Peter Eisentraut wrote:
> On Monday 11 August 2008 16:23:29 Jan Urbański wrote:
>> Often clients want their searches to be
>> accented-or-language-specific letters insensitive. So searching for
>> 'łódź' returns 'lodz'. So the use case is there (in fact, the lack of
>> such facility made me consider not upgrading particular client to 8.3...).
>
> These are valid ideas, but then please design a new function that addresses
> your use case in a well-defined way, and don't overload questionable old
> interfaces for new purposes.
>
> In the Unicode standard you can find well-defined methods to decompose
> characters into diacritic marks, and then you could strip them off. But this
> has nothing to do with ASCII or UTF8 or encodings. Cyrillic characters can
> have diacritic marks as well, for example.
OK, I was envisioning something like that:
http://search.cpan.org/~sburke/Text-Unidecode-0.04/lib/Text/Unidecode.pm
but now that I think of it, I can always just write a plperlu function
that uses that module. The only inconvenience is having to have plperlu
in the db, but I can live with that.
Postgres extensibility rocks and I rest my case.
Cheers,
Jan
--
Jan Urbanski
GPG key ID: E583D7D2
ouden estin