✨ feat: Added Docker push workflow, Dockerfile updates, and build scr…

…ipt (#807)

* ✨ feat: Added Docker push workflow, Dockerfile updates, and build script

add Docker Push workflow and stages for building and pushing Docker images with different variants for autoraghq/autorag, including handling production image tagging and updating Docker Hub description.

* ♻️ refactor: remove unnecessary tags and pull request configuration from Docker Push workflow

---------

Co-authored-by: Jeffrey (Dongkyu) Kim <[email protected]>

.github/workflows/docker-push.yml

@@ -0,0 +1,91 @@
+name: Docker Push
+
+on:
+  push:
+    branches: [ "main" ]
+
+env:
+  DOCKER_REPO: "autoraghq/autorag"
+
+jobs:
+  check-version:
+    runs-on: ubuntu-latest
+    outputs:
+      changed: ${{ steps.version_changed.outputs.changed }}
+    steps:
+    - name: Checkout code
+      uses: actions/[email protected]
+
+    - name: Get changed files
+      id: changed-files
+      uses: tj-actions/[email protected]
+
+    - name: Check for VERSION file change
+      id: version_changed
+      env:
+        ALL_CHANGED_FILES: ${{ steps.changed-files.outputs.all_changed_files }}
+      run: |
+        echo "changed=false" >> $GITHUB_OUTPUT
+        if echo "${ALL_CHANGED_FILES}" | grep -q 'VERSION'; then
+          echo "changed=true" >> $GITHUB_OUTPUT
+        fi
+
+  build-and-push:
+    needs: check-version
+    if: needs.check-version.outputs.changed == 'true'
+    runs-on: ubuntu-latest
+    strategy:
+      matrix:
+        variant: [all] # [all, ko, dev, parsing]
+    steps:
+    - name: Checkout code
+      uses: actions/[email protected]
+
+    - name: Read VERSION file
+      run: echo "VERSION=$(cat autorag/VERSION)" >> $GITHUB_ENV
+
+    - name: Set up Docker Buildx
+      uses: docker/[email protected]
+
+    - name: Login to Docker Hub
+      uses: docker/[email protected]
+      with:
+        username: ${{ secrets.DOCKER_HUB_USERNAME }}
+        password: ${{ secrets.DOCKER_HUB_ACCESS_TOKEN }}
+
+    - name: Extract metadata (tags, labels) for Docker
+      id: meta
+      uses: docker/[email protected]
+      with:
+        images: ${{ env.DOCKER_REPO }}
+        tags: |
+          type=raw,value=${{ env.VERSION }}-${{ matrix.variant }}
+          type=raw,value=${{ matrix.variant }},enable=${{ github.ref == format('refs/heads/{0}', 'main') }}
+          type=raw,value=latest-${{ matrix.variant }},enable=${{ github.ref == format('refs/heads/{0}', 'main') }}
+
+    - name: Build and push Docker image
+      uses: docker/[email protected]
+      with:
+        context: .
+        file: ./Dockerfile
+        push: true
+        # push: ${{ github.event_name != 'pull_request' && matrix.variant != 'test' }}
+        tags: ${{ env.DOCKER_REPO }}:${{ env.VERSION }}-${{matrix.variant}}
+        labels: ${{ steps.meta.outputs.labels }}
+        build-args: |
+          TARGET_STAGE=${{ matrix.variant }}
+
+    - name: Tag and push 'all' for production
+      run: |
+        docker pull ${{ env.DOCKER_REPO }}:${{ env.VERSION }}-${{matrix.variant}}
+        docker tag ${{ env.DOCKER_REPO }}:${{ env.VERSION }}-${{matrix.variant}} ${{ env.DOCKER_REPO }}:latest
+        docker push ${{ env.DOCKER_REPO }}:${{ env.VERSION }}-${{matrix.variant}}
+        docker push ${{ env.DOCKER_REPO }}:latest
+
+    # - name: Update Docker Hub description
+    #   uses: peter-evans/dockerhub-description@v3
+    #   with:
+    #     username: ${{ secrets.DOCKER_HUB_USERNAME }}
+    #     password: ${{ secrets.DOCKER_HUB_PASSWORD }}
+    #     repository: ${{ env.DOCKER_REPO }}
+    #     short-description: ${{ github.event.repository.description }}

Dockerfile

@@ -1,4 +1,4 @@
-# Base stage: Install dependencies
+# Base stage: Install common dependencies
 FROM python:3.10-slim AS base
 
 # Install system dependencies
@@ -19,27 +19,43 @@ WORKDIR /usr/src/app
 # Install Python dependencies
 RUN pip install --upgrade pip setuptools setuptools-scm
 COPY requirements.txt /usr/src/app/requirements.txt
-
 RUN pip install -r requirements.txt
 
 # Copy project files
 COPY . /usr/src/app
 RUN pip install -e ./
 
-# Test stage: Run tests if CI=true
+# Test stage
 FROM base AS test
-
-# Install testing dependencies
 RUN pip install pytest pytest-xdist
 
-# Run tests if CI is set to true
-RUN pytest -o log_cli=true --log-cli-level=INFO -n auto tests
+# Ko stage
+FROM base AS ko
+RUN pip install --no-cache-dir kiwipiepy>=0.18.0 konlpy
+ENTRYPOINT ["python", "-m", "autorag.cli"]
+
+# Dev stage
+FROM base AS dev
+RUN pip install --no-cache-dir ruff pre-commit
+ENTRYPOINT ["python", "-m", "autorag.cli"]
+
+# Parsing stage
+FROM base AS parsing
+RUN pip install --no-cache-dir PyMuPDF pdfminer.six pdfplumber unstructured jq "unstructured[pdf]" "PyPDF2<3.0" pdf2image
+ENTRYPOINT ["python", "-m", "autorag.cli"]
+
+# Parsing stage
+FROM base AS all
+# TODO
+ENTRYPOINT ["python", "-m", "autorag.cli"]
 
-# Production stage: Create final image for production
+# Production stage (includes all features)
 FROM base AS production
+RUN pip install --no-cache-dir \
+    kiwipiepy>=0.18.0 konlpy \
+    ruff pre-commit \
+    PyMuPDF pdfminer.six pdfplumber unstructured jq "unstructured[pdf]" "PyPDF2<3.0" pdf2image
 
 COPY projects /usr/src/app/projects
 
-# Set the entrypoint for the production application
 ENTRYPOINT ["python", "-m", "autorag.cli"]
-# ENTRYPOINT ["bash"]

autorag/VERSION

@@ -1 +1 @@
-0.3.3
+0.3.3-rc16

build_and_push.sh

@@ -0,0 +1,39 @@
+#!/bin/bash
+
+# Check if version is provided
+if [ "$#" -ne 1 ]; then
+    echo "Usage: $0 <version>"
+    exit 1
+fi
+
+VERSION=$1
+DOCKER_REPO="autoraghq/autorag"
+
+# Array of variants
+variants=("all" "ko" "dev" "parsing" "test")
+
+# Build and push for each variant
+for variant in "${variants[@]}"
+do
+    echo "Building $DOCKER_REPO:$VERSION-$variant"
+    docker build --build-arg TARGET_STAGE=$variant -t $DOCKER_REPO:$VERSION-$variant -f Dockerfile .
+
+    echo "Pushing $DOCKER_REPO:$VERSION-$variant"
+    docker push $DOCKER_REPO:$VERSION-$variant
+
+    # If it's a release build, also tag and push as latest for that variant
+    if [[ $VERSION =~ ^[0-9]+\.[0-9]+\.[0-9]+$ ]]; then
+        echo "Tagging and pushing $DOCKER_REPO:$variant"
+        docker tag $DOCKER_REPO:$VERSION-$variant $DOCKER_REPO:$variant
+        docker push $DOCKER_REPO:$variant
+    fi
+done
+
+# Special handling for 'production' as 'all'
+if [[ $VERSION =~ ^[0-9]+\.[0-9]+\.[0-9]+$ ]]; then
+    echo "Tagging and pushing $DOCKER_REPO:all"
+    docker tag $DOCKER_REPO:$VERSION-production $DOCKER_REPO:all
+    docker push $DOCKER_REPO:all
+fi
+
+echo "Build and push complete for all variants"