Re: [HACKERS] plpgsql problem

2001-04-17 Thread Stephan Szabo
My guess is that since userid and userID differ only in case, it's probably not actually using the aliased version and instead is using only the column one. The execute is different since you're effectively putting the *value* of userID into the query as opposed to the word. I'd suggest renaim

Re: [HACKERS] plpgsql problem

2001-04-17 Thread Tom Lane
Lehel Gyuro <[EMAIL PROTECTED]> writes: > -- the same applies to inline queries too. if issued with execute > -- everything is fine, but if the query has more than one arguments > -- the compiler dismisses all, except the last one This is more than slightly hard to believe. There are thousands o

[HACKERS] plpgsql problem

2001-04-17 Thread Lehel Gyuro
CREATE FUNCTION userHasAll (int4,int4) RETURNS boolean AS ' DECLARE row RECORD; kirakorow kirakok%ROWTYPE; userID ALIAS FOR $1; kirakoID ALIAS FOR $2; megvan int4:=0; kepdarabok INTEGER:=0; query text; BEGIN SELECT * INTO kirakorow FROM kirakok WHERE kirako_id=kirakoID; -- this wor