At 08:04 25/09/2011, haman...@t-online.de wrote:
>> select * from items where regexp_matches(items.code,'(ABC) (DE1)
>> (any_substring)')<>{};
>>
Hi Eduardo,
it is clear that scanning the table once with a list of matches will
outperform
rescanning the table for every string wanted. Now, my problem is
that the patterns are
dynamic as well. So if I could translate a table with one
column and a few thousand rows
into something like
regexp_matches(code,'string1|string2|.....string2781')
would ideally be a performant query. Unfortunately I have no idea
how I could achieve this
transformation inside the database. Doing it externally fails,
because any single query cannot
be more than so few characters.
You can create a plsql function and pass a setof text that do it.
Sorry but instead saying you What Must You Type, WMYT(c), i prefer
the How Should You Do way, HSYD(c). Note that you can get the same
results using other approachs (f.ex. using FTS described in chapter 12)
Check this topics:
Function
Creation http://www.postgresql.org/docs/9.0/static/sql-createfunction.html
Tutorial about Function
Creation http://www.adderpit.com/practical-postgresql/x10374.htm
HTH
--
Sent via pgsql-general mailing list (pgsql-general@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-general
