226099 - [RFE] Highlight "<![CDATA[" and matching "]]>" in View Source

Status: RESOLVED FIXED

(Core :: DOM: HTML Parser, enhancement)

Description

[David Leverton]

User-Agent:       Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.5) Gecko/20031007
Build Identifier: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.5) Gecko/20031007

The opening and closing markup of a CDATA section should be highlighted in the
View Source window.

Reproducible: Always

Steps to Reproduce:
1. Open XML/XHTML page containing a CDATA section.
2. View the source.

Actual Results:  
Source window opens, with the "<![CDATA[" and "]]>" markup in the same colour as
regular text, making it harder to distinguish.

Expected Results:  
"<![CDATA[" and "]]>" should be highlighted.

Comment 1

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

Can You give an example?
I see CDATA in different color than normal text.

Comment 2

[David Leverton]

Attachment: XHTML document containing CDATA section

Comment 3

[David Leverton]

Attachment: Screenshot of View Source window displaying previous attatchment

Comment 4

[Boris Zbarsky [:bzbarsky]]

Dup of the general "XML view source uses the HTML tokenizer" stuff.  The problem
is that the content of <script> tags is not tokenized as CDATA by the HTML
tokenizer.  Regular CDATA sections are.

Comment 5

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

*** This bug has been marked as a duplicate of 91240 ***

Comment 6

[Boris Zbarsky [:bzbarsky]]

Reopening, since we've decided there is no reasonable way to use the expat
tokenizer, so we'll fix the XML bugs one at a time.

Comment 7

[Blake Kaplan (:mrbkap) (inactive)]

bz, I'm not sure how possible this is. We could ignore <script> and <style> when
processing XML in the tokenzier (i.e., treat them as any other tags, as happens
in XML), but this sort of change scares me. Maybe restricting the behavior to
view-source + XML would be better.

Otherwise, this should be WONTFIX.

Comment 8

[Boris Zbarsky [:bzbarsky]]

Blake, the only time XML should be going through the HTML tokenizer is for view
source.  So you could just treat them normally for XML and assert that we're in
view source, I think.

Comment 9

[Blake Kaplan (:mrbkap) (inactive)]

Confirming, taking and moving to Parser (as that is where the changes are going
to have to take place).

Comment 10

[Blake Kaplan (:mrbkap) (inactive)]

Attachment: patch v1

While technically this doesn't fill the RFE (to only highlight the <![CDATA[
and the ]]>), this is as close as we're going to get until marked sections are
parsed properly.

Comment 11

[Boris Zbarsky [:bzbarsky]]

Comment on attachment 161725 [details] [diff] [review]
patch v1

r=bzbarsky assuming you've tested the effect this has on <script>, <style>,
<textarea> in XML documents...

Comment 12

[Blake Kaplan (:mrbkap) (inactive)]

Comment on attachment 161725 [details] [diff] [review]
patch v1

bz: I've just tested that they all still work. Of course, if this patch had
broken XML documents, something would have been seriously wrong (and the assert
would have told me).

rbs, could you sr?

Comment 13

[Boris Zbarsky [:bzbarsky]]

I meant on view-source of XML documents.  I agree that this should not affect
actual XML rendering in any way.  ;)

Comment 14

[rbs]

Comment on attachment 161725 [details] [diff] [review]
patch v1

sr=rbs

Comment 15

[Blake Kaplan (:mrbkap) (inactive)]

Marking fixed as timeless checked this in.