rdf3x

RDF-3X
(c) 2008 Thomas Neumann. Web site: http://www.mpi-inf.mpg.de/~neumann/rdf3x

This work is licensed under the Creative Commons
Attribution-Noncommercial-Share Alike 3.0 Unported License. To view a copy
of this license, visit http://creativecommons.org/licenses/by-nc-sa/3.0/
or send a letter to Creative Commons, 171 Second Street, Suite 300,
San Francisco, California, 94105, USA.


Building:
---------

RDF-3X must be build using GNU make (a Windows version is available from the
MinGW project) and a reasonable C++ compiler. Ideally a simple

   make

is enough, it will build the tree high-level executables (see below) in bin/.
To use a different C++ compiler, e.g. the Visual Studio compiler, use

   make CXX=cl

The program is reasonably portable, but a 64 bit operating system (and
compiler!) is highly recommended. RDF-3X will fail silently if data plus
intermediate results are larger than the virtual address space.

Using:
------

RDF-3X currently includes three high-level executables. The first (rdf3xload)
is used to build a new database from an turtle/ntriples input:

   rdf3xload db mydata.n3

The input file can be arbitrarily large, the rdf3xload spools to disk if
main memory is exhausted.
Note: The loader is currently hard-coded to use 1GB of memory. This can be
reduced by tweaking tools/rdf3xload/Sorter.cpp.

After loading the database can be queried with rdf3xquery:

   rdf3xquery db

The program shows a command prompt and accept SPARQL queries, one query per
line. Note: RDF-3X currently only supports "select" queries and relatively
simple FILTER predicates.
rdf3xquery accepts a file name as second parameter that can be used to
interpret a file with a single query.

The final program, rdf3xdump is mainly interesting for debugging purposes,
it dumps the database as turtle file:

   rdf3xdump db > mydump.n3