Page MenuHomePhabricator

Consider using IntlBreakIterator for the word counts shown on Special:Search results
Closed, DeclinedPublic

Description

This currently uses str_word_count() which is fast, but wrong on some languages (notable chinese, but probably others that don't use spaces). IntlBreakIterator uses icu to do the word breaking, and should give a better count. Still not sure if it's correct though.

Event Timeline

Deskana renamed this task from Consider using IntlBreakIteator for the word counts shown on Special:Search results to Consider using IntlBreakIterator for the word counts shown on Special:Search results.Oct 20 2016, 10:06 PM
Deskana added a subscriber: Deskana.

Not a priority right now, since it's a mostly cosmetic problem.

debt added a subscriber: debt.

Declining this, not sure how useful the word count really is to anyone, so spending time on this wouldn't be a good way to spend our time on a cosmetic 'fix'