AGENTS.md

AGENTS.md

Build/Test Commands

# Run the script
python compare_md_sim.py <file1.md> <file2.md>

# Install dependencies
pip install -r requirements.txt

Currently no automated tests exist. Future tests should use pytest.

Code Style Guidelines

• Imports: Organize as (stdlib, third-party, local). Use explicit imports, avoid import *.
• Formatting: Follow PEP 8. Use 4-space indentation.
• Naming:
  • Functions/variables: snake_case
  • Classes: PascalCase
  • Constants: UPPER_SNAKE_CASE
• Type hints: Not currently used; add when refactoring.
• Docstrings: Use triple-quoted format with description of parameters and return values (as shown in existing code).
• Error handling: Use try/except for external operations (file I/O, model loading); propagate context in error messages.
• Comments: Use for complex logic and non-obvious intent. Comments are in Chinese; maintain consistency with existing codebase.

Project Structure

• compare_md.py: Main script that compares semantic similarity of two Markdown files
• requirements.txt: Python dependencies (markdown-it-py, beautifulsoup4, sentence-transformers, torch)
• .gitignore: Standard Python ignore patterns