From: Gregory P. <gre...@gm...> - 2010-03-23 22:46:44
|
On Tue, Mar 23, 2010 at 8:26 PM, Gregory Petrosyan <gre...@gm...> wrote: > On Tue, Mar 23, 2010 at 3:53 PM, Johannes Weißl <ja...@mo...> wrote: >> On Tue, Mar 23, 2010 at 11:54:16AM +0300, Gregory Petrosyan wrote: >>> > 1. utf8 casefold (stolen from glib and converted to cmus style / functions) >>> > (already done) >>> >>> Looking at the source, it is full case folding, but not locale-sensitive >>> one (C + F as per http://www.unicode.org/Public/5.2.0/ucd/CaseFolding.txt), >>> e.g. ß -> SS? >> >> Yeah, but do you know any locale-sensitive one (besides libicu and >> towupper(), which we can't always use)? Isn't the full case mapping >> enough for just comparing strings? We can include the turkish table and >> do a locale-based check... > > I think using full case mapping is more than enough. BTW, filtering/searching for "bjork" should bring up "Björk". And this suggests that we need Unicode normalization, too. Gregory |