FWIW (non-technical readers)... The issue, I think, is that the \w escape sequence doesn't include characters that are outside what Unicode considered "word ch

FWIW (non-technical readers)... The issue, I think, is that the \w escape sequence doesn't include characters that are outside what Unicode considered "word characters" at the time it was originally implemented. So, non-western characters that are higher up in number don't get picked up by \w. The strategy here was to change the regex to mean, "accept any character that isn't something that's illegal for a file name".