kobotoolbox / kpi / 29012112022

81%

master: 76%

push

github

feat(bulkProcessing): add audio duration estimate to bulk transcription modal DEV-2255 (#7177)

### 📣 Summary
Adds an estimated total time for all selected submissions' audio
attachments in the bulk transcription modal.

### 💭 Notes
Implemented with parallel batched requests with exponential backoff if
there is a 504. Pseudo-code:
```
for each batch of ≤50 UIDs:
    attempt = 0
    while attempt < 3:
        POST /api/v2/assets/{asset_uid}/attachments/audio-duration/
        if 200 → store results, break
        if 504 → attempt++, wait, retry
        if other error → surface error, break
    if attempt == 3 → abort, surface error
aggregate total client-side
```

### 👀 Preview steps
**In order to test everything, you may want to have the following:**
- create ways to generate multiple batches. This can be done with:
    - `mockobo` to generate many submissions with audio files (> 500)
- or you can generate a few, and change the constant in
`BulkTranscriptionModal`
- return a 504
[here](https://github.com/kobotoolbox/kpi/blob/1e86db2d5/kpi/views/v2/attachment_audio_duration.py#L56-L57)
in order to test the retries + exponential backoff

1. ℹ️ have the above ready
2. set the number or rows displayed to 500
3. select all the rows and open the bulk transcription modal
5. 🟢 [on PR] notice that there is a `...` until the audio duration loads
& the duration is accurate when it does
6. generate the `504` in the backend code
7. Open the bulk transcription modal the say way again
8. 🟢 notice that in the network tab, there is maximum 3 tries then a
toast after the last try

---------

Co-authored-by: James Kiger <james.kiger@gmail.com>
Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com>
Co-authored-by: Leszek <smutnyleszek@gmail.com>

9293 of 12613 branches covered (73.68%)

30273 of 37124 relevant lines covered (81.55%)

5.52 hits per line