Manual Spellcheck Review & Approval #104
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| name: Manual Spellcheck & Auto PR | |
| on: | |
| workflow_dispatch: # Runs only when manually triggered | |
| permissions: | |
| contents: write # Needed to push changes to a new branch | |
| pull-requests: write # Needed to create a PR | |
| jobs: | |
| spellcheck: | |
| runs-on: ubuntu-latest | |
| steps: | |
| - name: Checkout Repository | |
| uses: actions/checkout@v4 | |
| - name: Install Dependencies | |
| run: | | |
| pip install codespell | |
| pip install fuzzywuzzy[speedup] | |
| - name: Verify Spellcheck Ignore List Exists | |
| run: | | |
| if [ ! -f .github/spellcheck-ignore.txt ]; then | |
| echo "Error: spellcheck-ignore.txt not found!" && exit 1 | |
| fi | |
| - name: Run Spellcheck and Apply Fixes | |
| run: | | |
| set -e # Exit on error | |
| # Run codespell on the full repo, save raw report, and prevent auto-fixes | |
| codespell --ignore-words=.github/spellcheck-ignore.txt \ | |
| --skip=".git,*.lock,*.json,*.yaml,*.yml,*.css,*.html" \ | |
| --quiet-level=2 > spellcheck_report_raw.txt || true | |
| # Process corrections with Python | |
| python3 <<EOF | |
| import re | |
| from fuzzywuzzy import process | |
| # Load spellcheck ignore list with case sensitivity | |
| ignore_list = {} | |
| with open(".github/spellcheck-ignore.txt", "r", encoding="utf-8") as f: | |
| for line in f: | |
| word = line.strip() | |
| ignore_list[word.lower()] = word # Store lowercase -> correct-case | |
| # Common word pairs and phrases to check for context-based correction | |
| common_phrases = { | |
| "identity provider": ["identiy provider", "identify provider"], | |
| "access token": ["access toekn", "acess token"], | |
| "user authentication": ["user authentification", "user authenthication"], | |
| "API gateway": ["API getway", "API gatway"] | |
| } | |
| # Function to check if a word is inside a code block, backticks, URL, or file reference | |
| def is_code_or_url_or_file(line): | |
| return re.search(r'`.*?`|https?://\S+|www\.\S+|/[\w./-]+', line) | |
| # Function to check if a word is part of a Markdown link | |
| def is_markdown_link(line, original): | |
| return re.search(r'\[.*?\]\(.*' + re.escape(original) + r'.*\)', line) | |
| # Function to determine if an ignore list word should be used | |
| def should_use_ignore_list(original, suggestion, line): | |
| best_match, score = process.extractOne(original, ignore_list.keys()) | |
| # Must be at least 90% similar to be considered a match | |
| if score < 90: | |
| return False | |
| # Reject if original contains best_match as a substring (e.g., "certifcate" vs "CE") | |
| if best_match in original and len(original) > len(best_match): | |
| return False | |
| # Enforce case-sensitive corrections for regular text, but NOT for file references/URLs/links | |
| if not is_code_or_url_or_file(line) and not is_markdown_link(line, original): | |
| return True if best_match in ignore_list else False | |
| # Allow case-insensitive corrections for code blocks, backticks, URLs, and markdown links | |
| return best_match.lower() in ignore_list | |
| # Function to apply context-based correction | |
| def apply_context_based_correction(line, original, suggestion): | |
| for correct_phrase, wrong_variants in common_phrases.items(): | |
| for wrong_phrase in wrong_variants: | |
| if wrong_phrase in line: | |
| return line.replace(wrong_phrase, correct_phrase) | |
| return line.replace(original, suggestion) | |
| # Process spellcheck output and apply fixes | |
| with open("spellcheck_report_raw.txt", "r", encoding="utf-8") as infile, open("spellcheck_report.txt", "w", encoding="utf-8") as outfile: | |
| for line in infile: | |
| match = re.match(r"(.*):(\d+): (\S+) ==> (\S+)", line) | |
| if match: | |
| file_path, line_number, original, suggestion = match.groups() | |
| corrected_word = suggestion | |
| # Read the line content from the file | |
| with open(file_path, "r", encoding="utf-8") as file: | |
| content_lines = file.readlines() | |
| context_line = content_lines[int(line_number) - 1].strip() | |
| # **Fix #1: Preserve case-sensitive ignored terms exactly** | |
| if original in ignore_list.values(): | |
| corrected_word = original # Use exact case from original text | |
| # **Fix #2: Use English dictionary for weak matches** | |
| elif should_use_ignore_list(original, suggestion, context_line): | |
| best_match, _ = process.extractOne(original, ignore_list.keys()) | |
| if not is_code_or_url_or_file(context_line) and not is_markdown_link(context_line, original): | |
| corrected_word = ignore_list[best_match] | |
| else: | |
| corrected_word = best_match.lower() # Keep it lowercase in URLs/links/files | |
| # **Fix #3: Apply context-based correction** | |
| corrected_line = apply_context_based_correction(context_line, original, corrected_word) | |
| # **Fix #4: Replace only the first occurrence of the word** | |
| corrected_line = re.sub(r'\b' + re.escape(original) + r'\b', corrected_word, corrected_line, count=1) | |
| # Write final output | |
| outfile.write(f"{file_path}:{line_number}: {original} ==> {corrected_word}\n") | |
| # Apply fix while maintaining case rules | |
| content_lines[int(line_number) - 1] = corrected_line + "\n" | |
| with open(file_path, "w", encoding="utf-8") as file: | |
| file.writelines(content_lines) | |
| EOF | |
| # Check if any files were modified | |
| if git status --porcelain | grep -q '^ M'; then | |
| echo "Corrections applied. Preparing to create PR." | |
| else | |
| echo "No spelling corrections found. Exiting." | |
| exit 0 | |
| fi | |
| - name: Create Pull Request with Corrections | |
| env: | |
| GITHUB_TOKEN: ${{ secrets.PAT_GITHUB_ACTIONS }} | |
| run: | | |
| git config --global user.name "github-actions[bot]" | |
| git config --global user.email "github-actions[bot]@users.noreply.github.com" | |
| BRANCH_NAME="spellcheck-fixes-$(date +%s)" | |
| git checkout -b $BRANCH_NAME | |
| # Commit the changes if there are any | |
| if [ -n "$(git status --porcelain)" ]; then | |
| git add . | |
| git commit -m "Spellcheck: Automatically fixed detected misspellings" | |
| git push origin $BRANCH_NAME | |
| # Create PR using GitHub CLI | |
| gh pr create \ | |
| --base main \ | |
| --head $BRANCH_NAME \ | |
| --title "Spellcheck Fixes" \ | |
| --body "This PR contains automatically applied spelling corrections." | |
| else | |
| echo "No changes detected. Skipping PR creation." | |
| fi |