over 1 year
TBS group has been releasing a popular series with the quality "CONVERT WEB" and "CONVERT 720p WEB". The quality is very poor and they do include a warning about this in their NFO. According to the scene rules  :
7.5) Converted video that has significant abnormalities (e.g. blended
frames jerky playback due to missing or duplicate frames) due to the
conversion and cannot be reversed to the native format must be tagged as
I would like the quality plugin to recognize the CONVERT tag to make it easier to avoid these low-quality rips, e.g. specify !convert in the quality settings. The only other option is to use the regexp plugin, but is not the best solution because "convert" could then match in the series name or group name, etc.
about 1 year
How do you think quality is parsed? Answer: Regex. The problems you mention with the regexp plugin will also be present in the quality parser. Therefore I don't think this will be implemented anytime soon.
about 1 year
I'm not familiar with the code, but doesn't the check for propers do something similar so that a keyword in the name or group name prevent triggering it? Couldn't you easily check in data_parts from utils/titles/series.py? Or if you want to make it a more generic check, when a quality keyword is found in the matches function in utils/qualities.py, it has access to the position of the keyword, so the check for CONVERT can be made after the position of the first found quality keyword.
That is not completely true. When the quality plugin parses the title, it has access to the position of each found quality keyword,. It can use the position of the first keyword to start looking for the CONVERT keyword. Doesn't the check for proper keywords do something similar so that a keyword in the title or group name prevent triggering it?
sorry, second paragraph was meant to be erased and there's no way to edit after posting.
If it's just one group using a non-standard quality tag, perhaps it would be easier to simply add a regex filter in your config that rejects anything matching those patterns?
Join the discussion!