On 25/01/15 01:19, maorellana92 wrote:
Hi,
I need to convert a large rdf file using jena, so I can't load it in the
main memory.
There is a way to read and write each triple directly in other rdf file?
Actually, i only have used FileManager.get().loadModel(String
filename) that return a Model m, then I do m.write(OutputStream out,
String lang) to change the model syntax, but this load the entire file.
I read about StreamRDF, but i have not found a good example. Do you have
any idea?
See RDFDataMgr for using StreamRDF and StreamRDFLib for ways to get a
StreamRDF.
If you just want to change the format of a large file, you can do:
<Jena>/bin/riot --out=LANG File1 File1 ...
(or riot.bat)
where LANG is a streamed output format. There various variants on, say,
Turtle, that are output streaming style. The default is Ntriples/NQuads.
(or use the usually tools and use a large heap)
Andy