[CITE-Forum] WCS 1.1.1 GetCapabilities: validation issue
Normand Savard
nsavard at mapgears.com
Fri Oct 2 08:47:32 EDT 2009
Jinsongdi Yu wrote:
> Dear Norm,
>
> Thanks for the valuable issues. Recently, I am also investigating into
> WCS coformance testing. I have read your report and check the test
> suite according to your report, and I also want to give some
> comments and questiongs on the issues.
>
>
> Issue8,
> wcs:GetCapabilities_Response_OperationsMetadata_Contents_CoverageSummary_Identifier_XML
> Does not seem to test the XML syntax.
> Issue 9:
> wcs:GetCapabilities_Response_OperationsMetadata_Contents_CoverageSummary_SupportedFormat_
> XML
> Does not seem to test the XML syntax.
> Issue 12:
> wcs:DescribeCoverage_Response_Identifier_XML
> Does not seem to test the XML syntax.
>
>
> do they validate by schema?
> E.g. in
> wcs:GetCapabilities_Response_OperationsMetadata_Contents_CoverageSummary_Identifier_XML
> wcs:get-capabilities-accept-versions is called, and it use the
> following parser to parse the schema.
> <ctlp:XMLValidatingParser>
> <ctlp:schemas>
> <ctlp:schema
> type="resource">xsd/ogc/wcs/1.1.1/wcsGetCapabilities.xsd</ctlp:schema>
> </ctlp:schemas>
> </ctlp:XMLValidatingParser>
> </ctl:request>
Yes you are right on this one, I missed the call to
wcs:get-capabilities-accept-versions.
>
> Issue 5:
> wcs:GetCapabilities_Response_OperationsMetadata_Post_Encoding
> Should be skipped if the server does not support 'POST' request.
>
> Yes, current CTL mix the madatory and optional test together, so do
> we also need something, like core test and extension test. E.g.
> Getcapabiltiy KVP test as core, Getcapabiltiy post test as extension.
>
Makes sense to me.
>
> issue3
> The 'AcceptVersions' is a list of preferred versions. There is no
> concept of lower/higher version in the specification. In fact, the
> server shall return an exception only if it does not support any
> version in the
> list.
>
> Yes, I can also not find the definition of lower/higher version.
>
Jinsongdi,
Thank you for you review. I put some comments inline.
Norm
More information about the CITE-Forum
mailing list