W3C home > Mailing lists > Public > public-xmlsec-maintwg@w3.org > July 2007

ACTION-62

From: Juan Carlos Cruellas <cruellas@ac.upc.edu>
Date: Tue, 17 Jul 2007 10:39:23 +0200
Message-ID: <469C803B.4030704@ac.upc.edu>
To: XMLSec <public-xmlsec-maintwg@w3.org>

Dear all,

This message tries to close action 62.

As you will have seen each testcase is described in a tabular form. Each 
table contains three columns.

First column contains the description of the test case. The second one 
contains the rationale for the test.
The third will contain, in its final version a link to a web page that 
could contain:

1. Details on the input document to be signed. I think that this is 
important for the two first test cases groups:
those ones dealing with inheritance of xml namespace attributes by 
canonicalization and those ones dealing with
implicit/explicit rules for canonicalization. I would say that all the 
participant should sign
the same xml documents, once agreed.

2. The XPath transformation that would oblige in these first two test 
cases group to get a part of the original
document.

3. Links to the signatures generated by each participant on the document.

A page of this type could look something like shown below:


Test case canXML11.xmllang#1-positive

Signed document:

<CanXML11Xmllang1Positive xmlns="http://www.ietf.org" 
xmlns:w3c="http://www.w3.org">

   <e1 xml:lang="ES">
      <e11 xmlns="" >
         <e111 />
      </e11>
      <e12 >
         <e121 />
      </e12>
   </e1>
   <e2 xmlns="">
      <e21 />
   </e2>
   
</CanXML11Xmllang1Positive>


XPath filter:

<!-- Here the text of the XPath transform used in a filter that selects 
e2 element. Also indication
of the namespace declarations used in the Xpath transform -->

Signatures:

<!-- And below a set of links to the signatures correctly
generated. Proposal for link names: TestcaseCode"_"ParticipantCode --->

canXML11.xmllang#1-positive_PARTICIPANT-1

canXML11.xmllang#1-positive_PARTICIPANT-2



I would say that in order to progress, if the test cases defined in the 
document that I produced are accepted, we should start writing both 
simple xml documents like the one above (taken from Canonical XML 1.1 
with some modifications) and the corresponding XPath expressions that 
make Xpath filtering to result in the document subset expected.


For xml:lang attribute, I would propose the former one

<CanXML11Xmllang1Positive xmlns="http://www.ietf.org" 
xmlns:w3c="http://www.w3.org">

   <e1 xml:lang="ES">
      <e11 xmlns="" >
         <e111 />
      </e11>
      <e12 >
         <e121 />
      </e12>
   </e1>
   <e2 xmlns="">
      <e21 />
   </e2>
   
</CanXML11Xmllang1Positive>

I think that we could define quite straightforwardly similar docs for 
xml:space, xml:id and xml:base.

Regards

Juan Carlos.




Regards

Juan Carlos.
Received on Tuesday, 17 July 2007 08:39:33 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 8 January 2008 14:22:00 GMT