W3C home > Mailing lists > Public > xml-dist-app@w3.org > December 2001

RE : Issue 171. Revised text to close issue

From: Murali Janakiraman <murali@roguewave.com>
Date: Thu, 13 Dec 2001 17:08:24 -0800
Message-ID: <F888C30C3021D411B9DA00B0D0209BE8014F44FC@cvo-exchange.roguewave.com>
To: "'xml-dist-app@w3.org'" <xml-dist-app@w3.org>

 I took an action item to provide a text to close issue 171. The initial
text provided for closing this issue went for some discussions in the WG and
I was asked to provide a revised text that clarifies the use of xsi:type in
addition to addressing issue 171.

Here is a quick reference to issue 171.

<issue_171>
 An example for illustration of the problem:
	<a xsi:type="string" id="1"    >3</a>
	<b xsi:type="int"    href="#1"/>
 The problem is: what is the type of b?
</issue_171>

Revised text is below. Thanks to Nova for his clarifications, suggestions
and text that was provided in a private conversation.

<Proposal>

The serialization rules of SOAP permit the use of xsi:type attribute to
represent the type of an element, and hence also the type of the
corresponding graph node.  The use of xsi:type to denote element types is
described in XML Schema Structures [1]. Although SOAP does not mandate the
use of any
particular schema notation, all SOAP processors MUST recognize xsi:type
references if present and MUST associate the corresponding type with the
deserialized nodes.

The use of xsi:type attribute on referencing accessors (I.e. those that
use the href="..." construction) is subtle and it is recommended
they be used with caution. Note that, per the rules of the XML Schema
Specification [1] for xsi:type, the existence of xsi:type attribute on
referencing accessors does not indicate the type of the referred element,
and hence does not affect the type of the corresponding node."

</Proposal>


[1]   http://www.w3.org/TR/xmlschema-1/
Received on Thursday, 13 December 2001 20:09:02 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 5 February 2014 22:28:13 UTC