Seth David Schoen writes: > That is, using the explicit list of characters in the backet > expression was three orders of magnitude faster! Using the > character class [:upper:] is also fast: > > schoen@sescenties:~$ time grep '[:upper:]' /usr/share/dict/words > /dev/null
I was wrong about that -- my grep doesn't have the patch from https://lists.gnu.org/archive/html/bug-grep/2010-09/msg00015.html so it didn't warn me that I wasn't actually using the character class I thought I was. In fact, using the character class [[:upper:]] is also _slow_, about as slow as using the range [A-Z]. -- Seth David Schoen <[email protected]> | Qué empresa fácil no pensar en http://www.loyalty.org/~schoen/ | un tigre, reflexioné. http://vitanuova.loyalty.org/ | -- Borges, El Zahir
