RE: Fixes to the test-suite

Sorry, the last change should have been CR138 instead.

The real change for CR143 is:
- change transferCoding to contentEncoding in the interchange format schemas
- similarly change the testcases (MessageTest-1G, 3G, 4G)
- similarly change the message-level assertion stylesheets
- update wsdl-xslt accordingly
- change MessageTest-1G so that it doesn't rely on gzip encoding support
(EchoString operation now has contentEncoding="".

Canon (and any implementer without gzip support) will show two failures in
MessageTest-4G and four failures in MessageTest-3G.

Jonathan Marsh - http://www.wso2.com - http://auburnmarshes.spaces.live.com
 

> -----Original Message-----
> From: Jonathan Marsh [mailto:jonathan@wso2.com]
> Sent: Thursday, February 08, 2007 4:34 PM
> To: 'Jonathan Marsh'; 'www-ws-desc'
> Subject: RE: Fixes to the test-suite
> 
> More changes for CR143.  I added a messageContentModel child element to
> the
> interfaceFaultComponent in the interchange format, fixed up wsdl-xslt to
> emit it properly, and refreshed the baselines accordingly.
> 
> Jonathan Marsh - http://www.wso2.com -
> http://auburnmarshes.spaces.live.com
> 
> 
> > -----Original Message-----
> > From: www-ws-desc-request@w3.org [mailto:www-ws-desc-request@w3.org] On
> > Behalf Of Jonathan Marsh
> > Sent: Thursday, February 08, 2007 3:50 PM
> > To: 'www-ws-desc'
> > Subject: RE: Fixes to the test-suite
> >
> >
> > I've done some changes re: CR117.
> >
> > I added two new testcases, LocationTemplate-2G and LocationTemplate-3G,
> > testing the encoding (or not) of location templates.  I split these up
> > because the rawest of the raw cases, namely replacing a template with a
> > full
> > uri, aren't possible to implement server-side by either WSO2 or Canon.
> > Even
> > if they were, the test framework is insufficient to check that the
> values
> > are correct.
> >
> > Therefore I've added only LocationTemplate-2G to the list of message
> > tests.
> > I suspect the one use of raw in that test may also need to be moved to
> > LocationTemplate-3G, but we'll see.
> >
> > I also updated the expected values for LocationTemplate-1G to match the
> > CR117 resolution.  Since all expected values are hand-calculated, it's
> > entirely possible you will discover some mistakes.
> >
> > Jonathan Marsh - http://www.wso2.com -
> > http://auburnmarshes.spaces.live.com
> >
> >
> > > -----Original Message-----
> > > From: www-ws-desc-request@w3.org [mailto:www-ws-desc-request@w3.org]
> On
> > > Behalf Of Jonathan Marsh
> > > Sent: Thursday, February 08, 2007 12:24 PM
> > > To: 'Youenn Fablet'; 'keith godwin chapman'
> > > Cc: 'www-ws-desc'
> > > Subject: RE: Fixes to the test-suite
> > >
> > >
> > > Done as indicated.
> > >
> > > More changes to follow shortly.
> > >
> > > Jonathan Marsh - http://www.wso2.com -
> > > http://auburnmarshes.spaces.live.com
> > >
> > > > -----Original Message-----
> > > > From: Youenn Fablet [mailto:youenn.fablet@crf.canon.fr]
> > > > Sent: Friday, February 02, 2007 5:22 AM
> > > > To: Jonathan Marsh; 'keith godwin chapman'
> > > > Cc: www-ws-desc
> > > > Subject: Fixes to the test-suite
> > > >
> > > >     Dear all,
> > > >
> > > > Here are some bugs that may need to be fixed in the test-suite wrt
> to
> > > > message exchange tests:
> > > >     1) MessageTest-2G: ignoreUncited=true is used for EchoName but
> the
> > > > MiddleName parameter is not optional in the schema and not present
> in
> > > > the location property. Making it optional in the schema may be a
> > simple
> > > > fix.
> > >
> > > MiddleNames are now optional.
> > >
> > > >     2) MessageTest-3G: #any is used for the Send function which is
> > bound
> > > > to a GET request. POST or PUT would be better suited.
> > >
> > > I moved whttp:method="GET" to GuaranteedFault instead.  Send will
> > default
> > > to
> > > PUT, without losing the only GET in the testcase.
> > >
> > > >     3) LocationTemplate-1G: ignoreUncited=true is used in the
> Escaping
> > > > binding. This binding does not allow the serialization of a
> mandatory
> > > > 'time' parameter. We could either set ignoreUncited to false or
> change
> > > > the dateSpace element definition so that it is nillable or
> > minOccurs=0.
> > > > These fixes should improve the exchange results, at least ours...
> > >
> > > I think I'll change the location to include a time, because the
> > > minOccurs=0
> > > maxOccurs=2 has already proved instructive.
> > >
> > > > Jonathan, if that is ok with you, can you update the test-suite?
> > > > Keith, can you send messages to our MessageTest-2G endpoint?
> > > >
> > > > Thanks,
> > > >     Youenn
> >

Received on Friday, 9 February 2007 02:49:00 UTC