Hi Bryan,
This bug snuck up and bit me again...Per your suggestion, here is a
reproducible script:
(your original email is at the bottom. I will log a Jira issue would be created
as this bug keeps coming back to haunt me)
CREATE TABLE "REVIEWS"."GEOFF__REVIEWS__REVIEW"
(
PK INTEGER PRIMARY KEY not null,
numstars BIGINT,
body VARCHAR(32672),
title VARCHAR(32672),
authoremail VARCHAR(32672)
);
INSERT INTO "REVIEWS"."GEOFF__REVIEWS__REVIEW"
(PK,numstars,body,title,authoremail) VALUES (0 /*not nullable*/,0,'s','s','s');
SELECT "review"."numstars"
FROM
"GEOFF__REVIEWS__REVIEW" AS "review"
WHERE
"review"."PK" = 1
ORDER BY
"review".PK
________________________________
From: Bryan Pendleton <[email protected]>
To: Derby Discussion <[email protected]>
Sent: Wednesday, July 2, 2008 5:53:12 PM
Subject: Re: why is column being returned?
Geoff hendrey wrote:
> Hi,
>
> I am doing something like:
>
> SELECT
> "A"."C1" AS "A.C1",
> "A"."C2" AS "A.C2"
> FROM
> ATABLE AS "A"
> WHERE
> "A"."C1" = "X"
> ORDER BY
> PK
>
> Does it make sense that PK should be returned in the result set? I expected
> ONLY "A.C1" and "A.C2" to be the column names returned by JDBC...but I see
> "PK" as well. It may be a bug in my code somewhere else, or is it possible
> that the ORDER BY is forcing "PK" to be returned?
I think that it *is* due to the ORDER BY, and I think that it is
incorrect behavior. I noticed this myself a while back and tried
to pin it down, but I was having trouble isolating a standalone
reproducible test case.
If you can construct a simple standalone test program which demonstrates
this behavior, I think it would be great if you could log it into JIRA
so we can track it and study it more.
thanks,
bryan
