You can start the tests by running
mvn verify
in the project directory.
The TEAM engine writes a couple of XML files that are not really human
readable. A report plugin converts these files into HTML pages. This is
done during the "site" phase:
mvn site
target/site/teamengine-report.html is the entry point for the TEAM
engine reports with links to details of every single test.
Currently 11 of 50 tests fail -- mostly related to URN problems.
Jesper
--
Dipl.-Inf. Jesper Zedlitz Dept. for Computer Science, CAU of Kiel
Room 1108 Communication Systems Research Group
Phone: +49-(0)431-880-7279
Christian-Albrechts-Platz 4 Fax: +49-(0)431-880-7615
24098 Kiel - Germany j...@informatik.uni-kiel.de
> What IDE are you using for coding ?
>
I am using Eclipse. There is a Maven plug-in for Eclipse
http://m2eclipse.codehaus.org/ With that plug-in I think you do not need
a separate Maven installation.
> If you could set up instructions of how to build the code etc..
>
Here are step-by-step instructions for the command line (you can skip
the first two steps if you already have Maven installed on your system):
1. Download Maven from http://maven.apache.org/download.html
2. extract the archive file: tar xvfj apache-maven-2.1.0-bin.tar.bz2
3. check out the source code from the Subversion repository:
svn co
https://oostethys.googlecode.com/svn/trunk/component/server/java/oostethys-java-server
4. change into project's directory cd oostethys-java-server
5. build the project: ../apache-maven-2.1.0/bin/mvn compile
Instead of "mvn compile" you can use "mvn jetty:run" to start a test
server of "mvn test" to run the unit tests, etc.
OGC document 07-092r1 "Definition identifier URNs in OGC namespace"
specifies the form URNs shall have:
urn:ogc:def:objectType:authority:version:code
However, for procedures we use URNs like
uri:mbari:org:mooring:m2
urn:mbari:org:device:1455
I think these are valid replacements:
urn:ogc:object:procedure:mbari.org:device:1455
urn:ogc:object:procedure:mbari.org:mooring:m2
This would fix TEAM test getCapabilities:core-SOS.GetCapabilities-ResponseContentsValidProcedure.1
The TEAM test
getCapabilities:core-SOS.GetCapabilities-ResponseContentsValidObservedProperty.1
finds fault with these URNs:
urn:ogc:phenomenon:time:iso8601
urn:ogc:phenomenon:latitude:wgs84
urn:ogc:phenomenon:longitude:wgs84
http://mmisw.org/ont/cf/parameter/depth
http://mmisw.org/ont/cf/parameter/sea_water_temperature
http://mmisw.org/ont/cf/parameter/sea_water_electrical_conductivity
http://mmisw.org/ont/cf/parameter/sea_water_salinity
I don't think either of the 'valid replacements' are valid unless they
are formally issued in some sense, but I am not clear on that subject.
John
John
--------------
John Graybeal <mailto:gray...@mbari.org> -- 831-775-1956
Monterey Bay Aquarium Research Institute
Marine Metadata Interoperability Project: http://marinemetadata.org
First of all, there are indeed some improper URNs that were
discovered, and bonus points to the team for writing something that
would detect those!
I may have lost the appropriate context here, so apologies if there
have been some decisions which I missed, or if the context for this
test is more local than all of OOSTethys. Some questions invited by
this test are:
1) Has it been decided only URNs are acceptable? (See Luis' note
below.) Note I can come up with a good 'valid HTTP' test pattern if
that helps; it was just cited in the semweb list.
2) Has it been decided only OGC URNs are acceptable? (There are many
non-OGC URNs that would fail that are perfectly valid URNs.)
3) As it stands this is a minimum test of validity. Would it be
possible to add testing against
(a) only the truly valid OGC URN 'categories'? (only about 6 or 8
URN patterns have been 'allocated' by OGC, each of them with a
different prefix -- is there any reason not to test against those
particular prefixes)
(b) only truly valid OGC URNs? (for most of those categories, there
is a vocabulary that can then be appended to the prefix to create the
full URN, and for the remainder there are rules; it would be
compelling to have a test for inclusion in those lists)
I recognize 3b is seriously out of scope of the testing mechanism;
Luis' suggestion for checking against a registry is perhaps also out
of scope, but it is a nice improvement on the problem. Of course, OGC
doesn't put its URNs into a registry, at least yet, so that won't work
yet either. And some URNs (including some from OGC) can only be
tested by rule, because they include variable values (like lat/lon
locations).
It will be nice to be really clear about what we are and are not
allowing, and testing against, so developers and others can stay
conforming.
John
On Jul 2, 2009, at 4:47 AM, Luis Bermudez wrote:
> Interesting. I think this "rule" should be relaxed to allow any URL.
> Or much better have a configuration to check if the URI really exists.
> For example pointing to a registry of terminologies.
>
> -luis
On Jul 1, 2009, at 4:10 AM, Jesper Zedlitz wrote:
> Jesper Zedlitz wrote:
>> Currently 11 of 50 tests fail -- mostly related to URN problems.
>>
> Here are some details about the URN problems:
>
> OGC document 07-092r1 "Definition identifier URNs in OGC namespace"
> specifies the form URNs shall have:
> urn:ogc:def:objectType:authority:version:code
>
> However, for procedures we use URNs like
> uri:mbari:org:mooring:m2
> urn:mbari:org:device:1455
>
> I think these are valid replacements:
> urn:ogc:object:procedure:mbari.org:device:1455
> urn:ogc:object:procedure:mbari.org:mooring:m2
> This would fix TEAM test getCapabilities:core-SOS.GetCapabilities-
> ResponseContentsValidProcedure.1
>
> The TEAM test
> getCapabilities:core-SOS.GetCapabilities-
> ResponseContentsValidObservedProperty.1
> finds fault with these URNs:
> urn:ogc:phenomenon:time:iso8601
> urn:ogc:phenomenon:latitude:wgs84
> urn:ogc:phenomenon:longitude:wgs84
> http://mmisw.org/ont/cf/parameter/depth
> http://mmisw.org/ont/cf/parameter/sea_water_temperature
> http://mmisw.org/ont/cf/parameter/sea_water_electrical_conductivity
> http://mmisw.org/ont/cf/parameter/sea_water_salinity
>
> Jesper
>
> --
> Dipl.-Inf. Jesper Zedlitz Dept. for Computer Science, CAU of
> Kiel
> Room 1108 Communication Systems Research Group
> Phone: +49-(0)431-880-7279
> Christian-Albrechts-Platz 4 Fax: +49-(0)431-880-7615
> 24098 Kiel - Germany j...@informatik.uni-kiel.de
> What IDE are you using for coding ?
>
I am using Eclipse. There is a Maven plug-in for Eclipse
http://m2eclipse.codehaus.org/ With that plug-in I think you do not need
a separate Maven installation.
> If you could set up instructions of how to build the code etc..
>
Here are step-by-step instructions for the command line (you can skip
the first two steps if you already have Maven installed on your system):
1. Download Maven from http://maven.apache.org/download.html
2. extract the archive file: tar xvfj apache-maven-2.1.0-bin.tar.bz2
3. check out the source code from the Subversion repository:
svn co
https://oostethys.googlecode.com/svn/trunk/component/server/java/oostethys-java-server
4. change into project's directory cd oostethys-java-server
5. build the project: ../apache-maven-2.1.0/bin/mvn compile
Instead of "mvn compile" you can use "mvn jetty:run" to start a test
server of "mvn test" to run the unit tests, etc.