is there a lightweight sparql server/endpoint i can setup perferably a nice DSL or something.
also looking for a federated search tool
everything i've seen so far is monolithic java stuff. i come from the agile world were all of this stuff seems really enterprisey
asked 11 Oct '10, 10:50
Here are some RDF frameworks with SPARQL support in non-Java languages...
If you're familiar with PHP and MySQL, then ARC2 might be what you're looking for. IIRC it doesn't offer full support for SPARQL 1.0, but offers a pretty good/useful subset, and also a useful subset of SPARQL Query.
If you're familiar with Perl, then the RDF::Query and RDF::Trine packages provide a solid RDF and SPARQL framework; with RDF::Endpoint providing an HTTP-accessible interface. (Most SQL database backends will work, including PostgreSQL, MySQL and SQLite. There are also some good in-memory stores, though these are not persistent!) It supports pretty much all of SPARQL 1.0 I think, and a good helping of 1.1 too.
RDFLib is probably the best RDF framework for Python. Its SPARQL implementation is not especially complete though. cwm is also pretty cool, offering full N3 support including rule-based reasoning. It implements SPARQL, but I don't know how complete its implementation is. It doesn't offer an HTTP-based endpoint certainly.
Redland (librdf) is written in C and includes libraptor (an RDF parser/serialiser library) and linrasqal (an RDF query library). Last time I checked, its SPARQL implementation was not especially good, but I've heard there have been recent major improvements. It has bindings for Perl, PHP, Python, Ruby and (experimentally) Lua.
If, rather than a framework for programming, all you want is a black box app you can throw triples/queries at, then you could try 4store, AllegroGraph or Virtuoso.
answered 11 Oct '10, 11:34
I would say that the easiest store to set up and start with is ARC, which is a PHP based RDF store, which also provides SPARQL Endpoint over the store. Everything you need to do is to extract the installation package somewhere on your Apache server and set up the configuration for how the store will use your MySQL database. Of course, you need to have PHP and MySQL set up. Your SPARQL Endpoint is then available on the path where you've instantiated the store with following lines:
If you are not working with PHP in your project, you can still use this store by sending the SPARQL queries over HTTP directly to the Endpoint.
answered 11 Oct '10, 11:51
There is also Talis platform, that is a cloud-based service and offers nice ways to insert/delete and query data. Also supports new SPARQL features such as COUNT etc. and the support people are great.
A free account may be obtained for academic/public data purposes. http://www.talis.com/platform/developers/contact/
It also manages content negotiation automatically.
answered 15 Oct '10, 16:27