thump.php seems to use its own hook for parsing image param string (Introduced dcbe8e7dd3), instead of using the media handlers parseParamString method. This is confusing, and I'm not sure why it does this.
Version: 1.22.0
Severity: normal
Description
Description
Details
Details
- Reference
- bz49629
Event Timeline
Comment Actions
At the moment, wfExtractThumbParams() just takes a string and then does stuff to that string -- it's very lightweight. I guess that's why they didn't go for fancy MIME detection/object instantiation which might be memory intensive.
Comment Actions
Also, the Apache code for redirecting requests for non-existent thumbs to thumb.php will need to be adjusted; AFAIK at the moment it includes a manual list of permutations.
Comment Actions
...or maybe not, if that's what rewrite.py does. Not sure to be honest, bit out of my depth.
Comment Actions
Related URL: https://gerrit.wikimedia.org/r/69019 (Gerrit Change I3e703c84ff84e30d8b96acf491318829767eb2ab)
Comment Actions
Related URL: https://gerrit.wikimedia.org/r/69027 (Gerrit Change I43a0e7d6f0e45536e4e24b72140157791f5f027f)
Comment Actions
Change 69027 merged by jenkins-bot:
Make thumb_handler.php use MediaHandler::parseParamString