Re: Fine Grained Acces Control & cursor

From: Tony <andrewst_at_onetel.net.uk>
Date: 7 Jun 2004 04:19:34 -0700
Message-ID: <c0e3f26e.0406070319.10afd4e1_at_posting.google.com>


thomas.kyte_at_oracle.com (Thomas Kyte) wrote in message news:<7b0834a8.0406060725.86a2686_at_posting.google.com>...
> for single row selects -- select .... INTO .... from .... is the *only
> way to go*.

I agree. But I do find that this gets me into trouble with the "PL/SQL police" on my project, who point me to the project PL/SQL standards that say thou shalt always use explicit cursors, supported by the usual myths plus some additional ones (quote: "use explicit cursors if you want to re-use a cursor. This will improve performance because there is a high chance that the cursor will execute pre-parsed SQL in the SGA").

Recently, I used a SELECT INTO in a situation where exactly one record should be found. However, during system testing TOO_MANY_ROWS was raised due to corrupt data caused by poor validation elsewhere in the system. Of course, rather than being thanked for finding a problem that needs solving, I was mildly scolded for using a SELECT INTO instead of opening a cursor and doing a single fetch like everyone else! Received on Mon Jun 07 2004 - 13:19:34 CEST

Original text of this message