226193 - Javascript exception when accessing document["config"] or document["domConfig"]

Status: VERIFIED FIXED

(Core :: DOM: Core & HTML, defect)

Description

[Hubert Kauker]

User-Agent:       Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.6a) Gecko/20031030
Build Identifier: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.6a) Gecko/20031030

"Component returned failure code: 0x80004001 (NS_ERROR_NOT_IMPLEMENTED)
[nsIDOM3Document.config]"

when the following javascript loop is run:

	for( var p in window.document )
	{
		try {
			var v = window.document[p];
		}
		catch(e) { 
			alert ( e.message );
		}
	}

This both happens before and after the onload event of the window.
At the point of exception we have p == "config".




Reproducible: Always

Steps to Reproduce:
1.Include above script fragment into SCRIPT element in HEAD of a document.
2.Optionally enclose the code by window.onload = function() { ... }
3.Load the document.


Actual Results:  
Alert() is shown containing message:

Component returned failure code: 0x80004001 (NS_ERROR_NOT_IMPLEMENTED)
[nsIDOM3Document.config]



Expected Results:  
Should have executed the loop without any visible action.

Comment 1

[Zibi Braniecki [:zbraniecki][:gandalf]]

Confirming with Linux Firebird 0.7+

Comment 2

[Zibi Braniecki [:zbraniecki][:gandalf]]

Attachment: testcase

Comment 3

[Zibi Braniecki [:zbraniecki][:gandalf]]

Comment on attachment 135926 [details]
testcase

This testcase throws error on "config" element in Mozilla Suite and on
"domConfig" element in Firebird

Comment 4

[Boris Zbarsky [:bzbarsky]]

It throws on domConfig on both, actually, if you use a recent suite or
something.  And that's correct -- you are accessing a DOM3 property that is in
fact not implemented....

Is there a reason here to make it not throw?

Anyway, not JS engine; the "Component returned failure code" means it's an error
in the Mozilla code somewhere.

Comment 5

[Zibi Braniecki [:zbraniecki][:gandalf]]

then why it can be iterated with for-in ?

Comment 6

[Boris Zbarsky [:bzbarsky]]

Because it _is_ defined as a property on the nsIDOM3Document interface.  It's
just one that throws if you try to access its getter.

Please do not undo my component changes.  Thank you.

Comment 7

[Zibi Braniecki [:zbraniecki][:gandalf]]

Ouch. Sorry - ctrl+r in Gecko leaves selects in old position.

So what kind of bug is it? 

The goal is to remove not implemented method/property from interface or to
implement it?!?

Comment 8

[Boris Zbarsky [:bzbarsky]]

The goal is to eventually implement this.

Comment 9

[Boris Zbarsky [:bzbarsky]]

*** Bug 239550 has been marked as a duplicate of this bug. ***

Comment 10

[Boris Zbarsky [:bzbarsky]]

*** Bug 262981 has been marked as a duplicate of this bug. ***

Comment 11

[WADA:World Anti-bad-Duping Agency]

*** Bug 302383 has been marked as a duplicate of this bug. ***

Comment 12

[Boris Zbarsky [:bzbarsky]]

*** Bug 295581 has been marked as a duplicate of this bug. ***

Comment 13

[Phil Ringnalda (:philor)]

*** Bug 360654 has been marked as a duplicate of this bug. ***

Comment 14

[Erik Fabert]

*** Bug 362132 has been marked as a duplicate of this bug. ***

Comment 15

[James Justin Harrell]

> Is there a reason here to make it not throw?

Iterating over document can throw an error, even if there's no attempt to use the property.

for each( var propValue in document ) { ... }
for( var [propName, propValue] in document ) { ... }

Comment 16

[Georg Maaß]

In XulRunner trunk: Mozilla/5.0 (Macintosh; U; PPC Mac OS X Mach-O; de-DE; rv:1.9a5pre) Gecko/20070525 this bug occures too.

Comment 17

[Tom S [:evilpie]]

domConfig was removed by Bug 615509.

Comment 18

[Georg Maaß]

Status should be set to RESOLVED WONT FIX as this bug is not invalid for the release it was filed for, but is obsolete for upcoming release. So status RESOLVED INVALID is not correct. Correct status is RESOLVED WONT FIX because the domConfig interface this bug belongs to will be removed in XulRunner 6.0.

Comment 19

[Tom S [:evilpie]]

George Maaß not sure this was explained in an other way to me. This Bug is invalid because it refers to something that doesn't exists anymore. I think it's not important when it will be removed from where. But i have no experience with the DOM component, so it might as well we Wontfix. (Actually i don't really care about the resolution thought)

Comment 20

[:Ms2ger (he/him; ⌚ UTC+1/+2)]

I'd argue it's fixed: accessing document.domConfig doesn't throw anymore.

Comment 21

[Biju]

The resolution should be WORKSFORME, see bug 385520
FIXED only when we exactly know which bug/patch fixed that.

Comment 22

[Boris Zbarsky [:bzbarsky]]

Biju, we do know that.  The bug number is even listed in comment 17.

Verified fixed.  Please no more noise in this bug!