GitHub user zyratlo added a comment to the discussion: Proposal: Audio 
Attribute Extractor, measuring audio files in a workflow

Thanks for writing this up. I don't personally work with audio files, but I can 
see the use cases that would need an operator like this.

The Python dependencies should be fine. I checked the licenses and they all 
look good (miniaudio is MIT, cffi is MIT-0, pycparser is BSD-3), which are all 
on the ASF Category A list. Same with the librosa-derived code, since ISC is 
also Category A. The PR reviewer should verify this though.

On PR size, we generally want PRs to be as small and modular as possible so 
they're easier to review. Yours is about 3.5k lines, so even if most of that is 
tests and other non-production code, it's still quite large. It would be good 
to split it up so that each PR still makes sense on its own. For example, the 
first PR could add the operator with decoding and the simpler attributes 
(waveform, rms, zcr), and then pitch, onsets and beats could follow. Beats is 
the only part that uses the librosa port, so that would also keep the licensing 
changes out of the first PR.

GitHub link: 
https://github.com/apache/texera/discussions/9276#discussioncomment-19403245

----
This is an automatically sent email for [email protected].
To unsubscribe, please send an email to: [email protected]

Reply via email to