351932 - DOMParser and XMLHttpRequest insert whitespace nodes while parsing XUL

Status: RESOLVED INVALID

(Firefox :: General, defect)

Description

[Fabian Dankof]

User-Agent:       Mozilla/5.0 (Windows; U; Windows NT 5.1; de; rv:1.8.0.6) Gecko/20060728 Firefox/1.5.0.6
Build Identifier: Mozilla/5.0 (Windows; U; Windows NT 5.1; de; rv:1.8.0.6) Gecko/20060728 Firefox/1.5.0.6

The problem occured while I tried to parse a XUL document which I got via XMLHttpRequest. The XUL document is retrieved correctly (req.responseText).
But if I try to parse this text with the DOMParser() object (as 'text/xml'),
the resulting DOM tree contains plenty of whitespace node.
Furthermore, XMLHttpRequest does the same when getting the retrieved conten
with req.responseXML.

The actual workaround for this problem is to write a JavaScript function that strips existing whitespace text nodes from the new tree.





Reproducible: Always

Steps to Reproduce:
1. Get XUL content via XMLHttpRequest
2. Parse text response with DOMParser(text,'text/xml')

Actual Results:  
view resulting tree in debugger --> contains lots of text nodes (type: #text)

Expected Results:  
a dom tree _without_ whitespace nodes because they're not part of the original XUL document.

Comment 1

[Boris Zbarsky [:bzbarsky]]

As filed, this is invalid.  The XML document involved _does_ contain whitespace nodes.  If you want this document parsed as non-XML (which is what the XUL parser does), you need to use non-XML tools.

I suppose we could have a "XUL" mode for DOMParser, if XUL even makes it possible.  Feel free to resummarize accordingly and move to Core:DOM.  It's pretty similar to the requests to have an HTML mode for it.

Comment 2

[Fabian Dankof]

Ok. Bug has been reported to Core:DOM. ID: #352052.