226477 - Bugzilla::User->in_group calls non-existent method $dbh->selectrow

Status: RESOLVED FIXED

(Bugzilla :: User Accounts, defect, P3)

Description

[Joe Green]

User-Agent:       Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.4) Gecko/20030703
Build Identifier: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.4) Gecko/20030703

The Bugzilla::User->in_group method (Bugzilla/User.pm) calls
$dbh->selectrow, which is not a DBI method.  Changing the call to
$dbh->selectrow_array fixes the problem.


Reproducible: Always

Steps to Reproduce:
I call this method directly from a script for importing bugs from another bug
system, so I'm not sure how to reproduce this with the standard distribution.
Still, I think the problem can be seen by simple inspection.
Actual Results:  
When the method is call, it fails with an error saying that DBI::db cannot
identify the method "selectrow".

Expected Results:  
The Bugzilla method should not have crashed.

Comment 1

[Joe Green]

Attachment: Patch to fix invalid $dbh-selectrow call in Bugzilla::User->in_group

This is the patch I use to fix this problem.

Comment 2

[Joe Green]

This happens in the 2.17.6 version of Bugzilla.

Comment 3

[Dave Miller [:justdave]]

I can't reproduce the error (I've been using $user->in_group() in some of my
stuff and it hasn't errored - then again, I'm calling $user->groups() earlier,
which caches them.

The DBI docs agree though.

Comment 4

[Dave Miller [:justdave]]

Comment on attachment 136068 [details] [diff] [review]
Patch to fix invalid $dbh-selectrow call in Bugzilla::User->in_group

actually this was a little more broken than that...  because it returns an
array, we need to qualify $res as an array element to receive the response.

my ($res) = $dbh->.....

Comment 5

[Joe Green]

> actually this was a little more broken than that...  because it returns an
> array, we need to qualify $res as an array element to receive the response.

In this case it may be OK without the array qualifier, since all that is wanted
is a present/not present result, assuming that no match returns a zero-length
array.

If it's not OK, there are 3 or 4 other occurrences of the same structure in
Bugzilla/User.pm.

Comment 6

[Bradley Baetz (:bbaetz)]

selectrow is an alias for selectrow_array, which in scalar context returns
either teh first or the last coulmn in the query. (its undefined which). As long
as the query only selects one column, its fine.

Comment 7

[Joe Green]

> selectrow is an alias for selectrow_array, which in scalar context returns
> either teh first or the last coulmn in the query. (its undefined which).

Can you give me a pointer to where the support for this is?  It failed on my
system as an unknown method, and I just downloaded DBI-1.38 from dbi.perl.org,
and I don't see it in there either.

Comment 8

[Bradley Baetz (:bbaetz)]

Hmm. I thought it was, but its fetch that is the alias for fetchrow_array, and
theres no equiv alias for select. Bleh. How did this ever work?

Comment 9

[Joel Peshkin]

Attachment: Replaced with a valid call

To both see the problem and test the fix, you can slide this code into a copy
of index.cgi


my $joel = Bugzilla::User->new(1);
print "content-type: text/plain\n\n";
print $joel->in_group("admin");
exit;

Comment 10

[Joel Peshkin]

Checking in Bugzilla/User.pm;
/cvsroot/mozilla/webtools/bugzilla/Bugzilla/User.pm,v  <--  User.pm
new revision: 1.18; previous revision: 1.17
done


Nice catch, Joe