Docs: Update version references to v0.22.0 in READMEs and docs (#11211)

### What problem does this PR solve?

- Update version tags in README files (including translations) from
v0.21.1 to v0.22.0
- Modify Docker image references and documentation to reflect new
version
- Update version badges and image descriptions
- Maintain consistency across all language variants of README files

### Type of change

- [x] Documentation Update

.github/workflows/release.yml

@@ -19,7 +19,7 @@ jobs:
     runs-on: [ "self-hosted", "ragflow-test" ]
     steps:
       - name: Ensure workspace ownership
-        run: echo "chown -R $USER $GITHUB_WORKSPACE" && sudo chown -R $USER $GITHUB_WORKSPACE
+        run: echo "chown -R ${USER} ${GITHUB_WORKSPACE}" && sudo chown -R ${USER} ${GITHUB_WORKSPACE}
 
       # https://github.com/actions/checkout/blob/v3/README.md
       - name: Check out code
@@ -31,37 +31,37 @@ jobs:
 
       - name: Prepare release body
         run: |
-          if [[ $GITHUB_EVENT_NAME == 'create' ]]; then
+          if [[ ${GITHUB_EVENT_NAME} == "create" ]]; then
             RELEASE_TAG=${GITHUB_REF#refs/tags/}
-            if [[ $RELEASE_TAG == 'nightly' ]]; then
+            if [[ ${RELEASE_TAG} == "nightly" ]]; then
               PRERELEASE=true
             else
               PRERELEASE=false
             fi
-            echo "Workflow triggered by create tag: $RELEASE_TAG"
+            echo "Workflow triggered by create tag: ${RELEASE_TAG}"
           else
             RELEASE_TAG=nightly
             PRERELEASE=true
             echo "Workflow triggered by schedule"
           fi
-          echo "RELEASE_TAG=$RELEASE_TAG" >> $GITHUB_ENV
-          echo "PRERELEASE=$PRERELEASE" >> $GITHUB_ENV
+          echo "RELEASE_TAG=${RELEASE_TAG}" >> ${GITHUB_ENV}
+          echo "PRERELEASE=${PRERELEASE}" >> ${GITHUB_ENV}
           RELEASE_DATETIME=$(date --rfc-3339=seconds)
-          echo Release $RELEASE_TAG created from $GITHUB_SHA at $RELEASE_DATETIME > release_body.md
+          echo Release ${RELEASE_TAG} created from ${GITHUB_SHA} at ${RELEASE_DATETIME} > release_body.md
 
       - name: Move the existing mutable tag
         # https://github.com/softprops/action-gh-release/issues/171
         run: |
           git fetch --tags
-          if [[ $GITHUB_EVENT_NAME == 'schedule' ]]; then
+          if [[ ${GITHUB_EVENT_NAME} == "schedule" ]]; then
             # Determine if a given tag exists and matches a specific Git commit.
             # actions/checkout@v4 fetch-tags doesn't work when triggered by schedule
-            if [ "$(git rev-parse -q --verify "refs/tags/$RELEASE_TAG")" = "$GITHUB_SHA" ]; then
-              echo "mutable tag $RELEASE_TAG exists and matches $GITHUB_SHA"
+            if [ "$(git rev-parse -q --verify "refs/tags/${RELEASE_TAG}")" = "${GITHUB_SHA}" ]; then
+              echo "mutable tag ${RELEASE_TAG} exists and matches ${GITHUB_SHA}"
             else
-              git tag -f $RELEASE_TAG $GITHUB_SHA
-              git push -f origin $RELEASE_TAG:refs/tags/$RELEASE_TAG
-              echo "created/moved mutable tag $RELEASE_TAG to $GITHUB_SHA"
+              git tag -f ${RELEASE_TAG} ${GITHUB_SHA}
+              git push -f origin ${RELEASE_TAG}:refs/tags/${RELEASE_TAG}
+              echo "created/moved mutable tag ${RELEASE_TAG} to ${GITHUB_SHA}"
             fi
           fi
 
@@ -87,7 +87,7 @@ jobs:
 
       - name: Build and push image
         run: |
-          echo ${{ secrets.DOCKERHUB_TOKEN }} | sudo docker login --username infiniflow --password-stdin
+          sudo docker login --username infiniflow --password-stdin <<< ${{ secrets.DOCKERHUB_TOKEN }}
           sudo docker build --build-arg NEED_MIRROR=1 -t infiniflow/ragflow:${RELEASE_TAG} -f Dockerfile .
           sudo docker tag infiniflow/ragflow:${RELEASE_TAG} infiniflow/ragflow:latest
           sudo docker push infiniflow/ragflow:${RELEASE_TAG}

.github/workflows/tests.yml

@@ -9,8 +9,11 @@ on:
       - 'docs/**'
       - '*.md'
       - '*.mdx'
-  pull_request:
-    types: [ labeled, synchronize, reopened ]
+  # The only difference between pull_request and pull_request_target is the context in which the workflow runs:
+  # — pull_request_target workflows use the workflow files from the default branch, and secrets are available.
+  # — pull_request workflows use the workflow files from the pull request branch, and secrets are unavailable.
+  pull_request_target:
+    types: [ synchronize, ready_for_review ]
     paths-ignore:
       - 'docs/**'
       - '*.md'
@@ -28,7 +31,7 @@ jobs:
     name: ragflow_tests
     # https://docs.github.com/en/actions/using-jobs/using-conditions-to-control-job-execution
     # https://github.com/orgs/community/discussions/26261
-    if: ${{ github.event_name != 'pull_request' || contains(github.event.pull_request.labels.*.name, 'ci') }}
+    if: ${{ github.event_name != 'pull_request_target' || contains(github.event.pull_request.labels.*.name, 'ci') }}
     runs-on: [ "self-hosted", "ragflow-test" ]
     steps:
       # https://github.com/hmarr/debug-action
@@ -37,19 +40,20 @@ jobs:
       - name: Ensure workspace ownership
         run: |
           echo "Workflow triggered by ${{ github.event_name }}"
-          echo "chown -R $USER $GITHUB_WORKSPACE" && sudo chown -R $USER $GITHUB_WORKSPACE
+          echo "chown -R ${USER} ${GITHUB_WORKSPACE}" && sudo chown -R ${USER} ${GITHUB_WORKSPACE}
 
       # https://github.com/actions/checkout/issues/1781
       - name: Check out code
         uses: actions/checkout@v4
         with:
+          ref: ${{ (github.event_name == 'pull_request' || github.event_name == 'pull_request_target') && format('refs/pull/{0}/merge', github.event.pull_request.number) || github.sha }}
           fetch-depth: 0
           fetch-tags: true
 
       - name: Check workflow duplication
-        if: ${{ !cancelled() && !failure() && (github.event_name != 'pull_request' || contains(github.event.pull_request.labels.*.name, 'ci')) }}
+        if: ${{ !cancelled() && !failure() }}
         run: |
-          if [[ "$GITHUB_EVENT_NAME" != "pull_request" && "$GITHUB_EVENT_NAME" != "schedule" ]]; then
+          if [[ ${GITHUB_EVENT_NAME} != "pull_request_target" && ${GITHUB_EVENT_NAME} != "schedule" ]]; then
             HEAD=$(git rev-parse HEAD)
             # Find a PR that introduced a given commit
             gh auth login --with-token <<< "${{ secrets.GITHUB_TOKEN }}"
@@ -67,14 +71,14 @@ jobs:
                   gh run cancel ${GITHUB_RUN_ID}
                   while true; do
                     status=$(gh run view ${GITHUB_RUN_ID} --json status -q .status)
-                    [ "$status" = "completed" ] && break
+                    [ "${status}" = "completed" ] && break
                     sleep 5
                   done
                   exit 1
                 fi
               fi
             fi
-          else
+          elif [[ ${GITHUB_EVENT_NAME} == "pull_request_target" ]]; then
             PR_NUMBER=${{ github.event.pull_request.number }}
             PR_SHA_FP=${RUNNER_WORKSPACE_PREFIX}/artifacts/${GITHUB_REPOSITORY}/PR_${PR_NUMBER}
             # Calculate the hash of the current workspace content
@@ -93,18 +97,18 @@ jobs:
 
       - name: Build ragflow:nightly
         run: |
-          RUNNER_WORKSPACE_PREFIX=${RUNNER_WORKSPACE_PREFIX:-$HOME}
+          RUNNER_WORKSPACE_PREFIX=${RUNNER_WORKSPACE_PREFIX:-${HOME}}
           RAGFLOW_IMAGE=infiniflow/ragflow:${GITHUB_RUN_ID}
-          echo "RAGFLOW_IMAGE=${RAGFLOW_IMAGE}" >> $GITHUB_ENV
+          echo "RAGFLOW_IMAGE=${RAGFLOW_IMAGE}" >> ${GITHUB_ENV}
           sudo docker pull ubuntu:22.04
           sudo DOCKER_BUILDKIT=1 docker build --build-arg NEED_MIRROR=1 -f Dockerfile -t ${RAGFLOW_IMAGE} .
-          if [[ "$GITHUB_EVENT_NAME" == "schedule" ]]; then
+          if [[ ${GITHUB_EVENT_NAME} == "schedule" ]]; then
             export HTTP_API_TEST_LEVEL=p3
           else
             export HTTP_API_TEST_LEVEL=p2
           fi
-          echo "HTTP_API_TEST_LEVEL=${HTTP_API_TEST_LEVEL}" >> $GITHUB_ENV
-          echo "RAGFLOW_CONTAINER=${GITHUB_RUN_ID}-ragflow-cpu-1" >> $GITHUB_ENV
+          echo "HTTP_API_TEST_LEVEL=${HTTP_API_TEST_LEVEL}" >> ${GITHUB_ENV}
+          echo "RAGFLOW_CONTAINER=${GITHUB_RUN_ID}-ragflow-cpu-1" >> ${GITHUB_ENV}
 
       - name: Start ragflow:nightly
         run: |
@@ -154,7 +158,7 @@ jobs:
           echo -e "COMPOSE_PROFILES=\${COMPOSE_PROFILES},tei-cpu" >> docker/.env
           echo -e "TEI_MODEL=BAAI/bge-small-en-v1.5" >> docker/.env
           echo -e "RAGFLOW_IMAGE=${RAGFLOW_IMAGE}" >> docker/.env
-          echo "HOST_ADDRESS=http://host.docker.internal:${SVR_HTTP_PORT}" >> $GITHUB_ENV
+          echo "HOST_ADDRESS=http://host.docker.internal:${SVR_HTTP_PORT}" >> ${GITHUB_ENV}
 
           sudo docker compose -f docker/docker-compose.yml -p ${GITHUB_RUN_ID} up -d
           uv sync --python 3.10 --only-group test --no-default-groups --frozen && uv pip install sdk/python
@@ -189,7 +193,8 @@ jobs:
       - name: Stop ragflow:nightly
         if: always()  # always run this step even if previous steps failed
         run: |
-          sudo docker compose -f docker/docker-compose.yml -p ${GITHUB_RUN_ID} down -v
+          sudo docker compose -f docker/docker-compose.yml -p ${GITHUB_RUN_ID} down -v || true
+          sudo docker ps -a --filter "label=com.docker.compose.project=${GITHUB_RUN_ID}" -q | xargs -r sudo docker rm -f
 
       - name: Start ragflow:nightly
         run: |
@@ -226,5 +231,9 @@ jobs:
       - name: Stop ragflow:nightly
         if: always()  # always run this step even if previous steps failed
         run: |
-          sudo docker compose -f docker/docker-compose.yml -p ${GITHUB_RUN_ID} down -v
-          sudo docker rmi -f ${RAGFLOW_IMAGE:-NO_IMAGE} || true
+          # Sometimes `docker compose down` fail due to hang container, heavy load etc. Need to remove such containers to release resources(for example, listen ports).
+          sudo docker compose -f docker/docker-compose.yml -p ${GITHUB_RUN_ID} down -v || true
+          sudo docker ps -a --filter "label=com.docker.compose.project=${GITHUB_RUN_ID}" -q | xargs -r sudo docker rm -f
+          if [[ -n ${RAGFLOW_IMAGE} ]]; then
+            sudo docker rmi -f ${RAGFLOW_IMAGE}
+          fi

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

README.md

@@ -22,7 +22,7 @@
         <img alt="Static Badge" src="https://img.shields.io/badge/Online-Demo-4e6b99">
     </a>
     <a href="https://hub.docker.com/r/infiniflow/ragflow" target="_blank">
-        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.21.1">
+        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.22.0">
     </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">
@@ -61,8 +61,7 @@
 - 🔎 [System Architecture](#-system-architecture)
 - 🎬 [Get Started](#-get-started)
 - 🔧 [Configurations](#-configurations)
-- 🔧 [Build a docker image without embedding models](#-build-a-docker-image-without-embedding-models)
-- 🔧 [Build a docker image including embedding models](#-build-a-docker-image-including-embedding-models)
+- 🔧 [Build a Docker image](#-build-a-docker-image)
 - 🔨 [Launch service from source for development](#-launch-service-from-source-for-development)
 - 📚 [Documentation](#-documentation)
 - 📜 [Roadmap](#-roadmap)
@@ -86,6 +85,7 @@ Try our demo at [https://demo.ragflow.io](https://demo.ragflow.io).
 
 ## 🔥 Latest Updates
 
+- 2025-11-12 Supports data synchronization from Confluence, AWS S3, Discord, Google Drive.
 - 2025-10-23 Supports MinerU & Docling as document parsing methods.
 - 2025-10-15 Supports orchestrable ingestion pipeline.
 - 2025-08-08 Supports OpenAI's latest GPT-5 series models.
@@ -93,7 +93,6 @@ Try our demo at [https://demo.ragflow.io](https://demo.ragflow.io).
 - 2025-05-23 Adds a Python/JavaScript code executor component to Agent.
 - 2025-05-05 Supports cross-language query.
 - 2025-03-19 Supports using a multi-modal model to make sense of images within PDF or DOCX files.
-- 2025-02-28 Combined with Internet search (Tavily), supports reasoning like Deep Research for any LLMs.
 - 2024-12-18 Upgrades Document Layout Analysis model in DeepDoc.
 - 2024-08-22 Support text to SQL statements through RAG.
 
@@ -189,25 +188,29 @@ releases! 🌟
 > All Docker images are built for x86 platforms. We don't currently offer Docker images for ARM64.
 > If you are on an ARM64 platform, follow [this guide](https://ragflow.io/docs/dev/build_docker_image) to build a Docker image compatible with your system.
 
-> The command below downloads the `v0.21.1-slim` edition of the RAGFlow Docker image. See the following table for descriptions of different RAGFlow editions. To download a RAGFlow edition different from `v0.21.1-slim`, update the `RAGFLOW_IMAGE` variable accordingly in **docker/.env** before using `docker compose` to start the server.
+> The command below downloads the `v0.22.0` edition of the RAGFlow Docker image. See the following table for descriptions of different RAGFlow editions. To download a RAGFlow edition different from `v0.22.0`, update the `RAGFLOW_IMAGE` variable accordingly in **docker/.env** before using `docker compose` to start the server.
 
 ```bash
    $ cd ragflow/docker
-   # Use CPU for embedding and DeepDoc tasks:
+   
+   # Optional: use a stable tag (see releases: https://github.com/infiniflow/ragflow/releases), e.g.: git checkout v0.22.0
+
+   # Use CPU for DeepDoc tasks:
    $ docker compose -f docker-compose.yml up -d
 
-   # To use GPU to accelerate embedding and DeepDoc tasks:
+   # To use GPU to accelerate DeepDoc tasks:
    # sed -i '1i DEVICE=gpu' .env
    # docker compose -f docker-compose.yml up -d
 ```
 
-| RAGFlow image tag | Image size (GB) | Has embedding models? | Stable?                    |
-| ----------------- | --------------- | --------------------- | -------------------------- |
-| v0.21.1           | &approx;9       | ✔️                  | Stable release             |
-| v0.21.1-slim      | &approx;2       | ❌                    | Stable release             |
-| nightly           | &approx;2       | ❌                    | _Unstable_ nightly build |
+> Note: Prior to `v0.22.0`, we provided both images with embedding models and slim images without embedding models. Details as follows:
 
-> Note: Starting with `v0.22.0`, we ship only the slim edition and no longer append the **-slim** suffix to the image tag.
+| RAGFlow image tag | Image size (GB) | Has embedding models? | Stable?                  |
+| ----------------- | --------------- | --------------------- | ------------------------ |
+| v0.21.1           | &approx;9       | ✔️                    | Stable release           |
+| v0.21.1-slim      | &approx;2       | ❌                    | Stable release           |
+
+> Starting with `v0.22.0`, we ship only the slim edition and no longer append the **-slim** suffix to the image tag.
 
 4. Check the server status after having the server up and running:
 
@@ -288,7 +291,7 @@ RAGFlow uses Elasticsearch by default for storing full text and vectors. To swit
 > [!WARNING]
 > Switching to Infinity on a Linux/arm64 machine is not yet officially supported.
 
-## 🔧 Build a Docker image without embedding models
+## 🔧 Build a Docker image
 
 This image is approximately 2 GB in size and relies on external LLM and embedding services.
 

README_id.md

@@ -22,7 +22,7 @@
         <img alt="Lencana Daring" src="https://img.shields.io/badge/Online-Demo-4e6b99">
     </a>
     <a href="https://hub.docker.com/r/infiniflow/ragflow" target="_blank">
-        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.21.1">
+        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.22.0">
     </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">
@@ -61,8 +61,7 @@
 - 🔎 [Arsitektur Sistem](#-arsitektur-sistem)
 - 🎬 [Mulai](#-mulai)
 - 🔧 [Konfigurasi](#-konfigurasi)
-- 🔧 [Membangun Image Docker tanpa Model Embedding](#-membangun-image-docker-tanpa-model-embedding)
-- 🔧 [Membangun Image Docker dengan Model Embedding](#-membangun-image-docker-dengan-model-embedding)
+- 🔧 [Membangun Image Docker](#-membangun-docker-image)
 - 🔨 [Meluncurkan aplikasi dari Sumber untuk Pengembangan](#-meluncurkan-aplikasi-dari-sumber-untuk-pengembangan)
 - 📚 [Dokumentasi](#-dokumentasi)
 - 📜 [Peta Jalan](#-peta-jalan)
@@ -86,6 +85,7 @@ Coba demo kami di [https://demo.ragflow.io](https://demo.ragflow.io).
 
 ## 🔥 Pembaruan Terbaru
 
+- 2025-11-12 Mendukung sinkronisasi data dari Confluence, AWS S3, Discord, Google Drive.
 - 2025-10-23 Mendukung MinerU & Docling sebagai metode penguraian dokumen.
 - 2025-10-15 Dukungan untuk jalur data yang terorkestrasi.
 - 2025-08-08 Mendukung model seri GPT-5 terbaru dari OpenAI.
@@ -93,7 +93,6 @@ Coba demo kami di [https://demo.ragflow.io](https://demo.ragflow.io).
 - 2025-05-23 Menambahkan komponen pelaksana kode Python/JS ke Agen.
 - 2025-05-05 Mendukung kueri lintas bahasa.
 - 2025-03-19 Mendukung penggunaan model multi-modal untuk memahami gambar di dalam file PDF atau DOCX.
-- 2025-02-28 dikombinasikan dengan pencarian Internet (TAVILY), mendukung penelitian mendalam untuk LLM apa pun.
 - 2024-12-18 Meningkatkan model Analisis Tata Letak Dokumen di DeepDoc.
 - 2024-08-22 Dukungan untuk teks ke pernyataan SQL melalui RAG.
 
@@ -187,25 +186,29 @@ Coba demo kami di [https://demo.ragflow.io](https://demo.ragflow.io).
 > Semua gambar Docker dibangun untuk platform x86. Saat ini, kami tidak menawarkan gambar Docker untuk ARM64.
 > Jika Anda menggunakan platform ARM64, [silakan gunakan panduan ini untuk membangun gambar Docker yang kompatibel dengan sistem Anda](https://ragflow.io/docs/dev/build_docker_image).
 
-> Perintah di bawah ini mengunduh edisi v0.21.1 dari gambar Docker RAGFlow. Silakan merujuk ke tabel berikut untuk deskripsi berbagai edisi RAGFlow. Untuk mengunduh edisi RAGFlow yang berbeda dari v0.21.1, perbarui variabel RAGFLOW_IMAGE di docker/.env sebelum menggunakan docker compose untuk memulai server.
+> Perintah di bawah ini mengunduh edisi v0.22.0 dari gambar Docker RAGFlow. Silakan merujuk ke tabel berikut untuk deskripsi berbagai edisi RAGFlow. Untuk mengunduh edisi RAGFlow yang berbeda dari v0.22.0, perbarui variabel RAGFLOW_IMAGE di docker/.env sebelum menggunakan docker compose untuk memulai server.
 
 ```bash
    $ cd ragflow/docker
-   # Use CPU for embedding and DeepDoc tasks:
+   
+   # Opsional: gunakan tag stabil (lihat releases: https://github.com/infiniflow/ragflow/releases), contoh: git checkout v0.22.0
+
+   # Use CPU for DeepDoc tasks:
    $ docker compose -f docker-compose.yml up -d
 
-   # To use GPU to accelerate embedding and DeepDoc tasks:
+   # To use GPU to accelerate DeepDoc tasks:
    # sed -i '1i DEVICE=gpu' .env
    # docker compose -f docker-compose.yml up -d
 ```
 
-| RAGFlow image tag | Image size (GB) | Has embedding models? | Stable?                    |
-| ----------------- | --------------- | --------------------- | -------------------------- |
-| v0.21.1           | &approx;9       | ✔️                  | Stable release             |
-| v0.21.1-slim      | &approx;2       | ❌                    | Stable release             |
-| nightly           | &approx;2       | ❌                    | _Unstable_ nightly build |
+> Catatan: Sebelum `v0.22.0`, kami menyediakan image dengan model embedding dan image slim tanpa model embedding. Detailnya sebagai berikut:
 
-> Catatan: Mulai dari `v0.22.0`, kami hanya menyediakan edisi slim dan tidak lagi menambahkan akhiran **-slim** pada tag image.
+| RAGFlow image tag | Image size (GB) | Has embedding models? | Stable?                  |
+| ----------------- | --------------- | --------------------- | ------------------------ |
+| v0.21.1           | &approx;9       | ✔️                    | Stable release           |
+| v0.21.1-slim      | &approx;2       | ❌                    | Stable release           |
+
+> Mulai dari `v0.22.0`, kami hanya menyediakan edisi slim dan tidak lagi menambahkan akhiran **-slim** pada tag image.
 
 1. Periksa status server setelah server aktif dan berjalan:
 
@@ -260,7 +263,7 @@ Pembaruan konfigurasi ini memerlukan reboot semua kontainer agar efektif:
 > $ docker compose -f docker-compose.yml up -d
 > ```
 
-## 🔧 Membangun Docker Image tanpa Model Embedding
+## 🔧 Membangun Docker Image
 
 Image ini berukuran sekitar 2 GB dan bergantung pada aplikasi LLM eksternal dan embedding.
 

README_ja.md

@@ -22,7 +22,7 @@
         <img alt="Static Badge" src="https://img.shields.io/badge/Online-Demo-4e6b99">
     </a>
     <a href="https://hub.docker.com/r/infiniflow/ragflow" target="_blank">
-        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.21.1">
+        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.22.0">
     </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">
@@ -66,6 +66,7 @@
 
 ## 🔥 最新情報
 
+- 2025-11-12 Confluence、AWS S3、Discord、Google Drive からのデータ同期をサポートします。
 - 2025-10-23 ドキュメント解析方法として MinerU と Docling をサポートします。
 - 2025-10-15 オーケストレーションされたデータパイプラインのサポート。
 - 2025-08-08 OpenAI の最新 GPT-5 シリーズモデルをサポートします。
@@ -73,7 +74,6 @@
 - 2025-05-23 エージェントに Python/JS コードエグゼキュータコンポーネントを追加しました。
 - 2025-05-05 言語間クエリをサポートしました。
 - 2025-03-19 PDFまたはDOCXファイル内の画像を理解するために、多モーダルモデルを使用することをサポートします。
-- 2025-02-28 インターネット検索 (TAVILY) と組み合わせて、あらゆる LLM の詳細な調査をサポートします。
 - 2024-12-18 DeepDoc のドキュメント レイアウト分析モデルをアップグレードします。
 - 2024-08-22 RAG を介して SQL ステートメントへのテキストをサポートします。
 
@@ -166,28 +166,32 @@
 > 現在、公式に提供されているすべての Docker イメージは x86 アーキテクチャ向けにビルドされており、ARM64 用の Docker イメージは提供されていません。
 > ARM64 アーキテクチャのオペレーティングシステムを使用している場合は、[このドキュメント](https://ragflow.io/docs/dev/build_docker_image)を参照して Docker イメージを自分でビルドしてください。
 
-> 以下のコマンドは、RAGFlow Docker イメージの v0.21.1 エディションをダウンロードします。異なる RAGFlow エディションの説明については、以下の表を参照してください。v0.21.1 とは異なるエディションをダウンロードするには、docker/.env ファイルの RAGFLOW_IMAGE 変数を適宜更新し、docker compose を使用してサーバーを起動してください。
+> 以下のコマンドは、RAGFlow Docker イメージの v0.22.0 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。v0.22.0 とは異なるエディションをダウンロードするには、docker/.env ファイルの RAGFLOW_IMAGE 変数を適宜更新し、docker compose を使用してサーバーを起動してください。
 
 ```bash
    $ cd ragflow/docker
-   # Use CPU for embedding and DeepDoc tasks:
+
+   # 任意: 安定版タグを利用 (一覧: https://github.com/infiniflow/ragflow/releases) 例: git checkout v0.22.0
+
+   # Use CPU for DeepDoc tasks:
    $ docker compose -f docker-compose.yml up -d
 
-   # To use GPU to accelerate embedding and DeepDoc tasks:
+   # To use GPU to accelerate DeepDoc tasks:
    # sed -i '1i DEVICE=gpu' .env
    # docker compose -f docker-compose.yml up -d
 ```
 
-| RAGFlow image tag | Image size (GB) | Has embedding models? | Stable?                    |
-| ----------------- | --------------- | --------------------- | -------------------------- |
-| v0.21.1           | &approx;9       | ✔️                  | Stable release             |
-| v0.21.1-slim      | &approx;2       | ❌                    | Stable release             |
-| nightly           | &approx;2       | ❌                    | _Unstable_ nightly build |
+> 注意：`v0.22.0` より前のバージョンでは、embedding モデルを含むイメージと、embedding モデルを含まない slim イメージの両方を提供していました。詳細は以下の通りです：
 
-> 注意：`v0.22.0` 以降、当プロジェクトでは slim エディションのみを提供し、イメージタグに **-slim** サフィックスを付けなくなりました。
+| RAGFlow image tag | Image size (GB) | Has embedding models? | Stable?                  |
+| ----------------- | --------------- | --------------------- | ------------------------ |
+| v0.21.1           | &approx;9       | ✔️                    | Stable release           |
+| v0.21.1-slim      | &approx;2       | ❌                    | Stable release           |
 
-1. サーバーを立ち上げた後、サーバーの状態を確認する:
+> `v0.22.0` 以降、当プロジェクトでは slim エディションのみを提供し、イメージタグに **-slim** サフィックスを付けなくなりました。
 
+   1. サーバーを立ち上げた後、サーバーの状態を確認する:   
+   
    ```bash
    $ docker logs -f docker-ragflow-cpu-1
    ```
@@ -259,7 +263,7 @@ RAGFlow はデフォルトで Elasticsearch を使用して全文とベクトル
    > Linux/arm64 マシンでの Infinity への切り替えは正式にサポートされていません。
    >
 
-## 🔧 ソースコードで Docker イメージを作成（埋め込みモデルなし）
+## 🔧 ソースコードで Docker イメージを作成
 
 この Docker イメージのサイズは約 1GB で、外部の大モデルと埋め込みサービスに依存しています。
 

README_ko.md

@@ -22,7 +22,7 @@
         <img alt="Static Badge" src="https://img.shields.io/badge/Online-Demo-4e6b99">
     </a>
     <a href="https://hub.docker.com/r/infiniflow/ragflow" target="_blank">
-        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.21.1">
+        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.22.0">
     </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">
@@ -67,6 +67,7 @@
 
 ## 🔥 업데이트
 
+- 2025-11-12 Confluence, AWS S3, Discord, Google Drive에서 데이터 동기화를 지원합니다.
 - 2025-10-23 문서 파싱 방법으로 MinerU 및 Docling을 지원합니다.
 - 2025-10-15 조정된 데이터 파이프라인 지원.
 - 2025-08-08 OpenAI의 최신 GPT-5 시리즈 모델을 지원합니다.
@@ -74,7 +75,6 @@
 - 2025-05-23 Agent에 Python/JS 코드 실행기 구성 요소를 추가합니다.
 - 2025-05-05 언어 간 쿼리를 지원합니다.
 - 2025-03-19 PDF 또는 DOCX 파일 내의 이미지를 이해하기 위해 다중 모드 모델을 사용하는 것을 지원합니다.
-- 2025-02-28 인터넷 검색(TAVILY)과 결합되어 모든 LLM에 대한 심층 연구를 지원합니다.
 - 2024-12-18 DeepDoc의 문서 레이아웃 분석 모델 업그레이드.
 - 2024-08-22 RAG를 통해 SQL 문에 텍스트를 지원합니다.
 
@@ -168,25 +168,29 @@
 > 모든 Docker 이미지는 x86 플랫폼을 위해 빌드되었습니다. 우리는 현재 ARM64 플랫폼을 위한 Docker 이미지를 제공하지 않습니다.
 > ARM64 플랫폼을 사용 중이라면, [시스템과 호환되는 Docker 이미지를 빌드하려면 이 가이드를 사용해 주세요](https://ragflow.io/docs/dev/build_docker_image).
 
-   > 아래 명령어는 RAGFlow Docker 이미지의 v0.21.1 버전을 다운로드합니다. 다양한 RAGFlow 버전에 대한 설명은 다음 표를 참조하십시오. v0.21.1과 다른 RAGFlow 버전을 다운로드하려면, docker/.env 파일에서 RAGFLOW_IMAGE 변수를 적절히 업데이트한 후 docker compose를 사용하여 서버를 시작하십시오.
+   > 아래 명령어는 RAGFlow Docker 이미지의 v0.22.0 버전을 다운로드합니다. 다양한 RAGFlow 버전에 대한 설명은 다음 표를 참조하십시오. v0.22.0과 다른 RAGFlow 버전을 다운로드하려면, docker/.env 파일에서 RAGFLOW_IMAGE 변수를 적절히 업데이트한 후 docker compose를 사용하여 서버를 시작하십시오.
 
    ```bash
    $ cd ragflow/docker
-   # Use CPU for embedding and DeepDoc tasks:
+   
+   # Optional: use a stable tag (see releases: https://github.com/infiniflow/ragflow/releases), e.g.: git checkout v0.22.0
+
+   # Use CPU for DeepDoc tasks:
    $ docker compose -f docker-compose.yml up -d
 
-   # To use GPU to accelerate embedding and DeepDoc tasks:
+   # To use GPU to accelerate DeepDoc tasks:
    # sed -i '1i DEVICE=gpu' .env
    # docker compose -f docker-compose.yml up -d
-   ```
+```
 
-   | RAGFlow image tag | Image size (GB) | Has embedding models? | Stable?                  |
-   | ----------------- | --------------- | --------------------- | ------------------------ |
-   | v0.21.1           | &approx;9       | ✔️                    | Stable release           |
-   | v0.21.1-slim      | &approx;2       | ❌                    | Stable release           |
-   | nightly           | &approx;2       | ❌                    | _Unstable_ nightly build |
+> 참고: `v0.22.0` 이전 버전에서는 embedding 모델이 포함된 이미지와 embedding 모델이 포함되지 않은 slim 이미지를 모두 제공했습니다. 자세한 내용은 다음과 같습니다:
 
-   > 참고: `v0.22.0`부터는 slim 에디션만 배포하며 이미지 태그에 **-slim** 접미사를 더 이상 붙이지 않습니다.
+| RAGFlow image tag | Image size (GB) | Has embedding models? | Stable?                  |
+| ----------------- | --------------- | --------------------- | ------------------------ |
+| v0.21.1           | &approx;9       | ✔️                    | Stable release           |
+| v0.21.1-slim      | &approx;2       | ❌                    | Stable release           |
+
+> `v0.22.0`부터는 slim 에디션만 배포하며 이미지 태그에 **-slim** 접미사를 더 이상 붙이지 않습니다.
 
 1. 서버가 시작된 후 서버 상태를 확인하세요:
 
@@ -253,7 +257,7 @@ RAGFlow 는 기본적으로 Elasticsearch 를 사용하여 전체 텍스트 및
    > [!WARNING]
    > Linux/arm64 시스템에서 Infinity로 전환하는 것은 공식적으로 지원되지 않습니다.
 
-## 🔧 소스 코드로 Docker 이미지를 컴파일합니다(임베딩 모델 포함하지 않음)
+## 🔧 소스 코드로 Docker 이미지를 컴파일합니다
 
 이 Docker 이미지의 크기는 약 1GB이며, 외부 대형 모델과 임베딩 서비스에 의존합니다.
 

README_pt_br.md

@@ -22,7 +22,7 @@
         <img alt="Badge Estático" src="https://img.shields.io/badge/Online-Demo-4e6b99">
     </a>
     <a href="https://hub.docker.com/r/infiniflow/ragflow" target="_blank">
-        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.21.1">
+        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.22.0">
     </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">
@@ -86,6 +86,7 @@ Experimente nossa demo em [https://demo.ragflow.io](https://demo.ragflow.io).
 
 ## 🔥 Últimas Atualizações
 
+- 12-11-2025 Suporta a sincronização de dados do Confluence, AWS S3, Discord e Google Drive.
 - 23-10-2025 Suporta MinerU e Docling como métodos de análise de documentos.
 - 15-10-2025 Suporte para pipelines de dados orquestrados.
 - 08-08-2025 Suporta a mais recente série GPT-5 da OpenAI.
@@ -93,7 +94,6 @@ Experimente nossa demo em [https://demo.ragflow.io](https://demo.ragflow.io).
 - 23-05-2025 Adicione o componente executor de código Python/JS ao Agente.
 - 05-05-2025 Suporte a consultas entre idiomas.
 - 19-03-2025 Suporta o uso de um modelo multi-modal para entender imagens dentro de arquivos PDF ou DOCX.
-- 28-02-2025 combinado com a pesquisa na Internet (T AVI LY), suporta pesquisas profundas para qualquer LLM.
 - 18-12-2024 Atualiza o modelo de Análise de Layout de Documentos no DeepDoc.
 - 22-08-2024 Suporta conversão de texto para comandos SQL via RAG.
 
@@ -186,25 +186,29 @@ Experimente nossa demo em [https://demo.ragflow.io](https://demo.ragflow.io).
 > Todas as imagens Docker são construídas para plataformas x86. Atualmente, não oferecemos imagens Docker para ARM64.
 > Se você estiver usando uma plataforma ARM64, por favor, utilize [este guia](https://ragflow.io/docs/dev/build_docker_image) para construir uma imagem Docker compatível com o seu sistema.
 
-    > O comando abaixo baixa a edição`v0.21.1` da imagem Docker do RAGFlow. Consulte a tabela a seguir para descrições de diferentes edições do RAGFlow. Para baixar uma edição do RAGFlow diferente da `v0.21.1`, atualize a variável `RAGFLOW_IMAGE` conforme necessário no **docker/.env** antes de usar `docker compose` para iniciar o servidor.
+    > O comando abaixo baixa a edição`v0.22.0` 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 `v0.22.0`, atualize a variável `RAGFLOW_IMAGE` conforme necessário no **docker/.env** antes de usar `docker compose` para iniciar o servidor.
 
 ```bash
    $ cd ragflow/docker
-   # Use CPU for embedding and DeepDoc tasks:
+   
+   # Opcional: use uma tag estável (veja releases: https://github.com/infiniflow/ragflow/releases), ex.: git checkout v0.22.0
+
+   # Use CPU for DeepDoc tasks:
    $ docker compose -f docker-compose.yml up -d
 
-   # To use GPU to accelerate embedding and DeepDoc tasks:
+   # To use GPU to accelerate DeepDoc tasks:
    # sed -i '1i DEVICE=gpu' .env
    # docker compose -f docker-compose.yml up -d
 ```
 
-| Tag da imagem RAGFlow | Tamanho da imagem (GB) | Possui modelos de incorporação? | Estável?                      |
-| --------------------- | ---------------------- | --------------------------------- | ------------------------------ |
-| v0.21.1               | &approx;9              | ✔️                              | Lançamento estável           |
-| v0.21.1-slim          | &approx;2              | ❌                                | Lançamento estável           |
-| nightly               | &approx;2              | ❌                                | Construção noturna instável |
+> Nota: Antes da `v0.22.0`, fornecíamos imagens com modelos de embedding e imagens slim sem modelos de embedding. Detalhes a seguir:
 
-    > Observação: A partir da`v0.22.0`, distribuímos apenas a edição slim e não adicionamos mais o sufixo **-slim** às tags das imagens.
+| RAGFlow image tag | Image size (GB) | Has embedding models? | Stable?                  |
+| ----------------- | --------------- | --------------------- | ------------------------ |
+| v0.21.1           | &approx;9       | ✔️                    | Stable release           |
+| v0.21.1-slim      | &approx;2       | ❌                    | Stable release           |
+
+> A partir da `v0.22.0`, distribuímos apenas a edição slim e não adicionamos mais o sufixo **-slim** às tags das imagens.
 
 4. Verifique o status do servidor após tê-lo iniciado:
 
@@ -274,9 +278,9 @@ O RAGFlow usa o Elasticsearch por padrão para armazenar texto completo e vetore
    ```
 
 > [!ATENÇÃO]
-> A mudança para o Infinity em uma máquina Linux/arm64 ainda não é oficialmente suportada.
+ > A mudança para o Infinity em uma máquina Linux/arm64 ainda não é oficialmente suportada.
 
-## 🔧 Criar uma imagem Docker sem modelos de incorporação
+## 🔧 Criar uma imagem Docker
 
 Esta imagem tem cerca de 2 GB de tamanho e depende de serviços externos de LLM e incorporação.
 

README_tzh.md

@@ -22,7 +22,7 @@
         <img alt="Static Badge" src="https://img.shields.io/badge/Online-Demo-4e6b99">
     </a>
     <a href="https://hub.docker.com/r/infiniflow/ragflow" target="_blank">
-        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.21.1">
+        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.22.0">
     </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">
@@ -85,6 +85,7 @@
 
 ## 🔥 近期更新
 
+- 2025-11-12 支援從 Confluence、AWS S3、Discord、Google Drive 進行資料同步。
 - 2025-10-23 支援 MinerU 和 Docling 作為文件解析方法。
 - 2025-10-15 支援可編排的資料管道。
 - 2025-08-08 支援 OpenAI 最新的 GPT-5 系列模型。
@@ -92,7 +93,6 @@
 - 2025-05-23 為 Agent 新增 Python/JS 程式碼執行器元件。
 - 2025-05-05 支援跨語言查詢。
 - 2025-03-19 PDF和DOCX中的圖支持用多模態大模型去解析得到描述.
-- 2025-02-28 結合網路搜尋（Tavily），對於任意大模型實現類似 Deep Research 的推理功能.
 - 2024-12-18 升級了 DeepDoc 的文檔佈局分析模型。
 - 2024-08-22 支援用 RAG 技術實現從自然語言到 SQL 語句的轉換。
 
@@ -185,25 +185,29 @@
 > 所有 Docker 映像檔都是為 x86 平台建置的。目前，我們不提供 ARM64 平台的 Docker 映像檔。
 > 如果您使用的是 ARM64 平台，請使用 [這份指南](https://ragflow.io/docs/dev/build_docker_image) 來建置適合您系統的 Docker 映像檔。
 
-> 執行以下指令會自動下載 RAGFlow slim Docker 映像 `v0.21.1`。請參考下表查看不同 Docker 發行版的說明。如需下載不同於 `v0.21.1` 的 Docker 映像，請在執行 `docker compose` 啟動服務之前先更新 **docker/.env** 檔案內的 `RAGFLOW_IMAGE` 變數。
+> 執行以下指令會自動下載 RAGFlow Docker 映像 `v0.22.0`。請參考下表查看不同 Docker 發行版的說明。如需下載不同於 `v0.22.0` 的 Docker 映像，請在執行 `docker compose` 啟動服務之前先更新 **docker/.env** 檔案內的 `RAGFLOW_IMAGE` 變數。
 
 ```bash
    $ cd ragflow/docker
-   # Use CPU for embedding and DeepDoc tasks:
+   
+   # 可選：使用穩定版標籤（查看發佈：https://github.com/infiniflow/ragflow/releases），例：git checkout v0.22.0
+
+   # Use CPU for DeepDoc tasks:
    $ docker compose -f docker-compose.yml up -d
 
-   # To use GPU to accelerate embedding and DeepDoc tasks:
+   # To use GPU to accelerate DeepDoc tasks:
    # sed -i '1i DEVICE=gpu' .env
    # docker compose -f docker-compose.yml up -d
 ```
 
-| RAGFlow image tag | Image size (GB) | Has embedding models? | Stable?                    |
-| ----------------- | --------------- | --------------------- | -------------------------- |
-| v0.21.1           | &approx;9       | ✔️                  | Stable release             |
-| v0.21.1-slim      | &approx;2       | ❌                    | Stable release             |
-| nightly           | &approx;2       | ❌                    | _Unstable_ nightly build |
+> 注意：在 `v0.22.0` 之前的版本，我們會同時提供包含 embedding 模型的映像和不含 embedding 模型的 slim 映像。具體如下：
 
-> 注意：自 `v0.22.0` 起，我們僅發佈 slim 版本，並且不再在映像標籤後附加 **-slim** 後綴。
+| RAGFlow image tag | Image size (GB) | Has embedding models? | Stable?                  |
+| ----------------- | --------------- | --------------------- | ------------------------ |
+| v0.21.1           | &approx;9       | ✔️                    | Stable release           |
+| v0.21.1-slim      | &approx;2       | ❌                    | Stable release           |
+
+> 從 `v0.22.0` 開始，我們只發佈 slim 版本，並且不再在映像標籤後附加 **-slim** 後綴。
 
 > [!TIP]
 > 如果你遇到 Docker 映像檔拉不下來的問題，可以在 **docker/.env** 檔案內根據變數 `RAGFLOW_IMAGE` 的註解提示選擇華為雲或阿里雲的對應映像。
@@ -285,7 +289,7 @@ RAGFlow 預設使用 Elasticsearch 儲存文字和向量資料. 如果要切換
 > [!WARNING]
 > Infinity 目前官方並未正式支援在 Linux/arm64 架構下的機器上運行.
 
-## 🔧 原始碼編譯 Docker 映像（不含 embedding 模型）
+## 🔧 原始碼編譯 Docker 映像
 
 本 Docker 映像大小約 2 GB 左右並且依賴外部的大模型和 embedding 服務。
 

README_zh.md

@@ -22,7 +22,7 @@
         <img alt="Static Badge" src="https://img.shields.io/badge/Online-Demo-4e6b99">
     </a>
     <a href="https://hub.docker.com/r/infiniflow/ragflow" target="_blank">
-        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.21.1">
+        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.22.0">
     </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">
@@ -85,6 +85,7 @@
 
 ## 🔥 近期更新
 
+- 2025-11-12 支持从 Confluence、AWS S3、Discord、Google Drive 进行数据同步。
 - 2025-10-23 支持 MinerU 和 Docling 作为文档解析方法。
 - 2025-10-15 支持可编排的数据管道。
 - 2025-08-08 支持 OpenAI 最新的 GPT-5 系列模型。
@@ -92,7 +93,6 @@
 - 2025-05-23 Agent 新增 Python/JS 代码执行器组件。
 - 2025-05-05 支持跨语言查询。
 - 2025-03-19 PDF 和 DOCX 中的图支持用多模态大模型去解析得到描述.
-- 2025-02-28 结合互联网搜索（Tavily），对于任意大模型实现类似 Deep Research 的推理功能.
 - 2024-12-18 升级了 DeepDoc 的文档布局分析模型。
 - 2024-08-22 支持用 RAG 技术实现从自然语言到 SQL 语句的转换。
 
@@ -186,25 +186,29 @@
 > 请注意，目前官方提供的所有 Docker 镜像均基于 x86 架构构建，并不提供基于 ARM64 的 Docker 镜像。
 > 如果你的操作系统是 ARM64 架构，请参考[这篇文档](https://ragflow.io/docs/dev/build_docker_image)自行构建 Docker 镜像。
 
-   > 运行以下命令会自动下载 RAGFlow slim Docker 镜像 `v0.21.1`。请参考下表查看不同 Docker 发行版的描述。如需下载不同于 `v0.21.1` 的 Docker 镜像，请在运行 `docker compose` 启动服务之前先更新 **docker/.env** 文件内的 `RAGFLOW_IMAGE` 变量。
+   > 运行以下命令会自动下载 RAGFlow Docker 镜像 `v0.22.0`。请参考下表查看不同 Docker 发行版的描述。如需下载不同于 `v0.22.0` 的 Docker 镜像，请在运行 `docker compose` 启动服务之前先更新 **docker/.env** 文件内的 `RAGFLOW_IMAGE` 变量。
 
    ```bash
    $ cd ragflow/docker
-   # Use CPU for embedding and DeepDoc tasks:
+   
+   # 可选：使用稳定版本标签（查看发布：https://github.com/infiniflow/ragflow/releases），例如：git checkout v0.22.0
+
+   # Use CPU for DeepDoc tasks:
    $ docker compose -f docker-compose.yml up -d
 
-   # To use GPU to accelerate embedding and DeepDoc tasks:
+   # To use GPU to accelerate DeepDoc tasks:
    # sed -i '1i DEVICE=gpu' .env
    # docker compose -f docker-compose.yml up -d
    ```
+   
+   > 注意：在 `v0.22.0` 之前的版本，我们会同时提供包含 embedding 模型的镜像和不含 embedding 模型的 slim 镜像。具体如下：
 
    | RAGFlow image tag | Image size (GB) | Has embedding models? | Stable?                  |
    | ----------------- | --------------- | --------------------- | ------------------------ |
    | v0.21.1           | &approx;9       | ✔️                    | Stable release           |
    | v0.21.1-slim      | &approx;2       | ❌                    | Stable release           |
-   | nightly           | &approx;2       | ❌                    | _Unstable_ nightly build |
 
-   > 注意：从 `v0.22.0` 开始，我们只发布 slim 版本，并且不再在镜像标签后附加 **-slim** 后缀。
+   > 从 `v0.22.0` 开始，我们只发布 slim 版本，并且不再在镜像标签后附加 **-slim** 后缀。
 
    > [!TIP]
    > 如果你遇到 Docker 镜像拉不下来的问题，可以在 **docker/.env** 文件内根据变量 `RAGFLOW_IMAGE` 的注释提示选择华为云或者阿里云的相应镜像。
@@ -284,7 +288,7 @@ RAGFlow 默认使用 Elasticsearch 存储文本和向量数据. 如果要切换
 > [!WARNING]
 > Infinity 目前官方并未正式支持在 Linux/arm64 架构下的机器上运行.
 
-## 🔧 源码编译 Docker 镜像（不含 embedding 模型）
+## 🔧 源码编译 Docker 镜像
 
 本 Docker 镜像大小约 2 GB 左右并且依赖外部的大模型和 embedding 服务。
 

admin/client/README.md

@@ -48,7 +48,7 @@ It consists of a server-side Service and a command-line client (CLI), both imple
 1.  Ensure the Admin Service is running.
 2.  Install ragflow-cli.
     ```bash
-    pip install ragflow-cli==0.21.1
+    pip install ragflow-cli==0.22.0
     ```
 3.  Launch the CLI client:
     ```bash

admin/client/admin_client.py

@@ -23,6 +23,7 @@ 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
+import getpass
 
 GRAMMAR = r"""
 start: command
@@ -51,6 +52,7 @@ sql_command: list_services
            | revoke_permission
            | alter_user_role
            | show_user_permission
+           | show_version
 
 // meta command definition
 meta_command: "\\" meta_command_name [meta_args]
@@ -92,6 +94,7 @@ FOR: "FOR"i
 RESOURCES: "RESOURCES"i
 ON: "ON"i
 SET: "SET"i
+VERSION: "VERSION"i
 
 list_services: LIST SERVICES ";"
 show_service: SHOW SERVICE NUMBER ";"
@@ -120,6 +123,8 @@ revoke_permission: REVOKE action_list ON identifier FROM ROLE identifier ";"
 alter_user_role: ALTER USER quoted_string SET ROLE identifier ";"
 show_user_permission: SHOW USER PERMISSION quoted_string ";"
 
+show_version: SHOW VERSION ";"
+
 action_list: identifier ("," identifier)*
 
 identifier: WORD
@@ -246,6 +251,9 @@ class AdminTransformer(Transformer):
         user_name = items[3]
         return {"type": "show_user_permission", "user_name": user_name}
 
+    def show_version(self, items):
+        return {"type": "show_version"}
+
     def action_list(self, items):
         return items
 
@@ -359,7 +367,7 @@ class AdminCLI(Cmd):
             if single_command:
                 admin_passwd = arguments['password']
             else:
-                admin_passwd = input(f"password for {self.admin_account}: ").strip()
+                admin_passwd = getpass.getpass(f"password for {self.admin_account}: ").strip()
             try:
                 self.admin_password = encrypt(admin_passwd)
                 response = self.session.post(url, json={'email': self.admin_account, 'password': self.admin_password})
@@ -370,7 +378,7 @@ class AdminCLI(Cmd):
                         self.session.headers.update({
                             'Content-Type': 'application/json',
                             'Authorization': response.headers['Authorization'],
-                            'User-Agent': 'RAGFlow-CLI/0.21.1'
+                            'User-Agent': 'RAGFlow-CLI/0.22.0'
                         })
                         print("Authentication successful.")
                         return True
@@ -384,6 +392,21 @@ class AdminCLI(Cmd):
                 print(str(e))
                 print(f"Can't access {self.host}, port: {self.port}")
 
+    def _format_service_detail_table(self, data):
+        if not any([isinstance(v, list) for v in data.values()]):
+            # normal table
+            return data
+        # handle task_executor heartbeats map, for example {'name': [{'done': 2, 'now': timestamp1}, {'done': 3, 'now': timestamp2}]
+        task_executor_list = []
+        for k, v in data.items():
+            # display latest status
+            heartbeats = sorted(v, key=lambda x: x["now"], reverse=True)
+            task_executor_list.append({
+                "task_executor_name": k,
+                **heartbeats[0],
+            })
+        return task_executor_list
+
     def _print_table_simple(self, data):
         if not data:
             print("No data to print")
@@ -555,6 +578,8 @@ class AdminCLI(Cmd):
                 self._alter_user_role(command_dict)
             case 'show_user_permission':
                 self._show_user_permission(command_dict)
+            case 'show_version':
+                self._show_version(command_dict)
             case 'meta':
                 self._handle_meta_command(command_dict)
             case _:
@@ -585,7 +610,8 @@ class AdminCLI(Cmd):
                 if isinstance(res_data['message'], str):
                     print(res_data['message'])
                 else:
-                    self._print_table_simple(res_data['message'])
+                    data = self._format_service_detail_table(res_data['message'])
+                    self._print_table_simple(data)
             else:
                 print(f"Service {res_data['service_name']} is down, {res_data['message']}")
         else:
@@ -622,7 +648,9 @@ class AdminCLI(Cmd):
         response = self.session.get(url)
         res_json = response.json()
         if response.status_code == 200:
-            self._print_table_simple(res_json['data'])
+            table_data = res_json['data']
+            table_data.pop('avatar')
+            self._print_table_simple(table_data)
         else:
             print(f"Fail to get user {user_name}, code: {res_json['code']}, message: {res_json['message']}")
 
@@ -695,7 +723,10 @@ class AdminCLI(Cmd):
         response = self.session.get(url)
         res_json = response.json()
         if response.status_code == 200:
-            self._print_table_simple(res_json['data'])
+            table_data = res_json['data']
+            for t in table_data:
+                t.pop('avatar')
+            self._print_table_simple(table_data)
         else:
             print(f"Fail to get all datasets of {user_name}, code: {res_json['code']}, message: {res_json['message']}")
 
@@ -707,7 +738,10 @@ class AdminCLI(Cmd):
         response = self.session.get(url)
         res_json = response.json()
         if response.status_code == 200:
-            self._print_table_simple(res_json['data'])
+            table_data = res_json['data']
+            for t in table_data:
+                t.pop('avatar')
+            self._print_table_simple(table_data)
         else:
             print(f"Fail to get all agents of {user_name}, code: {res_json['code']}, message: {res_json['message']}")
 
@@ -861,6 +895,16 @@ class AdminCLI(Cmd):
             print(
                 f"Fail to show user: {user_name_str} permission, code: {res_json['code']}, message: {res_json['message']}")
 
+    def _show_version(self, command):
+        print("show_version")
+        url = f'http://{self.host}:{self.port}/api/v1/admin/version'
+        response = self.session.get(url)
+        res_json = response.json()
+        if response.status_code == 200:
+            self._print_table_simple(res_json['data'])
+        else:
+            print(f"Fail to show version, code: {res_json['code']}, message: {res_json['message']}")
+
     def _handle_meta_command(self, command):
         meta_command = command['command']
         args = command.get('args', [])

admin/client/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "ragflow-cli"
-version = "0.21.1"
+version = "0.22.0"
 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" }

admin/server/admin_server.py

@@ -23,13 +23,15 @@ import traceback
 from werkzeug.serving import run_simple
 from flask import Flask
 from routes import admin_bp
-from api.utils.log_utils import init_root_logger
-from api.constants import SERVICE_CONF
-from api import settings
+from common.log_utils import init_root_logger
+from common.constants import SERVICE_CONF
+from common.config_utils import show_configs
+from common import settings
 from config import load_configurations, SERVICE_CONFIGS
 from auth import init_default_admin, setup_auth
 from flask_session import Session
 from flask_login import LoginManager
+from common.versions import get_ragflow_version
 
 stop_event = threading.Event()
 
@@ -51,6 +53,8 @@ if __name__ == '__main__':
         os.environ.get("MAX_CONTENT_LENGTH", 1024 * 1024 * 1024)
     )
     Session(app)
+    logging.info(f'RAGFlow version: {get_ragflow_version()}')
+    show_configs()
     login_manager = LoginManager()
     login_manager.init_app(app)
     settings.init_settings()
@@ -65,7 +69,7 @@ if __name__ == '__main__':
             port=9381,
             application=app,
             threaded=True,
-            use_reloader=True,
+            use_reloader=False,
             use_debugger=True,
         )
     except Exception:

admin/server/auth.py

@@ -23,17 +23,15 @@ 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.common.base64 import encode_to_base64
 from api.db.services import UserService
-from api.db import ActiveEnum, StatusEnum
+from common.constants import ActiveEnum, StatusEnum
 from api.utils.crypt import decrypt
-from api.utils import get_uuid
+from common.misc_utils import get_uuid
 from common.time_utils import current_timestamp, datetime_format, get_format_time
-from api.utils.api_utils import (
-    construct_response,
-)
+from common.connection_utils import construct_response
+from common import settings
 
 
 def setup_auth(login_manager):

admin/server/config.py

@@ -21,7 +21,7 @@ from enum import Enum
 
 from pydantic import BaseModel
 from typing import Any
-from api.utils.configs import read_config
+from common.config_utils import read_config
 from urllib.parse import urlparse
 
 
@@ -183,11 +183,13 @@ class RAGFlowServerConfig(BaseConfig):
 
 
 class TaskExecutorConfig(BaseConfig):
+    message_queue_type: str
 
     def to_dict(self) -> dict[str, Any]:
         result = super().to_dict()
         if 'extra' not in result:
             result['extra'] = dict()
+        result['extra']['message_queue_type'] = self.message_queue_type
         return result
 
 
@@ -299,6 +301,15 @@ def load_configurations(config_path: str) -> list[BaseConfig]:
                 id_count += 1
             case "admin":
                 pass
+            case "task_executor":
+                name: str = 'task_executor'
+                host: str = v.get('host', '')
+                port: int = v.get('port', 0)
+                message_queue_type: str = v.get('message_queue_type')
+                config = TaskExecutorConfig(id=id_count, name=name, host=host, port=port, message_queue_type=message_queue_type,
+                                            service_type="task_executor", detail_func_name="check_task_executor_alive")
+                configurations.append(config)
+                id_count += 1
             case _:
                 logging.warning(f"Unknown configuration key: {k}")
                 continue

admin/server/routes.py

@@ -24,6 +24,7 @@ from responses import success_response, error_response
 from services import UserMgr, ServiceMgr, UserServiceMgr
 from roles import RoleMgr
 from api.common.exceptions import AdminException
+from common.versions import get_ragflow_version
 
 admin_bp = Blueprint('admin', __name__, url_prefix='/api/v1/admin')
 
@@ -369,3 +370,13 @@ def get_user_permission(user_name: str):
         return success_response(res)
     except Exception as e:
         return error_response(str(e), 500)
+
+@admin_bp.route('/version', methods=['GET'])
+@login_required
+@check_admin_auth
+def show_version():
+    try:
+        res = {"version": get_ragflow_version()}
+        return success_response(res)
+    except Exception as e:
+        return error_response(str(e), 500)

admin/server/services.py

@@ -17,7 +17,7 @@
 
 import re
 from werkzeug.security import check_password_hash
-from api.db import ActiveEnum
+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
@@ -52,6 +52,7 @@ class UserMgr:
         result = []
         for user in users:
             result.append({
+                'avatar': user.avatar,
                 'email': user.email,
                 'language': user.language,
                 'last_login_time': user.last_login_time,
@@ -170,7 +171,8 @@ class UserServiceMgr:
         return [{
             'title': r['title'],
             'permission': r['permission'],
-            'canvas_category': r['canvas_category'].split('_')[0]
+            'canvas_category': r['canvas_category'].split('_')[0],
+            'avatar': r['avatar']
         } for r in res]
 
 
@@ -190,6 +192,10 @@ class ServiceMgr:
                     config_dict['status'] = 'timeout'
             except Exception:
                 config_dict['status'] = 'timeout'
+            if not config_dict['host']:
+                config_dict['host'] = '-'
+            if not config_dict['port']:
+                config_dict['port'] = '-'
             result.append(config_dict)
         return result
 

agent/canvas.py

@@ -26,7 +26,9 @@ 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 api.db.services.task_service import has_canceled
+from common.misc_utils import get_uuid, hash_str2int
+from common.exceptions import TaskCanceledException
 from rag.prompts.generator import chunks_format
 from rag.utils.redis_conn import REDIS_CONN
 
@@ -126,6 +128,7 @@ class Graph:
             self.components[k]["obj"].reset()
         try:
             REDIS_CONN.delete(f"{self.task_id}-logs")
+            REDIS_CONN.delete(f"{self.task_id}-cancel")
         except Exception as e:
             logging.exception(e)
 
@@ -153,6 +156,33 @@ class Graph:
     def get_tenant_id(self):
         return self._tenant_id
 
+    def get_value_with_variable(self,value: str) -> Any:
+        pat = re.compile(r"\{* *\{([a-zA-Z:0-9]+@[A-Za-z0-9_.]+|sys\.[A-Za-z0-9_.]+|env\.[A-Za-z0-9_.]+)\} *\}*")
+        out_parts = []
+        last = 0
+
+        for m in pat.finditer(value):
+            out_parts.append(value[last:m.start()])
+            key = m.group(1)
+            v = self.get_variable_value(key)
+            if v is None:
+                rep = ""
+            elif isinstance(v, partial):
+                buf = []
+                for chunk in v():
+                    buf.append(chunk)
+                rep = "".join(buf)
+            elif isinstance(v, str):
+                rep = v
+            else:
+                rep = json.dumps(v, ensure_ascii=False)
+
+            out_parts.append(rep)
+            last = m.end()
+
+        out_parts.append(value[last:])
+        return("".join(out_parts))
+
     def get_variable_value(self, exp: str) -> Any:
         exp = exp.strip("{").strip("}").strip(" ").strip("{").strip("}")
         if exp.find("@") < 0:
@@ -161,7 +191,43 @@ class Graph:
         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)
+        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
+
+    def is_canceled(self) -> bool:
+        return has_canceled(self.task_id)
+
+    def cancel_task(self) -> bool:
+        try:
+            REDIS_CONN.set(f"{self.task_id}-cancel", "x")
+        except Exception as e:
+            logging.exception(e)
+            return False
+        return True
 
 
 class Canvas(Graph):
@@ -187,7 +253,7 @@ class Canvas(Graph):
             "sys.conversation_turns": 0,
             "sys.files": []
         }
-            
+
         self.retrieval = self.dsl["retrieval"]
         self.memory = self.dsl.get("memory", [])
 
@@ -204,18 +270,19 @@ class Canvas(Graph):
             self.retrieval = []
             self.memory = []
         for k in self.globals.keys():
-            if isinstance(self.globals[k], str):
-                self.globals[k] = ""
-            elif isinstance(self.globals[k], int):
-                self.globals[k] = 0
-            elif isinstance(self.globals[k], float):
-                self.globals[k] = 0
-            elif isinstance(self.globals[k], list):
-                self.globals[k] = []
-            elif isinstance(self.globals[k], dict):
-                self.globals[k] = {}
-            else:
-                self.globals[k] = None
+            if k.startswith("sys."):
+                if isinstance(self.globals[k], str):
+                    self.globals[k] = ""
+                elif isinstance(self.globals[k], int):
+                    self.globals[k] = 0
+                elif isinstance(self.globals[k], float):
+                    self.globals[k] = 0
+                elif isinstance(self.globals[k], list):
+                    self.globals[k] = []
+                elif isinstance(self.globals[k], dict):
+                    self.globals[k] = {}
+                else:
+                    self.globals[k] = None
 
     def run(self, **kwargs):
         st = time.perf_counter()
@@ -225,6 +292,14 @@ class Canvas(Graph):
         for k, cpn in self.components.items():
             self.components[k]["obj"].reset(True)
 
+        if kwargs.get("webhook_payload"):
+            for k, cpn in self.components.items():
+                if self.components[k]["obj"].component_name.lower() == "webhook":
+                    for kk, vv in kwargs["webhook_payload"].items():
+                        self.components[k]["obj"].set_output(kk, vv)
+
+            self.components[k]["obj"].reset(True)
+
         for k in kwargs.keys():
             if k in ["query", "user_id", "files"] and kwargs[k]:
                 if k == "files":
@@ -250,18 +325,37 @@ class Canvas(Graph):
             self.path.append("begin")
             self.retrieval.append({"chunks": [], "doc_aggs": []})
 
+        if self.is_canceled():
+            msg = f"Task {self.task_id} has been canceled before starting."
+            logging.info(msg)
+            raise TaskCanceledException(msg)
+
         yield decorate("workflow_started", {"inputs": kwargs.get("inputs")})
         self.retrieval.append({"chunks": {}, "doc_aggs": {}})
 
         def _run_batch(f, t):
+            if self.is_canceled():
+                msg = f"Task {self.task_id} has been canceled during batch execution."
+                logging.info(msg)
+                raise TaskCanceledException(msg)
+
             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] and self.path[0].lower().find("userfillup") < 0:
+                                self.path.pop(i)
+                                t -= 1
+                                break
+                        else:
+                            thr.append(executor.submit(cpn.invoke, **cpn.get_input()))
+                            i += 1
                 for t in thr:
                     t.result()
 
@@ -291,6 +385,7 @@ 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])
@@ -385,9 +480,10 @@ class Canvas(Graph):
                 for c in path:
                     o = self.get_component_obj(c)
                     if o.component_name.lower() == "userfillup":
+                        o.invoke()
                         another_inputs.update(o.get_input_elements())
                         if o.get_param("enable_tips"):
-                            tips = o.get_param("tips")
+                            tips = o.output("tips")
                 self.path = path
                 yield decorate("user_inputs", {"inputs": another_inputs, "tips": tips})
                 return
@@ -401,6 +497,14 @@ class Canvas(Graph):
                            "created_at": st,
                        })
             self.history.append(("assistant", self.get_component_obj(self.path[-1]).output()))
+        elif "Task has been canceled" in self.error:
+            yield decorate("workflow_finished",
+                       {
+                           "inputs": kwargs.get("inputs"),
+                           "outputs": "Task has been canceled",
+                           "elapsed_time": time.perf_counter() - st,
+                           "created_at": st,
+                       })
 
     def is_reff(self, exp: str) -> bool:
         exp = exp.strip("{").strip("}")

agent/component/__init__.py

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

agent/component/agent_with_tools.py

@@ -27,7 +27,7 @@ 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.generator import next_step, COMPLETE_TASK, analyze_task, \
     citation_prompt, reflect, rank_memories, kb_prompt, citation_plus, full_question, message_fit_in
 from rag.utils.mcp_tool_call_conn import MCPToolCallSession, mcp_tool_metadata_to_openai_tool
@@ -139,6 +139,9 @@ class Agent(LLM, ToolBase):
 
     @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 20*60)))
     def _invoke(self, **kwargs):
+        if self.check_if_canceled("Agent processing"):
+            return
+
         if kwargs.get("user_prompt"):
             usr_pmt = ""
             if kwargs.get("reasoning"):
@@ -152,13 +155,20 @@ class Agent(LLM, ToolBase):
             self._param.prompts = [{"role": "user", "content": usr_pmt}]
 
         if not self.tools:
+            if self.check_if_canceled("Agent processing"):
+                return
             return LLM._invoke(self, **kwargs)
 
         prompt, msg, user_defined_prompt = self._prepare_prompt_variables()
 
         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
 
@@ -166,6 +176,8 @@ class Agent(LLM, ToolBase):
         use_tools = []
         ans = ""
         for delta_ans, tk in self._react_with_tools_streamly(prompt, msg, use_tools, user_defined_prompt):
+            if self.check_if_canceled("Agent processing"):
+                return
             ans += delta_ans
 
         if ans.find("**ERROR**") >= 0:
@@ -186,12 +198,16 @@ class Agent(LLM, ToolBase):
         answer_without_toolcall = ""
         use_tools = []
         for delta_ans,_ in self._react_with_tools_streamly(prompt, msg, use_tools, user_defined_prompt):
+            if self.check_if_canceled("Agent streaming"):
+                return
+
             if delta_ans.find("**ERROR**") >= 0:
                 if self.get_exception_default_value():
                     self.set_output("content", self.get_exception_default_value())
                     yield self.get_exception_default_value()
                 else:
                     self.set_output("_ERROR", delta_ans)
+                    return
             answer_without_toolcall += delta_ans
             yield delta_ans
 
@@ -266,6 +282,8 @@ class Agent(LLM, ToolBase):
             st = timer()
             txt = ""
             for delta_ans in self._gen_citations(entire_txt):
+                if self.check_if_canceled("Agent streaming"):
+                    return
                 yield delta_ans, 0
                 txt += delta_ans
 
@@ -281,6 +299,8 @@ class Agent(LLM, ToolBase):
         task_desc = analyze_task(self.chat_mdl, prompt, user_request, tool_metas, user_defined_prompt)
         self.callback("analyze_task", {}, task_desc, elapsed_time=timer()-st)
         for _ in range(self._param.max_rounds + 1):
+            if self.check_if_canceled("Agent streaming"):
+                return
             response, tk = next_step(self.chat_mdl, hist, tool_metas, task_desc, user_defined_prompt)
             # self.callback("next_step", {}, str(response)[:256]+"...")
             token_count += tk
@@ -328,6 +348,8 @@ Instructions:
 6. Focus on delivering VALUE with the information already gathered
 Respond immediately with your final comprehensive answer.
         """
+        if self.check_if_canceled("Agent final instruction"):
+            return
         append_user_content(hist, final_instruction)
 
         for txt, tkcnt in complete():

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"
@@ -393,7 +393,7 @@ class ComponentParamBase(ABC):
 class ComponentBase(ABC):
     component_name: str
     thread_limiter = trio.CapacityLimiter(int(os.environ.get('MAX_CONCURRENT_CHATS', 10)))
-    variable_ref_patt = r"\{* *\{([a-zA-Z:0-9]+@[A-Za-z:0-9_.-]+|sys\.[a-z_]+)\} *\}*"
+    variable_ref_patt = r"\{* *\{([a-zA-Z:0-9]+@[A-Za-z0-9_.]+|sys\.[A-Za-z0-9_.]+|env\.[A-Za-z0-9_.]+)\} *\}*"
 
     def __str__(self):
         """
@@ -417,6 +417,20 @@ class ComponentBase(ABC):
         self._param = param
         self._param.check()
 
+    def is_canceled(self) -> bool:
+        return self._canvas.is_canceled()
+
+    def check_if_canceled(self, message: str = "") -> bool:
+        if self.is_canceled():
+            task_id = getattr(self._canvas, 'task_id', 'unknown')
+            log_message = f"Task {task_id} has been canceled"
+            if message:
+                log_message += f" during {message}"
+            logging.info(log_message)
+            self.set_output("_ERROR", "Task has been canceled")
+            return True
+        return False
+
     def invoke(self, **kwargs) -> dict[str, Any]:
         self.set_output("_created_time", time.perf_counter())
         try:
@@ -514,6 +528,7 @@ class ComponentBase(ABC):
     def get_param(self, name):
         if hasattr(self._param, name):
             return getattr(self._param, name)
+        return None
 
     def debug(self, **kwargs):
         return self._invoke(**kwargs)
@@ -521,7 +536,7 @@ class ComponentBase(ABC):
     def get_parent(self) -> Union[object, None]:
         pid = self._canvas.get_component(self._id).get("parent_id")
         if not pid:
-            return
+            return None
         return self._canvas.get_component(pid)["obj"]
 
     def get_upstream(self) -> List[str]:
@@ -546,7 +561,7 @@ class ComponentBase(ABC):
 
     def exception_handler(self):
         if not self._param.exception_method:
-            return
+            return None
         return {
             "goto": self._param.exception_goto,
             "default_value": self._param.exception_default_value

agent/component/begin.py

@@ -37,7 +37,13 @@ class Begin(UserFillUp):
     component_name = "Begin"
 
     def _invoke(self, **kwargs):
+        if self.check_if_canceled("Begin processing"):
+            return
+
         for k, v in kwargs.get("inputs", {}).items():
+            if self.check_if_canceled("Begin processing"):
+                return
+
             if isinstance(v, dict) and v.get("type", "").lower().find("file") >=0:
                 if v.get("optional") and v.get("value", None) is None:
                     v = None

agent/component/categorize.py

@@ -18,10 +18,10 @@ 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
 
 
@@ -98,6 +98,9 @@ class Categorize(LLM, ABC):
 
     @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60)))
     def _invoke(self, **kwargs):
+        if self.check_if_canceled("Categorize processing"):
+            return
+
         msg = self._canvas.get_history(self._param.message_history_window_size)
         if not msg:
             msg = [{"role": "user", "content": ""}]
@@ -114,10 +117,18 @@ class Categorize(LLM, ABC):
 ---- Real Data ----
 {} →
 """.format(" | ".join(["{}: \"{}\"".format(c["role"].upper(), re.sub(r"\n", "", c["content"], flags=re.DOTALL)) for c in msg]))
+
+        if self.check_if_canceled("Categorize processing"):
+            return
+
         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)}")
         if ERROR_PREFIX in ans:
             raise Exception(ans)
+
+        if self.check_if_canceled("Categorize processing"):
+            return
+
         # Count the number of times each category appears in the answer.
         category_counts = {}
         for c in self._param.category_description.keys():

agent/component/data_operations.py

@@ -0,0 +1,203 @@
+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.query = []
+        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.query or [])
+            for k, o in self.get_input_elements_from_text(input_item).items()
+        }
+
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60)))
+    def _invoke(self, **kwargs):
+        self.input_objects=[]
+        inputs = getattr(self._param, "query", None)
+        if not isinstance(inputs, (list, tuple)):
+            inputs = [inputs]
+        for input_ref in inputs:
+            input_object=self._canvas.get_variable_value(input_ref)
+            self.set_input_value(input_ref, input_object)
+            if input_object is None:
+                continue
+            if isinstance(input_object,dict):
+                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 == "recursive_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"))
+        target = self._canvas.get_value_with_variable(target) or target
+        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] = self._canvas.get_value_with_variable(item.get("value")) or 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/fillup.py

@@ -13,7 +13,11 @@
 #  See the License for the specific language governing permissions and
 #  limitations under the License.
 #
-from agent.component.base import ComponentBase, ComponentParamBase
+import json
+import re
+from functools import partial
+
+from agent.component.base import ComponentParamBase, ComponentBase
 
 
 class UserFillUpParam(ComponentParamBase):
@@ -31,10 +35,35 @@ class UserFillUp(ComponentBase):
     component_name = "UserFillUp"
 
     def _invoke(self, **kwargs):
+        if self.check_if_canceled("UserFillUp processing"):
+            return
+
+        if self._param.enable_tips:
+            content = self._param.tips
+            for k, v in self.get_input_elements_from_text(self._param.tips).items():
+                v = v["value"]
+                ans = ""
+                if isinstance(v, partial):
+                    for t in v():
+                        ans += t
+                elif isinstance(v, list):
+                    ans = ",".join([str(vv) for vv in v])
+                elif not isinstance(v, str):
+                    try:
+                        ans = json.dumps(v, ensure_ascii=False)
+                    except Exception:
+                        pass
+                else:
+                    ans = v
+                if not ans:
+                    ans = ""
+                content = re.sub(r"\{%s\}"%k, ans, content)
+
+            self.set_output("tips", content)
         for k, v in kwargs.get("inputs", {}).items():
+            if self.check_if_canceled("UserFillUp processing"):
+                return
             self.set_output(k, v)
 
     def thoughts(self) -> str:
         return "Waiting for your input..."
-
-

agent/component/invoke.py

@@ -23,7 +23,7 @@ from abc import ABC
 import requests
 
 from agent.component.base import ComponentBase, ComponentParamBase
-from api.utils.api_utils import timeout
+from common.connection_utils import timeout
 from deepdoc.parser import HtmlParser
 
 
@@ -56,6 +56,9 @@ class Invoke(ComponentBase, ABC):
 
     @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 3)))
     def _invoke(self, **kwargs):
+        if self.check_if_canceled("Invoke processing"):
+            return
+
         args = {}
         for para in self._param.variables:
             if para.get("value"):
@@ -89,6 +92,9 @@ class Invoke(ComponentBase, ABC):
 
         last_e = ""
         for _ in range(self._param.max_retries + 1):
+            if self.check_if_canceled("Invoke processing"):
+                return
+
             try:
                 if method == "get":
                     response = requests.get(url=url, params=args, headers=headers, proxies=proxies, timeout=self._param.timeout)
@@ -121,6 +127,9 @@ class Invoke(ComponentBase, ABC):
 
                 return self.output("result")
             except Exception as e:
+                if self.check_if_canceled("Invoke processing"):
+                    return
+
                 last_e = e
                 logging.exception(f"Http request error: {e}")
                 time.sleep(self._param.delay_after_error)

agent/component/iteration.py

@@ -16,6 +16,13 @@
 from abc import ABC
 from agent.component.base import ComponentBase, ComponentParamBase
 
+"""
+class VariableModel(BaseModel):
+    data_type: Annotated[Literal["string", "number", "Object", "Boolean", "Array<string>", "Array<number>", "Array<object>", "Array<boolean>"], Field(default="Array<string>")]
+    input_mode: Annotated[Literal["constant", "variable"], Field(default="constant")]
+    value: Annotated[Any, Field(default=None)]
+    model_config = ConfigDict(extra="forbid")
+"""
 
 class IterationParam(ComponentParamBase):
     """
@@ -49,6 +56,9 @@ class Iteration(ComponentBase, ABC):
                 return cid
 
     def _invoke(self, **kwargs):
+        if self.check_if_canceled("Iteration processing"):
+            return
+
         arr = self._canvas.get_variable_value(self._param.items_ref)
         if not isinstance(arr, list):
             self.set_output("_ERROR", self._param.items_ref + " must be an array, but its type is "+str(type(arr)))

agent/component/iterationitem.py

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

agent/component/llm.py

@@ -21,12 +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 rag.prompts.generator import tool_call_summary, message_fit_in, citation_prompt
+from common.connection_utils import timeout
+from rag.prompts.generator import tool_call_summary, message_fit_in, citation_prompt, structured_output_prompt
 
 
 class LLMParam(ComponentParamBase):
@@ -207,6 +207,9 @@ class LLM(ComponentBase):
 
     @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60)))
     def _invoke(self, **kwargs):
+        if self.check_if_canceled("LLM processing"):
+            return
+
         def clean_formated_answer(ans: str) -> str:
             ans = re.sub(r"^.*</think>", "", ans, flags=re.DOTALL)
             ans = re.sub(r"^.*```json", "", ans, flags=re.DOTALL)
@@ -214,11 +217,18 @@ class LLM(ComponentBase):
 
         prompt, msg, _ = self._prepare_prompt_variables()
         error: str = ""
-
-        if self._param.output_structure:
-            prompt += "\nThe output MUST follow this JSON format:\n"+json.dumps(self._param.output_structure, ensure_ascii=False, indent=2)
-            prompt += "\nRedundant information is FORBIDDEN."
+        output_structure=None
+        try:
+            output_structure = self._param.outputs['structured']
+        except Exception:
+            pass
+        if output_structure:
+            schema=json.dumps(output_structure, ensure_ascii=False, indent=2)
+            prompt += structured_output_prompt(schema)
             for _ in range(self._param.max_retries+1):
+                if self.check_if_canceled("LLM processing"):
+                    return
+
                 _, msg = message_fit_in([{"role": "system", "content": prompt}, *msg], int(self.chat_mdl.max_length * 0.97))
                 error = ""
                 ans = self._generate(msg)
@@ -228,7 +238,7 @@ 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"})
@@ -239,11 +249,14 @@ 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
 
         for _ in range(self._param.max_retries+1):
+            if self.check_if_canceled("LLM processing"):
+                return
+
             _, msg = message_fit_in([{"role": "system", "content": prompt}, *msg], int(self.chat_mdl.max_length * 0.97))
             error = ""
             ans = self._generate(msg)
@@ -265,6 +278,9 @@ class LLM(ComponentBase):
         _, msg = message_fit_in([{"role": "system", "content": prompt}, *msg], int(self.chat_mdl.max_length * 0.97))
         answer = ""
         for ans in self._generate_streamly(msg):
+            if self.check_if_canceled("LLM streaming"):
+                return
+
             if ans.find("**ERROR**") >= 0:
                 if self.get_exception_default_value():
                     self.set_output("content", self.get_exception_default_value())
@@ -283,4 +299,4 @@ class LLM(ComponentBase):
 
     def thoughts(self) -> str:
         _, msg,_ = self._prepare_prompt_variables()
-        return "⌛Give me a moment—starting from: \n\n" + re.sub(r"(User's query:|[\\]+)", '', msg[-1]['content'], flags=re.DOTALL) + "\n\nI’ll figure out our best next move."
+        return "⌛Give me a moment—starting from: \n\n" + re.sub(r"(User's query:|[\\]+)", '', msg[-1]['content'], flags=re.DOTALL) + "\n\nI’ll figure out our best next move."

agent/component/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):
@@ -49,6 +49,9 @@ class MessageParam(ComponentParamBase):
 class Message(ComponentBase):
     component_name = "Message"
 
+    def get_input_elements(self) -> dict[str, Any]:
+        return self.get_input_elements_from_text("".join(self._param.content))
+
     def get_kwargs(self, script:str, kwargs:dict = {}, delimiter:str=None) -> tuple[str, dict[str, str | list | Any]]:
         for k,v in self.get_input_elements_from_text(script).items():
             if k in kwargs:
@@ -86,6 +89,9 @@ class Message(ComponentBase):
         all_content = ""
         cache = {}
         for r in re.finditer(self.variable_ref_patt, rand_cnt, flags=re.DOTALL):
+            if self.check_if_canceled("Message streaming"):
+                return
+
             all_content += rand_cnt[s: r.start()]
             yield rand_cnt[s: r.start()]
             s = r.end()
@@ -96,26 +102,33 @@ class Message(ComponentBase):
                 continue
 
             v = self._canvas.get_variable_value(exp)
-            if not v:
+            if v is None:
                 v = ""
             if isinstance(v, partial):
                 cnt = ""
                 for t in v():
+                    if self.check_if_canceled("Message streaming"):
+                        return
+
                     all_content += t
                     cnt += t
                     yield t
-
+                self.set_input_value(exp, cnt)
                 continue
             elif not isinstance(v, str):
                 try:
-                    v = json.dumps(v, ensure_ascii=False, indent=2)
+                    v = json.dumps(v, ensure_ascii=False)
                 except Exception:
                     v = str(v)
             yield v
+            self.set_input_value(exp, v)
             all_content += v
             cache[exp] = v
 
         if s < len(rand_cnt):
+            if self.check_if_canceled("Message streaming"):
+                return
+
             all_content += rand_cnt[s: ]
             yield rand_cnt[s: ]
 
@@ -129,6 +142,9 @@ class Message(ComponentBase):
 
     @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60)))
     def _invoke(self, **kwargs):
+        if self.check_if_canceled("Message processing"):
+            return
+
         rand_cnt = random.choice(self._param.content)
         if self._param.stream and not self._is_jinjia2(rand_cnt):
             self.set_output("content", partial(self._stream, rand_cnt))
@@ -141,6 +157,9 @@ class Message(ComponentBase):
         except Exception:
             pass
 
+        if self.check_if_canceled("Message processing"):
+            return
+
         for n, v in kwargs.items():
             content = re.sub(n, v, content)
 

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 {
@@ -58,17 +63,24 @@ class StringTransform(Message, ABC):
 
     @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60)))
     def _invoke(self, **kwargs):
+        if self.check_if_canceled("StringTransform processing"):
+            return
+
         if self._param.method == "split":
             self._split(kwargs.get("line"))
         else:
             self._merge(kwargs)
 
     def _split(self, line:str|None = None):
+        if self.check_if_canceled("StringTransform split processing"):
+            return
+
         var = self._canvas.get_variable_value(self._param.split_ref) if not line else line
         if not var:
             var = ""
         assert isinstance(var, str), "The input variable is not a string: {}".format(type(var))
         self.set_input_value(self._param.split_ref, var)
+
         res = []
         for i,s in enumerate(re.split(r"(%s)"%("|".join([re.escape(d) for d in self._param.delimiters])), var, flags=re.DOTALL)):
             if i % 2 == 1:
@@ -77,6 +89,9 @@ class StringTransform(Message, ABC):
         self.set_output("result", res)
 
     def _merge(self, kwargs:dict[str, str] = {}):
+        if self.check_if_canceled("StringTransform merge processing"):
+            return
+
         script = self._param.script
         script, kwargs = self.get_kwargs(script, kwargs, self._param.delimiters[0])
 

agent/component/switch.py

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

agent/component/varaiable_aggregator.py

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

agent/component/webhook.py

@@ -0,0 +1,38 @@
+#
+#  Copyright 2024 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+#
+from agent.component.base import ComponentParamBase, ComponentBase
+
+
+class WebhookParam(ComponentParamBase):
+
+    """
+    Define the Begin component parameters.
+    """
+    def __init__(self):
+        super().__init__()
+
+    def get_input_form(self) -> dict[str, dict]:
+        return getattr(self, "inputs")
+
+
+class Webhook(ComponentBase):
+    component_name = "Webhook"
+
+    def _invoke(self, **kwargs):
+        pass
+
+    def thoughts(self) -> str:
+        return ""

agent/templates/user_interaction.json

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

agent/test/client.py

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

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

agent/tools/base.py

@@ -20,7 +20,7 @@ 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.generator import kb_prompt
 from rag.utils.mcp_tool_call_conn import MCPToolCallSession
@@ -125,6 +125,9 @@ class ToolBase(ComponentBase):
         return self._param.get_meta()
 
     def invoke(self, **kwargs):
+        if self.check_if_canceled("Tool processing"):
+            return
+
         self.set_output("_created_time", time.perf_counter())
         try:
             res = self._invoke(**kwargs)
@@ -170,4 +173,4 @@ class ToolBase(ComponentBase):
         self.set_output("formalized_content", "\n".join(kb_prompt({"chunks": chunks, "doc_aggs": aggs}, 200000, True)))
 
     def thoughts(self) -> str:
-        return self._canvas.get_component_name(self._id) + " is running..."
+        return self._canvas.get_component_name(self._id) + " is running..."

agent/tools/code_exec.py

@@ -21,8 +21,8 @@ from strenum import StrEnum
 from typing import Optional
 from pydantic import BaseModel, Field, field_validator
 from agent.tools.base import ToolParamBase, ToolBase, ToolMeta
-from api import settings
-from api.utils.api_utils import timeout
+from common.connection_utils import timeout
+from common import settings
 
 
 class Language(StrEnum):
@@ -131,10 +131,14 @@ class CodeExec(ToolBase, ABC):
 
     @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60)))
     def _invoke(self, **kwargs):
+        if self.check_if_canceled("CodeExec processing"):
+            return
+
         lang = kwargs.get("lang", self._param.lang)
         script = kwargs.get("script", self._param.script)
         arguments = {}
         for k, v in self._param.arguments.items():
+
             if kwargs.get(k):
                 arguments[k] = kwargs[k]
                 continue
@@ -149,15 +153,28 @@ class CodeExec(ToolBase, ABC):
     def _execute_code(self, language: str, code: str, arguments: dict):
         import requests
 
+        if self.check_if_canceled("CodeExec execution"):
+            return
+
         try:
             code_b64 = self._encode_code(code)
             code_req = CodeExecutionRequest(code_b64=code_b64, language=language, arguments=arguments).model_dump()
         except Exception as e:
+            if self.check_if_canceled("CodeExec execution"):
+                return
+
             self.set_output("_ERROR", "construct code request error: " + str(e))
 
         try:
+            if self.check_if_canceled("CodeExec execution"):
+                return "Task has been canceled"
+
             resp = requests.post(url=f"http://{settings.SANDBOX_HOST}:9385/run", json=code_req, timeout=int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60)))
             logging.info(f"http://{settings.SANDBOX_HOST}:9385/run,  code_req: {code_req}, resp.status_code {resp.status_code}:")
+
+            if self.check_if_canceled("CodeExec execution"):
+                return "Task has been canceled"
+
             if resp.status_code != 200:
                 resp.raise_for_status()
             body = resp.json()
@@ -173,16 +190,25 @@ class CodeExec(ToolBase, ABC):
                 logging.info(f"http://{settings.SANDBOX_HOST}:9385/run -> {rt}")
                 if isinstance(rt, tuple):
                     for i, (k, o) in enumerate(self._param.outputs.items()):
+                        if self.check_if_canceled("CodeExec execution"):
+                            return
+
                         if k.find("_") == 0:
                             continue
                         o["value"] = rt[i]
                 elif isinstance(rt, dict):
                     for i, (k, o) in enumerate(self._param.outputs.items()):
+                        if self.check_if_canceled("CodeExec execution"):
+                            return
+
                         if k not in rt or k.find("_") == 0:
                             continue
                         o["value"] = rt[k]
                 else:
                     for i, (k, o) in enumerate(self._param.outputs.items()):
+                        if self.check_if_canceled("CodeExec execution"):
+                            return
+
                         if k.find("_") == 0:
                             continue
                         o["value"] = rt
@@ -190,6 +216,9 @@ class CodeExec(ToolBase, ABC):
                 self.set_output("_ERROR", "There is no response from sandbox")
 
         except Exception as e:
+            if self.check_if_canceled("CodeExec execution"):
+                return
+
             self.set_output("_ERROR", "Exception executing code: " + str(e))
 
         return self.output()

agent/tools/crawler.py

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

agent/tools/deepl.py

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

agent/tools/duckduckgo.py

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

agent/tools/email.py

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

agent/tools/exesql.py

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

agent/tools/github.py

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

agent/tools/google.py

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

agent/tools/googlescholar.py

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

agent/tools/jin10.py

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

agent/tools/pubmed.py

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

agent/tools/qweather.py

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

agent/tools/retrieval.py

@@ -13,17 +13,19 @@
 #  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 settings
+from common.connection_utils import timeout
 from rag.app.tag import label_question
 from rag.prompts.generator import cross_languages, kb_prompt, gen_meta_filter
 
@@ -80,8 +82,12 @@ class Retrieval(ToolBase, ABC):
 
     @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12)))
     def _invoke(self, **kwargs):
+        if self.check_if_canceled("Retrieval processing"):
+            return
+
         if not kwargs.get("query"):
             self.set_output("formalized_content", self._param.empty_response)
+            return
 
         kb_ids: list[str] = []
         for id in self._param.kb_ids:
@@ -120,7 +126,7 @@ 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)
@@ -131,7 +137,35 @@ class Retrieval(ToolBase, ABC):
                 if not doc_ids:
                     doc_ids = None
             elif self._param.meta_data_filter.get("method") == "manual":
-                doc_ids.extend(meta_filter(metas, self._param.meta_data_filter["manual"]))
+                filters=self._param.meta_data_filter["manual"]
+                for flt in filters:
+                    pat = re.compile(self.variable_ref_patt)
+                    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
 
@@ -154,9 +188,14 @@ class Retrieval(ToolBase, ABC):
                 rerank_mdl=rerank_mdl,
                 rank_feature=label_question(query, kbs),
             )
+            if self.check_if_canceled("Retrieval processing"):
+                return
+
             if self._param.toc_enhance:
                 chat_mdl = LLMBundle(self._canvas._tenant_id, LLMType.CHAT)
                 cks = settings.retriever.retrieval_by_toc(query, kbinfos["chunks"], [kb.tenant_id for kb in kbs], chat_mdl, self._param.top_n)
+                if self.check_if_canceled("Retrieval processing"):
+                    return
                 if cks:
                     kbinfos["chunks"] = cks
             if self._param.use_kg:
@@ -165,6 +204,8 @@ class Retrieval(ToolBase, ABC):
                                                        kb_ids,
                                                        embd_mdl,
                                                        LLMBundle(self._canvas.get_tenant_id(), LLMType.CHAT))
+                if self.check_if_canceled("Retrieval processing"):
+                    return
                 if ck["content_with_weight"]:
                     kbinfos["chunks"].insert(0, ck)
         else:
@@ -172,6 +213,8 @@ class Retrieval(ToolBase, ABC):
 
         if self._param.use_kg and kbs:
             ck = settings.kg_retriever.retrieval(query, [kb.tenant_id for kb in kbs], filtered_kb_ids, embd_mdl, LLMBundle(kbs[0].tenant_id, LLMType.CHAT))
+            if self.check_if_canceled("Retrieval processing"):
+                return
             if ck["content_with_weight"]:
                 ck["content"] = ck["content_with_weight"]
                 del ck["content_with_weight"]

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

agent/tools/tavily.py

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

agent/tools/tushare.py

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

agent/tools/wencai.py

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

agent/tools/wikipedia.py

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

agent/tools/yahoofinance.py

@@ -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):
@@ -74,15 +74,24 @@ class YahooFinance(ToolBase, ABC):
 
     @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 60)))
     def _invoke(self, **kwargs):
+        if self.check_if_canceled("YahooFinance processing"):
+            return
+
         if not kwargs.get("stock_code"):
             self.set_output("report", "")
             return ""
 
         last_e = ""
         for _ in range(self._param.max_retries+1):
+            if self.check_if_canceled("YahooFinance processing"):
+                return
+
             yohoo_res = []
             try:
                 msft = yf.Ticker(kwargs["stock_code"])
+                if self.check_if_canceled("YahooFinance processing"):
+                    return
+
                 if self._param.info:
                     yohoo_res.append("# Information:\n" + pd.Series(msft.info).to_markdown() + "\n")
                 if self._param.history:
@@ -100,6 +109,9 @@ class YahooFinance(ToolBase, ABC):
                 self.set_output("report", "\n\n".join(yohoo_res))
                 return self.output("report")
             except Exception as e:
+                if self.check_if_canceled("YahooFinance processing"):
+                    return
+
                 last_e = e
                 logging.exception(f"YahooFinance error: {e}")
                 time.sleep(self._param.delay_after_error)

api/apps/__init__.py

@@ -24,7 +24,7 @@ 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.json_encode import CustomJSONEncoder
@@ -33,7 +33,7 @@ from api.utils import commands
 from flask_mail import Mail
 from flask_session import Session
 from flask_login import LoginManager
-from api import settings
+from common import settings
 from api.utils.api_utils import server_error_response
 from api.constants import API_VERSION
 

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
@@ -32,21 +32,21 @@ 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 api.utils import get_uuid
+from common.misc_utils import get_uuid
+from common.constants import RetCode, VALID_TASK_STATUS, LLMType, ParserType, FileSource
 from api.utils.api_utils import server_error_response, get_data_error_result, get_json_result, validate_request, \
     generate_confirmation_token
 
 from api.utils.file_utils import filename_type, thumbnail
 from rag.app.tag import label_question
 from rag.prompts.generator import keyword_extraction
-from rag.utils.storage_factory import STORAGE_IMPL
 from 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 settings
 
 
 @manager.route('/new_token', methods=['POST'])  # noqa: F821
@@ -145,7 +145,7 @@ 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)
@@ -186,7 +186,7 @@ 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:
@@ -352,7 +352,7 @@ 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)
@@ -362,7 +362,7 @@ 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:
@@ -383,7 +383,7 @@ 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
@@ -399,12 +399,12 @@ 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"]
@@ -427,10 +427,10 @@ def upload():
                 message="This type of file has not been supported yet!")
 
         location = filename
-        while STORAGE_IMPL.obj_exist(kb_id, location):
+        while settings.STORAGE_IMPL.obj_exist(kb_id, location):
             location += "_"
         blob = request.files['file'].read()
-        STORAGE_IMPL.put(kb_id, location, blob)
+        settings.STORAGE_IMPL.put(kb_id, location, blob)
         doc = {
             "id": get_uuid(),
             "kb_id": kb.id,
@@ -466,10 +466,7 @@ def upload():
     if "run" in form_data.keys():
         if request.form.get("run").strip() == "1":
             try:
-                info = {"run": 1, "progress": 0}
-                info["progress_msg"] = ""
-                info["chunk_num"] = 0
-                info["token_num"] = 0
+                info = {"run": 1, "progress": 0, "progress_msg": "", "chunk_num": 0, "token_num": 0}
                 DocumentService.update_by_id(doc["id"], info)
                 # if str(req["run"]) == TaskStatus.CANCEL.value:
                 tenant_id = DocumentService.get_tenant_id(doc["id"])
@@ -496,17 +493,17 @@ 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)
@@ -519,7 +516,7 @@ 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
 
@@ -559,7 +556,7 @@ 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)
@@ -584,7 +581,7 @@ 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
@@ -637,7 +634,7 @@ 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)
@@ -651,7 +648,7 @@ 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
@@ -698,12 +695,12 @@ def document_rm():
             FileService.filter_delete([File.source_type == FileSource.KNOWLEDGEBASE, File.id == f2d[0].file_id])
             File2DocumentService.delete_by_document_id(doc_id)
 
-            STORAGE_IMPL.rm(b, n)
+            settings.STORAGE_IMPL.rm(b, n)
         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)
 
@@ -718,7 +715,7 @@ 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:
@@ -726,8 +723,7 @@ def completion_faq():
     if "quote" not in req:
         req["quote"] = True
 
-    msg = []
-    msg.append({"role": "user", "content": req["word"]})
+    msg = [{"role": "user", "content": req["word"]}]
     if not msg[-1].get("id"):
         msg[-1]["id"] = get_uuid()
     message_id = msg[-1]["id"]
@@ -791,7 +787,7 @@ def completion_faq():
                 if ans["reference"]["chunks"][chunk_idx]["img_id"]:
                     try:
                         bkt, nm = ans["reference"]["chunks"][chunk_idx]["img_id"].split("-")
-                        response = STORAGE_IMPL.get(bkt, nm)
+                        response = settings.STORAGE_IMPL.get(bkt, nm)
                         data_type_picture["url"] = base64.b64encode(response).decode('utf-8')
                         data.append(data_type_picture)
                         break
@@ -836,7 +832,7 @@ def completion_faq():
             if ans["reference"]["chunks"][chunk_idx]["img_id"]:
                 try:
                     bkt, nm = ans["reference"]["chunks"][chunk_idx]["img_id"].split("-")
-                    response = STORAGE_IMPL.get(bkt, nm)
+                    response = settings.STORAGE_IMPL.get(bkt, nm)
                     data_type_picture["url"] = base64.b64encode(response).decode('utf-8')
                     data.append(data_type_picture)
                     break
@@ -857,7 +853,7 @@ 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", [])
@@ -876,7 +872,7 @@ 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
@@ -895,5 +891,5 @@ 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/auth/github.py

@@ -34,7 +34,7 @@ class GithubOAuthClient(OAuthClient):
 
     def fetch_user_info(self, access_token, **kwargs):
         """
-        Fetch github user info.
+        Fetch GitHub user info.
         """
         user_info = {}
         try:

api/apps/auth/oidc.py

@@ -43,7 +43,8 @@ class OIDCClient(OAuthClient):
         self.jwks_uri = config['jwks_uri']
 
 
-    def _load_oidc_metadata(self, issuer):
+    @staticmethod
+    def _load_oidc_metadata(issuer):
         """
         Load OIDC metadata from `/.well-known/openid-configuration`.
         """

api/apps/canvas_app.py

@@ -25,7 +25,6 @@ from flask import request, Response
 from flask_login import login_required, current_user
 
 from agent.component import LLM
-from api import settings
 from api.db import CanvasCategory, FileType
 from api.db.services.canvas_service import CanvasTemplateService, UserCanvasService, API4ConversationService
 from api.db.services.document_service import DocumentService
@@ -34,8 +33,8 @@ from api.db.services.pipeline_operation_log_service import PipelineOperationLogS
 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 api.utils import get_uuid
+from common.constants import RetCode
+from common.misc_utils import get_uuid
 from api.utils.api_utils import get_json_result, server_error_response, validate_request, get_data_error_result
 from agent.canvas import Canvas
 from peewee import MySQLDatabase, PostgresqlDatabase
@@ -46,6 +45,7 @@ from api.utils.file_utils import filename_type, read_potential_broken_pdf
 from rag.flow.pipeline import Pipeline
 from rag.nlp import search
 from rag.utils.redis_conn import REDIS_CONN
+from common import settings
 
 
 @manager.route('/templates', methods=['GET'])  # noqa: F821
@@ -156,7 +156,7 @@ def run():
         return get_json_result(data={"message_id": task_id})
 
     try:
-        canvas = Canvas(cvs.dsl, current_user.id, req["id"])
+        canvas = Canvas(cvs.dsl, current_user.id)
     except Exception as e:
         return server_error_response(e)
 
@@ -168,8 +168,10 @@ def run():
 
             cvs.dsl = json.loads(str(canvas))
             UserCanvasService.update_by_id(req["id"], cvs.to_dict())
+
         except Exception as e:
             logging.exception(e)
+            canvas.cancel_task()
             yield "data:" + json.dumps({"code": 500, "message": str(e), "data": False}, ensure_ascii=False) + "\n\n"
 
     resp = Response(sse(), mimetype="text/event-stream")
@@ -177,6 +179,7 @@ def run():
     resp.headers.add_header("Connection", "keep-alive")
     resp.headers.add_header("X-Accel-Buffering", "no")
     resp.headers.add_header("Content-Type", "text/event-stream; charset=utf-8")
+    resp.call_on_close(lambda: canvas.cancel_task())
     return resp
 
 
@@ -410,27 +413,27 @@ def test_db_connect():
             ibm_db.close(conn)
             return get_json_result(data="Database Connection Successful!")
         elif req["db_type"] == 'trino':
-            def _parse_catalog_schema(db: str):
-                if not db:
+            def _parse_catalog_schema(db_name: str):
+                if not db_name:
                     return None, None
-                if "." in db:
-                    c, s = db.split(".", 1)
-                elif "/" in db:
-                    c, s = db.split("/", 1)
+                if "." in db_name:
+                    catalog_name, schema_name = db_name.split(".", 1)
+                elif "/" in db_name:
+                    catalog_name, schema_name = db_name.split("/", 1)
                 else:
-                    c, s = db, "default"
-                return c, s
+                    catalog_name, schema_name = db_name, "default"
+                return catalog_name, schema_name
             try:
                 import trino
                 import os
                 from trino.auth import BasicAuthentication
-            except Exception:
-                return server_error_response("Missing dependency 'trino'. Please install: pip install trino")
+            except Exception as e:
+                return server_error_response(f"Missing dependency 'trino'. Please install: pip install trino, detail: {e}")
 
             catalog, schema = _parse_catalog_schema(req["database"])
             if not catalog:
                 return server_error_response("For Trino, 'database' must be 'catalog.schema' or at least 'catalog'.")
-            
+
             http_scheme = "https" if os.environ.get("TRINO_USE_TLS", "0") == "1" else "http"
 
             auth = None
@@ -479,7 +482,6 @@ def getlistversion(canvas_id):
 @login_required
 def getversion( version_id):
     try:
-
         e, version = UserCanvasVersionService.get_by_id(version_id)
         if version:
             return get_json_result(data=version.to_dict())
@@ -546,11 +548,11 @@ def trace():
     cvs_id = request.args.get("canvas_id")
     msg_id = request.args.get("message_id")
     try:
-        bin = REDIS_CONN.get(f"{cvs_id}-{msg_id}-logs")
-        if not bin:
+        binary = REDIS_CONN.get(f"{cvs_id}-{msg_id}-logs")
+        if not binary:
             return get_json_result(data={})
 
-        return get_json_result(data=json.loads(bin.encode("utf-8")))
+        return get_json_result(data=json.loads(binary.encode("utf-8")))
     except Exception as e:
         logging.exception(e)
 
@@ -604,4 +606,4 @@ def download():
     id = request.args.get("id")
     created_by = request.args.get("created_by")
     blob = FileService.get_blob(created_by, id)
-    return flask.make_response(blob)
+    return flask.make_response(blob)

api/apps/chunk_app.py

@@ -21,8 +21,6 @@ import xxhash
 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
@@ -34,8 +32,9 @@ from rag.app.qa import beAdoc, rmPrefix
 from rag.app.tag import label_question
 from rag.nlp import rag_tokenizer, search
 from rag.prompts.generator import gen_meta_filter, cross_languages, keyword_extraction
-from rag.settings import PAGERANK_FLD
 from common.string_utils import remove_redundant_spaces
+from common.constants import RetCode, LLMType, ParserType, PAGERANK_FLD
+from common import settings
 
 
 @manager.route('/list', methods=['POST'])  # noqa: F821
@@ -83,7 +82,7 @@ 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)
 
 
@@ -115,7 +114,7 @@ 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)
 
 
@@ -200,7 +199,6 @@ def switch():
 @login_required
 @validate_request("chunk_ids", "doc_id")
 def rm():
-    from rag.utils.storage_factory import STORAGE_IMPL
     req = request.json
     try:
         e, doc = DocumentService.get_by_id(req["doc_id"])
@@ -214,8 +212,8 @@ def rm():
         chunk_number = len(deleted_chunk_ids)
         DocumentService.decrement_chunk_num(doc.id, doc.kb_id, 1, chunk_number, 0)
         for cid in deleted_chunk_ids:
-            if STORAGE_IMPL.obj_exist(doc.kb_id, cid):
-                STORAGE_IMPL.rm(doc.kb_id, cid)
+            if settings.STORAGE_IMPL.obj_exist(doc.kb_id, cid):
+                settings.STORAGE_IMPL.rm(doc.kb_id, cid)
         return get_json_result(data=True)
     except Exception as e:
         return server_error_response(e)
@@ -292,7 +290,7 @@ 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)
@@ -326,7 +324,7 @@ 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:
@@ -371,7 +369,7 @@ 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)
 
 

api/apps/connector_app.py

@@ -0,0 +1,295 @@
+#
+#  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 logging
+import time
+import uuid
+from html import escape
+from typing import Any
+
+from flask import make_response, request
+from flask_login import current_user, login_required
+from google_auth_oauthlib.flow import Flow
+
+from api.db import InputType
+from api.db.services.connector_service import ConnectorService, SyncLogsService
+from api.utils.api_utils import get_data_error_result, get_json_result, validate_request
+from common.constants import RetCode, TaskStatus
+from common.data_source.config import GOOGLE_DRIVE_WEB_OAUTH_REDIRECT_URI, DocumentSource
+from common.data_source.google_util.constant import GOOGLE_DRIVE_WEB_OAUTH_POPUP_TEMPLATE, GOOGLE_SCOPES
+from common.misc_utils import get_uuid
+from rag.utils.redis_conn import REDIS_CONN
+
+
+@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)
+    arr, total = SyncLogsService.list_sync_tasks(connector_id, int(req.get("page", 1)), int(req.get("page_size", 15)))
+    return get_json_result(data={"total": total, "logs": arr})
+
+
+@manager.route("/<connector_id>/resume", methods=["PUT"])  # noqa: F821
+@login_required
+def resume(connector_id):
+    req = request.json
+    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>/rebuild", methods=["PUT"])  # noqa: F821
+@login_required
+@validate_request("kb_id")
+def rebuild(connector_id):
+    req = request.json
+    err = ConnectorService.rebuild(req["kb_id"], connector_id, current_user.id)
+    if err:
+        return get_json_result(data=False, message=err, code=RetCode.SERVER_ERROR)
+    return get_json_result(data=True)
+
+
+@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)
+
+
+GOOGLE_WEB_FLOW_STATE_PREFIX = "google_drive_web_flow_state"
+GOOGLE_WEB_FLOW_RESULT_PREFIX = "google_drive_web_flow_result"
+WEB_FLOW_TTL_SECS = 15 * 60
+
+
+def _web_state_cache_key(flow_id: str) -> str:
+    return f"{GOOGLE_WEB_FLOW_STATE_PREFIX}:{flow_id}"
+
+
+def _web_result_cache_key(flow_id: str) -> str:
+    return f"{GOOGLE_WEB_FLOW_RESULT_PREFIX}:{flow_id}"
+
+
+def _load_credentials(payload: str | dict[str, Any]) -> dict[str, Any]:
+    if isinstance(payload, dict):
+        return payload
+    try:
+        return json.loads(payload)
+    except json.JSONDecodeError as exc:  # pragma: no cover - defensive
+        raise ValueError("Invalid Google credentials JSON.") from exc
+
+
+def _get_web_client_config(credentials: dict[str, Any]) -> dict[str, Any]:
+    web_section = credentials.get("web")
+    if not isinstance(web_section, dict):
+        raise ValueError("Google OAuth JSON must include a 'web' client configuration to use browser-based authorization.")
+    return {"web": web_section}
+
+
+def _render_web_oauth_popup(flow_id: str, success: bool, message: str):
+    status = "success" if success else "error"
+    auto_close = "window.close();" if success else ""
+    escaped_message = escape(message)
+    payload_json = json.dumps(
+        {
+            "type": "ragflow-google-drive-oauth",
+            "status": status,
+            "flowId": flow_id or "",
+            "message": message,
+        }
+    )
+    html = GOOGLE_DRIVE_WEB_OAUTH_POPUP_TEMPLATE.format(
+        heading="Authorization complete" if success else "Authorization failed",
+        message=escaped_message,
+        payload_json=payload_json,
+        auto_close=auto_close,
+    )
+    response = make_response(html, 200)
+    response.headers["Content-Type"] = "text/html; charset=utf-8"
+    return response
+
+
+@manager.route("/google-drive/oauth/web/start", methods=["POST"])  # noqa: F821
+@login_required
+@validate_request("credentials")
+def start_google_drive_web_oauth():
+    if not GOOGLE_DRIVE_WEB_OAUTH_REDIRECT_URI:
+        return get_json_result(
+            code=RetCode.SERVER_ERROR,
+            message="Google Drive OAuth redirect URI is not configured on the server.",
+        )
+
+    req = request.json or {}
+    raw_credentials = req.get("credentials", "")
+    try:
+        credentials = _load_credentials(raw_credentials)
+    except ValueError as exc:
+        return get_json_result(code=RetCode.ARGUMENT_ERROR, message=str(exc))
+
+    if credentials.get("refresh_token"):
+        return get_json_result(
+            code=RetCode.ARGUMENT_ERROR,
+            message="Uploaded credentials already include a refresh token.",
+        )
+
+    try:
+        client_config = _get_web_client_config(credentials)
+    except ValueError as exc:
+        return get_json_result(code=RetCode.ARGUMENT_ERROR, message=str(exc))
+
+    flow_id = str(uuid.uuid4())
+    try:
+        flow = Flow.from_client_config(client_config, scopes=GOOGLE_SCOPES[DocumentSource.GOOGLE_DRIVE])
+        flow.redirect_uri = GOOGLE_DRIVE_WEB_OAUTH_REDIRECT_URI
+        authorization_url, _ = flow.authorization_url(
+            access_type="offline",
+            include_granted_scopes="true",
+            prompt="consent",
+            state=flow_id,
+        )
+    except Exception as exc:  # pragma: no cover - defensive
+        logging.exception("Failed to create Google OAuth flow: %s", exc)
+        return get_json_result(
+            code=RetCode.SERVER_ERROR,
+            message="Failed to initialize Google OAuth flow. Please verify the uploaded client configuration.",
+        )
+
+    cache_payload = {
+        "user_id": current_user.id,
+        "client_config": client_config,
+        "created_at": int(time.time()),
+    }
+    REDIS_CONN.set_obj(_web_state_cache_key(flow_id), cache_payload, WEB_FLOW_TTL_SECS)
+
+    return get_json_result(
+        data={
+            "flow_id": flow_id,
+            "authorization_url": authorization_url,
+            "expires_in": WEB_FLOW_TTL_SECS,
+        }
+    )
+
+
+@manager.route("/google-drive/oauth/web/callback", methods=["GET"])  # noqa: F821
+def google_drive_web_oauth_callback():
+    state_id = request.args.get("state")
+    error = request.args.get("error")
+    error_description = request.args.get("error_description") or error
+
+    if not state_id:
+        return _render_web_oauth_popup("", False, "Missing OAuth state parameter.")
+
+    state_cache = REDIS_CONN.get(_web_state_cache_key(state_id))
+    if not state_cache:
+        return _render_web_oauth_popup(state_id, False, "Authorization session expired. Please restart from the main window.")
+
+    state_obj = json.loads(state_cache)
+    client_config = state_obj.get("client_config")
+    if not client_config:
+        REDIS_CONN.delete(_web_state_cache_key(state_id))
+        return _render_web_oauth_popup(state_id, False, "Authorization session was invalid. Please retry.")
+
+    if error:
+        REDIS_CONN.delete(_web_state_cache_key(state_id))
+        return _render_web_oauth_popup(state_id, False, error_description or "Authorization was cancelled.")
+
+    code = request.args.get("code")
+    if not code:
+        return _render_web_oauth_popup(state_id, False, "Missing authorization code from Google.")
+
+    try:
+        flow = Flow.from_client_config(client_config, scopes=GOOGLE_SCOPES[DocumentSource.GOOGLE_DRIVE])
+        flow.redirect_uri = GOOGLE_DRIVE_WEB_OAUTH_REDIRECT_URI
+        flow.fetch_token(code=code)
+    except Exception as exc:  # pragma: no cover - defensive
+        logging.exception("Failed to exchange Google OAuth code: %s", exc)
+        REDIS_CONN.delete(_web_state_cache_key(state_id))
+        return _render_web_oauth_popup(state_id, False, "Failed to exchange tokens with Google. Please retry.")
+
+    creds_json = flow.credentials.to_json()
+    result_payload = {
+        "user_id": state_obj.get("user_id"),
+        "credentials": creds_json,
+    }
+    REDIS_CONN.set_obj(_web_result_cache_key(state_id), result_payload, WEB_FLOW_TTL_SECS)
+    REDIS_CONN.delete(_web_state_cache_key(state_id))
+
+    return _render_web_oauth_popup(state_id, True, "Authorization completed successfully.")
+
+
+@manager.route("/google-drive/oauth/web/result", methods=["POST"])  # noqa: F821
+@login_required
+@validate_request("flow_id")
+def poll_google_drive_web_result():
+    req = request.json or {}
+    flow_id = req.get("flow_id")
+    cache_raw = REDIS_CONN.get(_web_result_cache_key(flow_id))
+    if not cache_raw:
+        return get_json_result(code=RetCode.RUNNING, message="Authorization is still pending.")
+
+    result = json.loads(cache_raw)
+    if result.get("user_id") != current_user.id:
+        return get_json_result(code=RetCode.PERMISSION_ERROR, message="You are not allowed to access this authorization result.")
+
+    REDIS_CONN.delete(_web_result_cache_key(flow_id))
+    return get_json_result(data={"credentials": result.get("credentials")})

api/apps/conversation_app.py

@@ -19,8 +19,6 @@ 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
@@ -31,6 +29,7 @@ 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.template import load_prompt
 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]

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

@@ -23,30 +23,31 @@ 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.db_models import File, Task
+from api.db import VALID_FILE_TYPES, FileType
+from api.db.db_models import Task
 from api.db.services import duplicate_name
 from api.db.services.document_service import DocumentService, doc_upload_and_parse
 from 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, queue_dataflow
+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, rag_tokenizer
-from rag.utils.storage_factory import STORAGE_IMPL
+from common import settings
 
 
 @manager.route("/upload", methods=["POST"])  # noqa: F821
@@ -55,29 +56,29 @@ 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!")
     if not check_kb_team_permission(kb, 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)
 
     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)
@@ -89,16 +90,16 @@ 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=settings.RetCode.AUTHENTICATION_ERROR)
+        return get_json_result(data=False, message="No authorization.", code=RetCode.AUTHENTICATION_ERROR)
 
     blob = html2pdf(url)
     if not blob:
@@ -117,9 +118,9 @@ def web_crawl():
             raise RuntimeError("This type of file has not been supported yet!")
 
         location = filename
-        while STORAGE_IMPL.obj_exist(kb_id, location):
+        while settings.STORAGE_IMPL.obj_exist(kb_id, location):
             location += "_"
-        STORAGE_IMPL.put(kb_id, location, blob)
+        settings.STORAGE_IMPL.put(kb_id, location, blob)
         doc = {
             "id": get_uuid(),
             "kb_id": kb.id,
@@ -155,12 +156,12 @@ 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:
@@ -210,13 +211,13 @@ 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))
@@ -259,6 +260,8 @@ def list_docs():
         for doc_item in docs:
             if doc_item["thumbnail"] and not doc_item["thumbnail"].startswith(IMG_BASE64_PREFIX):
                 doc_item["thumbnail"] = f"/v1/document/image/{kb_id}-{doc_item['thumbnail']}"
+            if doc_item.get("source_type"):
+                doc_item["source_type"] = doc_item["source_type"].split("/")[0]
 
         return get_json_result(data={"total": tol, "docs": docs})
     except Exception as e:
@@ -272,13 +275,13 @@ 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", "")
 
@@ -310,7 +313,7 @@ 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()))
 
@@ -320,7 +323,7 @@ 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)
@@ -343,7 +346,7 @@ 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:
@@ -385,50 +388,12 @@ 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)
-    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)
+    errors = FileService.delete_docs(doc_ids, current_user.id)
 
     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)
 
@@ -440,7 +405,7 @@ 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"]:
@@ -473,23 +438,7 @@ def run():
 
             if str(req["run"]) == TaskStatus.RUNNING.value:
                 doc = doc.to_dict()
-                doc["tenant_id"] = tenant_id
-
-                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)
-                if doc.get("pipeline_id", ""):
-                    queue_dataflow(tenant_id, flow_id=doc["pipeline_id"], task_id=get_uuid(), doc_id=id)
-                else:
-                    bucket, name = File2DocumentService.get_storage_address(doc_id=doc["id"])
-                    queue_tasks(doc, bucket, name, 0)
+                DocumentService.run(tenant_id, doc, kb_table_num_map)
 
         return get_json_result(data=True)
     except Exception as e:
@@ -502,15 +451,15 @@ 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"]:
@@ -553,7 +502,7 @@ def get(doc_id):
             return get_data_error_result(message="Document not found!")
 
         b, n = File2DocumentService.get_storage_address(doc_id=doc_id)
-        response = flask.make_response(STORAGE_IMPL.get(b, n))
+        response = flask.make_response(settings.STORAGE_IMPL.get(b, n))
 
         ext = re.search(r"\.([^.]+)$", doc.name.lower())
         ext = ext.group(1) if ext else None
@@ -575,7 +524,7 @@ def change_parser():
 
     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)
 
     e, doc = DocumentService.get_by_id(req["doc_id"])
     if not e:
@@ -629,7 +578,7 @@ def get_image(image_id):
         if len(arr) != 2:
             return get_data_error_result(message="Image not found.")
         bkt, nm = image_id.split("-")
-        response = flask.make_response(STORAGE_IMPL.get(bkt, nm))
+        response = flask.make_response(settings.STORAGE_IMPL.get(bkt, nm))
         response.headers.set("Content-Type", "image/JPEG")
         return response
     except Exception as e:
@@ -641,12 +590,12 @@ 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)
 
@@ -659,7 +608,7 @@ 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
@@ -692,13 +641,13 @@ 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)
@@ -712,18 +661,18 @@ 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

@@ -26,15 +26,15 @@ 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 api.db import FileType, FileSource
+from common.misc_utils import get_uuid
+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
-from rag.utils.storage_factory import STORAGE_IMPL
+from common import settings
 
 
 @manager.route('/upload', methods=['POST'])  # noqa: F821
@@ -49,21 +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))
-            if MAX_FILE_NUM_PER_USER > 0 and DocumentService.get_doc_count(current_user.id) >= MAX_FILE_NUM_PER_USER:
+            MAX_FILE_NUM_PER_USER: int = int(os.environ.get('MAX_FILE_NUM_PER_USER', 0))
+            if 0 < MAX_FILE_NUM_PER_USER <= DocumentService.get_doc_count(current_user.id):
                 return get_data_error_result( message="Exceed the maximum file number of a free user!")
 
             # split file name path
@@ -95,14 +95,14 @@ def upload():
             # file type
             filetype = filename_type(file_obj_names[file_len - 1])
             location = file_obj_names[file_len - 1]
-            while STORAGE_IMPL.obj_exist(last_folder.id, location):
+            while settings.STORAGE_IMPL.obj_exist(last_folder.id, location):
                 location += "_"
             blob = file_obj.read()
             filename = duplicate_name(
                 FileService.query,
                 name=file_obj_names[file_len - 1],
                 parent_id=last_folder.id)
-            STORAGE_IMPL.put(last_folder.id, location, blob)
+            settings.STORAGE_IMPL.put(last_folder.id, location, blob)
             file = {
                 "id": get_uuid(),
                 "parent_id": last_folder.id,
@@ -134,7 +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.")
@@ -245,7 +245,7 @@ def rm():
     def _delete_single_file(file):
         try:
             if file.location:
-                STORAGE_IMPL.rm(file.parent_id, file.location)
+                settings.STORAGE_IMPL.rm(file.parent_id, file.location)
         except Exception:
             logging.exception(f"Fail to remove object: {file.parent_id}/{file.location}")
 
@@ -279,7 +279,7 @@ def rm():
             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=settings.RetCode.AUTHENTICATION_ERROR)
+                return get_json_result(data=False, message="No authorization.", code=RetCode.AUTHENTICATION_ERROR)
 
             if file.source_type == FileSource.KNOWLEDGEBASE:
                 continue
@@ -306,14 +306,14 @@ def rename():
         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=settings.RetCode.AUTHENTICATION_ERROR)
+            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(
@@ -344,12 +344,12 @@ def get(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=settings.RetCode.AUTHENTICATION_ERROR)
+            return get_json_result(data=False, message='No authorization.', code=RetCode.AUTHENTICATION_ERROR)
 
-        blob = STORAGE_IMPL.get(file.parent_id, file.location)
+        blob = settings.STORAGE_IMPL.get(file.parent_id, file.location)
         if not blob:
             b, n = File2DocumentService.get_storage_address(file_id=file_id)
-            blob = STORAGE_IMPL.get(b, n)
+            blob = settings.STORAGE_IMPL.get(b, n)
 
         response = flask.make_response(blob)
         ext = re.search(r"\.([^.]+)$", file.name.lower())
@@ -376,7 +376,7 @@ def move():
 
         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!")
+            return get_data_error_result(message="Parent folder not found!")
 
         files = FileService.get_by_ids(file_ids)
         if not files:
@@ -387,14 +387,14 @@ def move():
         for file_id in file_ids:
             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!")
             if not check_file_team_permission(file, current_user.id):
                 return get_json_result(
                     data=False,
                     message="No authorization.",
-                    code=settings.RetCode.AUTHENTICATION_ERROR,
+                    code=RetCode.AUTHENTICATION_ERROR,
                 )
 
         def _move_entry_recursive(source_file_entry, dest_folder):
@@ -428,11 +428,11 @@ def move():
             filename = source_file_entry.name
 
             new_location = filename
-            while STORAGE_IMPL.obj_exist(dest_folder.id, new_location):
+            while settings.STORAGE_IMPL.obj_exist(dest_folder.id, new_location):
                 new_location += "_"
 
             try:
-                STORAGE_IMPL.move(old_parent_id, old_location, dest_folder.id, new_location)
+                settings.STORAGE_IMPL.move(old_parent_id, old_location, dest_folder.id, new_location)
             except Exception as storage_err:
                 raise RuntimeError(f"Move file failed at storage layer: {str(storage_err)}")
 

api/apps/kb_app.py

@@ -21,8 +21,8 @@ from flask import request
 from flask_login import login_required, current_user
 import numpy as np
 
-from api.db import LLMType
-from api.db.services import duplicate_name
+
+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
@@ -31,82 +31,33 @@ from api.db.services.pipeline_operation_log_service import PipelineOperationLogS
 from api.db.services.task_service import TaskService, GRAPH_RAPTOR_FAKE_DOC_ID
 from api.db.services.user_service import TenantService, UserTenantService
 from api.utils.api_utils import get_error_data_result, server_error_response, get_data_error_result, validate_request, not_allowed_parameters
-from api.utils import get_uuid
-from api.db import PipelineTaskType, StatusEnum, FileSource, VALID_FILE_TYPES, VALID_TASK_STATUS
+from api.db import VALID_FILE_TYPES
 from api.db.services.knowledgebase_service import KnowledgebaseService
 from api.db.db_models import File
 from api.utils.api_utils import get_json_result
-from api import settings
 from rag.nlp import search
 from api.constants import DATASET_NAME_LIMIT
-from rag.settings import PAGERANK_FLD
 from rag.utils.redis_conn import REDIS_CONN
-from rag.utils.storage_factory import STORAGE_IMPL
-from rag.utils.doc_store_conn import OrderByExpr  
-
+from rag.utils.doc_store_conn import OrderByExpr
+from common.constants import RetCode, PipelineTaskType, StatusEnum, VALID_TASK_STATUS, FileSource, LLMType, PAGERANK_FLD
+from common import settings
 
 @manager.route('/create', methods=['post'])  # noqa: F821
 @login_required
 @validate_request("name")
 def create():
     req = request.json
-    dataset_name = req["name"]
-    if not isinstance(dataset_name, str):
-        return get_data_error_result(message="Dataset name must be string.")
-    if dataset_name.strip() == "":
-        return get_data_error_result(message="Dataset name can't be empty.")
-    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}")
+    req = KnowledgebaseService.create_with_name(
+        name = req.pop("name", None),
+        tenant_id = current_user.id,
+        parser_id = req.pop("parser_id", None),
+        **req
+    )
 
-    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
-        if not req.get("parser_id"):
-            req["parser_id"] = "naive"
-        e, t = TenantService.get_by_id(current_user.id)
-        if not e:
-            return get_data_error_result(message="Tenant not found.")
-
-        req["parser_config"] = {
-            "layout_recognize": "DeepDOC",
-            "chunk_token_num": 512,
-            "delimiter": "\n",
-            "auto_keywords": 0,
-            "auto_questions": 0,
-            "html4excel": False,
-            "topn_tags": 3,
-            "raptor": {
-                "use_raptor": True,
-                "prompt": "Please summarize the following paragraphs. Be careful with the numbers, do not make things up. Paragraphs as following:\n      {cluster_content}\nThe above is the content you need to summarize.",
-                "max_token": 256,
-                "threshold": 0.1,
-                "max_cluster": 64,
-                "random_seed": 0
-            },
-            "graphrag": {
-                "use_graphrag": True,
-                "entity_types": [
-                    "organization",
-                    "person",
-                    "geo",
-                    "event",
-                    "category"
-                ],
-                "method": "light"
-            }
-        }
         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)
 
@@ -130,14 +81,14 @@ 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:
@@ -151,6 +102,10 @@ def update():
                 message="Duplicated knowledgebase name.")
 
         del req["kb_id"]
+        connectors = []
+        if "connectors" in req:
+            connectors = req["connectors"]
+            del req["connectors"]
         if not KnowledgebaseService.update_by_id(kb.id, req):
             return get_data_error_result()
 
@@ -167,8 +122,12 @@ def update():
         if not e:
             return get_data_error_result(
                 message="Database error (Knowledgebase rename)!")
+        errors = Connector2KbService.link_connectors(kb.id, [conn for conn in connectors], current_user.id)
+        if errors:
+            logging.error("Link KB errors: ", errors)
         kb = kb.to_dict()
         kb.update(req)
+        kb["connectors"] = connectors
 
         return get_json_result(data=kb)
     except Exception as e:
@@ -188,12 +147,14 @@ 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")
@@ -246,7 +207,7 @@ def rm():
         return get_json_result(
             data=False,
             message='No authorization.',
-            code=settings.RetCode.AUTHENTICATION_ERROR
+            code=RetCode.AUTHENTICATION_ERROR
         )
     try:
         kbs = KnowledgebaseService.query(
@@ -254,7 +215,7 @@ 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):
@@ -272,8 +233,8 @@ def rm():
         for kb in kbs:
             settings.docStoreConn.delete({"kb_id": kb.id}, search.index_name(kb.tenant_id), kb.id)
             settings.docStoreConn.deleteIdx(search.index_name(kb.tenant_id), kb.id)
-            if hasattr(STORAGE_IMPL, 'remove_bucket'):
-                STORAGE_IMPL.remove_bucket(kb.id)
+            if hasattr(settings.STORAGE_IMPL, 'remove_bucket'):
+                settings.STORAGE_IMPL.remove_bucket(kb.id)
         return get_json_result(data=True)
     except Exception as e:
         return server_error_response(e)
@@ -286,7 +247,7 @@ 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)
@@ -305,7 +266,7 @@ 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)
@@ -323,7 +284,7 @@ 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)
 
@@ -343,7 +304,7 @@ 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)
 
@@ -361,7 +322,7 @@ 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 = {
@@ -401,7 +362,7 @@ 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)
@@ -418,7 +379,7 @@ 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))
 
@@ -431,7 +392,7 @@ def get_basic_info():
         return get_json_result(
             data=False,
             message='No authorization.',
-            code=settings.RetCode.AUTHENTICATION_ERROR
+            code=RetCode.AUTHENTICATION_ERROR
         )
 
     basic_info = DocumentService.knowledgebase_basic_info(kb_id)
@@ -444,7 +405,7 @@ def get_basic_info():
 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=settings.RetCode.ARGUMENT_ERROR)
+        return get_json_result(data=False, message='Lack of "KB ID"', code=RetCode.ARGUMENT_ERROR)
 
     keywords = request.args.get("keywords", "")
 
@@ -488,7 +449,7 @@ def list_pipeline_logs():
 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=settings.RetCode.ARGUMENT_ERROR)
+        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))
@@ -522,7 +483,7 @@ def list_pipeline_dataset_logs():
 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=settings.RetCode.ARGUMENT_ERROR)
+        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", [])
@@ -537,7 +498,7 @@ def delete_pipeline_logs():
 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=settings.RetCode.ARGUMENT_ERROR)
+        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:
@@ -610,7 +571,7 @@ def trace_graphrag():
 
     ok, task = TaskService.get_by_id(task_id)
     if not ok:
-        return get_error_data_result(message="GraphRAG Task Not Found or Error Occurred")
+        return get_json_result(data={})
 
     return get_json_result(data=task.to_dict())
 
@@ -767,26 +728,30 @@ def delete_kb_task():
     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:
-            settings.docStoreConn.delete({"knowledge_graph_kwd": ["graph", "subgraph", "entity", "relation"]}, search.index_name(kb.tenant_id), kb_id)
             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)
+            settings.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)
+            settings.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")
 
-    def cancel_task(task_id):
-        REDIS_CONN.set(f"{task_id}-cancel", "x")
-    cancel_task(task_id)
 
     ok = KnowledgebaseService.update_by_id(kb_id, {kb_task_id_field: "", kb_task_finish_at: None})
     if not ok:
@@ -815,14 +780,14 @@ def check_embedding():
 
     def _to_1d(x):
         a = np.asarray(x, dtype=np.float32)
-        return a.reshape(-1)  
+        return a.reshape(-1)
 
     def _cos_sim(a, b, eps=1e-12):
         a = _to_1d(a)
         b = _to_1d(b)
         na = np.linalg.norm(a)
         nb = np.linalg.norm(b)
-        if na < eps or nb < eps: 
+        if na < eps or nb < eps:
             return 0.0
         return float(np.dot(a, b) / (na * nb))
 
@@ -860,7 +825,7 @@ def check_embedding():
                 indexNames=index_nm, knowledgebaseIds=[kb_id]
             )
             ids = docStoreConn.getChunkIds(res1)
-            if not ids: 
+            if not ids:
                 continue
 
             cid = ids[0]
@@ -904,7 +869,7 @@ def check_embedding():
             continue
 
         try:
-            qv, _ = emb_mdl.encode_queries(txt)  
+            qv, _ = emb_mdl.encode_queries(txt)
             sim = _cos_sim(qv, ck["vector"])
         except Exception:
             return get_error_data_result(message="embedding failure")
@@ -930,4 +895,6 @@ def check_embedding():
     }
     if summary["avg_cos_sim"] > 0.99:
         return get_json_result(data={"summary": summary, "results": results})
-    return get_json_result(code=settings.RetCode.NOT_EFFECTIVE, message="failed", data={"summary": summary, "results": results})
+    return get_json_result(code=RetCode.NOT_EFFECTIVE, message="failed", data={"summary": summary, "results": results})
+
+

api/apps/llm_app.py

@@ -21,19 +21,19 @@ 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.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.base64_image import test_image
+from api.utils.api_utils import get_json_result, get_allowed_llm_factories
+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 = [f.to_dict() for f in fac if f.name not in ["Youdao", "FastEmbed", "BAAI"]]
+        fac = get_allowed_llm_factories()
+        fac = [f.to_dict() for f in fac if f.name not in ["Youdao", "FastEmbed", "BAAI", "Builtin"]]
         llms = LLMService.get_all()
         mdl_types = {}
         for m in llms:
@@ -43,14 +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,
-                                                              LLMType.IMAGE2TEXT, LLMType.SPEECH2TEXT, LLMType.TTS]))
+            f["model_types"] = list(mdl_types.get(f["name"], [LLMType.CHAT, LLMType.EMBEDDING, LLMType.RERANK, LLMType.IMAGE2TEXT, LLMType.SPEECH2TEXT, LLMType.TTS]))
         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](
-                req["api_key"], llm.llm_name, base_url=req.get("base_url"))
+            mdl = EmbeddingModel[factory](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](
-                req["api_key"], llm.llm_name, base_url=req.get("base_url"), **extra)
+            mdl = ChatModel[factory](req["api_key"], llm.llm_name, base_url=req.get("base_url"), **extra)
             try:
-                m, tc = mdl.chat(None, [{"role": "user", "content": "Hello! How are you doing!"}],
-                                 {"temperature": 0.9, 'max_tokens': 50})
+                m, tc = mdl.chat(None, [{"role": "user", "content": "Hello! How are you doing!"}], {"temperature": 0.9, "max_tokens": 50})
                 if m.find("**ERROR**") >= 0:
                     raise Exception(m)
                 chat_passed = True
             except Exception as e:
-                msg += f"\nFail to access model({llm.fid}/{llm.llm_name}) using this api key." + str(
-                    e)
+                msg += f"\nFail to access model({llm.fid}/{llm.llm_name}) using this api key." + str(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](
-                req["api_key"], llm.llm_name, base_url=req.get("base_url"))
+            mdl = RerankModel[factory](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(
-                    e)
+                msg += f"\nFail to access model({llm.fid}/{llm.llm_name}) using this api key." + str(e)
         if any([embd_passed, chat_passed, rerank_passed]):
-            msg = ''
+            msg = ""
             break
 
     if msg:
         return get_data_error_result(message=msg)
 
-    llm_config = {
-        "api_key": req["api_key"],
-        "api_base": req.get("base_url", "")
-    }
+    llm_config = {"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
-        if not TenantLLMService.filter_update(
-                [TenantLLM.tenant_id == current_user.id,
-                 TenantLLM.llm_factory == factory,
-                 TenantLLM.llm_name == llm.llm_name],
-                llm_config):
+        llm_config["max_tokens"] = llm.max_tokens
+        if not TenantLLMService.filter_update([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 [f.name for f in get_allowed_llm_factories()]:
+        return get_data_error_result(message=f"LLM factory {factory} is not allowed")
+
     def apikey_json(keys):
         nonlocal req
         return json.dumps({k: req.get(k, "") for k in keys})
@@ -204,7 +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 = ""
@@ -212,10 +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](
-            key=llm['api_key'],
-            model_name=mdl_nm,
-            base_url=llm["api_base"])   
+        mdl = EmbeddingModel[factory](key=llm["api_key"], model_name=mdl_nm, base_url=llm["api_base"])
         try:
             arr, tc = mdl.encode(["Test if the api key is available"])
             if len(arr[0]) == 0:
@@ -225,42 +211,31 @@ 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!"}], {
-                "temperature": 0.9})
+            m, tc = mdl.chat(None, [{"role": "user", "content": "Hello! How are you doing!"}], {"temperature": 0.9})
             if not tc and m.find("**ERROR**:") >= 0:
                 raise Exception(m)
         except Exception as e:
-            msg += f"\nFail to access model({factory}/{mdl_nm})." + str(
-                e)
+            msg += f"\nFail to access model({factory}/{mdl_nm})." + str(e)
     elif llm["model_type"] == LLMType.RERANK:
         assert factory in RerankModel, f"RE-rank model from {factory} is not supported yet."
         try:
-            mdl = RerankModel[factory](
-                key=llm["api_key"],
-                model_name=mdl_nm,
-                base_url=llm["api_base"]
-            )
+            mdl = RerankModel[factory](key=llm["api_key"], model_name=mdl_nm, base_url=llm["api_base"])
             arr, tc = mdl.similarity("Hello~ RAGFlower!", ["Hi, there!", "Ohh, my friend!"])
             if len(arr) == 0:
                 raise Exception("Not known.")
         except KeyError:
             msg += f"{factory} dose not support this model({factory}/{mdl_nm})"
         except Exception as e:
-            msg += f"\nFail to access model({factory}/{mdl_nm})." + str(
-                e)
+            msg += f"\nFail to access model({factory}/{mdl_nm})." + str(e)
     elif llm["model_type"] == LLMType.IMAGE2TEXT.value:
         assert factory in CvModel, f"Image to text model from {factory} is not supported yet."
-        mdl = CvModel[factory](
-            key=llm["api_key"],
-            model_name=mdl_nm,
-            base_url=llm["api_base"]
-        )
+        mdl = CvModel[factory](key=llm["api_key"], model_name=mdl_nm, base_url=llm["api_base"])
         try:
             image_data = test_image
             m, tc = mdl.describe(image_data)
@@ -270,9 +245,7 @@ def add_llm():
             msg += f"\nFail to access model({factory}/{mdl_nm})." + str(e)
     elif llm["model_type"] == LLMType.TTS:
         assert factory in TTSModel, f"TTS model from {factory} is not supported yet."
-        mdl = TTSModel[factory](
-            key=llm["api_key"], model_name=mdl_nm, base_url=llm["api_base"]
-        )
+        mdl = TTSModel[factory](key=llm["api_key"], model_name=mdl_nm, base_url=llm["api_base"])
         try:
             for resp in mdl.tts("Hello~ RAGFlower!"):
                 pass
@@ -285,40 +258,46 @@ def add_llm():
     if msg:
         return get_data_error_result(message=msg)
 
-    if not TenantLLMService.filter_update(
-            [TenantLLM.tenant_id == current_user.id, TenantLLM.llm_factory == factory,
-             TenantLLM.llm_name == llm["llm_name"]], llm):
+    if not TenantLLMService.filter_update([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(
-        [TenantLLM.tenant_id == current_user.id, TenantLLM.llm_factory == req["llm_factory"],
-         TenantLLM.llm_name == req["llm_name"]])
+    TenantLLMService.filter_delete([TenantLLM.tenant_id == current_user.id, TenantLLM.llm_factory == req["llm_factory"], TenantLLM.llm_name == req["llm_name"]])
     return get_json_result(data=True)
 
 
-@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(
-        [TenantLLM.tenant_id == current_user.id, TenantLLM.llm_factory == req["llm_factory"]])
+    TenantLLMService.filter_delete([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 = {}
@@ -334,38 +313,31 @@ def my_llms():
                         break
 
                 if o_dict["llm_factory"] not in res:
-                    res[o_dict["llm_factory"]] = {
-                        "tags": factory_tags,
-                        "llm": []
-                    }
+                    res[o_dict["llm_factory"]] = {"tags": factory_tags, "llm": []}
 
-                res[o_dict["llm_factory"]]["llm"].append({
-                    "type": o_dict["model_type"],
-                    "name": o_dict["llm_name"],
-                    "used_token": o_dict["used_tokens"],
-                    "api_base": o_dict["api_base"] or "",
-                    "max_tokens": o_dict["max_tokens"] or 8192
-                })
+                res[o_dict["llm_factory"]]["llm"].append(
+                    {
+                        "type": o_dict["model_type"],
+                        "name": o_dict["llm_name"],
+                        "used_token": o_dict["used_tokens"],
+                        "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"]] = {
-                        "tags": o["tags"],
-                        "llm": []
-                    }
-                res[o["llm_factory"]]["llm"].append({
-                    "type": o["model_type"],
-                    "name": o["llm_name"],
-                    "used_token": o["used_tokens"]
-                })
+                    res[o["llm_factory"]] = {"tags": o["tags"], "llm": []}
+                res[o["llm_factory"]]["llm"].append({"type": o["model_type"], "name": o["llm_name"], "used_token": o["used_tokens"], "status": o["status"]})
 
         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 = ["FastEmbed", "Ollama", "Xinference", "LocalAI", "LM-Studio", "GPUStack"]
@@ -373,20 +345,20 @@ def list_app():
     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()
-                for m in llms if m.status == StatusEnum.VALID.value and m.fid not in weighted]
+        llms = [m.to_dict() for m in llms if m.status == StatusEnum.VALID.value and m.fid not in weighted and (m.fid == 'Builtin' or (m.llm_name + "@" + m.fid) in status)]
         for m in llms:
             m["available"] = m["fid"] in facts or m["llm_name"].lower() == "flag-embedding" or m["fid"] in self_deployed
-            if "tei-" in os.getenv("COMPOSE_PROFILES", "") and m["model_type"]==LLMType.EMBEDDING and m["fid"]=="Builtin" and m["llm_name"]==os.getenv('TEI_MODEL', ''):
+            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:
             if o.llm_name + "@" + o.llm_factory in llm_set:
                 continue
-            llms.append({"llm_name": o.llm_name, "model_type": o.model_type, "fid": o.llm_factory, "available": True})
+            llms.append({"llm_name": o.llm_name, "model_type": o.model_type, "fid": o.llm_factory, "available": True, "status": StatusEnum.VALID.value})
 
         res = {}
         for m in llms:

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/sdk/agents.py

@@ -15,15 +15,19 @@
 #
 
 import json
+import logging
 import time
 from typing import Any, cast
+
+from agent.canvas import Canvas
+from api.db import CanvasCategory
 from api.db.services.canvas_service import UserCanvasService
 from api.db.services.user_canvas_version import UserCanvasVersionService
-from api.settings import RetCode
-from api.utils import get_uuid
+from common.constants import RetCode
+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
+from flask import request, Response
 
 
 @manager.route('/agents', methods=['GET'])  # noqa: F821
@@ -127,3 +131,49 @@ def delete_agent(tenant_id: str, agent_id: str):
 
     UserCanvasService.delete_by_id(agent_id)
     return get_json_result(data=True)
+
+
+@manager.route('/webhook/<agent_id>', methods=['POST'])  # noqa: F821
+@token_required
+def webhook(tenant_id: str, agent_id: str):
+    req = request.json
+    if not UserCanvasService.accessible(req["id"], tenant_id):
+        return get_json_result(
+            data=False, message='Only owner of canvas authorized for this operation.',
+            code=RetCode.OPERATING_ERROR)
+
+    e, cvs = UserCanvasService.get_by_id(req["id"])
+    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)
+
+    if cvs.canvas_category == CanvasCategory.DataFlow:
+        return get_data_error_result(message="Dataflow can not be triggered by webhook.")
+
+    try:
+        canvas = Canvas(cvs.dsl, tenant_id, agent_id)
+    except Exception as e:
+        return get_json_result(
+            data=False, message=str(e),
+            code=RetCode.EXCEPTION_ERROR)
+
+    def sse():
+        nonlocal canvas
+        try:
+            for ans in canvas.run(query=req.get("query", ""), files=req.get("files", []), user_id=req.get("user_id", tenant_id), webhook_payload=req):
+                yield "data:" + json.dumps(ans, ensure_ascii=False) + "\n\n"
+
+            cvs.dsl = json.loads(str(canvas))
+            UserCanvasService.update_by_id(req["id"], cvs.to_dict())
+        except Exception as e:
+            logging.exception(e)
+            yield "data:" + json.dumps({"code": 500, "message": str(e), "data": False}, ensure_ascii=False) + "\n\n"
+
+    resp = Response(sse(), mimetype="text/event-stream")
+    resp.headers.add_header("Cache-control", "no-cache")
+    resp.headers.add_header("Connection", "keep-alive")
+    resp.headers.add_header("X-Accel-Buffering", "no")
+    resp.headers.add_header("Content-Type", "text/event-stream; charset=utf-8")
+    return resp

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")
@@ -167,7 +166,7 @@ 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"] = []

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,
@@ -50,7 +47,8 @@ from api.utils.validation_utils import (
     validate_and_parse_request_args,
 )
 from rag.nlp import search
-from rag.settings import PAGERANK_FLD
+from common.constants import PAGERANK_FLD
+from common import settings
 
 
 @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"
-                     ]
-              description: Chunking method.
+                     "picture", "presentation", "qa", "table", "tag"]
+              description: Chunking method; if omitted, defaults to "naive".
             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):
-            return get_error_operating_result(message=f"Dataset name '{req['name']}' already exists")
-
-        req["parser_config"] = get_parser_config(req["parser_id"], req["parser_config"])
-        req["id"] = get_uuid()
-        req["tenant_id"] = tenant_id
-        req["created_by"] = tenant_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
-
-        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:
+      if not KnowledgebaseService.save(**req):
+          return get_error_data_result()
+      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 Exception 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):
@@ -488,7 +484,7 @@ def knowledge_graph(tenant_id, dataset_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 = {
@@ -529,7 +525,7 @@ def delete_knowledge_graph(tenant_id, dataset_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"]},

api/apps/sdk/dify_retrieval.py

@@ -17,15 +17,14 @@ 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
-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 settings
 
 @manager.route('/dify/retrieval', methods=['POST'])  # noqa: F821
 @apikey_required
@@ -129,7 +128,7 @@ def retrieval(tenant_id):
 
         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)
@@ -179,7 +178,7 @@ 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

@@ -24,9 +24,8 @@ from flask import request, send_file
 from peewee import OperationalError
 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
@@ -41,8 +40,9 @@ from rag.app.qa import beAdoc, rmPrefix
 from rag.app.tag import label_question
 from rag.nlp import rag_tokenizer, search
 from rag.prompts.generator import cross_languages, keyword_extraction
-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 settings
 
 MAXIMUM_OF_UPLOADING_FILES = 256
 
@@ -127,13 +127,13 @@ 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 +145,7 @@ 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 +153,7 @@ 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 +253,12 @@ 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"]:
@@ -400,9 +400,9 @@ def download(tenant_id, dataset_id, document_id):
         return get_error_data_result(message=f"The dataset not own the document {document_id}.")
     # The process of downloading
     doc_id, doc_location = File2DocumentService.get_storage_address(doc_id=document_id)  # minio address
-    file_stream = STORAGE_IMPL.get(doc_id, doc_location)
+    file_stream = settings.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(
@@ -670,16 +670,16 @@ def delete(tenant_id, dataset_id):
             )
             File2DocumentService.delete_by_document_id(doc_id)
 
-            STORAGE_IMPL.rm(b, n)
+            settings.STORAGE_IMPL.rm(b, n)
             success_count += 1
         except Exception as e:
             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:
@@ -763,7 +763,7 @@ 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(
@@ -969,7 +969,7 @@ def list_chunks(tenant_id, dataset_id, document_id):
     if req.get("id"):
         chunk = settings.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):
@@ -1301,6 +1301,10 @@ 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:
@@ -1414,7 +1418,7 @@ 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.")
@@ -1505,6 +1509,6 @@ 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

@@ -26,13 +26,13 @@ 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
 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
+from common import settings
 
 
 @manager.route('/file/upload', methods=['POST'])  # noqa: F821
@@ -126,7 +126,7 @@ def upload(tenant_id):
 
             filetype = filename_type(file_obj_names[file_len - 1])
             location = file_obj_names[file_len - 1]
-            while STORAGE_IMPL.obj_exist(last_folder.id, location):
+            while settings.STORAGE_IMPL.obj_exist(last_folder.id, location):
                 location += "_"
             blob = file_obj.read()
             filename = duplicate_name(FileService.query, name=file_obj_names[file_len - 1], parent_id=last_folder.id)
@@ -142,7 +142,7 @@ def upload(tenant_id):
                 "size": len(blob),
             }
             file = FileService.insert(file)
-            STORAGE_IMPL.put(last_folder.id, location, blob)
+            settings.STORAGE_IMPL.put(last_folder.id, location, blob)
             file_res.append(file.to_json())
         return get_json_result(data=file_res)
     except Exception as e:
@@ -497,10 +497,10 @@ def rm(tenant_id):
                     e, file = FileService.get_by_id(inner_file_id)
                     if not e:
                         return get_json_result(message="File not found!", code=404)
-                    STORAGE_IMPL.rm(file.parent_id, file.location)
+                    settings.STORAGE_IMPL.rm(file.parent_id, file.location)
                 FileService.delete_folder_by_pf_id(tenant_id, file_id)
             else:
-                STORAGE_IMPL.rm(file.parent_id, file.location)
+                settings.STORAGE_IMPL.rm(file.parent_id, file.location)
                 if not FileService.delete(file):
                     return get_json_result(message="Database error (File removal)!", code=500)
 
@@ -614,10 +614,10 @@ def get(tenant_id, file_id):
         if not e:
             return get_json_result(message="Document not found!", code=404)
 
-        blob = STORAGE_IMPL.get(file.parent_id, file.location)
+        blob = settings.STORAGE_IMPL.get(file.parent_id, file.location)
         if not blob:
             b, n = File2DocumentService.get_storage_address(file_id=file_id)
-            blob = STORAGE_IMPL.get(b, n)
+            blob = settings.STORAGE_IMPL.get(b, n)
 
         response = flask.make_response(blob)
         ext = re.search(r"\.([^.]+)$", file.name)

api/apps/sdk/session.py

@@ -21,11 +21,9 @@ import tiktoken
 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 +33,14 @@ 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 rag.app.tag import label_question
 from rag.prompts.template import load_prompt
 from rag.prompts.generator import cross_languages, gen_meta_filter, keyword_extraction, chunks_format
-
+from common.constants import RetCode, LLMType, StatusEnum
+from common import settings
 
 @manager.route("/chats/<chat_id>/sessions", methods=["POST"])  # noqa: F821
 @token_required
@@ -412,7 +411,7 @@ 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,
@@ -430,7 +429,7 @@ 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,
@@ -959,7 +958,7 @@ 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))
@@ -996,7 +995,7 @@ def retrieval_test_embedded():
                     break
             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:
@@ -1034,7 +1033,7 @@ def retrieval_test_embedded():
     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)
 
 
@@ -1104,7 +1103,7 @@ def detail_share_embedded():
                 break
         else:
             return get_json_result(data=False, message="Has no permission for this operation.",
-                                   code=settings.RetCode.OPERATING_ERROR)
+                                   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

@@ -23,21 +23,20 @@ from api.db.db_models import APIToken
 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.api_utils import (
     get_json_result,
     get_data_error_result,
     server_error_response,
     generate_confirmation_token,
 )
-from api.versions import get_ragflow_version
+from common.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 settings
 
 
 @manager.route("/version", methods=["GET"])  # noqa: F821
@@ -112,15 +111,15 @@ def status():
 
     st = timer()
     try:
-        STORAGE_IMPL.health()
+        settings.STORAGE_IMPL.health()
         res["storage"] = {
-            "storage": STORAGE_IMPL_TYPE.lower(),
+            "storage": settings.STORAGE_IMPL_TYPE.lower(),
             "status": "green",
             "elapsed": "{:.1f}".format((timer() - st) * 1000.0),
         }
     except Exception as e:
         res["storage"] = {
-            "storage": STORAGE_IMPL_TYPE.lower(),
+            "storage": settings.STORAGE_IMPL_TYPE.lower(),
             "status": "red",
             "elapsed": "{:.1f}".format((timer() - st) * 1000.0),
             "error": str(e),

api/apps/tenant_app.py

@@ -17,16 +17,17 @@
 from flask import request
 from flask_login import login_required, current_user
 
-from api import settings
 from api.apps import smtp_mail_server
-from api.db import UserTenantRole, StatusEnum
+from api.db import UserTenantRole
 from api.db.db_models import UserTenant
 from api.db.services.user_service import UserTenantService, UserService
 
-from api.utils import get_uuid
+from common.constants import RetCode, StatusEnum
+from common.misc_utils import get_uuid
 from common.time_utils import delta_seconds
 from api.utils.api_utils import get_json_result, validate_request, server_error_response, get_data_error_result
 from api.utils.web_utils import send_invite_email
+from common import settings
 
 
 @manager.route("/<tenant_id>/user/list", methods=["GET"])  # noqa: F821
@@ -36,7 +37,7 @@ def user_list(tenant_id):
         return get_json_result(
             data=False,
             message='No authorization.',
-            code=settings.RetCode.AUTHENTICATION_ERROR)
+            code=RetCode.AUTHENTICATION_ERROR)
 
     try:
         users = UserTenantService.get_by_tenant_id(tenant_id)
@@ -55,7 +56,7 @@ def create(tenant_id):
         return get_json_result(
             data=False,
             message='No authorization.',
-            code=settings.RetCode.AUTHENTICATION_ERROR)
+            code=RetCode.AUTHENTICATION_ERROR)
 
     req = request.json
     invite_user_email = req["email"]
@@ -109,7 +110,7 @@ def rm(tenant_id, user_id):
         return get_json_result(
             data=False,
             message='No authorization.',
-            code=settings.RetCode.AUTHENTICATION_ERROR)
+            code=RetCode.AUTHENTICATION_ERROR)
 
     try:
         UserTenantService.filter_delete([UserTenant.tenant_id == tenant_id, UserTenant.user_id == user_id])

api/apps/user_app.py

@@ -26,7 +26,6 @@ from flask import redirect, request, session, make_response
 from flask_login import current_user, login_required, login_user, logout_user
 from werkzeug.security import check_password_hash, generate_password_hash
 
-from api import settings
 from api.apps.auth import get_auth_client
 from api.db import FileType, UserTenantRole
 from api.db.db_models import TenantLLM
@@ -35,9 +34,10 @@ from api.db.services.llm_service import get_init_tenant_llm
 from api.db.services.tenant_llm_service import TenantLLMService
 from api.db.services.user_service import TenantService, UserService, UserTenantService
 from common.time_utils import current_timestamp, datetime_format, get_format_time
-from api.utils import download_img, get_uuid
+from common.misc_utils import download_img, get_uuid
+from common.constants import RetCode
+from common.connection_utils import construct_response
 from api.utils.api_utils import (
-    construct_response,
     get_data_error_result,
     get_json_result,
     server_error_response,
@@ -57,6 +57,7 @@ from api.utils.web_utils import (
     hash_code,
     captcha_key,
 )
+from common import settings
 
 
 @manager.route("/login", methods=["POST", "GET"])  # noqa: F821
@@ -91,14 +92,14 @@ def login():
           type: object
     """
     if not request.json:
-        return get_json_result(data=False, code=settings.RetCode.AUTHENTICATION_ERROR, message="Unauthorized!")
+        return get_json_result(data=False, code=RetCode.AUTHENTICATION_ERROR, message="Unauthorized!")
 
     email = request.json.get("email", "")
     users = UserService.query(email=email)
     if not users:
         return get_json_result(
             data=False,
-            code=settings.RetCode.AUTHENTICATION_ERROR,
+            code=RetCode.AUTHENTICATION_ERROR,
             message=f"Email: {email} is not registered!",
         )
 
@@ -106,14 +107,14 @@ def login():
     try:
         password = decrypt(password)
     except BaseException:
-        return get_json_result(data=False, code=settings.RetCode.SERVER_ERROR, message="Fail to crypt password")
+        return get_json_result(data=False, code=RetCode.SERVER_ERROR, message="Fail to crypt password")
 
     user = UserService.query_user(email, password)
 
     if user and hasattr(user, 'is_active') and user.is_active == "0":
         return get_json_result(
             data=False,
-            code=settings.RetCode.FORBIDDEN,
+            code=RetCode.FORBIDDEN,
             message="This account has been disabled, please contact the administrator!",
         )
     elif user:
@@ -128,7 +129,7 @@ def login():
     else:
         return get_json_result(
             data=False,
-            code=settings.RetCode.AUTHENTICATION_ERROR,
+            code=RetCode.AUTHENTICATION_ERROR,
             message="Email and password do not match!",
         )
 
@@ -151,7 +152,7 @@ def get_login_channels():
         return get_json_result(data=channels)
     except Exception as e:
         logging.exception(e)
-        return get_json_result(data=[], message=f"Load channels failure, error: {str(e)}", code=settings.RetCode.EXCEPTION_ERROR)
+        return get_json_result(data=[], message=f"Load channels failure, error: {str(e)}", code=RetCode.EXCEPTION_ERROR)
 
 
 @manager.route("/login/<channel>", methods=["GET"])  # noqa: F821
@@ -535,7 +536,7 @@ def setting_user():
         if not check_password_hash(current_user.password, decrypt(request_data["password"])):
             return get_json_result(
                 data=False,
-                code=settings.RetCode.AUTHENTICATION_ERROR,
+                code=RetCode.AUTHENTICATION_ERROR,
                 message="Password error!",
             )
 
@@ -563,7 +564,7 @@ def setting_user():
         return get_json_result(data=True)
     except Exception as e:
         logging.exception(e)
-        return get_json_result(data=False, message="Update failure!", code=settings.RetCode.EXCEPTION_ERROR)
+        return get_json_result(data=False, message="Update failure!", code=RetCode.EXCEPTION_ERROR)
 
 
 @manager.route("/info", methods=["GET"])  # noqa: F821
@@ -693,7 +694,7 @@ def user_add():
         return get_json_result(
             data=False,
             message="User registration is disabled!",
-            code=settings.RetCode.OPERATING_ERROR,
+            code=RetCode.OPERATING_ERROR,
         )
 
     req = request.json
@@ -704,7 +705,7 @@ def user_add():
         return get_json_result(
             data=False,
             message=f"Invalid email address: {email_address}!",
-            code=settings.RetCode.OPERATING_ERROR,
+            code=RetCode.OPERATING_ERROR,
         )
 
     # Check if the email address is already used
@@ -712,7 +713,7 @@ def user_add():
         return get_json_result(
             data=False,
             message=f"Email: {email_address} has already registered!",
-            code=settings.RetCode.OPERATING_ERROR,
+            code=RetCode.OPERATING_ERROR,
         )
 
     # Construct user info data
@@ -747,7 +748,7 @@ def user_add():
         return get_json_result(
             data=False,
             message=f"User registration failure, error: {str(e)}",
-            code=settings.RetCode.EXCEPTION_ERROR,
+            code=RetCode.EXCEPTION_ERROR,
         )
 
 
@@ -847,11 +848,11 @@ def forget_get_captcha():
     """
     email = (request.args.get("email") or "")
     if not email:
-        return get_json_result(data=False, code=settings.RetCode.ARGUMENT_ERROR, message="email is required")
+        return get_json_result(data=False, code=RetCode.ARGUMENT_ERROR, message="email is required")
 
     users = UserService.query(email=email)
     if not users:
-        return get_json_result(data=False, code=settings.RetCode.DATA_ERROR, message="invalid email")
+        return get_json_result(data=False, code=RetCode.DATA_ERROR, message="invalid email")
 
     # Generate captcha text
     allowed = string.ascii_uppercase + string.digits
@@ -878,17 +879,17 @@ def forget_send_otp():
     captcha = (req.get("captcha") or "").strip()
 
     if not email or not captcha:
-        return get_json_result(data=False, code=settings.RetCode.ARGUMENT_ERROR, message="email and captcha required")
+        return get_json_result(data=False, code=RetCode.ARGUMENT_ERROR, message="email and captcha required")
 
     users = UserService.query(email=email)
     if not users:
-        return get_json_result(data=False, code=settings.RetCode.DATA_ERROR, message="invalid email")
+        return get_json_result(data=False, code=RetCode.DATA_ERROR, message="invalid email")
 
     stored_captcha = REDIS_CONN.get(captcha_key(email))
     if not stored_captcha:
-        return get_json_result(data=False, code=settings.RetCode.NOT_EFFECTIVE, message="invalid or expired captcha")
+        return get_json_result(data=False, code=RetCode.NOT_EFFECTIVE, message="invalid or expired captcha")
     if (stored_captcha or "").strip().lower() != captcha.lower():
-        return get_json_result(data=False, code=settings.RetCode.AUTHENTICATION_ERROR, message="invalid or expired captcha")
+        return get_json_result(data=False, code=RetCode.AUTHENTICATION_ERROR, message="invalid or expired captcha")
 
     # Delete captcha to prevent reuse
     REDIS_CONN.delete(captcha_key(email))
@@ -903,7 +904,7 @@ def forget_send_otp():
             elapsed = RESEND_COOLDOWN_SECONDS
         remaining = RESEND_COOLDOWN_SECONDS - elapsed
         if remaining > 0:
-            return get_json_result(data=False, code=settings.RetCode.NOT_EFFECTIVE, message=f"you still have to wait {remaining} seconds")
+            return get_json_result(data=False, code=RetCode.NOT_EFFECTIVE, message=f"you still have to wait {remaining} seconds")
 
     # Generate OTP (uppercase letters only) and store hashed
     otp = "".join(secrets.choice(string.ascii_uppercase) for _ in range(OTP_LENGTH))
@@ -928,9 +929,9 @@ def forget_send_otp():
                 ttl_min=ttl_min,
             )
         except Exception:
-            return get_json_result(data=False, code=settings.RetCode.SERVER_ERROR, message="failed to send email")
+            return get_json_result(data=False, code=RetCode.SERVER_ERROR, message="failed to send email")
         
-    return get_json_result(data=True, code=settings.RetCode.SUCCESS, message="verification passed, email sent")
+    return get_json_result(data=True, code=RetCode.SUCCESS, message="verification passed, email sent")
 
 
 @manager.route("/forget", methods=["POST"])  # noqa: F821
@@ -946,31 +947,31 @@ def forget():
     new_pwd2 = req.get("confirm_new_password")
 
     if not all([email, otp, new_pwd, new_pwd2]):
-        return get_json_result(data=False, code=settings.RetCode.ARGUMENT_ERROR, message="email, otp and passwords are required")
+        return get_json_result(data=False, code=RetCode.ARGUMENT_ERROR, message="email, otp and passwords are required")
 
     # For reset, passwords are provided as-is (no decrypt needed)
     if new_pwd != new_pwd2:
-        return get_json_result(data=False, code=settings.RetCode.ARGUMENT_ERROR, message="passwords do not match")
+        return get_json_result(data=False, code=RetCode.ARGUMENT_ERROR, message="passwords do not match")
 
     users = UserService.query(email=email)
     if not users:
-        return get_json_result(data=False, code=settings.RetCode.DATA_ERROR, message="invalid email")
+        return get_json_result(data=False, code=RetCode.DATA_ERROR, message="invalid email")
 
     user = users[0]
     # Verify OTP from Redis
     k_code, k_attempts, k_last, k_lock = otp_keys(email)
     if REDIS_CONN.get(k_lock):
-        return get_json_result(data=False, code=settings.RetCode.NOT_EFFECTIVE, message="too many attempts, try later")
+        return get_json_result(data=False, code=RetCode.NOT_EFFECTIVE, message="too many attempts, try later")
 
     stored = REDIS_CONN.get(k_code)
     if not stored:
-        return get_json_result(data=False, code=settings.RetCode.NOT_EFFECTIVE, message="expired otp")
+        return get_json_result(data=False, code=RetCode.NOT_EFFECTIVE, message="expired otp")
 
     try:
         stored_hash, salt_hex = str(stored).split(":", 1)
         salt = bytes.fromhex(salt_hex)
     except Exception:
-        return get_json_result(data=False, code=settings.RetCode.EXCEPTION_ERROR, message="otp storage corrupted")
+        return get_json_result(data=False, code=RetCode.EXCEPTION_ERROR, message="otp storage corrupted")
 
     # Case-insensitive verification: OTP generated uppercase
     calc = hash_code(otp.upper(), salt)
@@ -983,7 +984,7 @@ def forget():
         REDIS_CONN.set(k_attempts, attempts, OTP_TTL_SECONDS)
         if attempts >= ATTEMPT_LIMIT:
             REDIS_CONN.set(k_lock, int(time.time()), ATTEMPT_LOCK_SECONDS)
-        return get_json_result(data=False, code=settings.RetCode.AUTHENTICATION_ERROR, message="expired otp")
+        return get_json_result(data=False, code=RetCode.AUTHENTICATION_ERROR, message="expired otp")
 
     # Success: consume OTP and reset password
     REDIS_CONN.delete(k_code)
@@ -995,7 +996,7 @@ def forget():
         UserService.update_user_password(user.id, new_pwd)
     except Exception as e:
         logging.exception(e)
-        return get_json_result(data=False, code=settings.RetCode.EXCEPTION_ERROR, message="failed to reset password")
+        return get_json_result(data=False, code=RetCode.EXCEPTION_ERROR, message="failed to reset password")
 
     # Auto login (reuse login flow)
     user.access_token = get_uuid()

api/constants.py

@@ -17,8 +17,6 @@ NAME_LENGTH_LIMIT = 2**10
 
 IMG_BASE64_PREFIX = "data:image/png;base64,"
 
-SERVICE_CONF = "service_conf.yaml"
-
 API_VERSION = "v1"
 RAG_FLOW_SERVICE_NAME = "ragflow"
 REQUEST_WAIT_SEC = 2

api/db/__init__.py

@@ -13,21 +13,11 @@
 #  See the License for the specific language governing permissions and
 #  limitations under the License.
 #
-from enum import Enum
+
 from enum import IntEnum
 from strenum import StrEnum
 
 
-class StatusEnum(Enum):
-    VALID = "1"
-    INVALID = "0"
-
-
-class ActiveEnum(Enum):
-    ACTIVE = "1"
-    INACTIVE = "0"
-
-
 class UserTenantRole(StrEnum):
     OWNER = 'owner'
     ADMIN = 'admin'
@@ -56,76 +46,18 @@ class FileType(StrEnum):
 
 VALID_FILE_TYPES = {FileType.PDF, FileType.DOC, FileType.VISUAL, FileType.AURAL, FileType.VIRTUAL, FileType.FOLDER, FileType.OTHER}
 
-class LLMType(StrEnum):
-    CHAT = 'chat'
-    EMBEDDING = 'embedding'
-    SPEECH2TEXT = 'speech2text'
-    IMAGE2TEXT = 'image2text'
-    RERANK = 'rerank'
-    TTS    = 'tts'
 
-
-class ChatStyle(StrEnum):
-    CREATIVE = 'Creative'
-    PRECISE = 'Precise'
-    EVENLY = 'Evenly'
-    CUSTOM = 'Custom'
-
-
-class TaskStatus(StrEnum):
-    UNSTART = "0"
-    RUNNING = "1"
-    CANCEL = "2"
-    DONE = "3"
-    FAIL = "4"
-
-
-VALID_TASK_STATUS     = {TaskStatus.UNSTART, TaskStatus.RUNNING, TaskStatus.CANCEL, TaskStatus.DONE, TaskStatus.FAIL}
-
-
-class ParserType(StrEnum):
-    PRESENTATION = "presentation"
-    LAWS = "laws"
-    MANUAL = "manual"
-    PAPER = "paper"
-    RESUME = "resume"
-    BOOK = "book"
-    QA = "qa"
-    TABLE = "table"
-    NAIVE = "naive"
-    PICTURE = "picture"
-    ONE = "one"
-    AUDIO = "audio"
-    EMAIL = "email"
-    KG = "knowledge_graph"
-    TAG = "tag"
-
-
-class FileSource(StrEnum):
-    LOCAL = ""
-    KNOWLEDGEBASE = "knowledgebase"
-    S3 = "s3"
-
-
-class CanvasType(StrEnum):
-    ChatBot = "chatbot"
-    DocBot = "docbot"
+class InputType(StrEnum):
+    LOAD_STATE = "load_state"  # e.g. loading a current full state or a save state, such as from a file
+    POLL = "poll"  # e.g. calling an API to get all documents in the last hour
+    EVENT = "event"  # e.g. registered an endpoint as a listener, and processing connector events
+    SLIM_RETRIEVAL = "slim_retrieval"
 
 
 class CanvasCategory(StrEnum):
     Agent = "agent_canvas"
     DataFlow = "dataflow_canvas"
 
-VALID_CANVAS_CATEGORIES = {CanvasCategory.Agent, CanvasCategory.DataFlow}
-
-
-class MCPServerType(StrEnum):
-    SSE = "sse"
-    STREAMABLE_HTTP = "streamable-http"
-
-
-VALID_MCP_SERVER_TYPES = {MCPServerType.SSE, MCPServerType.STREAMABLE_HTTP}
-
 
 class PipelineTaskType(StrEnum):
     PARSE = "Parse"
@@ -138,4 +70,7 @@ class PipelineTaskType(StrEnum):
 VALID_PIPELINE_TASK_TYPES = {PipelineTaskType.PARSE, PipelineTaskType.DOWNLOAD, PipelineTaskType.RAPTOR, PipelineTaskType.GRAPH_RAG, PipelineTaskType.MINDMAP}
 
 
+PIPELINE_SPECIAL_PROGRESS_FREEZE_TASK_TYPES = {PipelineTaskType.RAPTOR.lower(), PipelineTaskType.GRAPH_RAG.lower(), PipelineTaskType.MINDMAP.lower()}
+
+
 KNOWLEDGEBASE_FOLDER_NAME=".knowledgebase"

api/db/db_models.py

@@ -21,6 +21,7 @@ import os
 import sys
 import time
 import typing
+from datetime import datetime, timezone
 from enum import Enum
 from functools import wraps
 
@@ -30,24 +31,15 @@ from peewee import InterfaceError, OperationalError, BigIntegerField, BooleanFie
 from playhouse.migrate import MySQLMigrator, PostgresqlMigrator, migrate
 from playhouse.pool import PooledMySQLDatabase, PooledPostgresqlDatabase
 
-from api import settings, utils
-from api.db import ParserType, SerializedType
+from api import utils
+from api.db import SerializedType
 from api.utils.json_encode import json_dumps, json_loads
 from api.utils.configs import deserialize_b64, serialize_b64
 
 from common.time_utils import current_timestamp, timestamp_to_date, date_string_to_timestamp
-
-
-def singleton(cls, *args, **kw):
-    instances = {}
-
-    def _singleton():
-        key = str(cls) + str(os.getpid())
-        if key not in instances:
-            instances[key] = cls(*args, **kw)
-        return instances[key]
-
-    return _singleton
+from common.decorator import singleton
+from common.constants import ParserType
+from common import settings
 
 
 CONTINUOUS_FIELD_TYPE = {IntegerField, FloatField, DateTimeField}
@@ -379,6 +371,7 @@ class RetryingPooledPostgresqlDatabase(PooledPostgresqlDatabase):
                     time.sleep(self.retry_delay * (2 ** attempt))
                 else:
                     raise
+        return None
 
 
 class PooledDatabase(Enum):
@@ -676,6 +669,7 @@ class LLMFactories(DataBaseModel):
     name = CharField(max_length=128, null=False, help_text="LLM factory name", primary_key=True)
     logo = TextField(null=True, help_text="llm logo base64")
     tags = CharField(max_length=255, null=False, help_text="LLM, Text Embedding, Image2Text, ASR", index=True)
+    rank = IntegerField(default=0, index=False)
     status = CharField(max_length=1, null=True, help_text="is it validate(0: wasted, 1: validate)", default="1", index=True)
 
     def __str__(self):
@@ -713,6 +707,7 @@ class TenantLLM(DataBaseModel):
     api_base = CharField(max_length=255, null=True, help_text="API Base")
     max_tokens = IntegerField(default=8192, index=True)
     used_tokens = IntegerField(default=0, index=True)
+    status = CharField(max_length=1, null=False, help_text="is it validate(0: wasted, 1: validate)", default="1", index=True)
 
     def __str__(self):
         return self.llm_name
@@ -1046,6 +1041,77 @@ class PipelineOperationLog(DataBaseModel):
         db_table = "pipeline_operation_log"
 
 
+class Connector(DataBaseModel):
+    id = CharField(max_length=32, primary_key=True)
+    tenant_id = CharField(max_length=32, null=False, index=True)
+    name = CharField(max_length=128, null=False, help_text="Search name", index=False)
+    source = CharField(max_length=128, null=False, help_text="Data source", index=True)
+    input_type = CharField(max_length=128, null=False, help_text="poll/event/..", index=True)
+    config = JSONField(null=False, default={})
+    refresh_freq = IntegerField(default=0, index=False)
+    prune_freq = IntegerField(default=0, index=False)
+    timeout_secs = IntegerField(default=3600, index=False)
+    indexing_start = DateTimeField(null=True, index=True)
+    status = CharField(max_length=16, null=True, help_text="schedule", default="schedule", index=True)
+
+    def __str__(self):
+        return self.name
+
+    class Meta:
+        db_table = "connector"
+
+
+class Connector2Kb(DataBaseModel):
+    id = CharField(max_length=32, primary_key=True)
+    connector_id = CharField(max_length=32, null=False, index=True)
+    kb_id = CharField(max_length=32, null=False, index=True)
+    auto_parse = CharField(max_length=1, null=False, default="1", index=False)
+
+    class Meta:
+        db_table = "connector2kb"
+
+
+class DateTimeTzField(CharField):
+    field_type = 'VARCHAR'
+
+    def db_value(self, value: datetime|None) -> str|None:
+        if value is not None:
+            if value.tzinfo is not None:
+                return value.isoformat()
+            else:
+                return value.replace(tzinfo=timezone.utc).isoformat()
+        return value
+
+    def python_value(self, value: str|None) -> datetime|None:
+        if value is not None:
+            dt = datetime.fromisoformat(value)
+            if dt.tzinfo is None:
+                import pytz
+                return dt.replace(tzinfo=pytz.UTC)
+            return dt
+        return value
+
+
+class SyncLogs(DataBaseModel):
+    id = CharField(max_length=32, primary_key=True)
+    connector_id = CharField(max_length=32, index=True)
+    status = CharField(max_length=128, null=False, help_text="Processing status", index=True)
+    from_beginning = CharField(max_length=1, null=True, help_text="", default="0", index=False)
+    new_docs_indexed = IntegerField(default=0, index=False)
+    total_docs_indexed = IntegerField(default=0, index=False)
+    docs_removed_from_index = IntegerField(default=0, index=False)
+    error_msg = TextField(null=False, help_text="process message", default="")
+    error_count = IntegerField(default=0, index=False)
+    full_exception_trace = TextField(null=True, help_text="process message", default="")
+    time_started = DateTimeField(null=True, index=True)
+    poll_range_start = DateTimeTzField(max_length=255, null=True, index=True)
+    poll_range_end = DateTimeTzField(max_length=255, null=True, index=True)
+    kb_id = CharField(max_length=32, null=False, index=True)
+
+    class Meta:
+        db_table = "sync_logs"
+
+
 def migrate_db():
     logging.disable(logging.ERROR)
     migrator = DatabaseMigrator[settings.DATABASE_TYPE.upper()].value(DB)
@@ -1214,4 +1280,16 @@ def migrate_db():
         migrate(migrator.alter_column_type("tenant_llm", "api_key", TextField(null=True, help_text="API KEY")))
     except Exception:
         pass
+    try:
+        migrate(migrator.add_column("tenant_llm", "status", CharField(max_length=1, null=False, help_text="is it validate(0: wasted, 1: validate)", default="1", index=True)))
+    except Exception:
+        pass
+    try:
+        migrate(migrator.add_column("connector2kb", "auto_parse", CharField(max_length=1, null=False, default="1", index=False)))
+    except Exception:
+        pass
+    try:
+        migrate(migrator.add_column("llm_factories", "rank", IntegerField(default=0, index=False)))
+    except Exception:
+        pass
     logging.disable(logging.NOTSET)

api/db/init_data.py

@@ -20,7 +20,7 @@ import time
 import uuid
 from copy import deepcopy
 
-from api.db import LLMType, UserTenantRole
+from api.db import UserTenantRole
 from api.db.db_models import init_database_tables as init_web_db, LLMFactories, LLM, TenantLLM
 from api.db.services import UserService
 from api.db.services.canvas_service import CanvasTemplateService
@@ -29,8 +29,9 @@ from api.db.services.knowledgebase_service import KnowledgebaseService
 from api.db.services.tenant_llm_service import LLMFactoriesService, TenantLLMService
 from api.db.services.llm_service import LLMService, LLMBundle, get_init_tenant_llm
 from api.db.services.user_service import TenantService, UserTenantService
-from api import settings
-from api.utils.file_utils import get_project_base_directory
+from common.constants import LLMType
+from common.file_utils import get_project_base_directory
+from common import settings
 from api.common.base64 import encode_to_base64
 
 
@@ -88,13 +89,7 @@ def init_superuser():
 
 
 def init_llm_factory():
-    try:
-        LLMService.filter_delete([(LLM.fid == "MiniMax" or LLM.fid == "Minimax")])
-        LLMService.filter_delete([(LLM.fid == "cohere")])
-        LLMFactoriesService.filter_delete([LLMFactories.name == "cohere"])
-    except Exception:
-        pass
-
+    LLMFactoriesService.filter_delete([1 == 1])
     factory_llm_infos = settings.FACTORY_LLM_INFOS
     for factory_llm_info in factory_llm_infos:
         info = deepcopy(factory_llm_info)

api/db/joint_services/user_account_service.py

@@ -16,9 +16,8 @@
 import logging
 import uuid
 
-from api import settings
 from api.utils.api_utils import group_by
-from api.db import FileType, UserTenantRole, ActiveEnum
+from api.db import FileType, UserTenantRole
 from api.db.services.api_service import APITokenService, API4ConversationService
 from api.db.services.canvas_service import UserCanvasService
 from api.db.services.conversation_service import ConversationService
@@ -35,9 +34,9 @@ from api.db.services.task_service import TaskService
 from api.db.services.tenant_llm_service import TenantLLMService
 from api.db.services.user_canvas_version import UserCanvasVersionService
 from api.db.services.user_service import TenantService, UserService, UserTenantService
-from rag.utils.storage_factory import STORAGE_IMPL
 from rag.nlp import search
-
+from common.constants import ActiveEnum
+from common import settings
 
 def create_new_user(user_info: dict) -> dict:
     """
@@ -158,8 +157,8 @@ def delete_user_data(user_id: str) -> dict:
             if kb_ids:
                 # step1.1.1 delete files in storage, remove bucket
                 for kb_id in kb_ids:
-                    if STORAGE_IMPL.bucket_exists(kb_id):
-                        STORAGE_IMPL.remove_bucket(kb_id)
+                    if settings.STORAGE_IMPL.bucket_exists(kb_id):
+                        settings.STORAGE_IMPL.remove_bucket(kb_id)
                 done_msg += f"- Removed {len(kb_ids)} dataset's buckets.\n"
                 # step1.1.2 delete file and document info in db
                 doc_ids = DocumentService.get_all_doc_ids_by_kb_ids(kb_ids)
@@ -218,7 +217,7 @@ def delete_user_data(user_id: str) -> dict:
                     if created_files:
                         # step2.1.1.1 delete file in storage
                         for f in created_files:
-                            STORAGE_IMPL.rm(f.parent_id, f.location)
+                            settings.STORAGE_IMPL.rm(f.parent_id, f.location)
                         done_msg += f"- Deleted {len(created_files)} uploaded file.\n"
                         # step2.1.1.2 delete file record
                         file_delete_res = FileService.delete_by_ids([f.id for f in created_files])

api/db/runtime_config.py

@@ -13,7 +13,7 @@
 #  See the License for the specific language governing permissions and
 #  limitations under the License.
 #
-from api.versions import get_ragflow_version
+from common.versions import get_ragflow_version
 from .reload_config_base import ReloadConfigBase
 
 

api/db/services/canvas_service.py

@@ -22,7 +22,7 @@ from api.db import CanvasCategory, TenantPermission
 from api.db.db_models import DB, CanvasTemplate, User, UserCanvas, API4Conversation
 from api.db.services.api_service import API4ConversationService
 from api.db.services.common_service import CommonService
-from api.utils import get_uuid
+from common.misc_utils import get_uuid
 from api.utils.api_utils import get_data_openai
 import tiktoken
 from peewee import fn
@@ -67,6 +67,7 @@ class UserCanvasService(CommonService):
         # will get all permitted agents, be cautious
         fields = [
             cls.model.id,
+            cls.model.avatar,
             cls.model.title,
             cls.model.permission,
             cls.model.canvas_type,
@@ -232,9 +233,9 @@ def completion(tenant_id, agent_id, session_id=None, **kwargs):
     API4ConversationService.append_message(conv["id"], conv)
 
 
-def completionOpenAI(tenant_id, agent_id, question, session_id=None, stream=True, **kwargs):
-    tiktokenenc = tiktoken.get_encoding("cl100k_base")
-    prompt_tokens = len(tiktokenenc.encode(str(question)))
+def completion_openai(tenant_id, agent_id, question, session_id=None, stream=True, **kwargs):
+    tiktoken_encoder = tiktoken.get_encoding("cl100k_base")
+    prompt_tokens = len(tiktoken_encoder.encode(str(question)))
     user_id = kwargs.get("user_id", "")
 
     if stream:
@@ -252,7 +253,7 @@ def completionOpenAI(tenant_id, agent_id, question, session_id=None, stream=True
                     try:
                         ans = json.loads(ans[5:])  # remove "data:"
                     except Exception as e:
-                        logging.exception(f"Agent OpenAI-Compatible completionOpenAI parse answer failed: {e}")
+                        logging.exception(f"Agent OpenAI-Compatible completion_openai parse answer failed: {e}")
                         continue
                 if ans.get("event") not in ["message", "message_end"]:
                     continue
@@ -261,7 +262,7 @@ def completionOpenAI(tenant_id, agent_id, question, session_id=None, stream=True
                 if ans["event"] == "message":
                     content_piece = ans["data"]["content"]
 
-                completion_tokens += len(tiktokenenc.encode(content_piece))
+                completion_tokens += len(tiktoken_encoder.encode(content_piece))
 
                 openai_data = get_data_openai(
                         id=session_id or str(uuid4()),
@@ -288,7 +289,7 @@ def completionOpenAI(tenant_id, agent_id, question, session_id=None, stream=True
                     content=f"**ERROR**: {str(e)}",
                     finish_reason="stop",
                     prompt_tokens=prompt_tokens,
-                    completion_tokens=len(tiktokenenc.encode(f"**ERROR**: {str(e)}")),
+                    completion_tokens=len(tiktoken_encoder.encode(f"**ERROR**: {str(e)}")),
                     stream=True
                 ),
                 ensure_ascii=False
@@ -318,7 +319,7 @@ def completionOpenAI(tenant_id, agent_id, question, session_id=None, stream=True
                 if ans.get("data", {}).get("reference", None):
                     reference.update(ans["data"]["reference"])
 
-            completion_tokens = len(tiktokenenc.encode(all_content))
+            completion_tokens = len(tiktoken_encoder.encode(all_content))
 
             openai_data = get_data_openai(
                 id=session_id or str(uuid4()),
@@ -340,7 +341,7 @@ def completionOpenAI(tenant_id, agent_id, question, session_id=None, stream=True
                 id=session_id or str(uuid4()),
                 model=agent_id,
                 prompt_tokens=prompt_tokens,
-                completion_tokens=len(tiktokenenc.encode(f"**ERROR**: {str(e)}")),
+                completion_tokens=len(tiktoken_encoder.encode(f"**ERROR**: {str(e)}")),
                 content=f"**ERROR**: {str(e)}",
                 finish_reason="stop",
                 param=None

api/db/services/common_service.py

@@ -19,7 +19,7 @@ import peewee
 from peewee import InterfaceError, OperationalError
 
 from api.db.db_models import DB
-from api.utils import get_uuid
+from common.misc_utils import get_uuid
 from common.time_utils import current_timestamp, datetime_format
 
 def retry_db_operation(func):
@@ -90,7 +90,7 @@ class CommonService:
         else:
             query_records = cls.model.select()
         if reverse is not None:
-            if not order_by or not hasattr(cls, order_by):
+            if not order_by or not hasattr(cls.model, order_by):
                 order_by = "create_time"
             if reverse is True:
                 query_records = query_records.order_by(cls.model.getter_by(order_by).desc())

api/db/services/connector_service.py

@@ -0,0 +1,283 @@
+#
+#  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 logging
+from datetime import datetime
+from typing import Tuple, List
+
+from anthropic import BaseModel
+from peewee import SQL, fn
+
+from api.db import InputType
+from api.db.db_models import Connector, SyncLogs, Connector2Kb, Knowledgebase
+from api.db.services.common_service import CommonService
+from api.db.services.document_service import DocumentService
+from api.db.services.file_service import FileService
+from common.misc_utils import get_uuid
+from common.constants import TaskStatus
+from common.time_utils import current_timestamp, timestamp_to_date
+
+
+class ConnectorService(CommonService):
+    model = Connector
+
+    @classmethod
+    def resume(cls, connector_id, status):
+        for c2k in Connector2KbService.query(connector_id=connector_id):
+            task = SyncLogsService.get_latest_task(connector_id, c2k.kb_id)
+            if not task:
+                if status == TaskStatus.SCHEDULE:
+                    SyncLogsService.schedule(connector_id, c2k.kb_id)
+                    ConnectorService.update_by_id(connector_id, {"status": status})
+                    return
+
+            if task.status == TaskStatus.DONE:
+                if status == TaskStatus.SCHEDULE:
+                    SyncLogsService.schedule(connector_id, c2k.kb_id, task.poll_range_end, total_docs_indexed=task.total_docs_indexed)
+                    ConnectorService.update_by_id(connector_id, {"status": status})
+                    return
+
+            task = task.to_dict()
+            task["status"] = status
+            SyncLogsService.update_by_id(task["id"], task)
+        ConnectorService.update_by_id(connector_id, {"status": status})
+
+    @classmethod
+    def list(cls, tenant_id):
+        fields = [
+            cls.model.id,
+            cls.model.name,
+            cls.model.source,
+            cls.model.status
+        ]
+        return list(cls.model.select(*fields).where(
+            cls.model.tenant_id == tenant_id
+        ).dicts())
+
+    @classmethod
+    def rebuild(cls, kb_id:str, connector_id: str, tenant_id:str):
+        e, conn = cls.get_by_id(connector_id)
+        if not e:
+            return
+        SyncLogsService.filter_delete([SyncLogs.connector_id==connector_id, SyncLogs.kb_id==kb_id])
+        docs = DocumentService.query(source_type=f"{conn.source}/{conn.id}", kb_id=kb_id)
+        err = FileService.delete_docs([d.id for d in docs], tenant_id)
+        SyncLogsService.schedule(connector_id, kb_id, reindex=True)
+        return err
+
+
+class SyncLogsService(CommonService):
+    model = SyncLogs
+
+    @classmethod
+    def list_sync_tasks(cls, connector_id=None, page_number=None, items_per_page=15) -> Tuple[List[dict], int]:
+        fields = [
+            cls.model.id,
+            cls.model.connector_id,
+            cls.model.kb_id,
+            cls.model.update_date,
+            cls.model.poll_range_start,
+            cls.model.poll_range_end,
+            cls.model.new_docs_indexed,
+            cls.model.total_docs_indexed,
+            cls.model.error_msg,
+            cls.model.full_exception_trace,
+            cls.model.error_count,
+            Connector.name,
+            Connector.source,
+            Connector.tenant_id,
+            Connector.timeout_secs,
+            Knowledgebase.name.alias("kb_name"),
+            Knowledgebase.avatar.alias("kb_avatar"),
+            Connector2Kb.auto_parse,
+            cls.model.from_beginning.alias("reindex"),
+            cls.model.status
+        ]
+        if not connector_id:
+            fields.append(Connector.config)
+            
+        query = cls.model.select(*fields)\
+            .join(Connector, on=(cls.model.connector_id==Connector.id))\
+            .join(Connector2Kb, on=(cls.model.kb_id==Connector2Kb.kb_id))\
+            .join(Knowledgebase, on=(cls.model.kb_id==Knowledgebase.id))
+
+        if connector_id:
+            query = query.where(cls.model.connector_id == connector_id)
+        else:
+            interval_expr = SQL("INTERVAL `t2`.`refresh_freq` MINUTE")
+            query = query.where(
+                Connector.input_type == InputType.POLL,
+                Connector.status == TaskStatus.SCHEDULE,
+                cls.model.status == TaskStatus.SCHEDULE,
+                cls.model.update_date < (fn.NOW() - interval_expr)
+            )
+
+        query = query.distinct().order_by(cls.model.update_time.desc())
+        totbal = query.count()
+        if page_number:
+            query = query.paginate(page_number, items_per_page)
+
+        return list(query.dicts()), totbal
+
+    @classmethod
+    def start(cls, id, connector_id):
+        cls.update_by_id(id, {"status": TaskStatus.RUNNING, "time_started": datetime.now().strftime('%Y-%m-%d %H:%M:%S') })
+        ConnectorService.update_by_id(connector_id, {"status": TaskStatus.RUNNING})
+
+    @classmethod
+    def done(cls, id, connector_id):
+        cls.update_by_id(id, {"status": TaskStatus.DONE})
+        ConnectorService.update_by_id(connector_id, {"status": TaskStatus.DONE})
+
+    @classmethod
+    def schedule(cls, connector_id, kb_id, poll_range_start=None, reindex=False, total_docs_indexed=0):
+        try:
+            if cls.model.select().where(cls.model.kb_id == kb_id, cls.model.connector_id == connector_id).count() > 100:
+                rm_ids = [m.id for m in cls.model.select(cls.model.id).where(cls.model.kb_id == kb_id, cls.model.connector_id == connector_id).order_by(cls.model.update_time.asc()).limit(70)]
+                deleted = cls.model.delete().where(cls.model.id.in_(rm_ids)).execute()
+                logging.info(f"[SyncLogService] Cleaned {deleted} old logs.")
+        except Exception as e:
+            logging.exception(e)
+
+        try:
+            e = cls.query(kb_id=kb_id, connector_id=connector_id, status=TaskStatus.SCHEDULE)
+            if e:
+                logging.warning(f"{kb_id}--{connector_id} has already had a scheduling sync task which is abnormal.")
+                return None
+            reindex = "1" if reindex else "0"
+            ConnectorService.update_by_id(connector_id, {"status": TaskStatus.SCHEDULE})
+            return cls.save(**{
+                "id": get_uuid(),
+                "kb_id": kb_id, "status": TaskStatus.SCHEDULE, "connector_id": connector_id,
+                "poll_range_start": poll_range_start, "from_beginning": reindex,
+                "total_docs_indexed": total_docs_indexed
+            })
+        except Exception as e:
+            logging.exception(e)
+            task = cls.get_latest_task(connector_id, kb_id)
+            if task:
+                cls.model.update(status=TaskStatus.SCHEDULE,
+                                 poll_range_start=poll_range_start,
+                                 error_msg=cls.model.error_msg + str(e),
+                                 full_exception_trace=cls.model.full_exception_trace + str(e)
+                                 ) \
+                .where(cls.model.id == task.id).execute()
+                ConnectorService.update_by_id(connector_id, {"status": TaskStatus.SCHEDULE})
+
+    @classmethod
+    def increase_docs(cls, id, min_update, max_update, doc_num, err_msg="", error_count=0):
+        cls.model.update(new_docs_indexed=cls.model.new_docs_indexed + doc_num,
+                         total_docs_indexed=cls.model.total_docs_indexed + doc_num,
+                         poll_range_start=fn.COALESCE(fn.LEAST(cls.model.poll_range_start,min_update), min_update),
+                         poll_range_end=fn.COALESCE(fn.GREATEST(cls.model.poll_range_end, max_update), max_update),
+                         error_msg=cls.model.error_msg + err_msg,
+                         error_count=cls.model.error_count + error_count,
+                         update_time=current_timestamp(),
+                         update_date=timestamp_to_date(current_timestamp())
+                         )\
+            .where(cls.model.id == id).execute()
+
+    @classmethod
+    def duplicate_and_parse(cls, kb, docs, tenant_id, src, auto_parse=True):
+        if not docs:
+            return None
+
+        class FileObj(BaseModel):
+            filename: str
+            blob: bytes
+
+            def read(self) -> bytes:
+                return self.blob
+
+        errs = []
+        files = [FileObj(filename=d["semantic_identifier"]+(f"{d['extension']}" if d["semantic_identifier"][::-1].find(d['extension'][::-1])<0 else ""), blob=d["blob"]) for d in docs]
+        doc_ids = []
+        err, doc_blob_pairs = FileService.upload_document(kb, files, tenant_id, src)
+        errs.extend(err)
+
+        kb_table_num_map = {}
+        for doc, _ in doc_blob_pairs:
+            doc_ids.append(doc["id"])
+            if not auto_parse or auto_parse == "0":
+                continue
+            DocumentService.run(tenant_id, doc, kb_table_num_map)
+
+        return errs, doc_ids
+
+    @classmethod
+    def get_latest_task(cls, connector_id, kb_id):
+        return cls.model.select().where(
+            cls.model.connector_id==connector_id,
+            cls.model.kb_id == kb_id
+        ).order_by(cls.model.update_time.desc()).first()
+
+
+class Connector2KbService(CommonService):
+    model = Connector2Kb
+
+    @classmethod
+    def link_connectors(cls, kb_id:str, connectors: list[dict], tenant_id:str):
+        arr = cls.query(kb_id=kb_id)
+        old_conn_ids = [a.connector_id for a in arr]
+        connector_ids = []
+        for conn in connectors:
+            conn_id = conn["id"]
+            connector_ids.append(conn_id)
+            if conn_id in old_conn_ids:
+                cls.filter_update([cls.model.connector_id==conn_id, cls.model.kb_id==kb_id], {"auto_parse": conn.get("auto_parse", "1")})
+                continue
+            cls.save(**{
+                "id": get_uuid(),
+                "connector_id": conn_id,
+                "kb_id": kb_id,
+                "auto_parse": conn.get("auto_parse", "1")   
+            })
+            SyncLogsService.schedule(conn_id, kb_id, reindex=True)
+
+        errs = []
+        for conn_id in old_conn_ids:
+            if conn_id in connector_ids:
+                continue
+            cls.filter_delete([cls.model.kb_id==kb_id, cls.model.connector_id==conn_id])
+            e, conn = ConnectorService.get_by_id(conn_id)
+            if not e:
+                continue
+            #SyncLogsService.filter_delete([SyncLogs.connector_id==conn_id, SyncLogs.kb_id==kb_id])
+            # Do not delete docs while unlinking.
+            SyncLogsService.filter_update([SyncLogs.connector_id==conn_id, SyncLogs.kb_id==kb_id, SyncLogs.status.in_([TaskStatus.SCHEDULE, TaskStatus.RUNNING])], {"status": TaskStatus.CANCEL})
+            #docs = DocumentService.query(source_type=f"{conn.source}/{conn.id}")
+            #err = FileService.delete_docs([d.id for d in docs], tenant_id)
+            #if err:
+            #    errs.append(err)
+        return "\n".join(errs)
+
+    @classmethod
+    def list_connectors(cls, kb_id):
+        fields = [
+            Connector.id,
+            Connector.source,
+            Connector.name,
+            cls.model.auto_parse,
+            Connector.status
+        ]
+        return list(cls.model.select(*fields)\
+                    .join(Connector, on=(cls.model.connector_id==Connector.id))\
+                    .where(
+                        cls.model.kb_id==kb_id
+                    ).dicts()
+        )
+
+
+

api/db/services/conversation_service.py

@@ -15,12 +15,12 @@
 #
 import time
 from uuid import uuid4
-from api.db import StatusEnum
+from common.constants import StatusEnum
 from api.db.db_models import Conversation, DB
 from api.db.services.api_service import API4ConversationService
 from api.db.services.common_service import CommonService
 from api.db.services.dialog_service import DialogService, chat
-from api.utils import get_uuid
+from common.misc_utils import get_uuid
 import json
 
 from rag.prompts.generator import chunks_format

api/db/services/dialog_service.py

@@ -25,8 +25,7 @@ import trio
 from langfuse import Langfuse
 from peewee import fn
 from agentic_reasoning import DeepResearcher
-from api import settings
-from api.db import LLMType, ParserType, StatusEnum
+from common.constants import LLMType, ParserType, StatusEnum
 from api.db.db_models import DB, Dialog
 from api.db.services.common_service import CommonService
 from api.db.services.document_service import DocumentService
@@ -41,9 +40,10 @@ from rag.app.tag import label_question
 from rag.nlp.search import index_name
 from rag.prompts.generator import chunks_format, citation_prompt, cross_languages, full_question, kb_prompt, keyword_extraction, message_fit_in, \
     gen_meta_filter, PROMPT_JINJA_ENV, ASK_SUMMARY
-from rag.utils import num_tokens_from_string
+from common.token_utils import num_tokens_from_string
 from rag.utils.tavily_conn import Tavily
 from common.string_utils import remove_redundant_spaces
+from common import settings
 
 
 class DialogService(CommonService):
@@ -293,12 +293,13 @@ def meta_filter(metas: dict, filters: list[dict]):
     def filter_out(v2docs, operator, value):
         ids = []
         for input, docids in v2docs.items():
-            try:
-                input = float(input)
-                value = float(value)
-            except Exception:
-                input = str(input)
-                value = str(value)
+            if operator in ["=", "≠", ">", "<", "≥", "≤"]:
+                try:
+                    input = float(input)
+                    value = float(value)
+                except Exception:
+                    input = str(input)
+                    value = str(value)
 
             for conds in [
                 (operator == "contains", str(value).lower() in str(input).lower()),
@@ -618,7 +619,12 @@ def chat(dialog, messages, stream=True, **kwargs):
 
 
 def use_sql(question, field_map, tenant_id, chat_mdl, quota=True, kb_ids=None):
-    sys_prompt = "You are a Database Administrator. You need to check the fields of the following tables based on the user's list of questions and write the SQL corresponding to the last question."
+    sys_prompt = """
+You are a Database Administrator. You need to check the fields of the following tables based on the user's list of questions and write the SQL corresponding to the last question. 
+Ensure that:
+1. Field names should not start with a digit. If any field name starts with a digit, use double quotes around it.
+2. Write only the SQL, no explanations or additional text.
+"""
     user_prompt = """
 Table name: {};
 Table of database fields are as follows:
@@ -639,6 +645,7 @@ Please write the SQL, only SQL, without any other explanations or text.
         sql = re.sub(r".*select ", "select ", sql.lower())
         sql = re.sub(r" +", " ", sql)
         sql = re.sub(r"([;；]|```).*", "", sql)
+        sql = re.sub(r"&", "and", sql)
         if sql[: len("select ")] != "select ":
             return None, None
         if not re.search(r"((sum|avg|max|min)\(|group by )", sql.lower()):

api/db/services/document_service.py

@@ -26,22 +26,20 @@ import trio
 import xxhash
 from peewee import fn, Case, JOIN
 
-from api import settings
 from api.constants import IMG_BASE64_PREFIX, FILE_NAME_LEN_LIMIT
-from api.db import FileType, LLMType, ParserType, StatusEnum, TaskStatus, UserTenantRole, CanvasCategory
+from api.db import PIPELINE_SPECIAL_PROGRESS_FREEZE_TASK_TYPES, FileType, UserTenantRole, CanvasCategory
 from api.db.db_models import DB, Document, Knowledgebase, Task, Tenant, UserTenant, File2Document, File, UserCanvas, \
     User
 from api.db.db_utils import bulk_insert_into_db
 from api.db.services.common_service import CommonService
 from api.db.services.knowledgebase_service import KnowledgebaseService
-from api.utils import get_uuid
+from common.misc_utils import get_uuid
 from common.time_utils import current_timestamp, get_format_time
+from common.constants import LLMType, ParserType, StatusEnum, TaskStatus, SVR_CONSUMER_GROUP_NAME
 from rag.nlp import rag_tokenizer, search
-from rag.settings import get_svr_queue_name, SVR_CONSUMER_GROUP_NAME
 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 import settings
 
 class DocumentService(CommonService):
     model = Document
@@ -317,11 +315,11 @@ class DocumentService(CommonService):
                 all_chunk_ids.extend(chunk_ids)
                 page += 1
             for cid in all_chunk_ids:
-                if STORAGE_IMPL.obj_exist(doc.kb_id, cid):
-                    STORAGE_IMPL.rm(doc.kb_id, cid)
+                if settings.STORAGE_IMPL.obj_exist(doc.kb_id, cid):
+                    settings.STORAGE_IMPL.rm(doc.kb_id, cid)
             if doc.thumbnail and not doc.thumbnail.startswith(IMG_BASE64_PREFIX):
-                if STORAGE_IMPL.obj_exist(doc.kb_id, doc.thumbnail):
-                    STORAGE_IMPL.rm(doc.kb_id, doc.thumbnail)
+                if settings.STORAGE_IMPL.obj_exist(doc.kb_id, doc.thumbnail):
+                    settings.STORAGE_IMPL.rm(doc.kb_id, doc.thumbnail)
             settings.docStoreConn.delete({"doc_id": doc.id}, search.index_name(tenant_id), doc.kb_id)
 
             graph_source = settings.docStoreConn.getFields(
@@ -374,12 +372,16 @@ class DocumentService(CommonService):
     def get_unfinished_docs(cls):
         fields = [cls.model.id, cls.model.process_begin_at, cls.model.parser_config, cls.model.progress_msg,
                   cls.model.run, cls.model.parser_id]
+        unfinished_task_query = Task.select(Task.doc_id).where(
+            (Task.progress >= 0) & (Task.progress < 1)
+        )
+
         docs = cls.model.select(*fields) \
             .where(
             cls.model.status == StatusEnum.VALID.value,
             ~(cls.model.type == FileType.VIRTUAL.value),
-            cls.model.progress < 1,
-            cls.model.progress > 0)
+            (((cls.model.progress < 1) & (cls.model.progress > 0)) |
+             (cls.model.id.in_(unfinished_task_query)))) # including unfinished tasks like GraphRAG, RAPTOR and Mindmap
         return list(docs.dicts())
 
     @classmethod
@@ -621,12 +623,17 @@ class DocumentService(CommonService):
 
     @classmethod
     @DB.connection_context()
-    def begin2parse(cls, docid):
-        cls.update_by_id(
-            docid, {"progress": random.random() * 1 / 100.,
-                    "progress_msg": "Task is queued...",
-                    "process_begin_at": get_format_time()
-                    })
+    def begin2parse(cls, doc_id, keep_progress=False):
+        info = {
+            "progress_msg": "Task is queued...",
+            "process_begin_at": get_format_time(),
+        }
+        if not keep_progress:
+            info["progress"] = random.random() * 1 / 100.
+            info["run"] = TaskStatus.RUNNING.value
+            # keep the doc in DONE state when keep_progress=True for GraphRAG, RAPTOR and Mindmap tasks
+
+        cls.update_by_id(doc_id, info)
 
     @classmethod
     @DB.connection_context()
@@ -685,8 +692,13 @@ class DocumentService(CommonService):
                 bad = 0
                 e, doc = DocumentService.get_by_id(d["id"])
                 status = doc.run  # TaskStatus.RUNNING.value
+                doc_progress = doc.progress if doc and doc.progress else 0.0
+                special_task_running = False
                 priority = 0
                 for t in tsks:
+                    task_type = (t.task_type or "").lower()
+                    if task_type in PIPELINE_SPECIAL_PROGRESS_FREEZE_TASK_TYPES:
+                        special_task_running = True
                     if 0 <= t.progress < 1:
                         finished = False
                     if t.progress == -1:
@@ -703,13 +715,15 @@ class DocumentService(CommonService):
                     prg = 1
                     status = TaskStatus.DONE.value
 
+                # only for special task and parsed docs and unfinised
+                freeze_progress = special_task_running and doc_progress >= 1 and not finished
                 msg = "\n".join(sorted(msg))
                 info = {
                     "process_duration": datetime.timestamp(
                         datetime.now()) -
                                        d["process_begin_at"].timestamp(),
                     "run": status}
-                if prg != 0:
+                if prg != 0 and not freeze_progress:
                     info["progress"] = prg
                 if msg:
                     info["progress_msg"] = msg
@@ -756,6 +770,14 @@ class DocumentService(CommonService):
             .where((cls.model.kb_id == kb_id) & (cls.model.run == TaskStatus.CANCEL))
             .scalar()
         )
+        downloaded = (
+            cls.model.select(fn.COUNT(1))
+            .where(
+                cls.model.kb_id == kb_id,
+                cls.model.source_type != "local"
+            )
+            .scalar()
+        )
 
         row = (
             cls.model.select(
@@ -792,8 +814,32 @@ class DocumentService(CommonService):
             "finished": int(row["finished"]),
             "failed": int(row["failed"]),
             "cancelled": int(cancelled),
+            "downloaded": int(downloaded)
         }
 
+    @classmethod
+    def run(cls, tenant_id:str, doc:dict, kb_table_num_map:dict):
+        from api.db.services.task_service import queue_dataflow, queue_tasks
+        from api.db.services.file2document_service import File2DocumentService
+
+        doc["tenant_id"] = tenant_id
+        doc_parser = doc.get("parser_id", ParserType.NAIVE)
+        if doc_parser == ParserType.TABLE:
+            kb_id = doc.get("kb_id")
+            if not kb_id:
+                return
+            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)
+        if doc.get("pipeline_id", ""):
+            queue_dataflow(tenant_id, flow_id=doc["pipeline_id"], task_id=get_uuid(), doc_id=doc["id"])
+        else:
+            bucket, name = File2DocumentService.get_storage_address(doc_id=doc["id"])
+            queue_tasks(doc, bucket, name, 0)
+
+
 def queue_raptor_o_graphrag_tasks(sample_doc_id, ty, priority, fake_doc_id="", doc_ids=[]):
     """
     You can provide a fake_doc_id to bypass the restriction of tasks at the knowledgebase level.
@@ -815,7 +861,7 @@ def queue_raptor_o_graphrag_tasks(sample_doc_id, ty, priority, fake_doc_id="", d
             "to_page": 100000000,
             "task_type": ty,
             "progress_msg":  datetime.now().strftime("%H:%M:%S") + " created task " + ty,
-            "begin_at": datetime.now(),
+            "begin_at": datetime.now().strftime("%Y-%m-%d %H:%M:%S"),
         }
 
     task = new_task()
@@ -827,13 +873,13 @@ def queue_raptor_o_graphrag_tasks(sample_doc_id, ty, priority, fake_doc_id="", d
 
     task["doc_id"] = fake_doc_id
     task["doc_ids"] = doc_ids
-    DocumentService.begin2parse(sample_doc_id["id"])
-    assert REDIS_CONN.queue_product(get_svr_queue_name(priority), message=task), "Can't access Redis. Please check the Redis' status."
+    DocumentService.begin2parse(sample_doc_id["id"], keep_progress=True)
+    assert REDIS_CONN.queue_product(settings.get_svr_queue_name(priority), message=task), "Can't access Redis. Please check the Redis' status."
     return task["id"]
 
 
 def get_queue_length(priority):
-    group_info = REDIS_CONN.queue_info(get_svr_queue_name(priority), SVR_CONSUMER_GROUP_NAME)
+    group_info = REDIS_CONN.queue_info(settings.get_svr_queue_name(priority), SVR_CONSUMER_GROUP_NAME)
     if not group_info:
         return 0
     return int(group_info.get("lag", 0) or 0)
@@ -915,7 +961,7 @@ def doc_upload_and_parse(conversation_id, file_objs, user_id):
             else:
                 d["image"].save(output_buffer, format='JPEG')
 
-            STORAGE_IMPL.put(kb.id, d["id"], output_buffer.getvalue())
+            settings.STORAGE_IMPL.put(kb.id, d["id"], output_buffer.getvalue())
             d["img_id"] = "{}-{}".format(kb.id, d["id"])
             d.pop("image", None)
             docs.append(d)
@@ -962,7 +1008,7 @@ def doc_upload_and_parse(conversation_id, file_objs, user_id):
                     "content_with_weight": mind_map,
                     "knowledge_graph_kwd": "mind_map"
                 })
-            except Exception as e:
+            except Exception:
                 logging.exception("Mind map generation error")
 
         vects = embedding(doc_id, [c["content_with_weight"] for c in cks])
@@ -981,4 +1027,3 @@ def doc_upload_and_parse(conversation_id, file_objs, user_id):
             doc_id, kb.id, token_counts[doc_id], chunk_counts[doc_id], 0)
 
     return [d["id"] for d, _ in files]
-

api/db/services/file2document_service.py

@@ -15,7 +15,7 @@
 #
 from datetime import datetime
 
-from api.db import FileSource
+from common.constants import FileSource
 from api.db.db_models import DB
 from api.db.db_models import File, File2Document
 from api.db.services.common_service import CommonService

api/db/services/file_service.py

@@ -21,16 +21,19 @@ from pathlib import Path
 from flask_login import current_user
 from peewee import fn
 
-from api.db import KNOWLEDGEBASE_FOLDER_NAME, FileSource, FileType, ParserType
-from api.db.db_models import DB, Document, File, File2Document, Knowledgebase
+from api.db import KNOWLEDGEBASE_FOLDER_NAME, FileType
+from api.db.db_models import DB, Document, File, File2Document, Knowledgebase, Task
 from api.db.services import duplicate_name
 from api.db.services.common_service import CommonService
 from api.db.services.document_service import DocumentService
 from api.db.services.file2document_service import File2DocumentService
-from api.utils import get_uuid
+from common.misc_utils import get_uuid
+from common.constants import TaskStatus, FileSource, ParserType
+from api.db.services.knowledgebase_service import KnowledgebaseService
+from api.db.services.task_service import TaskService
 from api.utils.file_utils import filename_type, read_potential_broken_pdf, thumbnail_img
 from rag.llm.cv_model import GptV4
-from rag.utils.storage_factory import STORAGE_IMPL
+from common import settings
 
 
 class FileService(CommonService):
@@ -420,7 +423,7 @@ class FileService(CommonService):
 
     @classmethod
     @DB.connection_context()
-    def upload_document(self, kb, file_objs, user_id):
+    def upload_document(self, kb, file_objs, user_id, src="local"):
         root_folder = self.get_root_folder(user_id)
         pf_id = root_folder["id"]
         self.init_knowledgebase_docs(pf_id, user_id)
@@ -437,13 +440,13 @@ class FileService(CommonService):
                     raise RuntimeError("This type of file has not been supported yet!")
 
                 location = filename
-                while STORAGE_IMPL.obj_exist(kb.id, location):
+                while settings.STORAGE_IMPL.obj_exist(kb.id, location):
                     location += "_"
 
                 blob = file.read()
                 if filetype == FileType.PDF.value:
                     blob = read_potential_broken_pdf(blob)
-                STORAGE_IMPL.put(kb.id, location, blob)
+                settings.STORAGE_IMPL.put(kb.id, location, blob)
 
                 doc_id = get_uuid()
 
@@ -451,7 +454,7 @@ class FileService(CommonService):
                 thumbnail_location = ""
                 if img is not None:
                     thumbnail_location = f"thumbnail_{doc_id}.png"
-                    STORAGE_IMPL.put(kb.id, thumbnail_location, img)
+                    settings.STORAGE_IMPL.put(kb.id, thumbnail_location, img)
 
                 doc = {
                     "id": doc_id,
@@ -462,6 +465,7 @@ class FileService(CommonService):
                     "created_by": user_id,
                     "type": filetype,
                     "name": filename,
+                    "source_type": src,
                     "suffix": Path(filename).suffix.lstrip("."),
                     "location": location,
                     "size": len(blob),
@@ -530,9 +534,54 @@ class FileService(CommonService):
     @staticmethod
     def get_blob(user_id, location):
         bname = f"{user_id}-downloads"
-        return STORAGE_IMPL.get(bname, location)
+        return settings.STORAGE_IMPL.get(bname, location)
 
     @staticmethod
     def put_blob(user_id, location, blob):
         bname = f"{user_id}-downloads"
-        return STORAGE_IMPL.put(bname, location, blob)
+        return settings.STORAGE_IMPL.put(bname, location, blob)
+
+    @classmethod
+    @DB.connection_context()
+    def delete_docs(cls, doc_ids, tenant_id):
+        root_folder = FileService.get_root_folder(tenant_id)
+        pf_id = root_folder["id"]
+        FileService.init_knowledgebase_docs(pf_id, tenant_id)
+        errors = ""
+        kb_table_num_map = {}
+        for doc_id in doc_ids:
+            try:
+                e, doc = DocumentService.get_by_id(doc_id)
+                if not e:
+                    raise Exception("Document not found!")
+                tenant_id = DocumentService.get_tenant_id(doc_id)
+                if not tenant_id:
+                    raise Exception("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):
+                    raise Exception("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:
+                    settings.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)
+
+        return errors

api/db/services/knowledgebase_service.py

@@ -17,11 +17,16 @@ from datetime import datetime
 
 from peewee import fn, JOIN
 
-from api.db import StatusEnum, TenantPermission
+from api.db import TenantPermission
 from api.db.db_models import DB, Document, Knowledgebase, User, UserTenant, UserCanvas
 from api.db.services.common_service import CommonService
 from common.time_utils import current_timestamp, datetime_format
-
+from api.db.services import duplicate_name
+from api.db.services.user_service import TenantService
+from common.misc_utils import get_uuid
+from common.constants import StatusEnum
+from api.constants import DATASET_NAME_LIMIT
+from api.utils.api_utils import get_parser_config, get_data_error_result
 
 class KnowledgebaseService(CommonService):
     """Service class for managing knowledge base operations.
@@ -87,7 +92,7 @@ class KnowledgebaseService(CommonService):
         # Returns:
         #     If all documents are parsed successfully, returns (True, None)
         #     If any document is not fully parsed, returns (False, error_message)
-        from api.db import TaskStatus
+        from common.constants import TaskStatus
         from api.db.services.document_service import DocumentService
 
         # Get knowledge base information
@@ -196,6 +201,7 @@ class KnowledgebaseService(CommonService):
         # will get all permitted kb, be cautious.
         fields = [
             cls.model.name,
+            cls.model.avatar,
             cls.model.language,
             cls.model.permission,
             cls.model.doc_num,
@@ -281,7 +287,7 @@ class KnowledgebaseService(CommonService):
             (cls.model.status == StatusEnum.VALID.value)
         ).dicts()
         if not kbs:
-            return
+            return None
         return kbs[0]
 
     @classmethod
@@ -363,6 +369,64 @@ class KnowledgebaseService(CommonService):
         #     List of all knowledge base IDs
         return [m["id"] for m in cls.model.select(cls.model.id).dicts()]
 
+
+    @classmethod
+    @DB.connection_context()
+    def create_with_name(
+        cls,
+        *,
+        name: str,
+        tenant_id: str,
+        parser_id: str | None = None,
+        **kwargs
+    ):
+        """Create a dataset (knowledgebase) by name with kb_app defaults.
+
+        This encapsulates the creation logic used in kb_app.create so other callers
+        (including RESTFul endpoints) can reuse the same behavior.
+
+        Returns:
+            (ok: bool, model_or_msg): On success, returns (True, Knowledgebase model instance);
+                                      on failure, returns (False, error_message).
+        """
+        # Validate name
+        if not isinstance(name, str):
+            return get_data_error_result(message="Dataset name must be string.")
+        dataset_name = name.strip()
+        if dataset_name == "":
+            return get_data_error_result(message="Dataset name can't be empty.")
+        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}")
+
+        # Deduplicate name within tenant
+        dataset_name = duplicate_name(
+            cls.query,
+            name=dataset_name,
+            tenant_id=tenant_id,
+            status=StatusEnum.VALID.value,
+        )
+
+        # Verify tenant exists
+        ok, _t = TenantService.get_by_id(tenant_id)
+        if not ok:
+            return False, "Tenant not found."
+
+        # Build payload
+        kb_id = get_uuid()
+        payload = {
+            "id": kb_id,
+            "name": dataset_name,
+            "tenant_id": tenant_id,
+            "created_by": tenant_id,
+            "parser_id": (parser_id or "naive"),
+            **kwargs
+        }
+
+        # Default parser_config (align with kb_app.create) — do not accept external overrides
+        payload["parser_config"] = get_parser_config(parser_id, kwargs.get("parser_config"))
+        return payload
+
+
     @classmethod
     @DB.connection_context()
     def get_list(cls, joined_tenant_ids, user_id,

api/db/services/llm_service.py

@@ -16,7 +16,7 @@
 import inspect
 import logging
 import re
-from rag.utils import num_tokens_from_string
+from common.token_utils import num_tokens_from_string
 from functools import partial
 from typing import Generator
 from api.db.db_models import LLM
@@ -29,7 +29,7 @@ class LLMService(CommonService):
 
 
 def get_init_tenant_llm(user_id):
-    from api import settings
+    from common import settings
     tenant_llm = []
 
     seen = set()