Re: Issue 11 in sadi: Registry error message is vague when input/output OWL class definitions can't be parsed

1 view
Skip to first unread message

sa...@googlecode.com

unread,
Dec 19, 2011, 11:57:35 AM12/19/11
to bu...@sadiframework.org
Updates:
Summary: Registry error message is vague when input/output OWL class
definitions can't be parsed
Owner: elmccar...@gmail.com
Labels: -Product-JavaAPI Product-Registry

Comment #2 on issue 11 by elmccar...@gmail.com: Registry error message is
vague when input/output OWL class definitions can't be parsed
http://code.google.com/p/sadi/issues/detail?id=11

It looks to me like the registry is sending a proper Accept header (which
really just means that the Jena library is sending a proper Accept header,
which I would hope would be the case…) The header is:

Accept: application/rdf+xml, application/xml; q=0.8, text/xml; q=0.7,
application/rss+xml; q=0.3, */*; q=0.2

…so there's no way the service should be sending N3.

Further, when I send this Accept header to a SADI service built with the
sadi-0.1.4-py2.6.egg from the Downloads tab above, I get the expected
response. Even further, I am able to register the service.

So the problem must be with the service you are trying to register. When I
send this header to your service, I get the expected response, but some
digging reveals that it is the URL for the output class that is not
respecting the Accept header (i.e.:
http://purl.org/twc/vocab/datafaqs#EvaluatedDataset is not available as
XML).

I'm surprised that doesn't result in a more explicit error message from the
registry (other problems with the input/output classes have better
messages) and I will fix that.

sa...@googlecode.com

unread,
Dec 19, 2011, 12:45:09 PM12/19/11
to bu...@sadiframework.org

Comment #3 on issue 11 by tim...@gmail.com: Registry error message is vague
when input/output OWL class definitions can't be parsed
http://code.google.com/p/sadi/issues/detail?id=11

elmccarthy,

Thanks for looking into this and finding a lead.

I'll fix the http://purl.org/twc/vocab/datafaqs#EvaluatedDataset Accept
issue.

Do you plan to clarify the Registry's error message, so that it indicates a
problem with the in/out class URIs?

Thanks,
Tim Lebo

sa...@googlecode.com

unread,
Dec 19, 2011, 12:56:14 PM12/19/11
to bu...@sadiframework.org

Comment #4 on issue 11 by tim...@gmail.com: Registry error message is vague
when input/output OWL class definitions can't be parsed
http://code.google.com/p/sadi/issues/detail?id=11

Although http://purl.org/twc/vocab/datafaqs#EvaluatedDataset now resolves
to RDF/XML, the Registry still reports the error:

com.hp.hpl.jena.shared.JenaException: org.xml.sax.SAXParseException: White
spaces are required between publicId and systemId.

Perhaps Jena isn't following the redirects...

(I'll hack the I/O class URIs on another service to check that)

bash-3.2$ curl -LI http://purl.org/twc/vocab/datafaqs#EvaluatedDataset
HTTP/1.1 302 Moved Temporarily
Date: Mon, 19 Dec 2011 17:52:12 GMT
Server: 1060 NetKernel v3.3 - Powered by Jetty
Location:
https://raw.github.com/timrdf/DataFAQs/master/ontology/datafaqs.ttl.owl
Content-Type: text/html; charset=iso-8859-1
X-Purl: 2.0; http://localhost:8080
Expires: Thu, 01 Jan 1970 00:00:00 GMT
Content-Length: 304

HTTP/1.1 200 OK
Server: nginx/1.0.4
Date: Mon, 19 Dec 2011 17:52:12 GMT
Content-Type: text/plain; charset=utf-8
Connection: keep-alive
Status: 200 OK
X-RateLimit-Limit: 100
X-Frame-Options: deny
X-RateLimit-Remaining: 100
X-Runtime: 11
Content-Length: 1
Cache-Control: no-cache

-Tim

Luke McCarthy

unread,
Dec 19, 2011, 1:19:34 PM12/19/11
to sadi...@googlegroups.com
Hi Tim,

Google's just decided to sign me out of my account and I can't seem to sign back in from the train that I'm on, so I'm replying here in the hopes that you'll notice.

Jena does follow redirects (seriously, why the disrespect for Jena? It's a mature library and not, as far as I can tell, written by idiots), but it's not following yours for some reason. It's actually not Jena, but Java in general that's having problems with your redirect, as the following attached code and output indicate (I've included a working redirect, too…)

Cheers,

Luke

RedirectTest.java
RedirectTest.output

sa...@googlecode.com

unread,
Dec 19, 2011, 2:22:17 PM12/19/11
to bu...@sadiframework.org

Comment #5 on issue 11 by elmccar...@gmail.com: Registry error message is
vague when input/output OWL class definitions can't be parsed
http://code.google.com/p/sadi/issues/detail?id=11

In reply to comment #3, I do plan to clarify the registry's error message,
as I already said in comment #2 (read all the way to the end…)

sa...@googlecode.com

unread,
Dec 19, 2011, 2:26:19 PM12/19/11
to bu...@sadiframework.org

Comment #6 on issue 11 by elmccar...@gmail.com: Registry error message is
vague when input/output OWL class definitions can't be parsed
http://code.google.com/p/sadi/issues/detail?id=11

Copied from sadi-dev list for completeness now that I'm able to sign into
Google again.

Jena does follow redirects (seriously, why the disrespect for Jena? It's a
mature library and not, as far as I can tell, written by idiots), but it's
not following yours for some reason. It's actually not Jena, but Java in
general that's having problems with your redirect, as the attached code and
output indicate (I've included a working redirect, too…)

Attachments:
RedirectTest.java 945 bytes
RedirectTest.output 1.7 KB

sa...@googlecode.com

unread,
Dec 19, 2011, 6:55:50 PM12/19/11
to bu...@sadiframework.org
Updates:
Status: Fixed

Comment #7 on issue 11 by elmccar...@gmail.com: Registry error message is

vague when input/output OWL class definitions can't be parsed
http://code.google.com/p/sadi/issues/detail?id=11

The error message from the registry now looks like this:

"error loading ontology for
http://purl.org/twc/vocab/datafaqs#EvaluatedDataset from
http://purl.org/twc/vocab/datafaqs caused by

com.hp.hpl.jena.shared.JenaException: org.xml.sax.SAXParseException: White
spaces are required between publicId and systemId."

I'm closing this issue because the outstanding redirect problem doesn't
seem to be due to anything on our side. Feel free to reopen it or followup
to either of the mailing lists if any new information comes to light.


Reply all
Reply to author
Forward
0 new messages