Hish Reporter
	Description • 19 years ago

User-Agent:       Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.8b4) Gecko/20050910 SeaMonkey/1.0a
Build Identifier: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.8b4) Gecko/20050910 SeaMonkey/1.0a

XMLHttpRequest fails to load an invalid XML file >6kb. 
This happens when the server sends contentType=text/xml and the file is not valid XML.
All version before Gecko 1.8 work fine here and return responseXML=null and responseText with the contents.
One could argue that this is not a bug but a feature because the server sends the wrong header. But the problem is that sometimes we have no control about foreign servers.

Reproducible: Always

Steps to Reproduce:

	Sébastien Auger
	Comment 1 • 19 years ago

If it's just failing without hanging the browser or crashing etc .. we should not be worried. Does it affect something else for you or this is not big deal ?

	Hish Reporter
	Comment 2 • 19 years ago

Why we shouldn't be worried?
a) That's a regression and 
b) we have no chance to load a document once the server sends text/xml by mistake

	Sébastien Auger
	Comment 3 • 19 years ago

(In reply to comment #2)
> Why we shouldn't be worried?
> a) That's a regression and 
> b) we have no chance to load a document once the server sends text/xml by
> mistake
> 

oh I see now, didn't understood very well your bug description.
but the build you're using is more than a month old, did you verified with newer builds?

	Nickolay_Ponomarev
	Comment 4 • 18 years ago

Could you attach / point to a testcase please? I couldn't reproduce the bug easily.

	Dão Gottwald [:dao]
	Comment 5 • 18 years ago

This is true for both 1.8 branch and trunk.

	Dão Gottwald [:dao]
	Comment 6 • 18 years ago

This is alert(r.responseXML.documentElement.textContent). r.responseXML should have been null.

	Dão Gottwald [:dao]
	Comment 7 • 18 years ago

Btw, responseText does not contain the error document but the real response.

	Hallvord R. M. Steen [:hallvors]
	Comment 8 • 11 years ago

Gecko has this old feature where XML parser outputs a document with error info if parsing fails. This is wrong per spec and other implementations

Test cases:
http://w3c-test.org/web-platform-tests/master/XMLHttpRequest/responsexml-media-type.htm
http://w3c-test.org/web-platform-tests/master/XMLHttpRequest/responsexml-non-well-formed.htm

	Thomas Wisniewski
	Comment 9 • 8 years ago

This bug seems like a dupe of bug 289714, which was just fixed. The web platform tests in comment 8 are now passing, and XHRs made by web content for invalid XML documents will now give a null responseXML instead of a <parsererror> document (and the parsing failure will be logged to the web console).