Re: tsearch2 punctuation question
От | Greg Sabino Mullane |
---|---|
Тема | Re: tsearch2 punctuation question |
Дата | |
Msg-id | 82ba8d8f860d58412e98d8643ac934a8@biglumber.com обсуждение исходный текст |
Ответ на | tsearch2 punctuation question (John DeSoi <desoi@pgedit.com>) |
Список | pgsql-general |
-----BEGIN PGP SIGNED MESSAGE----- Hash: RIPEMD160 > Is there a setting that allows me to specify that strings containing > the '/' should be parsed into separate words? As is, I can't find > 'apple' or 'orange'. No setting, I think you would have to mess with tsearch2 dictionaries. A far easier approach is to have your application simply split the words apart, or even write a wrapper function to do it for you within Postgres, e.g. CREATE OR REPLACE FUNCTION wordsplit(text) RETURNS text LANGUAGE plperl AS $_$ my $string = shift; $string =~ s/\W/ /g; return $string; $_$; SELECT to_tsvector(wordsplit('cat,dog apple/orange')); - -- Greg Sabino Mullane greg@turnstep.com PGP Key: 0x14964AC8 200704261140 http://biglumber.com/x/web?pk=2529DF6AB8F79407E94445B4BC9B906714964AC8 -----BEGIN PGP SIGNATURE----- iD8DBQFGMMikvJuQZxSWSsgRAwq4AKDJv4D6UDKZngU2vZt+cPgr6gGsnwCgmJET arG3n5+2pXxR+wedZ2LjZYU= =BPs4 -----END PGP SIGNATURE-----
В списке pgsql-general по дате отправления: