update-modelscope-docs.yml

name: Update ModelScope Documentation

on:
  # Scheduled trigger (daily at 18:00 UTC, which is 02:00 Beijing Time next day)
  schedule:
    - cron: '0 3 * * *'
  
  # Manual trigger (run via GitHub web interface)
  workflow_dispatch:
  
  # Trigger when model_list.txt or workflow files change
  push:
    paths:
      - 'docs/flagrelease_en/model_list.txt'
      - '.github/workflows/update-modelscope-docs.yml'
      - 'docs/scripts/download_readmes.py'

jobs:
  update-docs:
    runs-on: ubuntu-latest
    
    steps:
    - name: Checkout repository code
      uses: actions/checkout@v5
      with:
        fetch-depth: 0  # Fetch full history for proper git diff
    
    - name: Set up Python environment
      uses: actions/setup-python@v5
      with:
        python-version: '3.10'
        cache: 'pip'
        cache-dependency-path: 'requirements.txt'
    
    - name: Install Python dependencies
      run: |
        pip install -r requirements.txt
    - name: Generate model list
      run: |
        cd "$GITHUB_WORKSPACE"
        python docs/scripts/generate_model_list.py
      continue-on-error: true 
    
    - name: Download ModelScope documentation
      id: download
      run: |
        cd "$GITHUB_WORKSPACE"
        python docs/scripts/download_readmes.py
      continue-on-error: true  # Continue even if some downloads fail
    
    - name: Generate download report
      id: generate-report
      run: |
        # Create a detailed download report file with success and failure information
        # Store the report file in a tem directory to prevent it from pusing to the repo
        REPORT_FILE="/tmp/download_report_$(TZ='Asia/Shanghai' date +'%Y%m%d_%H%M%S').txt"
        echo "# ModelScope Documentation Update Report" > $REPORT_FILE
        echo "## Generated on: $(TZ='Asia/Shanghai' date '+%Y-%m-%d %H:%M:%S %Z')" >> $REPORT_FILE
        echo "## Workflow Run: [#${{ github.run_number }}](${{ github.server_url }}/${{ github.repository }}/actions/runs/${{ github.run_id }})" >> $REPORT_FILE
        echo "" >> $REPORT_FILE
        
        # Add all changes to the staging area to detect new files
        git add -N .
        
        # Count the number of updated files and get detailed information
        if [ -d "docs/flagrelease_en/model_readmes/" ]; then
          # Get the list of all changed files (both modified and added)
          CHANGED_FILES=$(git diff --name-status docs/flagrelease_en/model_readmes/ 2>/dev/null || echo "")
          if [ -z "$CHANGED_FILES" ]; then
            CHANGED_COUNT=0
          else
            CHANGED_COUNT=$(echo "$CHANGED_FILES" | wc -l)
          fi
          
          echo "## Summary" >> $REPORT_FILE
          echo "- **Total changes detected:** $CHANGED_COUNT files" >> $REPORT_FILE
          echo "" >> $REPORT_FILE
          
          if [ "$CHANGED_COUNT" -gt 0 ]; then
            echo "## Detailed Changes" >> $REPORT_FILE
            
            # Count added and modified files separately
            ADDED_FILES=$(echo "$CHANGED_FILES" | grep -c '^A' || echo "0")
            MODIFIED_FILES=$(echo "$CHANGED_FILES" | grep -c '^M' || echo "0")
            echo "- **Added files:** $ADDED_FILES" >> $REPORT_FILE
            echo "- **Modified files:** $MODIFIED_FILES" >> $REPORT_FILE
            echo "" >> $REPORT_FILE
            
            # List added files
            if [ "$ADDED_FILES" -gt 0 ]; then
              echo "### Added Files" >> $REPORT_FILE
              echo "$CHANGED_FILES" | grep '^A' | cut -f2 | while read file; do
                echo "- \`$file\`" >> $REPORT_FILE
              done
              echo "" >> $REPORT_FILE
            fi
            
            # List modified files
            if [ "$MODIFIED_FILES" -gt 0 ]; then
              echo "### Modified Files" >> $REPORT_FILE
              echo "$CHANGED_FILES" | grep '^M' | cut -f2 | while read file; do
                echo "- \`$file\`" >> $REPORT_FILE
              done
              echo "" >> $REPORT_FILE
            fi
          fi
        fi
        
        # Extract statistics from script output if available
        echo "## Model Download Statistics" >> $REPORT_FILE
        echo "- **Last run status:** ${{ job.status }}" >> $REPORT_FILE
        echo "- **Trigger type:** ${{ github.event_name }}" >> $REPORT_FILE
        if [ "${{ github.event_name }}" = "schedule" ]; then
          echo "- **Schedule:** Daily at 03:00 UTC (11:00 Beijing Time)" >> $REPORT_FILE
        fi
        echo "" >> $REPORT_FILE
        
        echo "## Next Steps" >> $REPORT_FILE
        echo "Please review the changes in the model documentation files above." >> $REPORT_FILE
        echo "After merging this PR, the documentation will be updated on the main branch." >> $REPORT_FILE
        
        echo "report_file=$REPORT_FILE" >> $GITHUB_OUTPUT
    
    - name: Check for file changes
      id: check-changes
      run: |
        # Check if any files have been modified or added
        git add -N .
        # Exclude the report file
        if git diff --quiet docs/flagrelease_en/model_readmes/ 2>/dev/null; then
          echo "No file changes detected"
          echo "has_changes=false" >> $GITHUB_OUTPUT
        else
          echo "File changes detected"
          echo "has_changes=true" >> $GITHUB_OUTPUT
        fi
    
    - name: Reset git configuration before creating PR
      if: steps.check-changes.outputs.has_changes == 'true'
      run: |
        # clear
        git config --local --unset-all http.https://github.com/.extraheader 2>/dev/null || true
        git config --global --unset-all http.https://github.com/.extraheader 2>/dev/null || true
      
    - name: Generate branch name and title
      id: branch-name
      if: steps.check-changes.outputs.has_changes == 'true'
      run: |
        BRANCH_NAME="update/modelscope-docs-$(TZ='Asia/Shanghai' date +'%Y%m%d-%H%M%S')"
        PR_TITLE="ModelScope Documentation Update - $(TZ='Asia/Shanghai' date +'%Y-%m-%d %H:%M')"
        echo "branch_name=$BRANCH_NAME" >> $GITHUB_OUTPUT
        echo "pr_title=$PR_TITLE" >> $GITHUB_OUTPUT
        echo "Generated branch name: $BRANCH_NAME"
        echo "Generated PR title: $PR_TITLE"
    
    - name: Create Pull Request
      if: steps.check-changes.outputs.has_changes == 'true'
      uses: peter-evans/create-pull-request@v5
      with:
        token: ${{ secrets.GITHUB_TOKEN }}
        commit-message: 'Auto-update ModelScope documentation [$(TZ=''Asia/Shanghai'' date +''%Y-%m-%d %H:%M'')]'
        branch: ${{ steps.branch-name.outputs.branch_name }}
        delete-branch: true
        title: ${{ steps.branch-name.outputs.pr_title }}
        body-path: ${{ steps.generate-report.outputs.report_file }}
        base: 'main'
        labels: 'documentation, auto-update, modelscope'
        draft: false
        # Remove the invalid 'paths' parameter and use 'add-paths' instead to specify which files to include
        add-paths: |
          docs/flagrelease_en/model_readmes/
          docs/flagrelease_en/model_list.txt