valentinedwv / ioostech

Automatically exported from code.google.com/p/ioostech

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

ows:ServiceType definition uncertainty and OGC TEAM issue

GoogleCodeExporter opened this issue · comments

The OGC hosted TEAM Engine fails IOOS SOS capabilities document schema 
validation with the error message "Attribute 'codespace' is not allowed to 
appear in element 'ows:ServiceType'". Because of that, the TEAM Engine stops 
testing, and never proceeds to other parts of the test.

The part of the IOOS SOS Capabilities document that causes the test failure is 
a ServiceType element in Service Identification section of the GetCapabilities 
template:

<ows:ServiceIdentification>
[...]
    <ows:ServiceType codeSpace="http://opengeospatial.net">OGC:SOS</ows:ServiceType>
    <ows:ServiceTypeVersion>1.0.0</ows:ServiceTypeVersion>
[...]
</ows:ServiceIdentification>

The reason for the issue is that <ows:ServiceType> is defined twice: in 
"owsGetCapabilities.xsd" as an <xsd:simpleType>, and in 
"owsServiceIdentification.xsd" as an <xsd:element> that belongs to a 
<ows:DescriptionType>. While the latter allows a complex service type name, 
e.g. "the values of the codeSpace URI and name and code string", the former 
allows just a string value that is the OWS type abbreviation.

Obviously, the OGC TEAM treats the ServiceType as the former, while IOOS 
implements the complex version of it, and this mismatch certainly plugs any OGC 
compliance testing of the IOOS implementation. Since both approaches are 
currently legal, I believe that the stringency of test has to be relaxed. 

Original issue reported on code.google.com by abir...@gmail.com on 26 Jun 2013 at 7:04

[deleted comment]

Original comment by abir...@gmail.com on 26 Jun 2013 at 7:10

  • Added labels: Milestone-Release1.0, Type-Other
Oddly enough, Oxygen indicates the same error, which may mean that 
ows:GetCapabilities schema takes preference over ows:ServiceIdentification one 
in this case. If that is right, it means that OGC is rather innocent, and the 
fault is on our side...

Original comment by abir...@gmail.com on 27 Jun 2013 at 3:11

OK, disregard the previous messages, the reason for the error was a simple 
typo: instead of "codespace" the attribute must be typed in a camelCase as 
"codeSpace". I am closing that issue now.

Original comment by abir...@gmail.com on 27 Jun 2013 at 3:52

Original comment by abir...@gmail.com on 27 Jun 2013 at 3:54

  • Changed state: Invalid