comment-check.yml

# name: Non-English Comments Check

# on:
#   pull_request:
#     branches:
#       - main
#   workflow_dispatch:

# jobs:
#   non-english-comments-check:
#     runs-on: ubuntu-latest

#     env:
#       # Directories to be excluded
#       EXCLUDE_DIRS: ".git docs tests scripts assets node_modules build"
#       # Files to be excluded (e.g., markdown, text files)
#       EXCLUDE_FILES: "*.md *.txt *.html *.css *.min.js *.mdx"

#     steps:
#       - uses: actions/checkout@v4

#       - name: Search for Non-English comments
#         run: |
#           set -e
#           # Define the regex pattern to match Chinese characters
#           pattern='[\p{Han}]'

#           # Process the directories to be excluded
#           exclude_dirs=""
#           for dir in $EXCLUDE_DIRS; do
#             exclude_dirs="$exclude_dirs --exclude-dir=$dir"
#           done

#           # Use grep to find all comments containing Non-English characters
#           grep -Pnr "$pattern" . $exclude_dirs | while read -r line ; do
#             file_name=$(echo "$line" | cut -d':' -f1)

#             # Check if the file matches any pattern in EXCLUDE_FILES
#             skip_file=false
#             for file_pattern in $EXCLUDE_FILES; do
#               # Use echo and grep to match the file name with the pattern
#               if echo "$file_name" | grep -q "$file_pattern"; then
#                 skip_file=true
#                 break
#               fi
#             done

#             # If the file matches an exclude pattern, skip it
#             if [ "$skip_file" = true ]; then
#               continue
#             fi

#             # Write the result to the non_english_comments.txt file
#             echo "$line" >> non_english_comments.txt
#           done || true

#       - name: Output non-English comments if found
#         id: check-comments
#         run: |
#           if [ -s non_english_comments.txt ]; then
#             echo "Non-English comments found in the following locations:"
#             cat non_english_comments.txt
#             comments=$(cat non_english_comments.txt | sed ':a;N;$!ba;s/\n/\\n/g')
#             echo "non_english_comments=$comments" >> $GITHUB_ENV
#             exit 1  # terminate the workflow
#           else
#             echo "No Non-English comments found."
#           fi

#       - name: Comment on PR if errors found
#         if: failure()  # This step runs only if the previous step fails
#         uses: peter-evans/create-or-update-comment@v4.0.0
#         with:
#           token: ${{ secrets.GITHUB_TOKEN }}  # GitHub token to post the comment
#           issue-number: ${{ github.event.pull_request.number }}  # PR number
#           edit-mode: replace  # This ensures that the comment is updated instead of creating a new one
#           body: |
#             ⚠️ Non-English comments were found in the following locations:
#             ```
#             $(while IFS= read -r line; do echo "$line"; done < non_english_comments.txt)
#             ```

name: Non-English Comments Check

on:
  pull_request:
    branches:
      - main
  workflow_dispatch:

jobs:
  non-english-comments-check:
    runs-on: ubuntu-latest

    env:
      # Directories to be excluded
      EXCLUDE_DIRS: ".git docs tests scripts assets node_modules build"
      # Files to be excluded
      EXCLUDE_FILES: "*.txt *.html *.css *.min.js *.mdx"

    steps:
      - uses: actions/checkout@v4
        with:
          fetch-depth: 0

      - name: Search for Non-English comments in PR files
        run: |
          set -e
          # Define the regex pattern to match Chinese characters
          pattern='[\p{Han}]'

          # Get the list of files changed in this PR compared to the base branch
          changed_files=$(git diff --name-only ${{ github.event.pull_request.base.sha }})

          # Process the directories to be excluded
          exclude_dirs=""
          for dir in $EXCLUDE_DIRS; do
            exclude_dirs="$exclude_dirs --exclude-dir=$dir"
          done

          # Process the file types to be excluded
          exclude_files=""
          for file in $EXCLUDE_FILES; do
            exclude_files="$exclude_files --exclude=$file"
          done

          # Check only the files modified in this PR for non-English comments
          for file in $changed_files; do
            # Skip files matching excluded patterns
            skip_file=false
            for file_pattern in $EXCLUDE_FILES; do
              if echo "$file" | grep -q "$file_pattern"; then
                skip_file=true
                break
              fi
            done

            # If the file matches an exclude pattern, skip it
            if [ "$skip_file" = true ]; then
              continue
            fi

            # Use grep to find all comments containing Non-English characters
            grep -Pn "$pattern" "$file" >> non_english_comments.txt || true
          done

      - name: Output non-English comments if found
        run: |
          if [ -s non_english_comments.txt ]; then
            echo "Non-English comments found in the following locations:"
            cat non_english_comments.txt
            exit 1  # terminate the workflow
          else
            echo "No Non-English comments found."
          fi