148674 - Boolean Charts don't work in Netpositive because '-' is sent as '%2D'

Status: RESOLVED FIXED

(Bugzilla :: Query/Bug List, defect)

Description

[timeless]

We need to treat this url:
http://bugzilla.mozilla.org/buglist.cgi?&product=Bugzilla&product=Webtools&
field0%2D0%2D0=groupset&type0%2D0%2D0=equals&value0%2D0%2D0=512
the same as we treat this url:
http://bugzilla.mozilla.org/buglist.cgi?&product=Bugzilla&product=Webtools&
field0-0-0=groupset&type0-0-0=equals&value0-0-0=512

User-Agent: Mozilla/3.0 (compatible; NetPositive/2.2)

Be lenient in what you accept.

Comment 1

[timeless]

Attachment: url_decode name

Comment 2

[J. Paul Reed [:preed]]

Comment on attachment 85984 [details] [diff] [review]
url_decode name

Looks like it should solve the symptoms of the problem.

Also, I looked at the code on the 2.14.1-BRANCH and it has the same problem;
justdave: should this be "wanted for 2.14.2"? This seems low-risk/useful enough
to include there as well.

-jpr

Comment 3

[Dave Miller [:justdave]]

-> patch author

Comment 4

[Dave Miller [:justdave]]

Comment on attachment 85984 [details] [diff] [review]
url_decode name

r= justdave

fix is pretty obvious, and I can't imagine it breaking anything.  I've posted
to the reviewers list regarding the suggestion of including this in 2.14.2.

Comment 5

[Bradley Baetz (:bbaetz)]

Are we sure that we can do this? ie is there somewhere where something won't be
escaped, but will look like it is? There shouldn't be, of course, but...

The real fix is to use CGI.pl

Comment 6

[Dave Miller [:justdave]]

But we are using CGI.pl.  In fact, that's what this patch is patching. :)

Comment 7

[Bradley Baetz (:bbaetz)]

Err, CGI.pm, rather.

You know, the standard perl module which should handle almost anything we throw
at it.... (Yes, I know we can't use it ATM for various reasons)

Comment 8

[Dave Miller [:justdave]]

For reference, here's the relevant chunk of code from CGI.pm:

sub parse_params {
    my($self,$tosplit) = @_;
    my(@pairs) = split(/[&;]/,$tosplit);
    my($param,$value);
    foreach (@pairs) {
        ($param,$value) = split('=',$_,2);
        $param = unescape($param);
        $value = unescape($value);
        $self->add_parameter($param);
        push (@{$self->{$param}},$value);
    }
}

Notice it does the unescape both on the name and the value.  So we're in the clear.

Comment 9

[Bradley Baetz (:bbaetz)]

Fair enough. OK, lets get this into the 2.16 branch and trunk - the general
consensus was that we didn't want this for 2.14.

Comment 10

[timeless]

fixed on HEAD and BUGZILLA-2_16-BRANCH

Comment 11

[Dave Miller [:justdave]]

OK, in light of:

http://bugzilla.mozilla.org/show_activity.cgi?id=44642

which timeless brought up on IRC, looks like this qualifies as a security bug
after all....

the group bits are defined in the form as "bit-1", "bit-2", "bit-512" etc.

If a member of one of those groups makes ANY changes to that bug with a browser
that encodes the param names this way, those become "bit%2D1" etc.  And since
the "bit-1" field isn't present in the form, it gets removed, thus making the
bug public without the user's intention to do so.

Comment 12

[timeless]

fixed on BUGZILLA-2_14_1-BRANCH

Comment 13

[Matthew Tuck [:CodeMachine]]

I don't really mind here, but just having the tag "security" doesn't make it
serious without the "exploitable" tag.  If this was the case, you should also
bring bug #130821 and bug #63018 onto the branch.