>Third, my oriiginal inquiry: > >SELECT * > FROM fm > WHERE name LIKE '%Juiian%' > OR info LIKE '%Julian%' > ORDER by name; > >returned the (correct) 6 rows.
BTW, if you or anyone else need a fuzzy compare function I have one that can help. It works with internally Unicode-unaccented versions of string and pattern. It also accept wildcards (but no heading '%'). It returns a Damerau-Levenshtein distance, so it's O(n*m) and therefore should only be used on _short_ strings (like names, cities). This is part of an extension library, along with Unicode-unaccented versions of LIKE and GLOB and a handful of similar Unicode strings functions. It doesn't use ICU but internal Unicode v5.1 tries. It's still in a beta stage and I'd like to get it tested by people using various languages to know how it behaves and also to surface bugs and issues. It was written for Windows and the collations function use one Windows call. For those that don't like that, it should compile easily without (untested). Drop me a note if you're willing to try it. _______________________________________________ sqlite-users mailing list sqlite-users@sqlite.org http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users
