In article <20051111141232.GH13177@svana.org>,
Martijn van Oosterhout <kleptog@svana.org> writes:
> It's even sillier than that:
> test=# SELECT substring ('1234567890' FOR 4::bigint);
> substring
> -----------
> (1 row)
> test=# SELECT substring ('1234567890' FOR 4::int);
> substring
> -----------
> 1234
> (1 row)
> Looking at the explain verbose make it look like it's using the wrong
> version of substring. It's using the oid 2074 one:
> test=# select oid, oid::regprocedure from pg_proc where proname =
> 'substring';
> oid | oid
> -------+-------------------------------------
> 936 | "substring"(text,integer,integer)
> 937 | "substring"(text,integer)
> 1680 | "substring"(bit,integer,integer)
> 1699 | "substring"(bit,integer)
> 2012 | "substring"(bytea,integer,integer)
> 2013 | "substring"(bytea,integer)
> 2073 | "substring"(text,text)
> 2074 | "substring"(text,text,text) <----
> 16579 | "substring"(citext,integer,integer)
> 16580 | "substring"(citext,integer)
> (10 rows)
> That substring is for regular expressions. Nasty, not sure how to deal
> with that one...
Ah, so it's using "substring (STRING from PATTERN for ESCAPE)"?
Yes, that explains the NULL. Looks like we're in the INT/BIGINT
confusion again...
> Have a nice day,
It's a nice day since I have a nice workaround for this misfeature :-)
