Skip to content

Latest commit

 

History

History
36 lines (25 loc) · 1.49 KB

CONTRIBUTING.md

File metadata and controls

36 lines (25 loc) · 1.49 KB

Contributing

Thanks for your interest in contributing!

Getting Started

  1. Fork the repo and clone your fork
  2. Install external tools: brew install ffmpeg wdiff ollama (macOS) or apt install ffmpeg wdiff + Ollama (Ubuntu/Debian)
  3. Install in editable mode with dev dependencies: pip install -e .[dev]
  4. Run tests: pytest -v

Making Changes

  • Create a feature branch from main
  • Add tests for new functionality
  • Run pytest -v and ensure all tests pass before submitting
  • Keep PRs focused — one feature or fix per PR

Code Organization

Source code lives in src/transcribe_critic/:

  • shared.py — Shared types (SpeechConfig, SpeechData) and utilities
  • transcriber.py — Pipeline orchestration, CLI, cost estimation, source survival analysis
  • download.py — Media downloading (yt-dlp)
  • transcription.py — Whisper transcription and multi-model ensembling
  • diarization.py — Speaker diarization (pyannote)
  • merge.py — Merge/alignment logic (wdiff, chunking, LLM adjudication)
  • slides.py — Slide extraction and vision analysis
  • output.py — Markdown generation

Tests are in tests/, organized by module: test_shared.py, test_transcriber.py, test_transcription.py, test_download.py, test_merge.py, test_output.py, test_slides.py, test_diarization.py.

Reporting Issues

Open an issue with steps to reproduce. For transcription quality issues, include the source URL and which merge path was used (structured vs. flat).