This is an archived snapshot of W3C's public bugzilla bug tracker, decommissioned in April 2019. Please see the home page for more details.

Bug 3992 - non-unix newlines confuse the parser
Summary: non-unix newlines confuse the parser
Status: RESOLVED FIXED
Alias: None
Product: Validator
Classification: Unclassified
Component: Parser (show other bugs)
Version: HEAD
Hardware: PC Linux
: P2 normal
Target Milestone: 0.8.0
Assignee: Olivier Thereaux
QA Contact: qa-dev tracking
URL: http://www.aadmm.de/en/mathml.xhtml
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2006-11-20 18:05 UTC by Matthias Mauch
Modified: 2007-04-02 06:31 UTC (History)
1 user (show)

See Also:


Attachments

Description Matthias Mauch 2006-11-20 18:05:50 UTC
It seems that W3C Unicorn have problems in following situation:

Profil: General Conformance Check
Address: http://www.aadmm.de/en/mathml.xhtml
Doctype: XHTML 1.1 plus MathML 2.0 plus SVG 1.1
Check failed...

If I test this page with http://validator.w3.org/
then I get passed validation "This Page Is Valid
XHTML 1.1 plus MathML 2.0 plus SVG 1.1"

Is this Doctype not available in Unicorn, because I can't
select it in the Profil "Validate Markup (HTML, etc.)"?
Comment 1 Olivier Thereaux 2006-11-21 00:58:00 UTC
This is a problem with the development version of the markup validator (which the unicorn test instance uses) with some types of documents, perhaps due to some specific line ending types.

Comment 2 Olivier Thereaux 2007-03-02 03:06:50 UTC
Björn, could this be a bug in S::P::O? 

It seems that check's preparse does find the doctype, and although the document given as a test case has the particularity of having the XML decl and doctype decl on the same line, I am not aware that it is forbidden.

I don't think it's a bug with check, nor does it look like a bug in OpenSP - see http://qa-dev.w3.org/wmvs/0.7/check?uri=http%3A%2F%2Fwww.aadmm.de%2Fen%2Fmathml.xhtml
so I'm suspecting a problem in S:P:O. 

Please look into it. If this is a bug in SPO, closing this bug with a pointer to the SPO bug tracking URI would be good.
Comment 3 Olivier Thereaux 2007-03-16 00:39:20 UTC
OK, I think I found and fixed the issue: newlines. Apparently the fact that the document mentioned above is using DOS newlines was the trigger. Converting the newlines to unix ones gave a more expected validation results.

I have a tentative fix 
http://lists.w3.org/Archives/Public/www-validator-cvs/2007Mar/0082.html
which hopefully does the right thing by converting newlines systematically.

Apparently it fixes the issue for the validator. 
I still think there might be some bug with SPO, though. 

Matthias, is it OK to save your mathml.xhtml document as test case in our test suite?

Thanks
Comment 4 Matthias Mauch 2007-03-30 17:15:04 UTC
> Matthias, is it OK to save your mathml.xhtml document as test case in our test
> suite?
Yes, it's ok. I have no problems if you save my document as test case. 
Comment 5 Olivier Thereaux 2007-04-02 06:31:01 UTC
(In reply to comment #4)
> Yes, it's ok. I have no problems if you save my document as test case. 

Wonderful. I've added the file as test case, and can close the bug. 
The version of the validator with the fix will go in beta test within days.

Thank you
olivier