mhgrove / Empire

JPA implementation for RDF
http://groups.google.com/group/empire-rdf/
Apache License 2.0
98 stars 34 forks source link

lmplement CriteriaBuilder #62

Open Riduidel opened 13 years ago

Riduidel commented 13 years ago

With Empire-RDF, there will be people adressing RDF triple database without the SPARQL knowledge. A good programmatic way to avoid the mental overhead of one more query language could be to implement in the EntityManagerImpl the support for the CriteraBuilder.

For more informations, one can take a look at

pratikpparikh commented 13 years ago

I agree with Riduidel as well. I just came across empire and I like the api but for me it is critical that Criteria Builder is part of it.

mhgrove commented 13 years ago

Empire does not use JPA2, which CriteriaBuilder is a part of. That's something I would like to support in the future, but I don't have a timetable for this, and it's probably not as easy as just dropping in a new jar.

pratikpparikh commented 13 years ago

I agree it is probably not as easy. If you can put together a design then i can help you implement it? Unfortunately implementation wise i am new to rdf and sparql world, but overall i am pretty sound.

Regards, Pratik Parikh On Thu, Nov 10, 2011 at 7:52 AM, Michael Grove < reply@reply.github.com>wrote:

Empire does not use JPA2, which CriteriaBuilder is a part of. That's something I would like to support in the future, but I don't have a timetable for this, and it's probably not as easy as just dropping in a new jar.


Reply to this email directly or view it on GitHub: https://github.com/mhgrove/Empire/issues/62#issuecomment-2694828

Pratik Parikh

mhgrove commented 13 years ago

I think you'd have to implement Criteria/CriteriaBuilder/CriteriaQuery over top of something like a query builder using some of the reflect stuff to look at the bean you are building the criteria on top of to figure out what properties to use and what not. The core of what's supported there maps to SPARQL well enough. There will be a few things that won't bin nicely into SPARQL1.0, such as in(Collection) that would have to be punted on for an initial implementation. That will be easy enough to support in SPARQL 1.1, but I don't have an timeline for that yet.

pratikpparikh commented 13 years ago

i will get back to you after analyzing the current implementation.

On Thu, Nov 10, 2011 at 9:35 AM, Michael Grove < reply@reply.github.com>wrote:

I think you'd have to implement Criteria/CriteriaBuilder/CriteriaQuery over top of something like a query builder using some of the reflect stuff to look at the bean you are building the criteria on top of to figure out what properties to use and what not. The core of what's supported there maps to SPARQL well enough. There will be a few things that won't bin nicely into SPARQL1.0, such as in(Collection) that would have to be punted on for an initial implementation. That will be easy enough to support in SPARQL 1.1, but I don't have an timeline for that yet.


Reply to this email directly or view it on GitHub: https://github.com/mhgrove/Empire/issues/62#issuecomment-2695776

Pratik Parikh