axis-java-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Jonny Rylands" <>
Subject RE: (soap encoding <-> XSD) mapping
Date Mon, 02 Aug 2004 21:46:36 GMT

Cool, thanks.

Just got a couple more questions if that's OK:

Sorry for my newbie level of understanding (!) but am I right in saying that
the XSD is the language to specify types (in WSDL) whereas SOAP encoding is
the 'language' to specify values (in encoded SOAP messages)?

Hence for a toolkit to fully support XSD when using SOAP encoded services,
you would need a full (Java/C++ etc) implementation XSD simple types and
their facets, plus the complex types allowed in the SOAP encoding spec
(arrays and sequences etc)? 

Jonny Rylands.

-----Original Message-----
From: Jim Murphy [] 
Sent: 02 August 2004 22:19
Subject: Re: (soap encoding <-> XSD) mapping

There is nothing stooping you from adding restriction facets to you WSDL 
- in fact it may be a good idea.  The problem is expecting SOAP toolkits 
to respect them.  This may not be that big a deal though...just don't 
rely on the toolkit providing any of this extra help for you.  If it 
helps minimize the client:faults you return because sombody actually 
took the time to interpret your schema more fully then cool.

Jim Murphy
Mindreef, Inc.

Jonny Rylands wrote:

> Hi
> I'm having great difficulty trying to figure out from the SOAP encoding
> section in the spec whether you're allowed to add constraining facets to
> derived XML schema types in the types section in your WSDL.
> e.g. are you allowed to derive the integer type (in the types section of
> WSDL document) and use any of the following constraining facets:
> totalDigits 
> fractionDigits 
> pattern 
> whiteSpace 
> enumeration 
> maxInclusive 
> maxExclusive 
> minInclusive 
> minExclusive
> Does anyone know the answer to this?
> I understand WSDL is generated for you in Axis, so I imagine if this were
> possible it wouldn't presently be supported in Axis (as there is no
> standardised way to specify the above in Java, without some form of
> metadata/javadoc) -- is this right?
> Many thanks for your time.
> Regards,
> Jonathan Rylands.

View raw message