Reply to thread

Message

<blockquote data-quote="rjburke377" data-source="post: 287185" data-attributes="member: 70884">I don't do the &quot;each movie in it's own folder&quot; thing. That was someone else in this thread.Your explanation is very good and expresses the need to have the matching algorithm use the extra info in the filename to match against the source being scraped.The current matching algorithm includes IMDB formatting characters like &quot;(year&lt;/sequence&gt;), (V), (TV), (VG), (TV series)&quot; when comparing to your filename. The filename is compared verbatim ... including the extension and stacking tags. Where the movie titiles are short compared to the extra verbiage this results in a lot of false ambiguity, or picking up the wrong entry.The situation gets more difficult if you have stacked files because your filename is ignored for all but the first stacked file scanned from the directory.My point was that you can improve the situation, a lot, by doing as you say: rename your files to better fit the implementation. Even so, the code does not provide a means to be 100% accurate even when you know exactly which IMDB entry you are trying to match.</blockquote>

[QUOTE="rjburke377, post: 287185, member: 70884"] I don't do the "each movie in it's own folder" thing. That was someone else in this thread. Your explanation is very good and expresses the need to have the matching algorithm use the extra info in the filename to match against the source being scraped. The current matching algorithm includes IMDB formatting characters like "(year</sequence>), (V), (TV), (VG), (TV series)" when comparing to your filename. The filename is compared verbatim ... including the extension and stacking tags. Where the movie titiles are short compared to the extra verbiage this results in a lot of false ambiguity, or picking up the wrong entry. The situation gets more difficult if you have stacked files because your filename is ignored for all but the first stacked file scanned from the directory. My point was that you can improve the situation, a lot, by doing as you say: rename your files to better fit the implementation. Even so, the code does not provide a means to be 100% accurate even when you know exactly which IMDB entry you are trying to match. [/QUOTE]

Verification