W3C home > Mailing lists > Public > public-xmlsec@w3.org > August 2009

Re: C14N 2.0 initial draft

From: <pratik.datta@oracle.com>
Date: Tue, 11 Aug 2009 00:32:35 -0700
Message-ID: <4A811E93.2080607@oracle.com>
To: Frederick Hirsch <frederick.hirsch@nokia.com>
CC: XMLSec WG Public List <public-xmlsec@w3.org>
I have incorporated most of these comments, except these
3) I retained the term "Canoncical form". Because "Canonical document" 
implies that the whole document is canonicalized.
9) I have left "xmlIdAncestors" as is  - I think it is useful to support 
C14n 1.0 as is, we can always remove it later on.
10) processing text outside the document root is required when we are 
asked to process the whole document.  In a DOM, the Document element has 
one Element child, and this Element child can have many Text and Comment 
siblings. So the regular tree walk will encounter all these text and 
comment nodes that are outside the document root.

I have also made these additional changes
a) Section 1.4.4 ,  added another requirement "Simplicity" as one of the 
goal for C14n 2.0
b) Section 1.4.3 added this line - This specification adds three 
features to improve robustness - a)remove leading and trailing 
whitespace from text nodes, b) allowing for qnames in content especially 
in the xsi:type attribute, c)rewriting prefixes
c) Section 1.4.1 added ths line - Also this algorithm visits each node 
exactly once, and it only visits the nodes that are being canonicalized.
d) Section 2.1  added this line - (Note: this model purposely does not 
support reinclusion, i.e. all the exclusions are applied after all the 
inclusions. So this is not like the XPath Filter 2 model where there is 
an ordered list of union, intesect and subtract operations)
e) Section 2.4 added new section "Processing for Streaming XML Parsers"
f) added References section

On 7/6/2009 4:24 PM, Frederick Hirsch wrote:
> Some initial quick questions/comments:
> 1) abstract, 1st para, does
>  "It also folds the 2.0 version of Exclusive Canonicalization into 
> same document"
> mean
> "It also incorporates an update to Exclusive Canonicalization, 
> effectively a 2.0 version,  as well."
> 2)abstract, last sentence
> did this mean 2.0?
> "Canonical XML Version 2.0 is applicable to XML 1.0. It is not defined 
> for XML 1.1."
> 3) Terminology
> would it be clearer to refer to "Canonical document" rather than 
> "Canonical form" to avoid confusion with "Canonical XML"?
> 4) Section 1.4
> add to 1st sentence, "with Canonical XML 1.0 and 1.1."
> 5) Section 1.4.1 change "However the C14N 1.x algorithms are only slow 
> if one follows the algorithm exactly as in the 1.x spec. "  to read 
> "However the C14N 1.x algorithms are only slow if one follows the 
> algorithm exactly as written in the 1.x spec without any attempt at 
> optimization."
> 6) Section 1.4.1, 2nd para
> change "So this specification restriction the input of the 
> canonicalization algorithm, so" to
> "This specification restricts the input of the canonicalization 
> algorithm so"
> 7) Section 2.1 Note: This input model...
> "of a the" to "of the"
> "all the known use cases" to "the essential use cases"
> 8) 2.2 table, entry trimTextNodes
> " then text nodes descendants of that element are not trimmed." to
> " then text nodes descendants of that element are not trimmed 
> regardless of the value of this parameter."
> 9) 2.2
> do we want "xmlIdAncestors"? Wasn't the behaviour a bug in C14N10? I 
> assume the goal is to allow a named parameter set for c14n10, but 
> given 1.1 not sure we want to support that.
> 10) 2.3.2, processTextNode
> how could you process a text node outside the document root (or 
> subtree root) in this processing model?
> 11) I assume at some needed material from C14N11 will be added to this 
> document
> e.g. join-URI-References description.
> Thanks for putting this draft together.
> regards, Frederick
> Frederick Hirsch
> Nokia
> On Jul 3, 2009, at 4:12 AM, ext pratik.datta@oracle.com wrote:
>> http://www.w3.org/2008/xmlsec/Drafts/c14n-20/Overview.html
Received on Tuesday, 11 August 2009 07:33:24 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:43:59 GMT