W3C home > Mailing lists > Public > public-css-testsuite@w3.org > April 2008

RE: Not compliant XHTML MS tests.

From: Arron Eicholz <Arron.Eicholz@microsoft.com>
Date: Fri, 4 Apr 2008 08:55:59 -0700
To: fantasai <fantasai.lists@inkedblade.net>
CC: Robert Stam <robert@tallcomponents.com>, "public-css-testsuite@w3.org" <public-css-testsuite@w3.org>
Message-ID: <66D5CB5D6AB0694592FAF5487C50368B191517E0D4@NA-EXMSG-C111.redmond.corp.microsoft.com>

fantasai wrote:

> As Anne points out, all of these tests need to validate as XHTML.
> If they are testing invalid CSS, then they shouldn't pass the
> *CSS* validator, but they do need to validate as XHTML. Robert's
> suggetion of using a CDATA section would be the easiest way to
> fix the tests where CSS data is tripping up the XML parser.

I agree that cases should validate but there are some exceptions. When the case is flagged HTMLonly it doesn't necessarily have to validate as XHTML does it? Those cases might not validate for XHTML. There could be all sorts of deprecated attributes and tags in there that might not allow XHTML to validate but would be perfectly fine in HTML.

Chapter_4\bad-selector.htm: Corrected with a CDATA
Chapter_4\invalid-decl-at-rule-001.htm: corrected with CDATA
Chapter_4\escaped-ident-001.htm: Corrected to use dash and underscore
Chapter_4\escaped-ident-spaces-001.htm: corrected to use dash and underscore
Chapter_16\white-space-009.htm: attribute corrected to be valid HTML but case is flagged as HTMLonly.

Side question... If an 'id' is invalid (i.e. "B&W?") should you still be able to match the invalid 'id' using CSS? 'id' attributes can't contain '&' or '?' so I am thinking there should be no way to match in this case since the 'id' is invalid but I'm want to confirm this.

--
Thanks,
Arron Eicholz

> -----Original Message-----
> From: fantasai [mailto:fantasai.lists@inkedblade.net]
> Sent: Thursday, April 03, 2008 5:11 PM
> To: Arron Eicholz
> Cc: Robert Stam; public-css-testsuite@w3.org
> Subject: Re: Not compliant XHTML MS tests.
>
> Arron Eicholz wrote:
> > Robert Stam wrote:
> >> Chapter_4\bad-selector.htm: do not include & in css style, because
> it will be
> >> an incorrect entity, use CDATA section.
> >> Chapter_4\invalid-decl-at-rule-001.htm: do not include & in css
> style, because
> >> it will be an incorrect entity, use CDATA section.
> >> Chapter_4\escaped-ident-001.htm: '&' character should be escaped in
> id.
> >> Chapter_4\escaped-ident-spaces-001.htm: '&' character should be
> escaped in id.
> >> Chapter_12\before-location-001.htm: incorrect '<' character.
> >> Chapter_14\Properties\Background-repeat\background-repeat-005.htm:
> </p> missing
> >> Chapter_16\white-space-009.htm: [nowrap] attribute may not be placed
> in XHTML.
> >> Chapter_16\white-space-011.htm: [wrap] attribute may not be placed
> in XHTML.
> >
> > Chapter_4\bad-selector.htm: is flagged as invalid and does not need
> to validate
> > Chapter_4\invalid-decl-at-rule-001.htm: is flagged as invalid and
> does not need to validate
> > Chapter_4\escaped-ident-001.htm: This case now flagged as invalid
> > Chapter_4\escaped-ident-spaces-001.htm: This case now flagged as
> invalid
> > Chapter_12\before-location-001.htm: '<' character is now properly
> changed to &lt;
> > Chapter_14\Properties\Background-repeat\background-repeat-005.htm:
> <p> tag is now closed properly
> > Chapter_16\white-space-009.htm: attribute corrected to be valid HTML
> but case is flagged as HTMLonly.
> > Chapter_16\white-space-011.htm: case correct to be valid. Attribute
> removed.
> >
> > These changes will be in our next release of test files. Thank you
> for submitting these issues.
>
> Hi Arron,
> As Anne points out, all of these tests need to validate as XHTML.
> If they are testing invalid CSS, then they shouldn't pass the
> *CSS* validator, but they do need to validate as XHTML. Robert's
> suggetion of using a CDATA section would be the easiest way to
> fix the tests where CSS data is tripping up the XML parser.
>
> Thanks~
>
> ~fantasai
Received on Friday, 4 April 2008 15:56:47 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 20 September 2010 17:51:56 GMT