OpenSAML user discussion

[Brent Putman <>]

This change has just been made on the java-xmltooling project trunk.

The Signer signObject and signObjects methods now throw an 
org.opensaml.xml.signature.SignatureException.  

Any calling code using the new code will have to be updated to reflect 
the new API.  Our java-opensaml2 project (and Shibboleth 2.0 java-idp 
project) have been updated appropriately

Thanks,
Brent



Chad La Joie wrote:
> Yes, I agree, go ahead and make the change.
>
> Brent Putman wrote:
> > Hmm, yeah, IMHO, I think this is a bug and it should probably throw 
> > an exception (probably of a new type, b/c we don't have an 
> > immediately obvious appropriate one).  I'll confer with Chad, b/c 
> > this will change the API substantially and calling code (ours and 
> > other people's) would have to change to handle.  But that seems the 
> > right thing to do.
> >
> >
> >
> >  wrote:
> > > Hi -    I was running some tests and noticed that my SAML output was 
> > > incorrect.  I was able to resolve my issue, but was curious as to 
> > > why code execution continued in a error state.  Checking the logs, I 
> > > saw an error regarding the signature process in 
> > > org.opensaml.xml.signature.Signer.signObject.  From looking at the 
> > > source, I noticed this code just writes the error to the log instead 
> > > of throwing an exception to the caller:
> > >
> > > lines 77-79 of org.opensaml.xml.signature.Signer.signObject:
> > >         } catch (XMLSecurityException e) {
> > >             log.error("An error occured computing the digital 
> > > signature", e);
> > >         }
> > >
> > > Is this behavior intentional?  It seems a bit odd to not alert the 
> > > caller of a critical error...
> > >
> > > Thanks!
> > > Dave
> > >