Mor1arty stops the system from garbling non-English filenames
Upload a document with Chinese characters in its name and it used to come through as gibberish. Not anymore.
When you upload a file, the software has to read the filename off the request. The library Mor1arty's fork relies on for that job was misreading any characters outside the basic English alphabet - Chinese names being the usual victim, but the same flaw would hit accented or other non-Latin text. The result: a document landing in the system under a scrambled, unreadable name.
The fix is small and surgical. Mor1arty re-reads the filename correctly before anything else touches it, so the original name survives intact all the way through. It's a known quirk of the underlying upload library rather than anything unique to Mike, which means the same patch drops cleanly into any fork built the same way. It's already merged.
Spotted something wrong? Or know the PR text has fresher detail than the writeup above?