W3C home > Mailing lists > Public > public-xml-processing-model-wg@w3.org > March 2007

Re: New review draft

From: Innovimax SARL <innovimax@gmail.com>
Date: Wed, 28 Mar 2007 16:03:59 +0200
Message-ID: <546c6c1c0703280703i4c0832c9o6f8ae55b3e51a404@mail.gmail.com>
To: "Norman Walsh" <Norman.Walsh@sun.com>
Cc: public-xml-processing-model-wg@w3.org

Dear,


Ok some remarks on components

== join sequences ==

<p:join-sequences>
  <p:input A/>
  <p:input B/>
  <p:input C/>
  <p:input D/>
</p:join-sequences>

equivalent to

<p:identity>
  <p:input name="source">
    A
    B
    C
    D
  </p:input>

A,B,C,D being p:document, p:pipe or p:inline

?
== p:load ==
1) "The load component has no inputs but takes a parameter that..."
In this case, it is a p:option, do we have a consistent story to
denominate them literally ?

2) I must admit that in that case, it seems odd to have p:option after p:output

== p:subsequences ==
s/It applies and XPath/It applies an XPath/

1) can you please tell what is the XPath expression for selecting head
(the first of the sequence) and tail (from second to the last) ?

== p:xinclude ==
It reminds me that we probably can add a p:map element to store
key/value pair, in that case key should be URI and value can be
p:document, p:pipe or p:inline. It won't solve the caching problem but
give some way to solve cases

== p:xslt ==
the sequence of document is a nice tool, but it is a sequence. It
means that the start of the second in the sequence is only available
when the end of the first has been reached.
It seems clear that the order of the documents is undefined

In the case of XSLT document it means that we have access to the rest
of the sequence iff :
* the precedent outputs are well formed

This means that if one output is not well formed, the other cannot be accessible

Which is not very suitable if we need to have access to small
independent output of the XSLT transformation

In this respect, i propose instead of
<p:output port="result" sequence="yes"/>
to have
<p:output port="result" sequence="no"/>
<p:output port="secondary" sequence="yes" />

== http request ==
Please replace "p:" by "c:" in the following

[[
For example, a form post would be formulated as:

<p:http-request method="post" href="http://www.example.com/form-action">
<p:entity content-type="application/x-www-form-urlencoded">
<p:body>
name=W3C&spec=XProc
</p:body>
</p:entity>
</p:http-request>

and if the response was an XHTML document, the response would be:

<p:http-response status="200">
<p:entity content-type="application/xhtml+xml">
<p:body>
<html xmlns="http://www.w3.org/1999/xhtml">
<head><title>OK</title></head>
<body><p>OK!</p></body>
</html>
</p:body>
</p:entity>
</p:http-response>
]]
==Relax NG Validate==
It means only the XML version of Relax'NG is usable


== XSLT 2 ==
i propose also instead of
<p:output port="result" sequence="yes"/>
to have
<p:output port="result" sequence="yes"/>
<p:output port="secondary" sequence="yes" />

== XQuery 1.0 ==
It means only the XQueryX is usable

Mohamed

-- 
Innovimax SARL
Consulting, Training & XML Development
9, impasse des Orteaux
75020 Paris
Tel : +33 8 72 475787
Fax : +33 1 4356 1746
http://www.innovimax.fr
RCS Paris 488.018.631
SARL au capital de 10.000 
Received on Wednesday, 28 March 2007 14:04:03 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 8 January 2008 14:21:50 GMT