Thanks for the advice Jens. My problem is the software I use (Embarcadero c++ builder) is more utf16 orientated. It does support utf8 but many of the built in functions return System.UnicodeString which is utf16.
From: Jens Alfke<mailto:j...@mooseyard.com> Sent: 02 February 2017 17:32 To: SQLite mailing list<mailto:sqlite-users@mailinglists.sqlite.org> Subject: Re: [sqlite] Help with custom collation Perhaps off-topic, but: UTF-16 is generally not recommended, unless you need to work with legacy APIs that require it. It has the same difficulties as UTF-8 (having to handle characters broken into multi-element sequences) but uses more RAM and isn’t downward compatible with ASCII. Also, since you’re less likely to run into multi-word character sequences in normal use [if you’re non-Asian], it’s easier to miss bugs in your multi-word character handling. (Tip: put some emoji in your test strings!) (A UTF-16 collator is also probably going to be slower than a UTF-8 one, since SQLite will need to copy/expand the UTF-8 strings from the database when it calls it. And yes, collators can easily become performance bottlenecks; I’ve seen it.) —Jens _______________________________________________ sqlite-users mailing list sqlite-users@mailinglists.sqlite.org http://mailinglists.sqlite.org/cgi-bin/mailman/listinfo/sqlite-users _______________________________________________ sqlite-users mailing list sqlite-users@mailinglists.sqlite.org http://mailinglists.sqlite.org/cgi-bin/mailman/listinfo/sqlite-users
