How cranish07/wavewright works
Wavewright currently implements an audio-dataset checker, not the broader cleaning and review toolkit described as planned. Its user workflow is to provide a directory or ZIP, inspect supported audio files, optionally match metadata, and generate quality reports. The sampled source supports separate hashing, FFprobe metadata inspection, FFmpeg signal analysis, and per-file checks; the excerpts do not show enough of the main routine to map its full dispatch or report-writing calls. Planned noise reduction, transcription review, and export are not implemented.
Main entry points: checker.py Per-file checks
Components 9
Audio inspection
- SHA-256 hashing
checker.py - Audio probing
checker.py - Signal analysis
checker.py
Dataset checks
- Metadata loading
checker.py - Per-file checks
checker.py
Runtime services
- Checker workflow
checker.py - Quality reports
checker.py
Other components
- Dataset user
- FFmpeg tools
Connections 8
- Dataset user → Checker workflow: provides dataset
README.md - Checker workflow → Quality reports: generates reports
checker.py - Per-file checks → SHA-256 hashing: hashes files
checker.py - Per-file checks → Audio probing: probes audio
checker.py - Per-file checks → Signal analysis: analyzes samples
checker.py - Per-file checks → Metadata loading: matches metadata
checker.py - Audio probing → FFmpeg tools: runs ffprobe
checker.py - Signal analysis → FFmpeg tools: decodes with
checker.py