Where is the HTTP Sparql endpoint configured?

[Tom Carroll]

Is this deployed by default? Where is the URL for this configured?

Thanks!

[Jordi Albornoz Mulligan]

Hi Tom,

On 4/27/2011 1:11 PM, Tom Carroll wrote:
> Is this deployed by default?

Yes. The URL by default is:
http://<hostname>/sparql

> Where is the URL for this configured?

If you'd like to change this, you can add a file to the 'config'
directory called sparql.properties with the following contents:
service.pid=org.openanzo.servlet.Sparql
org.openanzo.services.enabled=true
org.openanzo.servlet.contextPath=/sparql
org.openanzo.servlet.protectedPathSpec=/*
org.openanzo.servlet.authorizationType=BASIC
org.ops4j.pax.web.instanceId=osgi:org.ops4j.pax.web.instance/default

The org.openanzo.servlet.contextPath is the property that sets the URL
of the SPARQL endpoint.

Just remember that if you've previously started your server, a change to
the configuration files won't take effect unless you wipe the OSGi
configuration cache directories. Specifically those are found at:
openanzo-3.1.0\configuration\.settings
openanzo-3.1.0\configuration\org.eclipse.core.runtime
openanzo-3.1.0\configuration\org.eclipse.equinox.app
openanzo-3.1.0\configuration\org.eclipse.osgi

Just stop the server, delete all of those directories, then start the
server and your change should work.

--
Jordi Albornoz Mulligan
Founding Engineer - Cambridge Semantics
jo...@cambridgesemantics.com
(617) 401-7321

[Tom Carroll]

Thanks Jordi! 

Are there any special considerations to using this Glitter API? I have been trying some simple GETs like:

http://<hostname>/sparql?query=<urlencoded sparql query>

just by pasting that into the browser URL bar, and entering username/password in the browser's basic auth dialog.

When I have a syntactically bad query, I get nice error messages from Glitter, but when I have a query that seems correct, I get back an empty document. Am I missing something? Is there any documentation on this API? I found http://www.openanzo.org/projects/openanzo/wiki/SPARQLExtensions, but can't seem to find something about the basic use of the API.

[Lee Feigenbaum]

Hi Tom,

It is a standard implementation of the SPARQL Protocol. One thing that
might be causing confusion (hard to say without seeing the queries) is
that if no RDF dataset is specified, Glitter queries against an empty
dataset (and hence returns no results). This is different from the
behavior of some other SPARQL engines, which by default query against
all graphs in the system.

You can specify the graphs to query against using the FROM and FROM
NAMED clauses in your query, or by using the default-graph-uri and
named-graph-uri query string parameters.

You can also query against all the graphs in the system by using the
"magic" graph URIs:

http://openanzo.org/namedGraphs/reserved/namedGraphs/ALL
=> all named graphs
http://openanzo.org/namedGraphs/reserved/metadataGraphs/ALL
=> all metadata graphs
http://openanzo.org/namedGraphs/reserved/graphs/ALL
=> all named and metadata graphs

best,
Lee

> --
> You received this message because you are currently subscribed to the
> "OpenAnzo" group.
>
> To post to this group, send email to open...@googlegroups.com
> To unsubscribe from this group, send email to
> openanzo-u...@googlegroups.com
>
> For more options, visit this group at
> http://groups.google.com/group/openanzo?hl=en

--
Lee Feigenbaum
VP Technology & Client Services, Cambridge Semantics Inc
l...@cambridgesemantics.com
1-617-553-1060