- From: Sanjiva Weerawarana <sanjiva@watson.ibm.com>
- Date: Fri, 19 Nov 2004 19:33:35 +0600
- To: "Asir Vedamuthu" <asirv@webmethods.com>, "'Roberto Chinnici'" <Roberto.Chinnici@Sun.COM>, <www-ws-desc@w3.org>
Hi Asir, IIRC we rejected your item (2) below: the faults have to be *bound* only once per interface. However, we wanted to allow people to specific SOAP modules, features and properties for faults on a per-operation basis. Sanjiva. ----- Original Message ----- From: "Asir Vedamuthu" <asirv@webmethods.com> To: "'Roberto Chinnici'" <Roberto.Chinnici@Sun.COM>; <www-ws-desc@w3.org> Sent: Thursday, November 18, 2004 9:27 PM Subject: RE: Adding binding infault/outfault components (LC55) > > I read your proposal and have two questions, > > [1] it appears to me that binding fault reference and binding message > reference look alike to some extent. However, their component structures are > very different. And, their mappings from xml infoset are different too. Why > are these different? BTW, I like your proposed component structure. Would > you recommend upgrading binding message reference component structure to > > {message reference} REQUIRED - A Message Reference component > {features} OPTIONAL - A set of Feature components > {properties} OPTIONAL - A set of Property components > > [2] in light of your proposal, is it worth retaining Binding Fault > component? "wsoap:code", "wsoap:subcodes" and "whttp:code" and their > corresponding properties can be attached to the proposed Binding Fault > Reference component. > > Regards, > Asir S Vedamuthu > asirv at webmethods dot com > http://www.webmethods.com/ > > -----Original Message----- > From: www-ws-desc-request@w3.org [mailto:www-ws-desc-request@w3.org] On > Behalf Of Roberto Chinnici > Sent: Tuesday, November 16, 2004 9:47 PM > To: www-ws-desc@w3.org > Subject: Adding binding infault/outfault components (LC55) > > > > I had an action item to to write up the addition of infault and outfault > at the binding level plus modifications of the component model. (LC55) > > Define a Binding Fault Reference component with the following properties: > > {fault reference} REQUIRED - A Fault Reference component. > {features} OPTIONAL - A set of Feature components. > {properties} OPTIONAL - A set of Property components. > > The pseudo-schema for a binding operation would be updated to look like > this: > > <operation ref="xs:QName" > > <documentation />? > > <input messageLabel="xs:NCName"? > > <documentation />? > > <feature ... />* > > <property ... />* > </input>* > > <output messageLabel="xs:NCName"? > > <documentation />? > > <feature ... />* > > <property ... />* > </output>* > > <infault ref="xs:QName" messageLabel="xs:NCName"?> > <documentation />? > > <feature ... />* > > <property ... />* > </infault>* > > <outfault ref="xs:QName" messageLabel="xs:NCName"?> > <documentation />? > > <feature ... />* > > <property ... />* > </outfault>* > > <feature ... />* > > <property ... />* > </operation>* > > The mapping of a binding infault > > <infault ref="xs:QName" messageLabel="xs:NCName"?> > <documentation />? > > <feature ... />* > > <property ... />* > </infault>* > > to a Binding Fault Reference component BFR would be as follows: > > (the notation C.{P} denotes property {P} of component C) > > 1. start with the Binding Operation BO; > 2. BO.{operation reference} is an Interface Operation component I; > 3. I.{fault references} is a set of Fault Reference components; > 4. the value of BFR.{fault reference} is the unique element FR of I.{fault > references} such that > a. FR.{fault reference}.{name} == the value of the @ref attribute of > wsdl:infault > b. FR.{message label} == the value of the @message label of > wsdl:infault (*) > c. FR.{direction} == 'in' > > (*) For consistency with the mapping rules for the Fault Reference > component, the @message > attribute is optional provided that there is only one message in the MEP > used by I whose > corresponding fault has the 'in' direction (of course, taking the fault rule > used by the > MEP into account). > > Similarly for a binding outfault, with 'out' in place of 'in'. > > In part 3, we'd extend the pseudo-schema so as to allow wsoap:module inside > the binding infault/outfault elements: > > <operation ref="xs:QName" > whttp:location="xs:anyURI"?? > whttp:transferCodingDefault="xs:string"?? > > wsoap:mep="xs:anyURI"? > wsoap:action="xs:anyURI"? > > <documentation />? > > <wsoap:module ... />* > > <input messageLabel="xs:NCName"? > whttp:transferCoding="xs:string"?? > > <documentation />? > <wsoap:module ... />* > <feature ... />* > <property ... />* > </input>* > > <output messageLabel="xs:NCName"? > whttp:transferCoding="xs:string"?? > > <documentation />? > <wsoap:module ... />* > <feature ... />* > <property ... />* > </output>* > > <infault ref="xs:QName" messageLabel="xs:NCName"? > whttp:transferCoding="xs:string"?? > > <documentation />? > <wsoap:module ... />* > <feature ... />* > <property ... />* > </infault>* > > <outfault ref="xs:QName" messageLabel="xs:NCName"? > whttp:transferCoding="xs:string"?? > > <documentation />? > <wsoap:module ... />* > <feature ... />* > <property ... />* > </outfault>* > > <feature ... />* > <property ... />* > </operation>* > > Section 2.6.2 would be amended so that the {soap modules} property becomes > applicable to Binding Fault Reference components. > > Roberto > > -- > Roberto Chinnici > Java Web Services > Sun Microsystems, Inc. > roberto.chinnici@sun.com
Received on Friday, 19 November 2004 19:48:09 UTC