Keeps track of recorded audio files using a FileBackend.
What about the API? Do we pass audio when creating the metadata, or is this an API of it's own where one first create the Recording metadata, retrieve the identity and then store the audio using this identity?
| Subject | Author | Repo | Branch | Lines +/- | |
|---|---|---|---|---|---|
| Introduce RecordingStore | Karl Wettin (WMSE) | mediawiki/extensions/WikispeechSpeechDataCollector | master | +366 -0 |
| Status | Subtype | Assigned | Task | ||
|---|---|---|---|---|---|
| Resolved | • kalle | T266537 ☂ Initial SDC data API | |||
| Declined | None | T267108 Implement RecordingStore |
Change 642402 had a related patch set uploaded (by Karl Wettin (WMSE); owner: Karl Wettin (WMSE)):
[mediawiki/extensions/WikispeechSpeechDataCollector@master] Introduce RecordingStore
Note to self: set as declined/invalidated and document how we decided to go for MCR and point to those tasks.
Change 642402 abandoned by Karl Wettin (WMSE):
[mediawiki/extensions/WikispeechSpeechDataCollector@master] Introduce RecordingStore
Reason:
We have opted to use file namespace rather than an ad hoc file backend.
We are moving forward using file namespace (see T273038) rather than an ad hoc file backend.