swap-scala inference, Jena and testing

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
2 messages Options
Reply | Threaded
Open this post in threaded view
|

swap-scala inference, Jena and testing

Danny Ayers
Hi Dan,

Great to see your recent blog posts about this project [1] - although
I'm reasonably familiar with Python, in the past I've found the
original swap code a wee bit opaque to grok what's going on.

Anyhow, just a thought - for testing e.g. RDFS inference it should be
straightforward to strap in Jena libs (and associated tests) to your
Scala code - Jena has pretty sophisticated & mature inference
capabilities [2].

I found it fairly trivial to reuse Jena libs in Scala, as in [3]
(using a TDB store, heavy use of SPARQL - alas my code is still mostly
Java written in Scala syntax rather than using Scala idioms - lot of
learning to do :)

Incidentally, not sure what you'll make of map/territory in Jena
(seems a little blurred, but works ok pragmatically - hmm, surely when
represented in code everything is "map" anyway..?):
[[
public interface Resource
extends RDFNode
...
RDFNode -
Interface covering RDF resources and literals. Allows probing whether
a node is a literal/[blank, URI]resource, moving nodes from model to
model, and viewing them as different Java types using the .as()
polymorphism.
]]

Cheers,
Danny.

[1] http://code.google.com/p/swap-scala/
[2] http://jena.sourceforge.net/inference/
[3] http://www.knowledgeforge.net/project/gradino/

--
http://danny.ayers.name

Reply | Threaded
Open this post in threaded view
|

Re: swap-scala inference, Jena and testing

Dan Connolly
On Wed, 2010-04-28 at 10:07 +0200, Danny Ayers wrote:

> Hi Dan,
>
> Great to see your recent blog posts about this project [1] - although
> I'm reasonably familiar with Python, in the past I've found the
> original swap code a wee bit opaque to grok what's going on.
>
> Anyhow, just a thought - for testing e.g. RDFS inference it should be
> straightforward to strap in Jena libs (and associated tests) to your
> Scala code - Jena has pretty sophisticated & mature inference
> capabilities [2].

Jena probably has everything that swap-scala has and more.
But the whole point of the swap-scala exercise is to think things thru
by implementing them, and strapping on Jena would defeat
that point.

> Incidentally, not sure what you'll make of map/territory in Jena
> (seems a little blurred,

yes, it's typical of the blurred APIs.

>  but works ok pragmatically

well, sort of... but I think it's a source of quite
a bit of confusion.

>  - hmm, surely when
> represented in code everything is "map" anyway..?):
> [[
> public interface Resource
> extends RDFNode
> ...
> RDFNode -
> Interface covering RDF resources and literals. Allows probing whether
> a node is a literal/[blank, URI]resource, moving nodes from model to
> model, and viewing them as different Java types using the .as()
> polymorphism.
> ]]
>
> Cheers,
> Danny.
>
> [1] http://code.google.com/p/swap-scala/
> [2] http://jena.sourceforge.net/inference/
> [3] http://www.knowledgeforge.net/project/gradino/
>


--
Dan Connolly, W3C http://www.w3.org/People/Connolly/
gpg D3C2 887B 0F92 6005 C541  0875 0F91 96DE 6E52 C29E