- From: <paul.downey@bt.com>
- Date: Fri, 28 Apr 2006 13:10:05 +0100
- To: <gcowe@origoservices.com>, <public-xsd-databinding@w3.org>
Thanks George, I think it's good (from at least a versioning perspective) that these tools are able to cope with more elements being sent, but possibly not so good that tools may allow an application to send more items than constrained by the schema used to generate / map the data. So based upon this evidence, I propose we cite finite maxOccurs values as being one of our basic patterns I also propose adding a design consideration to flag that tools are at liberty (quite likelto represent the repeated item without necessarily presenting the finite restriction or limitation. Also, it may be possible, in some tools, to sucessfully send or receive more items than defined in the schema. I note that nobody seems to have questioned the value of maxOccurs="0" :-) Paul -----Original Message----- From: public-xsd-databinding-request@w3.org on behalf of George Cowe Sent: Fri 4/28/2006 8:59 AM To: public-xsd-databinding@w3.org Subject: Re: ISSUE-32: element repetitions other than \"0\", \"1\" and \"unbounded\" I have done some testing for the element repetitions other than 0, 1 and unbounded. The tools I tested this against were 1. Apache Axis 1.3 2. IBM Rational Application Developer v6 3. Microsoft .Net 2.0 My schema snippet <xsd:sequence> <xsd:element ref="ce:valuation_currency" minOccurs="0"/> <xsd:element ref="ce:valuation_request" minOccurs="0" maxOccurs="2"/> </xsd:sequence> All of the tools I tested against ignored the fact that only a maximum of 2 <ce:valuation_request> elements were required and basically assumed a value of "unbounded". George
Received on Friday, 28 April 2006 12:11:12 UTC