tballison commented on PR #2953:
URL: https://github.com/apache/tika/pull/2953#issuecomment-4980711144
Small nits from my agent:
```
1. Audio.HAS_DRM javadoc overstates the implementation. It says "…or a
protection scheme ('sinf') box," but the
code only flags the drms/enca sample-entry formats — it never
independently detects a standalone sinf. In
practice enca entries always carry sinf, so coverage is fine, but the doc
implies a check that isn't there. Trim
the sinf mention or reword to "protected sample entry formats."
2. Multi-track asymmetry. BITRATE is last-track-wins (documented), but
HAS_DRM is any-track-wins (only ever set
to true). Defensible — arguably more correct for DRM — but the two
properties describe "the last track" and "any
track" respectively. Worth a one-line note in the PR/javadoc if you want
it explicit.
3. M4A with avgBitrate=0 gets no BITRATE even when maxBitrate>0 (some
encoders only set max). Acceptable given
the property is defined as average; just a coverage gap, not a bug.
4. Vorbis nominal-only ⇒ is-variable-bitrate=true is aggressive but
correct in practice (nominal-only is
quality-mode VBR). Fine as-i
```
Not sure you need to address any of them. The add vs set, number 2 is ...
not clear what the right answer is?
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
