search

Willy-JL / F95Checker

GNU General Public License v3.0
113 stars 17 forks source link

Fix some developer parsing strings #22

Closed just-Addict closed 1 year ago

just-Addict commented 2 years ago

The developer of Intimate Relations isn't properly detected.

The thread uses Developer/Publisher: instead of just Developer: which causes F95Checker to trip and showing the slash as developer name.

Willy-JL commented 2 years ago

I can't test right now because I'm not home but I've improved the developer and analogous string detection in 9.4 beta, can you try to see if it's fixed with the latest build from the GitHub actions tab?

just-Addict commented 2 years ago

I tested with 9.4 beta 514 and it now lists the developer as "Unknown" after doing a full recheck, and your guess in the thread reply is correct, the slash in between isn't enclosed in the bold tags. both the developer and publisher texts are enclosed in their own set of bold tags

<b>Developer</b>/<b>publisher:</b> PTOLEMY

Willy-JL commented 2 years ago

I might have a better plan for this. Only search for strings that start with "Developer" (and analogous) in order to catch also stuff that is not bolded properly, then add all following text except links except the first one (sometimes the developer name itself is a link) until a "br" line break, then remove all known "developer" and analogous strings from the beginning of the plain text (so in this games case only "Developer" would match but also "/publisher" would get removed as plain text), remove the known social text (patreon an the like) and finally strip all parentheses / separators ()/-| (they sometimes are around the developer social links). Will have to see how well this works but I'm hopeful

Willy-JL commented 1 year ago

Looks like im gonna backtrack on what i proposed before, thats just too much. What ill do is fix a few small oversights with the system i had made that had it working suboptimally and lead to the problem you were facing