Skip to content

CHANGELOG.md

Latest commit

 

History

History
83 lines (67 loc) · 3.31 KB

CHANGELOG.md

File metadata and controls

83 lines (67 loc) · 3.31 KB

Changelog

All notable changes to this project will be documented in this file.

The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.

[v0.10.0]

Added

Audio Transcription Module

  • New retrocast transcription command group for audio-to-text transcription
  • Multiple backend support with automatic detection:
    • MLX Whisper: Optimized for Apple Silicon (M1/M2/M3) Macs
    • faster-whisper: CUDA GPU and CPU support for Linux/Windows
  • Whisper model support (tiny, base, small, medium, large)
  • Multiple output formats: TXT, JSON, SRT (subtitles), VTT (WebVTT)
  • Content-based deduplication using SHA256 hashing
  • Full-text search across transcribed content using SQLite FTS5
  • Rich CLI with progress bars and colored output

Transcription CLI Commands

  • retrocast transcription process PATH... - Transcribe audio files
  • retrocast transcription backends list - List available backends with status
  • retrocast transcription backends test BACKEND - Test specific backend availability
  • retrocast transcription search QUERY - Full-text search across transcriptions
  • retrocast transcription summary - Show overall transcription statistics
  • retrocast transcription podcasts list - List podcasts with transcriptions
  • retrocast transcription podcasts summary [PODCAST] - Detailed podcast statistics
  • retrocast transcription episodes list - Paginated episode listing with filters
  • retrocast transcription episodes summary - Aggregate episode statistics

Transcription Database Schema

  • New transcriptions table for transcription metadata
  • New transcription_segments table for timestamped text segments
  • FTS5 full-text search enabled on segment text
  • Content hash indexing for duplicate detection

Developer Documentation

  • Comprehensive user guide: docs/TRANSCRIPTION.md
  • Developer documentation: docs/TRANSCRIPTION_DEVELOPER.md
  • Architecture diagrams and backend implementation guide

Changed

  • Enhanced pyproject.toml with optional transcription dependencies:
    • transcription-mlx for Apple Silicon
    • transcription-cuda for NVIDIA GPU
    • transcription-cpu for CPU-only
    • transcription-diarization for speaker diarization (future)
  • Added poe tasks for transcription backend installation:
    • poe install:transcription-mlx
    • poe install:transcription-cuda
    • poe install:transcription-cpu

Technical Details

  • 126 tests passing with comprehensive coverage
  • Type-checked with ty type checker
  • Formatted with ruff and black
  • Class-based backend architecture with abstract base classes
  • Strategy pattern for backend selection
  • Platform-specific dependency management via PEP 508 markers

[0.10.0] - Initial Release

Added

  • Core Overcast data extraction pipeline
  • SQLite database storage with sqlite-utils
  • Authentication with Overcast API
  • OPML import/export
  • Feed and episode metadata extraction
  • Transcript download from podcast feeds
  • Episode download database with full-text search
  • HTML output generation
  • Integration with podcast-archiver for downloads
  • Chapter marker extraction
  • Datasette compatibility