Currently in included_rx() we search for keyword without end boundary, so keyword TS (from filter_telesync) will match movie called Tsunami and will block it. e.g. https://www.themoviedb.org/search/movie?query=tsunami
Same for other "small" keywords: cam, tc, scr.
So I updated included_rx to search whole keyword to avoid false matches.
Since we add spaces around title (value = ' ' + value.lower() + ' ') we can simply use \W+ - it will match end of word and end of line. (initially i wanted to use (\W+|$) or (?:\W+|$))
Currently in
included_rx()we search for keyword without end boundary, so keywordTS(from filter_telesync) will match movie called Tsunami and will block it. e.g. https://www.themoviedb.org/search/movie?query=tsunami Same for other "small" keywords: cam, tc, scr.So I updated included_rx to search whole keyword to avoid false matches. Since we add spaces around title (
value = ' ' + value.lower() + ' ') we can simply use\W+- it will match end of word and end of line. (initially i wanted to use(\W+|$)or(?:\W+|$))In the past it also was matched with end boundary - https://github.com/elgatito/script.elementum.burst/commit/67e7e0d86d7b4cf5510ed661d92c59d1669d0adf#diff-78cc99018f5874a1a9641c20d0e5b1d595c3ffa33528e7c7c49fad8e111a0268L68 e.g. it was
_ts_where_was replaced by space and then searched in title.I tested it with several torrents - i do not see any regression, but i see previously incorrectly blocked results.
Online regexp tests: https://regex101.com/r/nOI6P7/1 before:
after:
so now it does not match words like "camera" and "tsumani".
there is still small issue with filters line
trailerandline, since those are quite common words, but i guess we can't fix this easily.