route("/trace_mindmap", methods=["GET"])  # noqa: F821
+@login_required
+def trace_mindmap():
+    kb_id = request.args.get("kb_id", "")
+    if not kb_id:
+        return get_error_data_result(message='Lack of "KB ID"')
+
+    ok, kb = KnowledgebaseService.get_by_id(kb_id)
+    if not ok:
+        return get_error_data_result(message="Invalid Knowledgebase ID")
+
+    task_id = kb.mindmap_task_id
+    if not task_id:
+        return get_json_result(data={})
+
+    ok, task = TaskService.get_by_id(task_id)
+    if not ok:
+        return get_error_data_result(message="Mindmap Task Not Found or Error Occurred")
+
+    return get_json_result(data=task.to_dict())
+
+
+@manager.route("/unbind_task", methods=["DELETE"])  # noqa: F821
+@login_required
+def delete_kb_task():
+    kb_id = request.args.get("kb_id", "")
+    if not kb_id:
+        return get_error_data_result(message='Lack of "KB ID"')
+    ok, kb = KnowledgebaseService.get_by_id(kb_id)
+    if not ok:
+        return get_json_result(data=True)
+
+    pipeline_task_type = request.args.get("pipeline_task_type", "")
+    if not pipeline_task_type or pipeline_task_type not in [PipelineTaskType.GRAPH_RAG, PipelineTaskType.RAPTOR, PipelineTaskType.MINDMAP]:
+        return get_error_data_result(message="Invalid task type")
+
+    def cancel_task(task_id):
+        REDIS_CONN.set(f"{task_id}-cancel", "x")
+
+    match pipeline_task_type:
+        case PipelineTaskType.GRAPH_RAG:
+            kb_task_id_field = "graphrag_task_id"
+            task_id = kb.graphrag_task_id
+            kb_task_finish_at = "graphrag_task_finish_at"
+            cancel_task(task_id)
+            globals.docStoreConn.delete({"knowledge_graph_kwd": ["graph", "subgraph", "entity", "relation"]}, search.index_name(kb.tenant_id), kb_id)
+        case PipelineTaskType.RAPTOR:
+            kb_task_id_field = "raptor_task_id"
+            task_id = kb.raptor_task_id
+            kb_task_finish_at = "raptor_task_finish_at"
+            cancel_task(task_id)
+            globals.docStoreConn.delete({"raptor_kwd": ["raptor"]}, search.index_name(kb.tenant_id), kb_id)
+        case PipelineTaskType.MINDMAP:
+            kb_task_id_field = "mindmap_task_id"
+            task_id = kb.mindmap_task_id
+            kb_task_finish_at = "mindmap_task_finish_at"
+            cancel_task(task_id)
+        case _:
+            return get_error_data_result(message="Internal Error: Invalid task type")
+
+
+    ok = KnowledgebaseService.update_by_id(kb_id, {kb_task_id_field: "", kb_task_finish_at: None})
+    if not ok:
+        return server_error_response(f"Internal error: cannot delete task {pipeline_task_type}")
+
+    return get_json_result(data=True)
+
+@manager.route("/check_embedding", methods=["post"])  # noqa: F821
+@login_required
+def check_embedding():
+
+    def _guess_vec_field(src: dict) -> str | None:
+        for k in src or {}:
+            if k.endswith("_vec"):
+                return k
+        return None
+
+    def _as_float_vec(v):
+        if v is None:
+            return []
+        if isinstance(v, str):
+            return [float(x) for x in v.split("\t") if x != ""]
+        if isinstance(v, (list, tuple, np.ndarray)):
+            return [float(x) for x in v]
+        return []
+
+    def _to_1d(x):
+        a = np.asarray(x, dtype=np.float32)
+        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: 
+            return 0.0
+        return float(np.dot(a, b) / (na * nb))
+
+    def sample_random_chunks_with_vectors(
+        docStoreConn,
+        tenant_id: str,
+        kb_id: str,
+        n: int = 5,
+        base_fields=("docnm_kwd","doc_id","content_with_weight","page_num_int","position_int","top_int"),
+    ):
+        index_nm = search.index_name(tenant_id)
+
+        res0 = docStoreConn.search(
+            selectFields=[], highlightFields=[],
+            condition={"kb_id": kb_id, "available_int": 1},
+            matchExprs=[], orderBy=OrderByExpr(),
+            offset=0, limit=1,
+            indexNames=index_nm, knowledgebaseIds=[kb_id]
+        )
+        total = docStoreConn.getTotal(res0)
+        if total <= 0:
+            return []
+
+        n = min(n, total)
+        offsets = sorted(random.sample(range(total), n))
+        out = []
+
+        for off in offsets:
+            res1 = docStoreConn.search(
+                selectFields=list(base_fields),
+                highlightFields=[],
+                condition={"kb_id": kb_id, "available_int": 1},
+                matchExprs=[], orderBy=OrderByExpr(),
+                offset=off, limit=1,
+                indexNames=index_nm, knowledgebaseIds=[kb_id]
+            )
+            ids = docStoreConn.getChunkIds(res1)
+            if not ids: 
+                continue
+
+            cid = ids[0]
+            full_doc = docStoreConn.get(cid, index_nm, [kb_id]) or {}
+            vec_field = _guess_vec_field(full_doc)
+            vec = _as_float_vec(full_doc.get(vec_field))
+
+            out.append({
+                "chunk_id": cid,
+                "kb_id": kb_id,
+                "doc_id": full_doc.get("doc_id"),
+                "doc_name": full_doc.get("docnm_kwd"),
+                "vector_field": vec_field,
+                "vector_dim": len(vec),
+                "vector": vec,
+                "page_num_int": full_doc.get("page_num_int"),
+                "position_int": full_doc.get("position_int"),
+                "top_int": full_doc.get("top_int"),
+                "content_with_weight": full_doc.get("content_with_weight") or "",
+            })
+        return out
+    req = request.json
+    kb_id = req.get("kb_id", "")
+    embd_id = req.get("embd_id", "")
+    n = int(req.get("check_num", 5))
+    _, kb = KnowledgebaseService.get_by_id(kb_id)
+    tenant_id = kb.tenant_id
+
+    emb_mdl = LLMBundle(tenant_id, LLMType.EMBEDDING, embd_id)
+    samples = sample_random_chunks_with_vectors(globals.docStoreConn, tenant_id=tenant_id, kb_id=kb_id, n=n)
+
+    results, eff_sims = [], []
+    for ck in samples:
+        txt = (ck.get("content_with_weight") or "").strip()
+        if not txt:
+            results.append({"chunk_id": ck["chunk_id"], "reason": "no_text"})
+            continue
+
+        if not ck.get("vector"):
+            results.append({"chunk_id": ck["chunk_id"], "reason": "no_stored_vector"})
+            continue
+
+        try:
+            qv, _ = emb_mdl.encode_queries(txt)  
+            sim = _cos_sim(qv, ck["vector"])
+        except Exception:
+            return get_error_data_result(message="embedding failure")
+
+        eff_sims.append(sim)
+        results.append({
+            "chunk_id": ck["chunk_id"],
+            "doc_id": ck["doc_id"],
+            "doc_name": ck["doc_name"],
+            "vector_field": ck["vector_field"],
+            "vector_dim": ck["vector_dim"],
+            "cos_sim": round(sim, 6),
+        })
+
+    summary = {
+        "kb_id": kb_id,
+        "model": embd_id,
+        "sampled": len(samples),
+        "valid": len(eff_sims),
+        "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),
+    }
+    if summary["avg_cos_sim"] > 0.99:
+        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)
+

api/apps/llm_app.py

@@ -15,24 +15,24 @@
 #
 import logging
 import json
+import os
 from flask import request
 from flask_login 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 import settings
 from api.utils.api_utils import server_error_response, get_data_error_result, validate_request
-from api.db import StatusEnum, LLMType
+from common.constants import StatusEnum, LLMType
 from api.db.db_models import TenantLLM
-from api.utils.api_utils import get_json_result
+from api.utils.api_utils import get_json_result, get_allowed_llm_factories
-from api.utils.base64_image import test_image
+from rag.utils.base64_image import test_image
 from rag.llm import EmbeddingModel, ChatModel, RerankModel, CvModel, TTSModel
 
 
-@manager.route('/factories', methods=['GET'])  # noqa: F821
+@manager.route("/factories", methods=["GET"])  # noqa: F821
 @login_required
 def factories():
     try:
-        fac = LLMFactoriesService.get_all()
+        fac = get_allowed_llm_factories()
         fac = [f.to_dict() for f in fac if f.name not in ["Youdao", "FastEmbed", "BAAI"]]
         llms = LLMService.get_all()
         mdl_types = {}
@@ -43,14 +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,
+            f["model_types"] = list(mdl_types.get(f["name"], [LLMType.CHAT, LLMType.EMBEDDING, LLMType.RERANK, LLMType.IMAGE2TEXT, LLMType.SPEECH2TEXT, LLMType.TTS]))
-                                                              LLMType.IMAGE2TEXT, LLMType.SPEECH2TEXT, LLMType.TTS]))
         return get_json_result(data=fac)
     except Exception as e:
         return server_error_response(e)
 
 
-@manager.route('/set_api_key', methods=['POST'])  # noqa: F821
+@manager.route("/set_api_key", methods=["POST"])  # noqa: F821
 @login_required
 @validate_request("llm_factory", "api_key")
 def set_api_key():
@@ -63,8 +62,7 @@ def set_api_key():
     for llm in LLMService.query(fid=factory):
         if not embd_passed and llm.model_type == LLMType.EMBEDDING.value:
             assert factory in EmbeddingModel, f"Embedding model from {factory} is not supported yet."
-            mdl = EmbeddingModel[factory](
+            mdl = EmbeddingModel[factory](req["api_key"], llm.llm_name, base_url=req.get("base_url"))
-                req["api_key"], llm.llm_name, base_url=req.get("base_url"))
             try:
                 arr, tc = mdl.encode(["Test if the api key is available"])
                 if len(arr[0]) == 0:
@@ -74,52 +72,40 @@ def set_api_key():
                 msg += f"\nFail to access embedding model({llm.llm_name}) using this api key." + str(e)
         elif not chat_passed and llm.model_type == LLMType.CHAT.value:
             assert factory in ChatModel, f"Chat model from {factory} is not supported yet."
-            mdl = ChatModel[factory](
+            mdl = ChatModel[factory](req["api_key"], llm.llm_name, base_url=req.get("base_url"), **extra)
-                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!"}],
+                m, tc = mdl.chat(None, [{"role": "user", "content": "Hello! How are you doing!"}], {"temperature": 0.9, "max_tokens": 50})
-                                 {"temperature": 0.9, 'max_tokens': 50})
                 if m.find("**ERROR**") >= 0:
                     raise Exception(m)
                 chat_passed = True
             except Exception as e:
-                msg += f"\nFail to access model({llm.fid}/{llm.llm_name}) using this api key." + str(
+                msg += f"\nFail to access model({llm.fid}/{llm.llm_name}) using this api key." + str(e)
-                    e)
         elif not rerank_passed and llm.model_type == LLMType.RERANK:
             assert factory in RerankModel, f"Re-rank model from {factory} is not supported yet."
-            mdl = RerankModel[factory](
+            mdl = RerankModel[factory](req["api_key"], llm.llm_name, base_url=req.get("base_url"))
-                req["api_key"], llm.llm_name, base_url=req.get("base_url"))
             try:
                 arr, tc = mdl.similarity("What's the weather?", ["Is it sunny today?"])
                 if len(arr) == 0 or tc == 0:
                     raise Exception("Fail")
                 rerank_passed = True
-                logging.debug(f'passed model rerank {llm.llm_name}')
+                logging.debug(f"passed model rerank {llm.llm_name}")
             except Exception as e:
-                msg += f"\nFail to access model({llm.fid}/{llm.llm_name}) using this api key." + str(
+                msg += f"\nFail to access model({llm.fid}/{llm.llm_name}) using this api key." + str(e)
-                    e)
         if any([embd_passed, chat_passed, rerank_passed]):
-            msg = ''
+            msg = ""
             break
 
     if msg:
         return get_data_error_result(message=msg)
 
-    llm_config = {
+    llm_config = {"api_key": req["api_key"], "api_base": req.get("base_url", "")}
-        "api_key": req["api_key"],
-        "api_base": req.get("base_url", "")
-    }
     for n in ["model_type", "llm_name"]:
         if n in req:
             llm_config[n] = req[n]
 
     for llm in LLMService.query(fid=factory):
-        llm_config["max_tokens"]=llm.max_tokens
+        llm_config["max_tokens"] = llm.max_tokens
-        if not TenantLLMService.filter_update(
+        if not TenantLLMService.filter_update([TenantLLM.tenant_id == current_user.id, TenantLLM.llm_factory == factory, TenantLLM.llm_name == llm.llm_name], llm_config):
-                [TenantLLM.tenant_id == current_user.id,
-                 TenantLLM.llm_factory == factory,
-                 TenantLLM.llm_name == llm.llm_name],
-                llm_config):
             TenantLLMService.save(
                 tenant_id=current_user.id,
                 llm_factory=factory,
@@ -127,13 +113,13 @@ def set_api_key():
                 model_type=llm.model_type,
                 api_key=llm_config["api_key"],
                 api_base=llm_config["api_base"],
-                max_tokens=llm_config["max_tokens"]
+                max_tokens=llm_config["max_tokens"],
             )
 
     return get_json_result(data=True)
 
 
-@manager.route('/add_llm', methods=['POST'])  # noqa: F821
+@manager.route("/add_llm", methods=["POST"])  # noqa: F821
 @login_required
 @validate_request("llm_factory")
 def add_llm():
@@ -142,6 +128,9 @@ def add_llm():
     api_key = req.get("api_key", "x")
     llm_name = req.get("llm_name")
 
+    if factory not in get_allowed_llm_factories():
+        return get_data_error_result(message=f"LLM factory {factory} is not allowed")
+
     def apikey_json(keys):
         nonlocal req
         return json.dumps({k: req.get(k, "") for k in keys})
@@ -194,6 +183,9 @@ def add_llm():
     elif factory == "Azure-OpenAI":
         api_key = apikey_json(["api_key", "api_version"])
 
+    elif factory == "OpenRouter":
+        api_key = apikey_json(["api_key", "provider_order"])
+
     llm = {
         "tenant_id": current_user.id,
         "llm_factory": factory,
@@ -201,7 +193,7 @@ def add_llm():
         "llm_name": llm_name,
         "api_base": req.get("api_base", ""),
         "api_key": api_key,
-        "max_tokens": req.get("max_tokens")
+        "max_tokens": req.get("max_tokens"),
     }
 
     msg = ""
@@ -209,10 +201,7 @@ def add_llm():
     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](
+        mdl = EmbeddingModel[factory](key=llm["api_key"], model_name=mdl_nm, base_url=llm["api_base"])
-            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:
@@ -222,54 +211,41 @@ def add_llm():
     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'],
+            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!"}], {
+            m, tc = mdl.chat(None, [{"role": "user", "content": "Hello! How are you doing!"}], {"temperature": 0.9})
-                "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(
+            msg += f"\nFail to access model({factory}/{mdl_nm})." + str(e)
-                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](
+            mdl = RerankModel[factory](key=llm["api_key"], model_name=mdl_nm, base_url=llm["api_base"])
-                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(
+            msg += f"\nFail to access model({factory}/{mdl_nm})." + str(e)
-                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](
+        mdl = CvModel[factory](key=llm["api_key"], model_name=mdl_nm, base_url=llm["api_base"])
-            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 m and not tc:
+            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](
+        mdl = TTSModel[factory](key=llm["api_key"], model_name=mdl_nm, base_url=llm["api_base"])
-            key=llm["api_key"], model_name=mdl_nm, base_url=llm["api_base"]
-        )
         try:
             for resp in mdl.tts("Hello~ RAGFlower!"):
                 pass
@@ -282,40 +258,46 @@ def add_llm():
     if msg:
         return get_data_error_result(message=msg)
 
-    if not TenantLLMService.filter_update(
+    if not TenantLLMService.filter_update([TenantLLM.tenant_id == current_user.id, TenantLLM.llm_factory == factory, TenantLLM.llm_name == llm["llm_name"]], llm):
-            [TenantLLM.tenant_id == current_user.id, TenantLLM.llm_factory == factory,
-             TenantLLM.llm_name == llm["llm_name"]], llm):
         TenantLLMService.save(**llm)
 
     return get_json_result(data=True)
 
 
-@manager.route('/delete_llm', methods=['POST'])  # noqa: F821
+@manager.route("/delete_llm", methods=["POST"])  # noqa: F821
 @login_required
 @validate_request("llm_factory", "llm_name")
 def delete_llm():
     req = request.json
-    TenantLLMService.filter_delete(
+    TenantLLMService.filter_delete([TenantLLM.tenant_id == current_user.id, TenantLLM.llm_factory == req["llm_factory"], TenantLLM.llm_name == req["llm_name"]])
-        [TenantLLM.tenant_id == current_user.id, TenantLLM.llm_factory == req["llm_factory"],
-         TenantLLM.llm_name == req["llm_name"]])
     return get_json_result(data=True)
 
 
-@manager.route('/delete_factory', methods=['POST'])  # noqa: F821
+@manager.route("/enable_llm", methods=["POST"])  # noqa: F821
+@login_required
+@validate_request("llm_factory", "llm_name")
+def enable_llm():
+    req = 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"))}
+    )
+    return get_json_result(data=True)
+
+
+@manager.route("/delete_factory", methods=["POST"])  # noqa: F821
 @login_required
 @validate_request("llm_factory")
 def delete_factory():
     req = request.json
-    TenantLLMService.filter_delete(
+    TenantLLMService.filter_delete([TenantLLM.tenant_id == current_user.id, TenantLLM.llm_factory == req["llm_factory"]])
-        [TenantLLM.tenant_id == current_user.id, TenantLLM.llm_factory == req["llm_factory"]])
     return get_json_result(data=True)
 
 
-@manager.route('/my_llms', methods=['GET'])  # noqa: F821
+@manager.route("/my_llms", methods=["GET"])  # noqa: F821
 @login_required
 def my_llms():
     try:
-        include_details = request.args.get('include_details', 'false').lower() == 'true'
+        include_details = request.args.get("include_details", "false").lower() == "true"
 
         if include_details:
             res = {}
@@ -331,51 +313,46 @@ def my_llms():
                         break
 
                 if o_dict["llm_factory"] not in res:
-                    res[o_dict["llm_factory"]] = {
+                    res[o_dict["llm_factory"]] = {"tags": factory_tags, "llm": []}
-                        "tags": factory_tags,
-                        "llm": []
-                    }
 
-                res[o_dict["llm_factory"]]["llm"].append({
+                res[o_dict["llm_factory"]]["llm"].append(
-                    "type": o_dict["model_type"],
+                    {
-                    "name": o_dict["llm_name"],
+                        "type": o_dict["model_type"],
-                    "used_token": o_dict["used_tokens"],
+                        "name": o_dict["llm_name"],
-                    "api_base": o_dict["api_base"] or "",
+                        "used_token": o_dict["used_tokens"],
-                    "max_tokens": o_dict["max_tokens"] or 8192
+                        "api_base": o_dict["api_base"] or "",
-                })
+                        "max_tokens": o_dict["max_tokens"] or 8192,
+                        "status": o_dict["status"] or "1",
+                    }
+                )
         else:
             res = {}
             for o in TenantLLMService.get_my_llms(current_user.id):
                 if o["llm_factory"] not in res:
-                    res[o["llm_factory"]] = {
+                    res[o["llm_factory"]] = {"tags": o["tags"], "llm": []}
-                        "tags": o["tags"],
+                res[o["llm_factory"]]["llm"].append({"type": o["model_type"], "name": o["llm_name"], "used_token": o["used_tokens"], "status": o["status"]})
-                        "llm": []
-                    }
-                res[o["llm_factory"]]["llm"].append({
-                    "type": o["model_type"],
-                    "name": o["llm_name"],
-                    "used_token": o["used_tokens"]
-                })
 
         return get_json_result(data=res)
     except Exception as e:
         return server_error_response(e)
 
 
-@manager.route('/list', methods=['GET'])  # noqa: F821
+@manager.route("/list", methods=["GET"])  # noqa: F821
 @login_required
 def list_app():
-    self_deployed = ["Youdao", "FastEmbed", "BAAI", "Ollama", "Xinference", "LocalAI", "LM-Studio", "GPUStack"]
+    self_deployed = ["FastEmbed", "Ollama", "Xinference", "LocalAI", "LM-Studio", "GPUStack"]
-    weighted = ["Youdao", "FastEmbed", "BAAI"] if settings.LIGHTEN != 0 else []
+    weighted = []
     model_type = request.args.get("model_type")
     try:
         objs = TenantLLMService.query(tenant_id=current_user.id)
-        facts = set([o.to_dict()["llm_factory"] for o in objs if o.api_key])
+        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()
+        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]
-                for m in llms if m.status == StatusEnum.VALID.value and m.fid not in weighted]
         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", ""):
+                m["available"] = True
 
         llm_set = set([m["llm_name"] + "@" + m["fid"] for m in llms])
         for o in objs:

api/apps/mcp_server_app.py

@@ -16,13 +16,12 @@
 from flask import Response, request
 from flask_login import current_user, login_required
 
-from api.db import VALID_MCP_SERVER_TYPES
 from api.db.db_models import MCPServer
 from api.db.services.mcp_server_service import MCPServerService
 from api.db.services.user_service import TenantService
-from api.settings import RetCode
+from common.constants import RetCode, VALID_MCP_SERVER_TYPES
 
-from api.utils import get_uuid
+from common.misc_utils import get_uuid
 from api.utils.api_utils import get_data_error_result, get_json_result, server_error_response, validate_request, \
     get_mcp_tools
 from api.utils.web_utils import get_float, safe_json_parse

api/apps/plugin_app.py

@@ -1,8 +1,26 @@
+#
+#  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.
+#
+
+
 from flask import Response
 from flask_login import login_required
 from api.utils.api_utils import get_json_result
 from plugin import GlobalPluginManager
 
+
 @manager.route('/llm_tools', methods=['GET'])  # noqa: F821
 @login_required
 def llm_tools() -> Response:

api/apps/sdk/agent.py

@@ -19,12 +19,13 @@ import time
 from typing import Any, cast
 from api.db.services.canvas_service import UserCanvasService
 from api.db.services.user_canvas_version import UserCanvasVersionService
-from api.settings import RetCode
+from common.constants import RetCode
-from api.utils import get_uuid
+from common.misc_utils import get_uuid
 from api.utils.api_utils import get_data_error_result, get_error_data_result, get_json_result, token_required
 from api.utils.api_utils import get_result
 from flask import request
 
+
 @manager.route('/agents', methods=['GET'])  # noqa: F821
 @token_required
 def list_agents(tenant_id):
@@ -41,7 +42,7 @@ def list_agents(tenant_id):
         desc = False
     else:
         desc = True
-    canvas = UserCanvasService.get_list(tenant_id,page_number,items_per_page,orderby,desc,id,title)
+    canvas = UserCanvasService.get_list(tenant_id, page_number, items_per_page, orderby, desc, id, title)
     return get_result(data=canvas)
 
 
@@ -93,7 +94,7 @@ def update_agent(tenant_id: str, agent_id: str):
             req["dsl"] = json.dumps(req["dsl"], ensure_ascii=False)
 
         req["dsl"] = json.loads(req["dsl"])
-    
+
     if req.get("title") is not None:
         req["title"] = req["title"].strip()
 

api/apps/sdk/chat.py

@@ -17,13 +17,12 @@ import logging
 
 from flask import request
 
-from api import settings
-from api.db import StatusEnum
 from api.db.services.dialog_service import DialogService
 from api.db.services.knowledgebase_service import KnowledgebaseService
 from api.db.services.tenant_llm_service import TenantLLMService
 from api.db.services.user_service import TenantService
-from api.utils import get_uuid
+from common.misc_utils import get_uuid
+from common.constants import RetCode, StatusEnum
 from api.utils.api_utils import check_duplicate_ids, get_error_data_result, get_result, token_required
 
 
@@ -45,7 +44,7 @@ def create(tenant_id):
     embd_ids = [TenantLLMService.split_model_name_and_factory(kb.embd_id)[0] for kb in kbs]  # remove vendor suffix for comparison
     embd_count = list(set(embd_ids))
     if len(embd_count) > 1:
-        return get_result(message='Datasets use different embedding models."', code=settings.RetCode.AUTHENTICATION_ERROR)
+        return get_result(message='Datasets use different embedding models."', code=RetCode.AUTHENTICATION_ERROR)
     req["kb_ids"] = ids
     # llm
     llm = req.get("llm")
@@ -167,8 +166,10 @@ def update(tenant_id, chat_id):
         embd_ids = [TenantLLMService.split_model_name_and_factory(kb.embd_id)[0] for kb in kbs]  # remove vendor suffix for comparison
         embd_count = list(set(embd_ids))
         if len(embd_count) > 1:
-            return get_result(message='Datasets use different embedding models."', code=settings.RetCode.AUTHENTICATION_ERROR)
+            return get_result(message='Datasets use different embedding models."', code=RetCode.AUTHENTICATION_ERROR)
         req["kb_ids"] = ids
+    else:
+        req["kb_ids"] = []
     llm = req.get("llm")
     if llm:
         if "model_name" in llm:

api/apps/sdk/dataset.py

@@ -20,20 +20,17 @@ import os
 import json
 from flask import request
 from peewee import OperationalError
-from api import settings
-from api.db import FileSource, StatusEnum
 from api.db.db_models import File
 from api.db.services.document_service import DocumentService
 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.user_service import TenantService
-from api.utils import get_uuid
+from common.constants import RetCode, FileSource, StatusEnum
 from api.utils.api_utils import (
     deep_merge,
     get_error_argument_result,
     get_error_data_result,
-    get_error_operating_result,
     get_error_permission_result,
     get_parser_config,
     get_result,
@@ -51,6 +48,7 @@ from api.utils.validation_utils import (
 )
 from rag.nlp import search
 from rag.settings import PAGERANK_FLD
+from common import globals
 
 
 @manager.route("/datasets", methods=["POST"])  # noqa: F821
@@ -80,29 +78,28 @@ def create(tenant_id):
           properties:
             name:
               type: string
-              description: Name of the dataset.
+              description: Dataset name (required).
             avatar:
               type: string
-              description: Base64 encoding of the avatar.
+              description: Optional base64-encoded avatar image.
             description:
               type: string
-              description: Description of the dataset.
+              description: Optional dataset description.
             embedding_model:
               type: string
-              description: Embedding model Name.
+              description: Optional embedding model name; if omitted, the tenant's default embedding model is used.
             permission:
               type: string
               enum: ['me', 'team']
-              description: Dataset permission.
+              description: Visibility of the dataset (private to me or shared with team).
             chunk_method:
               type: string
               enum: ["naive", "book", "email", "laws", "manual", "one", "paper",
-                     "picture", "presentation", "qa", "table", "tag"
+                     "picture", "presentation", "qa", "table", "tag"]
-                     ]
+              description: Chunking method; if omitted, defaults to "naive".
-              description: Chunking method.
             parser_config:
               type: object
-              description: Parser configuration.
+              description: Optional parser configuration; server-side defaults will be applied.
     responses:
       200:
         description: Successful operation.
@@ -117,44 +114,43 @@ def create(tenant_id):
     # |----------------|-------------|
     # | embedding_model| embd_id     |
     # | chunk_method   | parser_id   |
+
     req, err = validate_and_parse_json_request(request, CreateDatasetReq)
     if err is not None:
         return get_error_argument_result(err)
+    
+    req = KnowledgebaseService.create_with_name(
+        name = req.pop("name", None),
+        tenant_id = tenant_id,
+        parser_id = req.pop("parser_id", None),
+        **req
+    )
+
+    # Insert embedding model(embd id)
+    ok, t = TenantService.get_by_id(tenant_id)
+    if not ok:
+        return get_error_permission_result(message="Tenant not found")
+    if not req.get("embd_id"):
+        req["embd_id"] = t.embd_id
+    else:
+        ok, err = verify_embedding_availability(req["embd_id"], tenant_id)
+        if not ok:
+            return err
+
 
     try:
-        if KnowledgebaseService.get_or_none(name=req["name"], tenant_id=tenant_id, status=StatusEnum.VALID.value):
+      if not KnowledgebaseService.save(**req):
-            return get_error_operating_result(message=f"Dataset name '{req['name']}' already exists")
+          return get_error_data_result()
-
+      ok, k = KnowledgebaseService.get_by_id(req["id"])
-        req["parser_config"] = get_parser_config(req["parser_id"], req["parser_config"])
+      if not ok:
-        req["id"] = get_uuid()
+        return get_error_data_result(message="Dataset created failed")
-        req["tenant_id"] = tenant_id
+    
-        req["created_by"] = tenant_id
+      response_data = remap_dictionary_keys(k.to_dict())
-
+      return get_result(data=response_data)
-        ok, t = TenantService.get_by_id(tenant_id)
+    except Exception as e:
-        if not ok:
-            return get_error_permission_result(message="Tenant not found")
-
-        if not req.get("embd_id"):
-            req["embd_id"] = t.embd_id
-        else:
-            ok, err = verify_embedding_availability(req["embd_id"], tenant_id)
-            if not ok:
-                return err
-
-        if not KnowledgebaseService.save(**req):
-            return get_error_data_result(message="Create dataset error.(Database error)")
-
-        ok, k = KnowledgebaseService.get_by_id(req["id"])
-        if not ok:
-            return get_error_data_result(message="Dataset created failed")
-
-        response_data = remap_dictionary_keys(k.to_dict())
-        return get_result(data=response_data)
-    except OperationalError as e:
         logging.exception(e)
         return get_error_data_result(message="Database operation failed")
 
-
 @manager.route("/datasets", methods=["DELETE"])  # noqa: F821
 @token_required
 def delete(tenant_id):
@@ -215,7 +211,8 @@ def delete(tenant_id):
                     continue
                 kb_id_instance_pairs.append((kb_id, kb))
             if len(error_kb_ids) > 0:
-                return get_error_permission_result(message=f"""User '{tenant_id}' lacks permission for datasets: '{", ".join(error_kb_ids)}'""")
+                return get_error_permission_result(
+                    message=f"""User '{tenant_id}' lacks permission for datasets: '{", ".join(error_kb_ids)}'""")
 
         errors = []
         success_count = 0
@@ -232,7 +229,8 @@ def delete(tenant_id):
                     ]
                 )
                 File2DocumentService.delete_by_document_id(doc.id)
-            FileService.filter_delete([File.source_type == FileSource.KNOWLEDGEBASE, File.type == "folder", File.name == kb.name])
+            FileService.filter_delete(
+                [File.source_type == FileSource.KNOWLEDGEBASE, File.type == "folder", File.name == kb.name])
             if not KnowledgebaseService.delete_by_id(kb_id):
                 errors.append(f"Delete dataset error for {kb_id}")
                 continue
@@ -329,7 +327,8 @@ def update(tenant_id, dataset_id):
     try:
         kb = KnowledgebaseService.get_or_none(id=dataset_id, tenant_id=tenant_id)
         if kb is None:
-            return get_error_permission_result(message=f"User '{tenant_id}' lacks permission for dataset '{dataset_id}'")
+            return get_error_permission_result(
+                message=f"User '{tenant_id}' lacks permission for dataset '{dataset_id}'")
 
         if req.get("parser_config"):
             req["parser_config"] = deep_merge(kb.parser_config, req["parser_config"])
@@ -341,7 +340,8 @@ def update(tenant_id, dataset_id):
             del req["parser_config"]
 
         if "name" in req and req["name"].lower() != kb.name.lower():
-            exists = KnowledgebaseService.get_or_none(name=req["name"], tenant_id=tenant_id, status=StatusEnum.VALID.value)
+            exists = KnowledgebaseService.get_or_none(name=req["name"], tenant_id=tenant_id,
+                                                      status=StatusEnum.VALID.value)
             if exists:
                 return get_error_data_result(message=f"Dataset name '{req['name']}' already exists")
 
@@ -349,7 +349,8 @@ def update(tenant_id, dataset_id):
             if not req["embd_id"]:
                 req["embd_id"] = kb.embd_id
             if kb.chunk_num != 0 and req["embd_id"] != kb.embd_id:
-                return get_error_data_result(message=f"When chunk_num ({kb.chunk_num}) > 0, embedding_model must remain {kb.embd_id}")
+                return get_error_data_result(
+                    message=f"When chunk_num ({kb.chunk_num}) > 0, embedding_model must remain {kb.embd_id}")
             ok, err = verify_embedding_availability(req["embd_id"], tenant_id)
             if not ok:
                 return err
@@ -359,10 +360,12 @@ def update(tenant_id, dataset_id):
                 return get_error_argument_result(message="'pagerank' can only be set when doc_engine is elasticsearch")
 
             if req["pagerank"] > 0:
-                settings.docStoreConn.update({"kb_id": kb.id}, {PAGERANK_FLD: req["pagerank"]}, search.index_name(kb.tenant_id), kb.id)
+                globals.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)
+                globals.docStoreConn.update({"exists": PAGERANK_FLD}, {"remove": PAGERANK_FLD},
+                                             search.index_name(kb.tenant_id), kb.id)
 
         if not KnowledgebaseService.update_by_id(kb.id, req):
             return get_error_data_result(message="Update dataset error.(Database error)")
@@ -454,7 +457,7 @@ def list_datasets(tenant_id):
                 return get_error_permission_result(message=f"User '{tenant_id}' lacks permission for dataset '{name}'")
 
         tenants = TenantService.get_joined_tenants_by_user_id(tenant_id)
-        kbs = KnowledgebaseService.get_list(
+        kbs, total = KnowledgebaseService.get_list(
             [m["tenant_id"] for m in tenants],
             tenant_id,
             args["page"],
@@ -468,19 +471,20 @@ def list_datasets(tenant_id):
         response_data_list = []
         for kb in kbs:
             response_data_list.append(remap_dictionary_keys(kb))
-        return get_result(data=response_data_list)
+        return get_result(data=response_data_list, total=total)
     except OperationalError as e:
         logging.exception(e)
         return get_error_data_result(message="Database operation failed")
 
+
 @manager.route('/datasets/<dataset_id>/knowledge_graph', methods=['GET'])  # noqa: F821
 @token_required
-def knowledge_graph(tenant_id,dataset_id):
+def knowledge_graph(tenant_id, dataset_id):
     if not KnowledgebaseService.accessible(dataset_id, tenant_id):
         return get_result(
             data=False,
             message='No authorization.',
-            code=settings.RetCode.AUTHENTICATION_ERROR
+            code=RetCode.AUTHENTICATION_ERROR
         )
     _, kb = KnowledgebaseService.get_by_id(dataset_id)
     req = {
@@ -489,9 +493,9 @@ def knowledge_graph(tenant_id,dataset_id):
     }
 
     obj = {"graph": {}, "mind_map": {}}
-    if not settings.docStoreConn.indexExist(search.index_name(kb.tenant_id), dataset_id):
+    if not globals.docStoreConn.indexExist(search.index_name(kb.tenant_id), dataset_id):
         return get_result(data=obj)
-    sres = settings.retrievaler.search(req, search.index_name(kb.tenant_id), [dataset_id])
+    sres = globals.retriever.search(req, search.index_name(kb.tenant_id), [dataset_id])
     if not len(sres.ids):
         return get_result(data=obj)
 
@@ -507,21 +511,24 @@ def knowledge_graph(tenant_id,dataset_id):
     if "nodes" in obj["graph"]:
         obj["graph"]["nodes"] = sorted(obj["graph"]["nodes"], key=lambda x: x.get("pagerank", 0), reverse=True)[:256]
         if "edges" in obj["graph"]:
-            node_id_set = { o["id"] for o in obj["graph"]["nodes"] }
+            node_id_set = {o["id"] for o in obj["graph"]["nodes"]}
-            filtered_edges = [o for o in obj["graph"]["edges"] if o["source"] != o["target"] and o["source"] in node_id_set and o["target"] in node_id_set]
+            filtered_edges = [o for o in obj["graph"]["edges"] if
+                              o["source"] != o["target"] and o["source"] in node_id_set and o["target"] in node_id_set]
             obj["graph"]["edges"] = sorted(filtered_edges, key=lambda x: x.get("weight", 0), reverse=True)[:128]
     return get_result(data=obj)
 
+
 @manager.route('/datasets/<dataset_id>/knowledge_graph', methods=['DELETE'])  # noqa: F821
 @token_required
-def delete_knowledge_graph(tenant_id,dataset_id):
+def delete_knowledge_graph(tenant_id, dataset_id):
     if not KnowledgebaseService.accessible(dataset_id, tenant_id):
         return get_result(
             data=False,
             message='No authorization.',
-            code=settings.RetCode.AUTHENTICATION_ERROR
+            code=RetCode.AUTHENTICATION_ERROR
         )
     _, kb = KnowledgebaseService.get_by_id(dataset_id)
-    settings.docStoreConn.delete({"knowledge_graph_kwd": ["graph", "subgraph", "entity", "relation"]}, search.index_name(kb.tenant_id), dataset_id)
+    globals.docStoreConn.delete({"knowledge_graph_kwd": ["graph", "subgraph", "entity", "relation"]},
+                                 search.index_name(kb.tenant_id), dataset_id)
 
     return get_result(data=True)

api/apps/sdk/dify_retrieval.py

@@ -1,4 +1,4 @@
-    #
+#
 #  Copyright 2024 The InfiniFlow Authors. All Rights Reserved.
 #
 #  Licensed under the Apache License, Version 2.0 (the "License");
@@ -17,7 +17,6 @@ import logging
 
 from flask import request, jsonify
 
-from api.db import LLMType
 from api.db.services.document_service import DocumentService
 from api.db.services.knowledgebase_service import KnowledgebaseService
 from api.db.services.llm_service import LLMBundle
@@ -25,12 +24,96 @@ from api import settings
 from api.utils.api_utils import validate_request, build_error_result, apikey_required
 from rag.app.tag import label_question
 from api.db.services.dialog_service import meta_filter, convert_conditions
-
+from common.constants import RetCode, LLMType
+from common import globals
 
 @manager.route('/dify/retrieval', methods=['POST'])  # noqa: F821
 @apikey_required
 @validate_request("knowledge_id", "query")
 def retrieval(tenant_id):
+    """
+    Dify-compatible retrieval API
+    ---
+    tags:
+      - SDK
+    security:
+      - ApiKeyAuth: []
+    parameters:
+      - in: body
+        name: body
+        required: true
+        schema:
+          type: object
+          required:
+            - knowledge_id
+            - query
+          properties:
+            knowledge_id:
+              type: string
+              description: Knowledge base ID
+            query:
+              type: string
+              description: Query text
+            use_kg:
+              type: boolean
+              description: Whether to use knowledge graph
+              default: false
+            retrieval_setting:
+              type: object
+              description: Retrieval configuration
+              properties:
+                score_threshold:
+                  type: number
+                  description: Similarity threshold
+                  default: 0.0
+                top_k:
+                  type: integer
+                  description: Number of results to return
+                  default: 1024
+            metadata_condition:
+              type: object
+              description: Metadata filter condition
+              properties:
+                conditions:
+                  type: array
+                  items:
+                    type: object
+                    properties:
+                      name:
+                        type: string
+                        description: Field name
+                      comparison_operator:
+                        type: string
+                        description: Comparison operator
+                      value:
+                        type: string
+                        description: Field value
+    responses:
+      200:
+        description: Retrieval succeeded
+        schema:
+          type: object
+          properties:
+            records:
+              type: array
+              items:
+                type: object
+                properties:
+                  content:
+                    type: string
+                    description: Content text
+                  score:
+                    type: number
+                    description: Similarity score
+                  title:
+                    type: string
+                    description: Document title
+                  metadata:
+                    type: object
+                    description: Metadata info
+      404:
+        description: Knowledge base or document not found
+    """
     req = request.json
     question = req["query"]
     kb_id = req["knowledge_id"]
@@ -38,24 +121,24 @@ def retrieval(tenant_id):
     retrieval_setting = req.get("retrieval_setting", {})
     similarity_threshold = float(retrieval_setting.get("score_threshold", 0.0))
     top = int(retrieval_setting.get("top_k", 1024))
-    metadata_condition = req.get("metadata_condition",{})
+    metadata_condition = req.get("metadata_condition", {})
     metas = DocumentService.get_meta_by_kbs([kb_id])
- 
+
     doc_ids = []
     try:
 
         e, kb = KnowledgebaseService.get_by_id(kb_id)
         if not e:
-            return build_error_result(message="Knowledgebase not found!", code=settings.RetCode.NOT_FOUND)
+            return build_error_result(message="Knowledgebase not found!", code=RetCode.NOT_FOUND)
 
         embd_mdl = LLMBundle(kb.tenant_id, LLMType.EMBEDDING.value, llm_name=kb.embd_id)
         print(metadata_condition)
-        print("after",convert_conditions(metadata_condition))
+        # print("after", convert_conditions(metadata_condition))
         doc_ids.extend(meta_filter(metas, convert_conditions(metadata_condition)))
-        print("doc_ids",doc_ids)
+        # print("doc_ids", doc_ids)
         if not doc_ids and metadata_condition is not None:
             doc_ids = ['-999']
-        ranks = settings.retrievaler.retrieval(
+        ranks = globals.retriever.retrieval(
             question,
             embd_mdl,
             kb.tenant_id,
@@ -70,17 +153,17 @@ def retrieval(tenant_id):
         )
 
         if use_kg:
-            ck = settings.kg_retrievaler.retrieval(question,
+            ck = settings.kg_retriever.retrieval(question,
-                                                   [tenant_id],
+                                                 [tenant_id],
-                                                   [kb_id],
+                                                 [kb_id],
-                                                   embd_mdl,
+                                                 embd_mdl,
-                                                   LLMBundle(kb.tenant_id, LLMType.CHAT))
+                                                 LLMBundle(kb.tenant_id, LLMType.CHAT))
             if ck["content_with_weight"]:
                 ranks["chunks"].insert(0, ck)
 
         records = []
         for c in ranks["chunks"]:
-            e, doc = DocumentService.get_by_id( c["doc_id"])
+            e, doc = DocumentService.get_by_id(c["doc_id"])
             c.pop("vector", None)
             meta = getattr(doc, 'meta_fields', {})
             meta["doc_id"] = c["doc_id"]
@@ -96,9 +179,7 @@ def retrieval(tenant_id):
         if str(e).find("not_found") > 0:
             return build_error_result(
                 message='No chunk found! Check the chunk status please!',
-                code=settings.RetCode.NOT_FOUND
+                code=RetCode.NOT_FOUND
             )
         logging.exception(e)
-        return build_error_result(message=str(e), code=settings.RetCode.SERVER_ERROR)
+        return build_error_result(message=str(e), code=RetCode.SERVER_ERROR)
-
-

api/apps/sdk/doc.py

@@ -26,7 +26,7 @@ from pydantic import BaseModel, Field, validator
 
 from api import settings
 from api.constants import FILE_NAME_LEN_LIMIT
-from api.db import FileSource, FileType, LLMType, ParserType, TaskStatus
+from api.db import FileType
 from api.db.db_models import File, Task
 from api.db.services.document_service import DocumentService
 from api.db.services.file2document_service import File2DocumentService
@@ -40,9 +40,11 @@ from api.utils.api_utils import check_duplicate_ids, construct_json_result, get_
 from rag.app.qa import beAdoc, rmPrefix
 from rag.app.tag import label_question
 from rag.nlp import rag_tokenizer, search
-from rag.prompts import cross_languages, keyword_extraction
+from rag.prompts.generator import cross_languages, keyword_extraction
-from rag.utils import rmSpace
 from rag.utils.storage_factory import STORAGE_IMPL
+from common.string_utils import remove_redundant_spaces
+from common.constants import RetCode, LLMType, ParserType, TaskStatus, FileSource
+from common import globals
 
 MAXIMUM_OF_UPLOADING_FILES = 256
 
@@ -127,13 +129,13 @@ def upload(dataset_id, tenant_id):
                     description: Processing status.
     """
     if "file" not in request.files:
-        return get_error_data_result(message="No file part!", code=settings.RetCode.ARGUMENT_ERROR)
+        return get_error_data_result(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_result(message="No file selected!", code=settings.RetCode.ARGUMENT_ERROR)
+            return get_result(message="No file selected!", code=RetCode.ARGUMENT_ERROR)
         if len(file_obj.filename.encode("utf-8")) > FILE_NAME_LEN_LIMIT:
-            return get_result(message=f"File name must be {FILE_NAME_LEN_LIMIT} bytes or less.", code=settings.RetCode.ARGUMENT_ERROR)
+            return get_result(message=f"File name must be {FILE_NAME_LEN_LIMIT} bytes or less.", code=RetCode.ARGUMENT_ERROR)
     """
     # total size
     total_size = 0
@@ -145,7 +147,7 @@ def upload(dataset_id, tenant_id):
     if total_size > MAX_TOTAL_FILE_SIZE:
         return get_result(
             message=f"Total file size exceeds 10MB limit! ({total_size / (1024 * 1024):.2f} MB)",
-            code=settings.RetCode.ARGUMENT_ERROR,
+            code=RetCode.ARGUMENT_ERROR,
         )
     """
     e, kb = KnowledgebaseService.get_by_id(dataset_id)
@@ -153,7 +155,7 @@ def upload(dataset_id, tenant_id):
         raise LookupError(f"Can't find the dataset with ID {dataset_id}!")
     err, files = FileService.upload_document(kb, file_objs, tenant_id)
     if err:
-        return get_result(message="\n".join(err), code=settings.RetCode.SERVER_ERROR)
+        return get_result(message="\n".join(err), code=RetCode.SERVER_ERROR)
     # rename key's name
     renamed_doc_list = []
     for file in files:
@@ -253,12 +255,12 @@ def update_doc(tenant_id, dataset_id, document_id):
         if len(req["name"].encode("utf-8")) > FILE_NAME_LEN_LIMIT:
             return get_result(
                 message=f"File name must be {FILE_NAME_LEN_LIMIT} bytes or less.",
-                code=settings.RetCode.ARGUMENT_ERROR,
+                code=RetCode.ARGUMENT_ERROR,
             )
         if pathlib.Path(req["name"].lower()).suffix != pathlib.Path(doc.name.lower()).suffix:
             return get_result(
                 message="The extension of file can't be changed",
-                code=settings.RetCode.ARGUMENT_ERROR,
+                code=RetCode.ARGUMENT_ERROR,
             )
         for d in DocumentService.query(name=req["name"], kb_id=doc.kb_id):
             if d.name == req["name"]:
@@ -306,7 +308,7 @@ def update_doc(tenant_id, dataset_id, document_id):
             )
             if not e:
                 return get_error_data_result(message="Document not found!")
-            settings.docStoreConn.delete({"doc_id": doc.id}, search.index_name(tenant_id), dataset_id)
+            globals.docStoreConn.delete({"doc_id": doc.id}, search.index_name(tenant_id), dataset_id)
 
     if "enabled" in req:
         status = int(req["enabled"])
@@ -315,7 +317,7 @@ def update_doc(tenant_id, dataset_id, document_id):
                 if not DocumentService.update_by_id(doc.id, {"status": str(status)}):
                     return get_error_data_result(message="Database error (Document update)!")
 
-                settings.docStoreConn.update({"doc_id": doc.id}, {"available_int": status}, search.index_name(kb.tenant_id), doc.kb_id)
+                globals.docStoreConn.update({"doc_id": doc.id}, {"available_int": status}, search.index_name(kb.tenant_id), doc.kb_id)
                 return get_result(data=True)
             except Exception as e:
                 return server_error_response(e)
@@ -402,7 +404,7 @@ def download(tenant_id, dataset_id, document_id):
     doc_id, doc_location = File2DocumentService.get_storage_address(doc_id=document_id)  # minio address
     file_stream = STORAGE_IMPL.get(doc_id, doc_location)
     if not file_stream:
-        return construct_json_result(message="This file is empty.", code=settings.RetCode.DATA_ERROR)
+        return construct_json_result(message="This file is empty.", code=RetCode.DATA_ERROR)
     file = BytesIO(file_stream)
     # Use send_file with a proper filename and MIME type
     return send_file(
@@ -458,7 +460,7 @@ def list_docs(dataset_id, tenant_id):
         required: false
         default: true
         description: Order in descending.
-    - in: query
+      - in: query
         name: create_time_from
         type: integer
         required: false
@@ -470,6 +472,20 @@ def list_docs(dataset_id, tenant_id):
         required: false
         default: 0
         description: Unix timestamp for filtering documents created before this time. 0 means no filter.
+      - in: query
+        name: suffix
+        type: array
+        items:
+          type: string
+        required: false
+        description: Filter by file suffix (e.g., ["pdf", "txt", "docx"]).
+      - in: query
+        name: run
+        type: array
+        items:
+          type: string
+        required: false
+        description: Filter by document run status. Supports both numeric ("0", "1", "2", "3", "4") and text formats ("UNSTART", "RUNNING", "CANCEL", "DONE", "FAIL").
       - in: header
         name: Authorization
         type: string
@@ -512,63 +528,62 @@ def list_docs(dataset_id, tenant_id):
                     description: Processing status.
     """
     if not KnowledgebaseService.accessible(kb_id=dataset_id, user_id=tenant_id):
-        return get_error_data_result(message=f"You don't own the dataset {dataset_id}. ")
+      return get_error_data_result(message=f"You don't own the dataset {dataset_id}. ")
-    id = request.args.get("id")
-    name = request.args.get("name")
 
-    if id and not DocumentService.query(id=id, kb_id=dataset_id):
+    q = request.args
-        return get_error_data_result(message=f"You don't own the document {id}.")
+    document_id = q.get("id")  
+    name        = q.get("name")
+
+    if document_id and not DocumentService.query(id=document_id, kb_id=dataset_id):
+        return get_error_data_result(message=f"You don't own the document {document_id}.")
     if name and not DocumentService.query(name=name, kb_id=dataset_id):
         return get_error_data_result(message=f"You don't own the document {name}.")
 
-    page = int(request.args.get("page", 1))
+    page        = int(q.get("page", 1))
-    keywords = request.args.get("keywords", "")
+    page_size   = int(q.get("page_size", 30))  
-    page_size = int(request.args.get("page_size", 30))
+    orderby     = q.get("orderby", "create_time")
-    orderby = request.args.get("orderby", "create_time")
+    desc        = str(q.get("desc", "true")).strip().lower() != "false"
-    if request.args.get("desc") == "False":
+    keywords    = q.get("keywords", "")
-        desc = False
-    else:
-        desc = True
-    docs, tol = DocumentService.get_list(dataset_id, page, page_size, orderby, desc, keywords, id, name)
 
-    create_time_from = int(request.args.get("create_time_from", 0))
+    # filters - align with OpenAPI parameter names
-    create_time_to = int(request.args.get("create_time_to", 0))
+    suffix               = q.getlist("suffix") 
+    run_status           = q.getlist("run")   
+    create_time_from     = int(q.get("create_time_from", 0))  
+    create_time_to       = int(q.get("create_time_to", 0))    
 
+    # map run status (accept text or numeric) - align with API parameter
+    run_status_text_to_numeric = {"UNSTART": "0", "RUNNING": "1", "CANCEL": "2", "DONE": "3", "FAIL": "4"}
+    run_status_converted = [run_status_text_to_numeric.get(v, v) for v in run_status]
+
+    docs, total = DocumentService.get_list(
+        dataset_id, page, page_size, orderby, desc, keywords, document_id, name, suffix, run_status_converted
+    )
+
+    # time range filter (0 means no bound)
     if create_time_from or create_time_to:
-        filtered_docs = []
+        docs = [
-        for doc in docs:
+            d for d in docs
-            doc_create_time = doc.get("create_time", 0)
+            if (create_time_from == 0 or d.get("create_time", 0) >= create_time_from)
-            if (create_time_from == 0 or doc_create_time >= create_time_from) and (create_time_to == 0 or doc_create_time <= create_time_to):
+            and (create_time_to == 0 or d.get("create_time", 0) <= create_time_to)
-                filtered_docs.append(doc)
+        ]
-        docs = filtered_docs
 
-    # rename key's name
+    # rename keys + map run status back to text for output
-    renamed_doc_list = []
     key_mapping = {
         "chunk_num": "chunk_count",
-        "kb_id": "dataset_id",
+        "kb_id": "dataset_id", 
         "token_num": "token_count",
         "parser_id": "chunk_method",
     }
-    run_mapping = {
+    run_status_numeric_to_text = {"0": "UNSTART", "1": "RUNNING", "2": "CANCEL", "3": "DONE", "4": "FAIL"}
-        "0": "UNSTART",
-        "1": "RUNNING",
-        "2": "CANCEL",
-        "3": "DONE",
-        "4": "FAIL",
-    }
-    for doc in docs:
-        renamed_doc = {}
-        for key, value in doc.items():
-            if key == "run":
-                renamed_doc["run"] = run_mapping.get(str(value))
-            new_key = key_mapping.get(key, key)
-            renamed_doc[new_key] = value
-            if key == "run":
-                renamed_doc["run"] = run_mapping.get(value)
-        renamed_doc_list.append(renamed_doc)
-    return get_result(data={"total": tol, "docs": renamed_doc_list})
 
+    output_docs = []
+    for d in docs:
+        renamed_doc = {key_mapping.get(k, k): v for k, v in d.items()}
+        if "run" in d:
+            renamed_doc["run"] = run_status_numeric_to_text.get(str(d["run"]), d["run"])
+        output_docs.append(renamed_doc)
+
+    return get_result(data={"total": total, "docs": output_docs})
 
 @manager.route("/datasets/<dataset_id>/documents", methods=["DELETE"])  # noqa: F821
 @token_required
@@ -663,10 +678,10 @@ def delete(tenant_id, dataset_id):
             errors += str(e)
 
     if not_found:
-        return get_result(message=f"Documents not found: {not_found}", code=settings.RetCode.DATA_ERROR)
+        return get_result(message=f"Documents not found: {not_found}", code=RetCode.DATA_ERROR)
 
     if errors:
-        return get_result(message=errors, code=settings.RetCode.SERVER_ERROR)
+        return get_result(message=errors, code=RetCode.SERVER_ERROR)
 
     if duplicate_messages:
         if success_count > 0:
@@ -741,7 +756,7 @@ def parse(tenant_id, dataset_id):
             return get_error_data_result("Can't parse document that is currently being processed")
         info = {"run": "1", "progress": 0, "progress_msg": "", "chunk_num": 0, "token_num": 0}
         DocumentService.update_by_id(id, info)
-        settings.docStoreConn.delete({"doc_id": id}, search.index_name(tenant_id), dataset_id)
+        globals.docStoreConn.delete({"doc_id": id}, search.index_name(tenant_id), dataset_id)
         TaskService.filter_delete([Task.doc_id == id])
         e, doc = DocumentService.get_by_id(id)
         doc = doc.to_dict()
@@ -750,7 +765,7 @@ def parse(tenant_id, dataset_id):
         queue_tasks(doc, bucket, name, 0)
         success_count += 1
     if not_found:
-        return get_result(message=f"Documents not found: {not_found}", code=settings.RetCode.DATA_ERROR)
+        return get_result(message=f"Documents not found: {not_found}", code=RetCode.DATA_ERROR)
     if duplicate_messages:
         if success_count > 0:
             return get_result(
@@ -821,7 +836,7 @@ def stop_parsing(tenant_id, dataset_id):
             return get_error_data_result("Can't stop parsing document with progress at 0 or 1")
         info = {"run": "2", "progress": 0, "chunk_num": 0}
         DocumentService.update_by_id(id, info)
-        settings.docStoreConn.delete({"doc_id": doc[0].id}, search.index_name(tenant_id), dataset_id)
+        globals.docStoreConn.delete({"doc_id": doc[0].id}, search.index_name(tenant_id), dataset_id)
         success_count += 1
     if duplicate_messages:
         if success_count > 0:
@@ -954,9 +969,9 @@ def list_chunks(tenant_id, dataset_id, document_id):
 
     res = {"total": 0, "chunks": [], "doc": renamed_doc}
     if req.get("id"):
-        chunk = settings.docStoreConn.get(req.get("id"), search.index_name(tenant_id), [dataset_id])
+        chunk = globals.docStoreConn.get(req.get("id"), search.index_name(tenant_id), [dataset_id])
         if not chunk:
-            return get_result(message=f"Chunk not found: {dataset_id}/{req.get('id')}", code=settings.RetCode.NOT_FOUND)
+            return get_result(message=f"Chunk not found: {dataset_id}/{req.get('id')}", code=RetCode.NOT_FOUND)
         k = []
         for n in chunk.keys():
             if re.search(r"(_vec$|_sm_|_tks|_ltks)", n):
@@ -981,13 +996,13 @@ def list_chunks(tenant_id, dataset_id, document_id):
         res["chunks"].append(final_chunk)
         _ = Chunk(**final_chunk)
 
-    elif settings.docStoreConn.indexExist(search.index_name(tenant_id), dataset_id):
+    elif globals.docStoreConn.indexExist(search.index_name(tenant_id), dataset_id):
-        sres = settings.retrievaler.search(query, search.index_name(tenant_id), [dataset_id], emb_mdl=None, highlight=True)
+        sres = globals.retriever.search(query, search.index_name(tenant_id), [dataset_id], emb_mdl=None, highlight=True)
         res["total"] = sres.total
         for id in sres.ids:
             d = {
                 "id": id,
-                "content": (rmSpace(sres.highlight[id]) if question and id in sres.highlight else sres.field[id].get("content_with_weight", "")),
+                "content": (remove_redundant_spaces(sres.highlight[id]) if question and id in sres.highlight else sres.field[id].get("content_with_weight", "")),
                 "document_id": sres.field[id]["doc_id"],
                 "docnm_kwd": sres.field[id]["docnm_kwd"],
                 "important_keywords": sres.field[id].get("important_kwd", []),
@@ -1106,7 +1121,7 @@ def add_chunk(tenant_id, dataset_id, document_id):
     v, c = embd_mdl.encode([doc.name, req["content"] 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), dataset_id)
+    globals.docStoreConn.insert([d], search.index_name(tenant_id), dataset_id)
 
     DocumentService.increment_chunk_num(doc.id, doc.kb_id, c, 1, 0)
     # rename keys
@@ -1187,7 +1202,7 @@ def rm_chunk(tenant_id, dataset_id, document_id):
     if "chunk_ids" in req:
         unique_chunk_ids, duplicate_messages = check_duplicate_ids(req["chunk_ids"], "chunk")
         condition["id"] = unique_chunk_ids
-    chunk_number = settings.docStoreConn.delete(condition, search.index_name(tenant_id), dataset_id)
+    chunk_number = globals.docStoreConn.delete(condition, search.index_name(tenant_id), dataset_id)
     if chunk_number != 0:
         DocumentService.decrement_chunk_num(document_id, dataset_id, 1, chunk_number, 0)
     if "chunk_ids" in req and chunk_number != len(unique_chunk_ids):
@@ -1259,7 +1274,7 @@ def update_chunk(tenant_id, dataset_id, document_id, chunk_id):
         schema:
           type: object
     """
-    chunk = settings.docStoreConn.get(chunk_id, search.index_name(tenant_id), [dataset_id])
+    chunk = globals.docStoreConn.get(chunk_id, search.index_name(tenant_id), [dataset_id])
     if chunk is None:
         return get_error_data_result(f"Can't find this chunk {chunk_id}")
     if not KnowledgebaseService.accessible(kb_id=dataset_id, user_id=tenant_id):
@@ -1288,6 +1303,10 @@ def update_chunk(tenant_id, dataset_id, document_id, chunk_id):
         d["question_tks"] = rag_tokenizer.tokenize("\n".join(req["questions"]))
     if "available" in req:
         d["available_int"] = int(req["available"])
+    if "positions" in req:
+        if not isinstance(req["positions"], list):
+            return get_error_data_result("`positions` should be a list")
+        d["position_int"] = req["positions"]
     embd_id = DocumentService.get_embd_id(document_id)
     embd_mdl = TenantLLMService.model_instance(tenant_id, LLMType.EMBEDDING.value, embd_id)
     if doc.parser_id == ParserType.QA:
@@ -1300,7 +1319,7 @@ def update_chunk(tenant_id, dataset_id, document_id, chunk_id):
     v, c = embd_mdl.encode([doc.name, d["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": chunk_id}, d, search.index_name(tenant_id), dataset_id)
+    globals.docStoreConn.update({"id": chunk_id}, d, search.index_name(tenant_id), dataset_id)
     return get_result()
 
 
@@ -1401,7 +1420,7 @@ def retrieval_test(tenant_id):
     if len(embd_nms) != 1:
         return get_result(
             message='Datasets use different embedding models."',
-            code=settings.RetCode.DATA_ERROR,
+            code=RetCode.DATA_ERROR,
         )
     if "question" not in req:
         return get_error_data_result("`question` is required.")
@@ -1446,7 +1465,7 @@ def retrieval_test(tenant_id):
             chat_mdl = LLMBundle(kb.tenant_id, LLMType.CHAT)
             question += keyword_extraction(chat_mdl, question)
 
-        ranks = settings.retrievaler.retrieval(
+        ranks = globals.retriever.retrieval(
             question,
             embd_mdl,
             tenant_ids,
@@ -1462,7 +1481,7 @@ def retrieval_test(tenant_id):
             rank_feature=label_question(question, kbs),
         )
         if use_kg:
-            ck = settings.kg_retrievaler.retrieval(question, [k.tenant_id for k in kbs], kb_ids, embd_mdl, LLMBundle(kb.tenant_id, LLMType.CHAT))
+            ck = settings.kg_retriever.retrieval(question, [k.tenant_id for k in kbs], kb_ids, embd_mdl, LLMBundle(kb.tenant_id, LLMType.CHAT))
             if ck["content_with_weight"]:
                 ranks["chunks"].insert(0, ck)
 
@@ -1492,6 +1511,6 @@ def retrieval_test(tenant_id):
         if str(e).find("not_found") > 0:
             return get_result(
                 message="No chunk found! Check the chunk status please!",
-                code=settings.RetCode.DATA_ERROR,
+                code=RetCode.DATA_ERROR,
             )
         return server_error_response(e)

api/apps/sdk/files.py

@@ -1,13 +1,32 @@
+#
+#  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.
+#
+
+
 import pathlib
 import re
 
 import flask
 from flask import request
+from pathlib import Path
 
 from api.db.services.document_service import DocumentService
 from api.db.services.file2document_service import File2DocumentService
+from api.db.services.knowledgebase_service import KnowledgebaseService
 from api.utils.api_utils import server_error_response, token_required
-from api.utils import get_uuid
+from common.misc_utils import get_uuid
 from api.db import FileType
 from api.db.services import duplicate_name
 from api.db.services.file_service import FileService
@@ -15,7 +34,8 @@ from api.utils.api_utils import get_json_result
 from api.utils.file_utils import filename_type
 from rag.utils.storage_factory import STORAGE_IMPL
 
-@manager.route('/file/upload', methods=['POST']) # noqa: F821
+
+@manager.route('/file/upload', methods=['POST'])  # noqa: F821
 @token_required
 def upload(tenant_id):
     """
@@ -42,22 +62,22 @@ def upload(tenant_id):
           type: object
           properties:
             data:
-            type: array
+              type: array
-            items:
+              items:
-              type: object
+                type: object
-              properties:
+                properties:
-                id:
+                  id:
-                  type: string
+                    type: string
-                  description: File ID
+                    description: File ID
-                name:
+                  name:
-                  type: string
+                    type: string
-                  description: File name
+                    description: File name
-                size:
+                  size:
-                  type: integer
+                    type: integer
-                  description: File size in bytes
+                    description: File size in bytes
-                type:
+                  type:
-                  type: string
+                    type: string
-                  description: File type (e.g., document, folder)
+                    description: File type (e.g., document, folder)
     """
     pf_id = request.form.get("parent_id")
 
@@ -81,26 +101,28 @@ def upload(tenant_id):
             return get_json_result(data=False, message="Can't find this folder!", code=404)
 
         for file_obj in file_objs:
-            # 文件路径处理
+            # Handle file path
             full_path = '/' + file_obj.filename
             file_obj_names = full_path.split('/')
             file_len = len(file_obj_names)
 
-            # 获取文件夹路径ID
+            # Get folder path ID
             file_id_list = FileService.get_id_list_by_id(pf_id, file_obj_names, 1, [pf_id])
             len_id_list = len(file_id_list)
 
-            # 创建文件夹结构
+            # Crete file folder
             if file_len != len_id_list:
                 e, file = FileService.get_by_id(file_id_list[len_id_list - 1])
                 if not e:
                     return get_json_result(data=False, message="Folder not found!", code=404)
-                last_folder = FileService.create_folder(file, file_id_list[len_id_list - 1], file_obj_names, len_id_list)
+                last_folder = 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])
                 if not e:
                     return get_json_result(data=False, message="Folder not found!", code=404)
-                last_folder = FileService.create_folder(file, file_id_list[len_id_list - 2], file_obj_names, len_id_list)
+                last_folder = FileService.create_folder(file, file_id_list[len_id_list - 2], file_obj_names,
+                                                        len_id_list)
 
             filetype = filename_type(file_obj_names[file_len - 1])
             location = file_obj_names[file_len - 1]
@@ -127,7 +149,7 @@ def upload(tenant_id):
         return server_error_response(e)
 
 
-@manager.route('/file/create', methods=['POST']) # noqa: F821
+@manager.route('/file/create', methods=['POST'])  # noqa: F821
 @token_required
 def create(tenant_id):
     """
@@ -205,7 +227,7 @@ def create(tenant_id):
         return server_error_response(e)
 
 
-@manager.route('/file/list', methods=['GET']) # noqa: F821
+@manager.route('/file/list', methods=['GET'])  # noqa: F821
 @token_required
 def list_files(tenant_id):
     """
@@ -297,7 +319,7 @@ def list_files(tenant_id):
         return server_error_response(e)
 
 
-@manager.route('/file/root_folder', methods=['GET']) # noqa: F821
+@manager.route('/file/root_folder', methods=['GET'])  # noqa: F821
 @token_required
 def get_root_folder(tenant_id):
     """
@@ -333,7 +355,7 @@ def get_root_folder(tenant_id):
         return server_error_response(e)
 
 
-@manager.route('/file/parent_folder', methods=['GET']) # noqa: F821
+@manager.route('/file/parent_folder', methods=['GET'])  # noqa: F821
 @token_required
 def get_parent_folder():
     """
@@ -378,7 +400,7 @@ def get_parent_folder():
         return server_error_response(e)
 
 
-@manager.route('/file/all_parent_folder', methods=['GET']) # noqa: F821
+@manager.route('/file/all_parent_folder', methods=['GET'])  # noqa: F821
 @token_required
 def get_all_parent_folders(tenant_id):
     """
@@ -426,7 +448,7 @@ def get_all_parent_folders(tenant_id):
         return server_error_response(e)
 
 
-@manager.route('/file/rm', methods=['POST']) # noqa: F821
+@manager.route('/file/rm', methods=['POST'])  # noqa: F821
 @token_required
 def rm(tenant_id):
     """
@@ -500,7 +522,7 @@ def rm(tenant_id):
         return server_error_response(e)
 
 
-@manager.route('/file/rename', methods=['POST']) # noqa: F821
+@manager.route('/file/rename', methods=['POST'])  # noqa: F821
 @token_required
 def rename(tenant_id):
     """
@@ -540,7 +562,8 @@ def rename(tenant_id):
         if not e:
             return get_json_result(message="File not found!", code=404)
 
-        if file.type != FileType.FOLDER.value and pathlib.Path(req["name"].lower()).suffix != pathlib.Path(file.name.lower()).suffix:
+        if file.type != FileType.FOLDER.value and pathlib.Path(req["name"].lower()).suffix != pathlib.Path(
+                file.name.lower()).suffix:
             return get_json_result(data=False, message="The extension of file can't be changed", code=400)
 
         for existing_file in FileService.query(name=req["name"], pf_id=file.parent_id):
@@ -560,9 +583,9 @@ def rename(tenant_id):
         return server_error_response(e)
 
 
-@manager.route('/file/get/<file_id>', methods=['GET']) # noqa: F821
+@manager.route('/file/get/<file_id>', methods=['GET'])  # noqa: F821
 @token_required
-def get(tenant_id,file_id):
+def get(tenant_id, file_id):
     """
     Download a file.
     ---
@@ -608,7 +631,7 @@ def get(tenant_id,file_id):
         return server_error_response(e)
 
 
-@manager.route('/file/mv', methods=['POST']) # noqa: F821
+@manager.route('/file/mv', methods=['POST'])  # noqa: F821
 @token_required
 def move(tenant_id):
     """
@@ -666,3 +689,72 @@ def move(tenant_id):
         return get_json_result(data=True)
     except Exception as e:
         return server_error_response(e)
+
+
+@manager.route('/file/convert', methods=['POST'])  # noqa: F821
+@token_required
+def convert(tenant_id):
+    req = request.json
+    kb_ids = req["kb_ids"]
+    file_ids = req["file_ids"]
+    file2documents = []
+
+    try:
+        files = FileService.get_by_ids(file_ids)
+        files_set = dict({file.id: file for file in files})
+        for file_id in file_ids:
+            file = files_set[file_id]
+            if not file:
+                return get_json_result(message="File not found!", code=404)
+            file_ids_list = [file_id]
+            if file.type == FileType.FOLDER.value:
+                file_ids_list = FileService.get_all_innermost_file_ids(file_id, [])
+            for id in file_ids_list:
+                informs = File2DocumentService.get_by_file_id(id)
+                # delete
+                for inform in informs:
+                    doc_id = inform.document_id
+                    e, doc = DocumentService.get_by_id(doc_id)
+                    if not e:
+                        return get_json_result(message="Document not found!", code=404)
+                    tenant_id = DocumentService.get_tenant_id(doc_id)
+                    if not tenant_id:
+                        return get_json_result(message="Tenant not found!", code=404)
+                    if not DocumentService.remove_document(doc, tenant_id):
+                        return get_json_result(
+                            message="Database error (Document removal)!", code=404)
+                File2DocumentService.delete_by_file_id(id)
+
+                # insert
+                for kb_id in kb_ids:
+                    e, kb = KnowledgebaseService.get_by_id(kb_id)
+                    if not e:
+                        return get_json_result(
+                            message="Can't find this knowledgebase!", code=404)
+                    e, file = FileService.get_by_id(id)
+                    if not e:
+                        return get_json_result(
+                            message="Can't find this file!", code=404)
+
+                    doc = DocumentService.insert({
+                        "id": get_uuid(),
+                        "kb_id": kb.id,
+                        "parser_id": FileService.get_parser(file.type, file.name, kb.parser_id),
+                        "parser_config": kb.parser_config,
+                        "created_by": tenant_id,
+                        "type": file.type,
+                        "name": file.name,
+                        "suffix": Path(file.name).suffix.lstrip("."),
+                        "location": file.location,
+                        "size": file.size
+                    })
+                    file2document = File2DocumentService.insert({
+                        "id": get_uuid(),
+                        "file_id": id,
+                        "document_id": doc.id,
+                    })
+
+                    file2documents.append(file2document.to_json())
+        return get_json_result(data=file2documents)
+    except Exception as e:
+        return server_error_response(e)

api/apps/sdk/session.py

@@ -22,10 +22,9 @@ from flask import Response, jsonify, request
 
 from agent.canvas import Canvas
 from api import settings
-from api.db import LLMType, StatusEnum
 from api.db.db_models import APIToken
 from api.db.services.api_service import API4ConversationService
-from api.db.services.canvas_service import UserCanvasService, completionOpenAI
+from api.db.services.canvas_service import UserCanvasService, completion_openai
 from api.db.services.canvas_service import completion as agent_completion
 from api.db.services.conversation_service import ConversationService, iframe_completion
 from api.db.services.conversation_service import completion as rag_completion
@@ -35,13 +34,14 @@ from api.db.services.knowledgebase_service import KnowledgebaseService
 from api.db.services.llm_service import LLMBundle
 from api.db.services.search_service import SearchService
 from api.db.services.user_service import UserTenantService
-from api.utils import get_uuid
+from common.misc_utils import get_uuid
-from api.utils.api_utils import check_duplicate_ids, get_data_openai, get_error_data_result, get_json_result, get_result, server_error_response, token_required, validate_request
+from api.utils.api_utils import check_duplicate_ids, get_data_openai, get_error_data_result, get_json_result, \
+    get_result, server_error_response, token_required, validate_request
 from rag.app.tag import label_question
-from rag.prompts import chunks_format
+from rag.prompts.template import load_prompt
-from rag.prompts.prompt_template import load_prompt
+from rag.prompts.generator import cross_languages, gen_meta_filter, keyword_extraction, chunks_format
-from rag.prompts.prompts import cross_languages, gen_meta_filter, keyword_extraction
+from common.constants import RetCode, LLMType, StatusEnum
-
+from common import globals
 
 @manager.route("/chats/<chat_id>/sessions", methods=["POST"])  # noqa: F821
 @token_required
@@ -89,7 +89,8 @@ def create_agent_session(tenant_id, agent_id):
     canvas.reset()
 
     cvs.dsl = json.loads(str(canvas))
-    conv = {"id": session_id, "dialog_id": cvs.id, "user_id": user_id, "message": [{"role": "assistant", "content": canvas.get_prologue()}], "source": "agent", "dsl": cvs.dsl}
+    conv = {"id": session_id, "dialog_id": cvs.id, "user_id": user_id,
+            "message": [{"role": "assistant", "content": canvas.get_prologue()}], "source": "agent", "dsl": cvs.dsl}
     API4ConversationService.save(**conv)
     conv["agent_id"] = conv.pop("dialog_id")
     return get_result(data=conv)
@@ -280,7 +281,7 @@ def chat_completion_openai_like(tenant_id, chat_id):
                     reasoning_match = re.search(r"<think>(.*?)</think>", answer, flags=re.DOTALL)
                     if reasoning_match:
                         reasoning_part = reasoning_match.group(1)
-                        content_part = answer[reasoning_match.end() :]
+                        content_part = answer[reasoning_match.end():]
                     else:
                         reasoning_part = ""
                         content_part = answer
@@ -325,7 +326,8 @@ def chat_completion_openai_like(tenant_id, chat_id):
             response["choices"][0]["delta"]["content"] = None
             response["choices"][0]["delta"]["reasoning_content"] = None
             response["choices"][0]["finish_reason"] = "stop"
-            response["usage"] = {"prompt_tokens": len(prompt), "completion_tokens": token_used, "total_tokens": len(prompt) + token_used}
+            response["usage"] = {"prompt_tokens": len(prompt), "completion_tokens": token_used,
+                                 "total_tokens": len(prompt) + token_used}
             if need_reference:
                 response["choices"][0]["delta"]["reference"] = chunks_format(last_ans.get("reference", []))
                 response["choices"][0]["delta"]["final_content"] = last_ans.get("answer", "")
@@ -410,11 +412,11 @@ def agents_completion_openai_compatibility(tenant_id, agent_id):
     stream = req.pop("stream", False)
     if stream:
         resp = Response(
-            completionOpenAI(
+            completion_openai(
                 tenant_id,
                 agent_id,
                 question,
-                session_id=req.get("session_id", req.get("id", "") or req.get("metadata", {}).get("id", "")),
+                session_id=req.pop("session_id", req.get("id", "")) or req.get("metadata", {}).get("id", ""),
                 stream=True,
                 **req,
             ),
@@ -428,11 +430,11 @@ def agents_completion_openai_compatibility(tenant_id, agent_id):
     else:
         # For non-streaming, just return the response directly
         response = next(
-            completionOpenAI(
+            completion_openai(
                 tenant_id,
                 agent_id,
                 question,
-                session_id=req.get("session_id", req.get("id", "") or req.get("metadata", {}).get("id", "")),
+                session_id=req.pop("session_id", req.get("id", "")) or req.get("metadata", {}).get("id", ""),
                 stream=False,
                 **req,
             )
@@ -560,7 +562,8 @@ def list_agent_session(tenant_id, agent_id):
         desc = True
     # dsl defaults to True in all cases except for False and false
     include_dsl = request.args.get("dsl") != "False" and request.args.get("dsl") != "false"
-    total, convs = API4ConversationService.get_list(agent_id, tenant_id, page_number, items_per_page, orderby, desc, id, user_id, include_dsl)
+    total, convs = API4ConversationService.get_list(agent_id, tenant_id, page_number, items_per_page, orderby, desc, id,
+                                                    user_id, include_dsl)
     if not convs:
         return get_result(data=[])
     for conv in convs:
@@ -582,7 +585,8 @@ def list_agent_session(tenant_id, agent_id):
                 if message_num != 0 and messages[message_num]["role"] != "user":
                     chunk_list = []
                     # Add boundary and type checks to prevent KeyError
-                    if chunk_num < len(conv["reference"]) and conv["reference"][chunk_num] is not None and isinstance(conv["reference"][chunk_num], dict) and "chunks" in conv["reference"][chunk_num]:
+                    if chunk_num < len(conv["reference"]) and conv["reference"][chunk_num] is not None and isinstance(
+                            conv["reference"][chunk_num], dict) and "chunks" in conv["reference"][chunk_num]:
                         chunks = conv["reference"][chunk_num]["chunks"]
                         for chunk in chunks:
                             # Ensure chunk is a dictionary before calling get method
@@ -640,13 +644,16 @@ def delete(tenant_id, chat_id):
 
     if errors:
         if success_count > 0:
-            return get_result(data={"success_count": success_count, "errors": errors}, message=f"Partially deleted {success_count} sessions with {len(errors)} errors")
+            return get_result(data={"success_count": success_count, "errors": errors},
+                              message=f"Partially deleted {success_count} sessions with {len(errors)} errors")
         else:
             return get_error_data_result(message="; ".join(errors))
 
     if duplicate_messages:
         if success_count > 0:
-            return get_result(message=f"Partially deleted {success_count} sessions with {len(duplicate_messages)} errors", data={"success_count": success_count, "errors": duplicate_messages})
+            return get_result(
+                message=f"Partially deleted {success_count} sessions with {len(duplicate_messages)} errors",
+                data={"success_count": success_count, "errors": duplicate_messages})
         else:
             return get_error_data_result(message=";".join(duplicate_messages))
 
@@ -692,13 +699,16 @@ def delete_agent_session(tenant_id, agent_id):
 
     if errors:
         if success_count > 0:
-            return get_result(data={"success_count": success_count, "errors": errors}, message=f"Partially deleted {success_count} sessions with {len(errors)} errors")
+            return get_result(data={"success_count": success_count, "errors": errors},
+                              message=f"Partially deleted {success_count} sessions with {len(errors)} errors")
         else:
             return get_error_data_result(message="; ".join(errors))
 
     if duplicate_messages:
         if success_count > 0:
-            return get_result(message=f"Partially deleted {success_count} sessions with {len(duplicate_messages)} errors", data={"success_count": success_count, "errors": duplicate_messages})
+            return get_result(
+                message=f"Partially deleted {success_count} sessions with {len(duplicate_messages)} errors",
+                data={"success_count": success_count, "errors": duplicate_messages})
         else:
             return get_error_data_result(message=";".join(duplicate_messages))
 
@@ -731,7 +741,9 @@ def ask_about(tenant_id):
             for ans in ask(req["question"], req["kb_ids"], uid):
                 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"
+            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(stream(), mimetype="text/event-stream")
@@ -883,7 +895,9 @@ def begin_inputs(agent_id):
         return get_error_data_result(f"Can't find agent by ID: {agent_id}")
 
     canvas = Canvas(json.dumps(cvs.dsl), objs[0].tenant_id)
-    return get_result(data={"title": cvs.title, "avatar": cvs.avatar, "inputs": canvas.get_component_input_form("begin"), "prologue": canvas.get_prologue(), "mode": canvas.get_mode()})
+    return get_result(
+        data={"title": cvs.title, "avatar": cvs.avatar, "inputs": canvas.get_component_input_form("begin"),
+              "prologue": canvas.get_prologue(), "mode": canvas.get_mode()})
 
 
 @manager.route("/searchbots/ask", methods=["POST"])  # noqa: F821
@@ -912,7 +926,9 @@ def ask_about_embedded():
             for ans in 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"
+            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(stream(), mimetype="text/event-stream")
@@ -943,7 +959,7 @@ def retrieval_test_embedded():
         kb_ids = [kb_ids]
     if not kb_ids:
         return get_json_result(data=False, message='Please specify dataset firstly.',
-                               code=settings.RetCode.DATA_ERROR)
+                               code=RetCode.DATA_ERROR)
     doc_ids = req.get("doc_ids", [])
     similarity_threshold = float(req.get("similarity_threshold", 0.0))
     vector_similarity_weight = float(req.get("vector_similarity_weight", 0.3))
@@ -979,7 +995,8 @@ def retrieval_test_embedded():
                     tenant_ids.append(tenant.tenant_id)
                     break
             else:
-                return get_json_result(data=False, message="Only owner of knowledgebase authorized for this operation.", code=settings.RetCode.OPERATING_ERROR)
+                return get_json_result(data=False, message="Only owner of knowledgebase authorized for this operation.",
+                                       code=RetCode.OPERATING_ERROR)
 
         e, kb = KnowledgebaseService.get_by_id(kb_ids[0])
         if not e:
@@ -999,11 +1016,13 @@ def retrieval_test_embedded():
             question += keyword_extraction(chat_mdl, question)
 
         labels = label_question(question, [kb])
-        ranks = settings.retrievaler.retrieval(
+        ranks = globals.retriever.retrieval(
-            question, embd_mdl, tenant_ids, kb_ids, page, size, similarity_threshold, vector_similarity_weight, top, doc_ids, rerank_mdl=rerank_mdl, highlight=req.get("highlight"), rank_feature=labels
+            question, embd_mdl, tenant_ids, kb_ids, page, size, similarity_threshold, vector_similarity_weight, top,
+            doc_ids, rerank_mdl=rerank_mdl, highlight=req.get("highlight"), rank_feature=labels
         )
         if use_kg:
-            ck = settings.kg_retrievaler.retrieval(question, tenant_ids, kb_ids, embd_mdl, LLMBundle(kb.tenant_id, LLMType.CHAT))
+            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)
 
@@ -1014,7 +1033,8 @@ def retrieval_test_embedded():
         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=settings.RetCode.DATA_ERROR)
+            return get_json_result(data=False, message="No chunk found! Check the chunk status please!",
+                                   code=RetCode.DATA_ERROR)
         return server_error_response(e)
 
 
@@ -1083,7 +1103,8 @@ def detail_share_embedded():
             if SearchService.query(tenant_id=tenant.tenant_id, id=search_id):
                 break
         else:
-            return get_json_result(data=False, message="Has no permission for this operation.", code=settings.RetCode.OPERATING_ERROR)
+            return get_json_result(data=False, message="Has no permission for this operation.",
+                                   code=RetCode.OPERATING_ERROR)
 
         search = SearchService.get_detail(search_id)
         if not search:

api/apps/search_app.py

@@ -17,14 +17,13 @@
 from flask import request
 from flask_login import current_user, login_required
 
-from api import settings
 from api.constants import DATASET_NAME_LIMIT
-from api.db import StatusEnum
 from api.db.db_models import DB
 from api.db.services import duplicate_name
 from api.db.services.search_service import SearchService
 from api.db.services.user_service import TenantService, UserTenantService
-from api.utils import get_uuid
+from common.misc_utils import get_uuid
+from common.constants import RetCode, StatusEnum
 from api.utils.api_utils import get_data_error_result, get_json_result, not_allowed_parameters, server_error_response, validate_request
 
 
@@ -82,12 +81,12 @@ def update():
 
     search_id = req["search_id"]
     if not SearchService.accessible4deletion(search_id, current_user.id):
-        return get_json_result(data=False, message="No authorization.", code=settings.RetCode.AUTHENTICATION_ERROR)
+        return get_json_result(data=False, message="No authorization.", code=RetCode.AUTHENTICATION_ERROR)
 
     try:
         search_app = SearchService.query(tenant_id=tenant_id, id=search_id)[0]
         if not search_app:
-            return get_json_result(data=False, message=f"Cannot find search {search_id}", code=settings.RetCode.DATA_ERROR)
+            return get_json_result(data=False, message=f"Cannot find search {search_id}", code=RetCode.DATA_ERROR)
 
         if req["name"].lower() != search_app.name.lower() and len(SearchService.query(name=req["name"], tenant_id=tenant_id, status=StatusEnum.VALID.value)) >= 1:
             return get_data_error_result(message="Duplicated search name.")
@@ -129,7 +128,7 @@ def detail():
             if SearchService.query(tenant_id=tenant.tenant_id, id=search_id):
                 break
         else:
-            return get_json_result(data=False, message="Has no permission for this operation.", code=settings.RetCode.OPERATING_ERROR)
+            return get_json_result(data=False, message="Has no permission for this operation.", code=RetCode.OPERATING_ERROR)
 
         search = SearchService.get_detail(search_id)
         if not search:
@@ -178,7 +177,7 @@ def rm():
     req = request.get_json()
     search_id = req["search_id"]
     if not SearchService.accessible4deletion(search_id, current_user.id):
-        return get_json_result(data=False, message="No authorization.", code=settings.RetCode.AUTHENTICATION_ERROR)
+        return get_json_result(data=False, message="No authorization.", code=RetCode.AUTHENTICATION_ERROR)
 
     try:
         if not SearchService.delete_by_id(search_id):

api/apps/system_app.py

@@ -24,7 +24,6 @@ from api.db.services.api_service import APITokenService
 from api.db.services.knowledgebase_service import KnowledgebaseService
 from api.db.services.user_service import UserTenantService
 from api import settings
-from api.utils import current_timestamp, datetime_format
 from api.utils.api_utils import (
     get_json_result,
     get_data_error_result,
@@ -32,10 +31,15 @@ from api.utils.api_utils import (
     generate_confirmation_token,
 )
 from api.versions import get_ragflow_version
+from common.time_utils import current_timestamp, datetime_format
 from rag.utils.storage_factory import STORAGE_IMPL, STORAGE_IMPL_TYPE
 from timeit import default_timer as timer
 
 from rag.utils.redis_conn import REDIS_CONN
+from flask import jsonify
+from api.utils.health_utils import run_health_checks
+from common import globals
+
 
 @manager.route("/version", methods=["GET"])  # noqa: F821
 @login_required
@@ -97,7 +101,7 @@ def status():
     res = {}
     st = timer()
     try:
-        res["doc_engine"] = settings.docStoreConn.health()
+        res["doc_engine"] = globals.docStoreConn.health()
         res["doc_engine"]["elapsed"] = "{:.1f}".format((timer() - st) * 1000.0)
     except Exception as e:
         res["doc_engine"] = {
@@ -159,7 +163,7 @@ def status():
         task_executors = REDIS_CONN.smembers("TASKEXE")
         now = datetime.now().timestamp()
         for task_executor_id in task_executors:
-            heartbeats = REDIS_CONN.zrangebyscore(task_executor_id, now - 60*30, now)
+            heartbeats = REDIS_CONN.zrangebyscore(task_executor_id, now - 60 * 30, now)
             heartbeats = [json.loads(heartbeat) for heartbeat in heartbeats]
             task_executor_heartbeats[task_executor_id] = heartbeats
     except Exception:
@@ -169,6 +173,17 @@ def status():
     return get_json_result(data=res)
 
 
+@manager.route("/healthz", methods=["GET"])  # noqa: F821
+def healthz():
+    result, all_ok = run_health_checks()
+    return jsonify(result), (200 if all_ok else 500)
+
+
+@manager.route("/ping", methods=["GET"]) # noqa: F821
+def ping():
+    return "pong", 200
+
+
 @manager.route("/new_token", methods=["POST"])  # noqa: F821
 @login_required
 def new_token():
@@ -203,8 +218,8 @@ def new_token():
         tenant_id = [tenant for tenant in tenants if tenant.role == 'owner'][0].tenant_id
         obj = {
             "tenant_id": tenant_id,
-            "token": generate_confirmation_token(tenant_id),
+            "token": generate_confirmation_token(),
-            "beta": generate_confirmation_token(generate_confirmation_token(tenant_id)).replace("ragflow-", "")[:32],
+            "beta": generate_confirmation_token().replace("ragflow-", "")[:32],
             "create_time": current_timestamp(),
             "create_date": datetime_format(datetime.now()),
             "update_time": None,
@@ -260,7 +275,8 @@ def token_list():
         objs = [o.to_dict() for o in objs]
         for o in objs:
             if not o["beta"]:
-                o["beta"] = generate_confirmation_token(generate_confirmation_token(tenants[0].tenant_id)).replace("ragflow-", "")[:32]
+                o["beta"] = generate_confirmation_token().replace(
+                    "ragflow-", "")[:32]
                 APITokenService.filter_update([APIToken.tenant_id == tenant_id, APIToken.token == o["token"]], o)
         return get_json_result(data=objs)
     except Exception as e: