cocoon-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Kirk Woerner" <k...@stoneseeker.com>
Subject RE: [C1][XInclude][PATCH] - XInclude DOM2 problem
Date Wed, 15 Nov 2000 05:40:48 GMT
um, I THINK that's what I'm saying.  It does seem a bit preposterous doesn't
it?  But it works (which is something:)  I also got an email from someone
else who had the same problem and same solution.  I'm not sure whether the
namespace info is lost.  I just know that using those NS functions don't
work and checking the attributes using getAttribute does work.  I actually
did look about in the Xalan code for some indication of how elements were
created but got confused and frustrated.  I understand the cocoon structure
pretty well from playing around in it, but not Xalan.

I actually didn't check the code out from CVS, just hacked the source that
came with 1.8.  I do have a CVS client though so I'll see what I can do...

>
>On Tue, 14 Nov 2000, Kirk Woerner wrote:
>
>> This is a little ugly, but it works and I think it's valid.  The bug
>> appears when attempting to process xinclude elements that were generated
>> from a stylesheet.  The DOM nodes generated from within xsl are
>apparently
>> not DOM II so the checks for "getAttributeNS()" in scanForXInclude() do
>> not work when they should.
>
>woah - just so i'm sure we're on the same level here - you're saying that
>when a page has been generated by the XSLT processor, the namespace URI
>information is lost, so the scanForXInclude method fails because it can't
>find the attributes href and parse in the xinclude namespace? and your
>suggested workaround is to look for attributes named xinclude:href
>instead? hmm. i don't like it, but i can't see any alternative - but it
>strikes me as _very_ strange that DOMs generated by Xalan aren't DOM2
>compliant. Scott, any other xalan people, can y'all verify that this is or
>is not the case? we're talking xalan-1.2 here i reckon.
>
>oh, as far as lack of diff goes - your cvs client should be able to
>produce a unified diff. please check and see - retyping patches by hand is
>kinda frustrating.
>
>- donald
>


Mime
View raw message