261108 - using <enter> to submit a form triggers Warning ``reference to undefined property event.button'' [xs] in file ``chrome://communicator/content/contentAreaClick.js'', line 189, character 0.

Status: RESOLVED FIXED

(Core :: DOM: Events, defect)

Description

[timeless]

Warning ``reference to undefined property event.button'' [xs] in file
``chrome://communicator/content/contentAreaClick.js'', line 189, character 0.

   // Called whenever the user clicks in the content area,
    172   // except when left-clicking on links (special case)
    173   // should always return true for click to go through
    174   function contentAreaClick(event) 
 -  175   {
 -  176     var isKeyPress = (event.type == "keypress");
 -  177     var href = hrefForClickEvent(event);
 -  178     if (href) {
 -  179       if (isKeyPress) {
 -  180         openNewTabWith(href, true, event.shiftKey);
 -  181         event.preventBubble();
    182       }
    183       else {
 -  184         handleLinkClick(event, href, null);
    185       }
 -  186       return true;
    187     }
    188 
 -  189     if (pref && !isKeyPress && event.button == 1 &&
 -  190

#0: function contentAreaClick(event=UIEvent:{0}) in
<chrome://communicator/content/contentAreaClick.js> line 189
#1: function onclick(event=UIEvent:{0}) in
<chrome://navigator/content/navigator.xul> line 1

0001: event instanceof Components.interfaces.nsIDOMMouseEvent
$[1] = [boolean] false
0001: event.type
$[2] = [string] "click"

enter triggered submits are evil magical creatures that pretend to be clicks but
aren't.

the question is, should they be real click events, or should i change 176 to use
 !(event instanceof Components.interfaces.nsIDOMMouseEvent).

I'm presuming the latter, hence the assignment and component, however if dom
wants to change, then i'm willing to do that too...

Comment 1

[timeless]

Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.8a4) Gecko/20040922

Comment 2

[[e-mail]]

Attachment: Test Case

Comment 3

[[e-mail]]

Attachment: Proposed patch

Everyone else creates an nsMouseEvent, this must have been an oversight.

Comment 4

[[e-mail]]

Comment on attachment 159834 [details] [diff] [review]
Proposed patch

Could whomever gets here first give r+sr ;-)

Comment 5

[Johnny Stenback (:jst)]

Comment on attachment 159834 [details] [diff] [review]
Proposed patch

r+sr=jst

Comment 6

[Brendan Eich [:brendan]]

Want this for branches, jst and I think.  Nice safe fix.

/be

Comment 7

[Brendan Eich [:brendan]]

Comment on attachment 159834 [details] [diff] [review]
Proposed patch

a=[e-mail] for the branches -- who will check this in?

/be

Comment 8

[Johnny Stenback (:jst)]

Fix landed on aviary and 1.7 branches.

Comment 9

[Serge Gautherie (:sgautherie)]

Comment on attachment 159834 [details] [diff] [review]
Proposed patch

'brendan: approval1.8a4-':
What about v1.8a5 Trunk then ?

Comment 10

[[e-mail]]

Fix checked in, thanks for reminding me.