12721 - Client-side image maps with spaces in name ignore spaces

Status: VERIFIED INVALID

(Core :: Layout, defect, P3)

Description

[Nicholas Cull]

This bug follows on from bug 11453.

Tested on:

Mozilla 1999082708, M10 build, Windows NT 4 Workstation.

Expected bahaviour:

For the test case attached, the client-side image map should not work. The map
name contains a space in the middle of it, while the usemap attribute of the
image does not contain this space.

Current bahaviour:

The image map works as if nothing is wrong.

Other browsers tested:

Opera 3.60 and Internet Explorer 5 behave as expected.

Other comments:

Be careful not to re-open bug 11453 when fixing this.

Comment 1

[Nicholas Cull]

Attachment: Testcase showing client-side map working unexpectedly.

Comment 2

[Mike Shaver (:shaver -- probably not reading bugmail closely)]

This isn't a DOM/vidur bug.  Calgon -- ahem, chofmann -- take me away!

Comment 3

[chris hofmann]

not sure if I understand why this is bad...
looks like a pretty good graceful failure.
maybe pnunn can figure out what to do.

Comment 4

[pnunn]

Let me get this straight.
The complaint is that the image map works,
right?
-pn

Comment 5

[pnunn]

This apparently is a *feature* implemented in
nsImageFrame::GetImageMap(). usemap.StripWhitespace() is
called for a NAV quirk. See line 563 in nsImageFrame.cpp.

This isn't really an imglib problem. I'd be happy to take
this one, if you are short handed. Looks like Kipp did alot
of the work in this area.
Reassign back to me if you'd like.

-pn

Comment 6

[troy]

Yes, that would be a big help if you could take this one. Thanks Pam

Comment 7

[pnunn]

The question now is do we support the Nav quirk or not...
I have no strong feelings on the subject either way.
If communicator ignored spaces before, what is the advantage
to suddenly start paying attention to them?
-pn

Comment 8

[pnunn]

Can anyone give me a reason to
not support the Nav quirk of ignoring
spaces in image map names?

Convince me or I mark the bug invalid.
-p

Comment 9

[pnunn]

I'm closing as invalid.

Comment 10

[Chris Petersen]

Marking verified invalid per last comments.

Comment 11

[Heikki Toivonen (remove -bugzilla when emailing directly)]

This bug has wrong resolution. This bug is not INVALID, it should be WONTFIX.

Aside from that, I am wondering if this should be fixed for strict mode. What 
does IE do? Adding harishd to CC.

Comment 12

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/integration/autoland/rev/3b30c4bdfe46
[fenix] For https://github.com/mozilla-mobile/fenix/issues/12721 - Remove SessionManager usages in TabCounter