Hacker News new | ask | show | jobs
by jan6 2074 days ago
I especially hate the way gnu utils treat character groups like [:alnum:] which, if you have utf8 version of a locale, matches characters in ANY unicode language set... so if you, say, grep for numbers in /dev/urandom with a shell script, you can end up with, say, some characters that are technically numbers but in a totally different language, say, arabic, or hebrew...