check sample rate and lower if too high #2986
Conversation
eukreign
added
area: blobs/files
type: improvement
lbry/file_analysis.py
Outdated
| log.debug(" Detected audio codec is %s", codec) | ||
| if not {"aac", "mp3", "flac", "vorbis", "opus"}.intersection(codec.split(",")): | ||
| return "Audio codec is not in the approved list of AAC, FLAC, MP3, Vorbis, and Opus. " \ | ||
| f"Actual: {codec} [{stream['codec_long_name']}]" | ||
| if sample_rate > "48000": |
There was a problem hiding this comment.
What are the possible other sample_rate values? This comparison will produce an incorrect result if the length of the string is different even if one rate is higher than the other rate... since it's doing string comparisons, not integer comparisons.
There was a problem hiding this comment.
@Death916 can you address the above? Maybe we can make the check more robust in case it does not exist.
There was a problem hiding this comment.
@tzarebczan @eukreign
Sorry was busy for a bit. I changed the checks to int on my local now. i originally just figured it wouldn't matter since the values were already strings from the json.
What do you mean by it does not exist though?
I was mostly just trying to address the issue mentioned specifically for this PR, but if you could elaborate on what you mean it shouldn't be a problem to do. Kind of late here so I could be missing something obvious. Do you mean if it cant find a sample rate to just give it one that's acceptable?
There was a problem hiding this comment.
@Death916 doing string comparisons leads to unexpected behavior:
>>> "50" > "48000"
True
>>> "3000000000" > "48000"
False
>>> "a" > "48000"
True
what are the possible values that sample_rate can be? can it be an empty string? can it be something other than an integer (perhaps a float? what else?)?
once we have a better idea of possible range of values then at least for the integer/float case the two sides of the > equality need to be converted to integers or floats.
added volume_filter example deleted _get_volume_filter
New pull request with same commit as this one #2933
Just had to clean up some git stuff.
Issue #2906