414637 - createContextualFragment() throws NS_ERROR_NOT_AVAILABLE

Status: RESOLVED FIXED

(Core :: DOM: Core & HTML, defect, P3)

Description

[Nick Nassar]

User-Agent:       Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.8.1.11) Gecko/20071127 Firefox/2.0.0.11
Build Identifier: XULRunner CVS trunk 01/29/2008

Calls to nsIDOMRange::createContextualFragment() on a HTML document throw the following exception.

[Exception... "Component returned failure code: 0x80040111 (NS_ERROR_NOT_AVAILABLE) [nsIDOMNSRange.createContextualFragment]"  nsresult: "0x80040111 (NS_ERROR_NOT_AVAILABLE)"  location: "JS frame :: file:///e:/miro/tv/platform/windows-xul/dist/components/jsbridge.js :: anonymous :: line 519"  data: no]

Reproducible: Always

Steps to Reproduce:
    var elt = document.getElementById(id);  //id is a string defined earlier
    var r = document.createRange();
    r.selectNode(document.documentElement);
    var frag = r.createContextualFragment(xml); //xml is a string defined earlier

Actual Results:  
throws NS_ERROR_NOT_AVAILABLE

Expected Results:  
Return the fragment

This is happening for me when I upgrade Miro to XULRunner 1.9 from CVS trunk. The above link is from someone else experiencing the same problem.

Comment 1

[Olli Pettay [:smaug][[e-mail]]]

A testcase uploaded using "Add an attachment" would be great.

Comment 2

[Olli Pettay [:smaug][[e-mail]]]

Attachment: testcase

Something to do with .documentElement

Comment 3

[Olli Pettay [:smaug][[e-mail]]]

mStartParent is null when selecting .documentElement.
Jonas, can you remember something related to this immediately?

Comment 4

[Olli Pettay [:smaug][[e-mail]]]

Er, no, sorry. It is doc which I QI'd to nsIContent in nsContentUtils::CreateContextualFragment.
That of doesn't work.

Comment 5

[Olli Pettay [:smaug][[e-mail]]]

s/which I/which is/

Comment 6

[Olli Pettay [:smaug][[e-mail]]]

regression range 2006-10-19 - 2006-10-21.
-> Bug 357445

Comment 7

[Olli Pettay [:smaug][[e-mail]]]

Attachment: testcase2

So on 1.8 .startContainer is htmlelement, on 1.9 it is htmldocument.
What 1.9 does is afaik right. That is what Opera and Safari returns.

(Safari crashes with this testcase.)

Comment 8

[Olli Pettay [:smaug][[e-mail]]]

Attachment: allow contextfragments when container is document

I think we should do something like this.

Comment 9

[Jonas Sicking (:sicking) No longer reading bugmail consistently]

Comment on attachment 300380 [details] [diff] [review]
allow contextfragments when container is document

Looks good

Comment 10

[Nick Nassar]

The patch works for me

Comment 11

[Mike Schroepfer]

Did you want this for b3?

Comment 12

[Olli Pettay [:smaug][[e-mail]]]

I was thinking after b3. Not so serious, or at least this feature isn't
apparently used very often. So just to get this fixed in FF3.

Comment 13

[Reed Loden [:reed]]

Checking in content/base/src/nsContentUtils.cpp;
/cvsroot/mozilla/content/base/src/nsContentUtils.cpp,v  <--  nsContentUtils.cpp
new revision: 1.272; previous revision: 1.271
done

Comment 14

[Reed Loden [:reed]]

*** 42217 ERROR FAIL | range.createContextualFragment() should throw when range node is DocType |  | /tests/parser/htmlparser/tests/mochitest/test_bug358797.html

Backed out.

Comment 15

[Olli Pettay [:smaug][[e-mail]]]

Bah, sorry.
I'll update either the testcase or patch when back on my dev.machine.

Comment 16

[Olli Pettay [:smaug][[e-mail]]]

I think changing the testcase is ok, it should just check that 
browser doesn't crash when creating contextual fragment using doctype.

Comment 17

[Olli Pettay [:smaug][[e-mail]]]

I changed the testcase.