Re: PL/SQL problem with Updates that specify varchar

pat.kelly_at_sanchez.com (Beelzehubub) wrote in message news:<df698d61.0408051102.905f683_at_posting.google.com>...
> Hi,
> I'm having a bit of trouble with an update performed by a PL/SQL
> routine that is updating more rows than it should be. Here's some of
> the code:
>
> Note that ZENT.APPID is NUMBER and ZENT.UCLS is VARCHAR(12)
>
> OPEN APPAccessCursor;
> LOOP
> FETCH APPAccessCursor INTO appid,ucls;
> EXIT WHEN APPAccessCursor%NOTFOUND;
>
> UPDATE ZENT SET "ZENT"."ACCLEVEL"='N' WHERE "ZENT"."UCLS"=ucls AND
> "ZENT"."APPKEY"=appid;
>
> END LOOP;
> CLOSE APPAccessCursor;
>
> appid and ucls are populated successfully by the fetch but when the
> update is performed, all rows matching appid are updated and the value
> of ucls is essentially ignored. I've tried =, like, and in for the
> ucls portion of the update statement and they all produce the same
> result. What am I doing wrong?

By any chance does the table zent has a column named ucls. If so you are comparing the column to itself, which always matches. In this case rename the plsql variable to v_ucls.

That is all I can think of with the code segment provided. HTH -- Mark D Powell -- Received on Fri Aug 06 2004 - 22:26:35 CEST