343870 - Line breaks normalized twice

Status: RESOLVED FIXED

(Core :: XML, defect, P3)

Description

[Henri Sivonen (:hsivonen)]

Content sink & expat driver tamper with line breaks that come from expat. This is wrong. It is the job of the XML processor to normalize line breaks. However, if there is 
 in the source, it should be expanded as a CR in the DOM and the DOM builder shouldn't normalize it to LF.

Steps to reproduce:
1) Load http://hsivonen.iki.fi/test/newline-normalization.xhtml

Actual results:
The last test fails.

Expected results:
The four cases should all pass.

Comment 1

[Henri Sivonen (:hsivonen)]

Taking.

Comment 2

[Robert Sayre]

Rumor has it that bug 18333 needs to be fixed first.

Comment 3

[Steuard Jensen]

For the record, it sounds like this bug would have to be fixed in order for the folks clamoring for the ability to add line breaks to tooltips to get their wish (I'm not necessarily one of them, mind you), at least when writing XML.  For details, see bug 358452.  That bug is written in the context of HTML, but should apply just as much to XHTML (a fix for this bug would play the same role as an ideal fix for bug 322270 in the HTML context).

Comment 4

[Boris Zbarsky [:bzbarsky]]

Attachment: Henri's testcase

Comment 5

[Boris Zbarsky [:bzbarsky]]

This has pretty much nothing to do with bug 358452, as far as I  can tell.

We should probably just fix this.  It'll basically involve removing that code in the expat driver and nothing else, right?

Comment 6

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

Blocking since we have to give peterv some blocker too, he'd feel left out otherwise :)

Comment 7

[Henri Sivonen (:hsivonen)]

There's also double normalization going on in nsXMLContentSink in addition to the expat driver:
http://mxr.mozilla.org/seamonkey/source/content/xml/document/src/nsXMLContentSink.cpp#1492

Comment 8

[Boris Zbarsky [:bzbarsky]]

Yeah, we should remove that too.  At least if expat performs the normalization for us.  Does it?

Comment 9

[Peter Van der Beken [:peterv]]

Attachment: v1

Comment 10

[Peter Van der Beken [:peterv]]

Attachment: v2

Comment 11

[Boris Zbarsky [:bzbarsky]]

Comment on attachment 267744 [details] [diff] [review]
v2

Looks good!

Comment 12

[Peter Van der Beken [:peterv]]

Attachment: Henri's testcase as mochitest