W3C home > Mailing lists > Public > www-xml-xinclude-comments@w3.org > July 2005

Re: element() scheme should return Element Node only or any kind of node?

From: Bill Han <bill.han@oracle.com>
Date: Thu, 21 Jul 2005 15:52:23 -0700
Message-ID: <42E02727.3090900@oracle.com>
To: Elliotte Harold <elharo@metalab.unc.edu>
CC: Paul Grosso <pgrosso@arbortext.com>, www-xml-xinclude-comments@w3.org, Oleg Tkachenko <oleg@tkachenko.com>, Jonathan Marsh <jmarsh@microsoft.com>

So, your metafallbacktestwithxpointer.xml test case is okay, but 
metafallbacktest5.xml is wrong.
The differnce between these two (or by href attr) is one has 
<xi:fallback>fallback text</xi:fallback>, another one has 
<xi:fallback><test>some data</test></xi:fallback> Your element(/1/1) 
returns "fallback text" string and <test>some data</test> element 
respectively.
I don't know what you meant by "catalog". I downloaded test suite from 
http://www.w3.org/XML/Test/XInclude/ (110032004 version).

Thanks,
- Bill


Elliotte Harold wrote:

> Bill Han wrote:
>
>> Hi,
>>   Question: element() scheme should return
>>    1. Element Node only?
>>    2. Any kind of node?
>>    3. Any kind of node except for whitespace node plus ...?
>
>
> element nodes only
>
>
>>   Backgroud: in Harold test suite, metafallbacktest5.xml, 
>> element(/1/1) returns Text Node ("fallback text" string).
>
>
> I'm not sure it does, Could you please cite the IDs of the tests 
> you're running? The only place I see the file you cite is in harold-26 
> which is an error test:
>
>     <testcase id="harold-26" href="metafallbacktest5.xml"
>                   type="error">
>             <contributor>Elliotte Rusty Harold</contributor>
>             <section 
> resource="http://www.w3.org/TR/xinclude/#fallback">4.4</section>
>       <date qualifier="created">31-Aug-2004</date>
>             <description>The XPointer does not select anything in the 
> acquired infoset, but does select something in the source 
> infoset.</description>   
>         </testcase>
>
> That is, it should not generate any output at all.
>
> It sounds like you're not consulting the catalog. The file names and 
> the presence or absence of particular files in the output directory do 
> not indicate the correct output. The expected output for processing 
> foo.xml is not necessarily foo.xml; it may be bar.xml or nothing at 
> all. Furthermore bar.xml may be the output for foo.xml, bar.xml, and 
> several other tests. The only way to tell is to look in the catalog file.
>
>
Received on Thursday, 21 July 2005 22:52:42 GMT

This archive was generated by hypermail 2.2.0 + w3c-0.30 : Thursday, 21 July 2005 22:52:46 GMT