OpenSAML user discussion

[Chad La Joie <>]

Yeah, sorry.  Looks like I introduced a bug.  I moved the directory 
containing the configuration files up one level, out of the test 
directory, so that it was clearer (I hope) that these are useful default 
config files, however when I did that I forgot to change the path in the 
 test case.  I just checked in the fix.  Take a look at the 
SAMLObjectTestCaseConfigInitializer class and you'll see all I did was 
remove the "/conf" part of the path.

Manuela Stanica wrote:
> Hi Chad,
>
> I followed closely the instructions, but I get the error: Unable to 
> configure OpenSAML: org.opensaml.xml.parse.XMLParserException: Unable to 
> read XML source.
> I remember that when I tried yesterday, there was another error message 
> related to the saml2md prefix, saying something like the prefix was not 
> declared. But today it seems that that xml document can't be read at 
> all. Has anything changed, do you know what could be the cause?
> Thank you.
>
> Manuela
>
> Chad La Joie schrieb:
> > This is now documented.
> >
> > https://authdev.it.ohio-state.edu/twiki/bin/view/Shibboleth/OSTwoUsrManJavaLibIntro 
> >
> >
> > Chad La Joie wrote:
> > > The developers will need to load the configuration files  within 
> > > their application but the configuration files themselves will be 
> > > provided.
> > >
> > > Take a look at 
> > > org.opensaml.common.SAMLObjectTestCaseConfigInitializer in the test 
> > > source.  You can see the static initializer (lines 60 - 88) how the 
> > > test cases loaded the configuration.  This is another area that I 
> > > have to do the documentation on which I'll be doing on Friday.
> > >
> > > Manuela Stanica wrote:
> > > > Chad La Joie schrieb:
> > > > > Make sure you have the latest XMLTooling and OpenSAML code.  My 
> > > > > guess is that your XMLTooling code is out of date.
> > > > Thank you, that did help :). Indeed I had forgotten to check out the 
> > > > XMLTooling code, as it compiled well and I supposed that it wouldn't 
> > > > be modified.
> > > > Now I no longer get any errors with the openSAML code either.
> > > >
> > > > As far as my implementation code is concerned, it is equivalent to 
> > > > the example that appears in the openSAML 2 user's manual. on the 
> > > > wiki page. However, where it still gets stuck is at the 
> > > > unmarshallerFactory.getUnmarshaller(metadataRoot) method, as it 
> > > > eventually tries to fetch the appropriate unmarshaller from the 
> > > > 'unmarshallers' Map, which is empty...
> > > > Is this map going to be populated with the unmarshaller entities 
> > > > corresponding to SAML 2 metadata types? Or are the developers 
> > > > supposed to do this manually by means of the 
> > > > registerUnmarshaller(QName key, Unmarshaller unmarshaller) method in 
> > > > UnmarshallerFactory? (this would seem like a lot of work...)
> > > >
> > > > After having manually registered the EntityDescriptorUnmarshaller in 
> > > > this way, the unmarshall() method encountered the same problem while 
> > > > trying to find the EntityDescriptorBuilder from the 'builders' Map, 
> > > > which is empty as well. I also inserted code to register this 
> > > > builder, but then of course the same happens while trying to 
> > > > unmarshall all the child elements of the given document.
> > > > Should all these unmarshallers and builders really be registered 
> > > > manually or are they to be found in the corresponding maps at some 
> > > > point?
> > > >
> > > > Thanks a lot for your help.
> > > >
> > > > (And thank you too Tom, I will take a look at the GridShib code you 
> > > > mentioned to see if there is anything that could apply to the 
> > > > project I work on (eduGAIN AAI infrastructure in GEANT2)).
> > > >
> > > > Manuela

--
Chad La Joie             2052-C Harris Bldg
OIS-Middleware           202.687.0124