Re: [dict-beta] Index.c and dictfmt.c don't agree on locale specific sorting
Diego Nehab <[email protected]> Thu, 21 Oct 2004 11:33:59 -0700 (PDT)
| Newsgroups | gmane.comp.web.wml |
|---|---|
| Message-ID | <[email protected]> |
Hi, The problem is that sort uses strcoll directly on the words, whereas index.c uses strcoll to order individual characters and than implements the lexicographic comparison using the ordering of the characters. Index.c is wrong because it will sort accented characters as if they were different from characters with no accent. All words starting with 'á' will come after words starting with 'a'. That is, "ábaco" will come after "azeite", which is wrong under pt_BR (or en_US for that matter). Naturally, the "sort" utility doesn't have this problem. The best thing would be to fix index.c, but as I pointed out, this is more complicated than fixing dictfmt to avoid calling "sort", matching the ordering expected by index.c instead. Let me know if you need more info. []s, Diego. _______________________________ Do you Yahoo!? Declare Yourself - Register online to vote today! http://vote.yahoo.com --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected] _____________________________________________________________________ Website META Language (WML) http://thewml.org User Support Mailing List [email protected] Automated List Manager [email protected]