252190 - editusers.cgi: Use of uninitialized value on line 178 for Bugzilla 2.18rc1

Status: RESOLVED FIXED

(Bugzilla :: Administration, task, P2)

Description

[Vincent Botteman]

User-Agent:       Mozilla/4.0 (compatible; MSIE 5.01; Windows NT 5.0; .NET CLR 1.1.4322)
Build Identifier: 

When editing an user which is not a member of all groups, an error is written 
in the Apache log file because the variable $checked is uninitialized at line 
178 (print "<INPUT TYPE=HIDDEN NAME=\"oldgroup_$groupid\" 
VALUE=\"$checked\">\n";). An error is written for each group of which the user 
is not a member. In Bugzilla 2.17.6 the variable $checked contains the 
value "0" in case the user is not a member but in 2.18rc1 the variable is NULL.

Reproducible: Always
Steps to Reproduce:
1. Go to the User administration page
2. Search for some normal users (no administrators) or create a new one
3. Edit the user by clicking on the e-mail address

Actual Results:  
In the Apache log file appears the error:
editusers.cgi: Use of uninitialized value in concatenation (.) or string at 
C:/bugzilla-2.18rc1/editusers.cgi line 178., referer: 
http://texas.vdscomputing.net/editusers.cgi?
action=list&matchstr=@vds&matchtype=substr

Expected Results:  
In case the user is not a member of a group then the variable $checked should 
contain the value "0".

Installation of Bugzilla 2.18rc1 on Windows 2000 according to the "Byron Jones' 
Bugzilla Installation Guide for Windows" http://www.glob.com.au/bugzilla .

Comment 1

[Vincent Botteman]

If you change the query on line 140 from
                "MAX(grant_type = " . GRANT_DIRECT . "), " .
                "MAX(grant_type = " . GRANT_DERIVED . "), " .
                "MAX(grant_type = " . GRANT_REGEXP . ") " .
to
                "COALESCE(MAX(grant_type = " . GRANT_DIRECT . "), '0')," .
                "COALESCE(MAX(grant_type = " . GRANT_DERIVED . "), '0')," .
                "COALESCE(MAX(grant_type = " . GRANT_REGEXP . "), '0') " .
then I think this problem is resolved.

Comment 2

[Joel Peshkin]

so it is...

Comment 3

[Joel Peshkin]

Attachment: patch

While we're at it, we may as well say what we mean

Comment 4

[Vlad Dascalu]

Comment on attachment 153856 [details] [diff] [review]
patch

So we have this problem for $isderived and $isregexp as well.

Comment 5

[Joel Peshkin]

checked in

Comment 6

[Olav Vitters]

*** Bug 258630 has been marked as a duplicate of this bug. ***