[Laszlo-dev] [Laszlo-checkins] r14496 - openlaszlo/trunk/WEB-INF/lps/lfc/kernel/dhtml

[P T Withington]

Yup.  Henry and Maynard reported the same.  I'll send a fix in a minute.

On 2009-08-19, at 16:55EDT, Max Carlson wrote:

> I think this change has a mistake in it.  In some cases, I get an  
> exception in Firebug when mousing around the compose window in webtop:
>
> $1.mouseevent is not a function
> anonymous(Object owner=view name: icon uid=511 __LZdiv=div.lzdiv,  
> "onmouseout")LFCdhtml.js (line 8784)
> anonymous("onmouseout", Object owner=view name: icon uid=511  
> __LZdiv=div.lzdiv)LFCdhtml.js (line 8807)
> anonymous("onmouseout", Object owner=view name: icon uid=511  
> __LZdiv=div.lzdiv)LFCdhtml.js (line 1438)
> anonymous(mouseout clientX=289, clientY=92)LFCdhtml.js (line 1665)
> [Break on this error] }};$1.mouseevent($2);if($2=="onmousedown") 
> {\nLFCdhtml.js (line 8784)
>
> I tracked this down to this line in LzSprite.js#div.onmouseout =  
> function(e) which was put in by you as part of r14496:
>
> LzMouseKernel.__sendEvent('onmouseout', e.target.owner);
>
> You can't rely on the event object having a target, much less an  
> owner that's a view.  This is causing an exception later when the  
> ModeManager tries to call the mouseevent() method on a view that  
> doesn't exist. Also, you may want to use the 'el' property from the  
> top of the method for this to work in IE.
>
> Please test for valid target/owner properties before sending the  
> event - and make sure it works in IE as well.  Remember, the mouse  
> may be moving over divs that aren't owned by OL, e.g. the iframe  
> that contains the rich text editor.
>
> ptw at openlaszlo.org wrote:
>> Author: ptw
>> Date: 2009-08-17 14:04:01 -0700 (Mon, 17 Aug 2009)
>> New Revision: 14496
>> Modified:
>>   openlaszlo/trunk/WEB-INF/lps/lfc/kernel/dhtml/LzMouseKernel.js
>>   openlaszlo/trunk/WEB-INF/lps/lfc/kernel/dhtml/LzSprite.js
>> Log:
>> Change 20090817-ptw-i by ptw at repo-man.home on 2009-08-17 10:37:40 EDT
>>    in /Users/ptw/OpenLaszlo/trunk-2
>>    for http://svn.openlaszlo.org/openlaszlo/trunk
>> Summary: Correct mouseover/mouseout behavior for clickdivs
>> Bugs Fixed: LPP-8375 onmouseout even on HTML component is firing  
>> onmouseover, not onmouseout
>> Technical Reviewer: max (Message-ID: <4A899FA5.4070706 at laszlosystems.com 
>> >)
>> QA Reviewer: lhenrywilkins at laszlosystems.com (pending)
>> Details:
>>    If the fix_clickable quirk is on, ignore mouseover/mouseout events
>>    that get sent when clickdivs are disabled, because they are
>>    redundant.  Similarly, when clickdivs are re-enabled by a global
>>    mouseout event, synthesize a mouseout to the now re-enabled view.
>> Tests:
>>    Test case from bug report
>> Modified: openlaszlo/trunk/WEB-INF/lps/lfc/kernel/dhtml/ 
>> LzMouseKernel.js
>> ===================================================================
>> --- openlaszlo/trunk/WEB-INF/lps/lfc/kernel/dhtml/LzMouseKernel.js	 
>> 2009-08-17 18:28:41 UTC (rev 14495)
>> +++ openlaszlo/trunk/WEB-INF/lps/lfc/kernel/dhtml/LzMouseKernel.js	 
>> 2009-08-17 21:04:01 UTC (rev 14496)
>> @@ -204,8 +204,10 @@
>>         }
>>     }
>>     ,__cachedSelection: null
>> +    ,__globalClickable: true
>>     ,setGlobalClickable: function (isclickable){
>>         var el = document.getElementById('lzcanvasclickdiv');
>> +        this.__globalClickable = isclickable;
>>         el.style.display = isclickable ? '' : 'none';
>>     }
>>     ,__sendMouseMove: function(e, offsetx, offsety) {
>> Modified: openlaszlo/trunk/WEB-INF/lps/lfc/kernel/dhtml/LzSprite.js
>> ===================================================================
>> --- openlaszlo/trunk/WEB-INF/lps/lfc/kernel/dhtml/LzSprite.js	 
>> 2009-08-17 18:28:41 UTC (rev 14495)
>> +++ openlaszlo/trunk/WEB-INF/lps/lfc/kernel/dhtml/LzSprite.js	 
>> 2009-08-17 21:04:01 UTC (rev 14496)
>> @@ -163,6 +163,7 @@
>>                     }
>>                 }
>>                 if (mousein) {
>> +                    var wasClickable =  
>> LzMouseKernel.__globalClickable;
>>                     if (quirks.fix_ie_clickable) {
>>                         LzMouseKernel.setGlobalClickable(true);
>>                     }
>> @@ -177,6 +178,13 @@
>>                     LzMouseKernel.setMouseControl(true);
>>                     LzMouseKernel.__resetMouse();
>>                     this.mouseisover = true;
>> +                    if (quirks.fix_clickable && (! wasClickable)  
>> && LzMouseKernel.__globalClickable) {
>> +                        // NOTE: [2008-08-17 ptw] (LPP-8375) Forward
>> +                        // the event to the view, if it would have
>> +                        // gotten it without the quirk clickability
>> +                        // diddling
>> +                        LzMouseKernel.__sendEvent('onmouseout',  
>> e.target.owner);
>> +                    }
>>                 } else {
>>                     if (quirks.focus_on_mouseover) {
>>                         if (LzInputTextSprite.prototype.__lastshown  
>> == null) {
>> @@ -1370,6 +1378,21 @@
>>             }
>>         }
>> +        if (this.quirks.fix_clickable && (!  
>> LzMouseKernel.__globalClickable)) {
>> +            // NOTE: [2008-08-17 ptw] (LPP-8375) When the mouse goes
>> +            // over an html clickdiv, globalClickable gets disabled,
>> +            // which generates a mouseout -- we want to ignore that.
>> +            // Simlutaneously, the mouse enters the associated  
>> iframe,
>> +            // which will forward a mouseover to us, but we already
>> +            // got one, so, we want to ignore that too.  The global
>> +            // mouseout handler will synthesize a mouseout event for
>> +            // the html sprite when the mouse leaves the iframe and
>> +            // re-enables the clickdiv.
>> +            if ((eventname == 'onmouseout') || (eventname ==  
>> 'onmouseover')) {
>> +                return;
>> +            }
>> +        }
>> +
>>         LzMouseKernel.__sendEvent(eventname, this.owner);
>>     }
>> }
>> _______________________________________________
>> Laszlo-checkins mailing list
>> Laszlo-checkins at openlaszlo.org
>> http://www.openlaszlo.org/mailman/listinfo/laszlo-checkins
>
> -- 
> Regards,
> Max Carlson
> OpenLaszlo.org