Sep 4, 2017: Our paper on RDFIO and rdf2smw was just published! If you use rdf2smw in scientific work, please cite:
Lampa S, Willighagen E, Kohonen P, King A, Vrandečić D, Grafström R, Spjuth O
RDFIO: extending Semantic MediaWiki for interoperable biomedical data management
Journal of Biomedical Semantics. 8:35 (2017). DOI: 10.1186/s13326-017-0136-y.

Sep 30, 2016: rdf2smw was covered in a talk at SMWCon in Frankfurt, Sep 2016. See: Talk page, Slides, Video.

A commandline tool to convert from RDF triples to Semantic MediaWiki facts in MediaWiki XML export format to be used with MediaWiki's built-in XML import feature.

This allows you to quickly and simply populate a Semantic MediaWiki page structure, from an RDF data file.

It is written in Go for better performance than PHP. The latest version processes triples into pages in the order of ~55K triples/sec converted into ~13K pages/sec on an 2014 i5 Haswell dual core processor, to give an idea.

rdf2smw is very similar to the RDF import function in the RDFIO Semantic MediaWiki extension, but takes another approach: Whereas RDFIO converts RDF to wiki pages and imports them in the same go, rdf2smw first converts RDF to an XML file outside of PHP (for better performance), and then importing using MediaWiki's built-in import function.

Status: The tool is pretty much feature complete, including ability to write facts via template calls if a categorization (via owl:Class or rdf:type) of the subject can be done. What is lacking is more options to fine-tune things. Right now you'll have to modify the source code yourself if you need any customization. Hope to address this in the near future.

The tool itself does not have any dependencies, apart from a unix-like operating system. For importing the generated XML dump file to make sense though, you will need a web server, PHP, MediaWiki and Semantic MediaWiki.

An automated virtualbox generation script (so valled "vagrant box"), with all of this, plus the RDFIO extension, can be found here, and is highly recommended, if you don't have a MediaWiki / SemanticMediawiki installation already!

For linux 64 bit:

1. Download the file rdf2smw_linux64.gz, on the latest release.
2. Unpack it with: gunzip rdf2smw_linux64.gz
3. Call it, on the commandline (see the usage section below).

Call the rdf2smw binary, specifying a file with triples in n-triples or turtle format, with the --in flag, and an output file in XML format with the --out flag, like so:

./rdf2smw --in triples.nt --out semantic_mediawiki_pages.xml

In addition to the specified output file, there will be separate files for templates and properties, named similar to the main output file, but replacing .xml with _templates.xml and _properties.xml respectively.

These XML files can then be imported into MediaWiki / Semantic MediaWiki, via the importDump.php maintenance script, located in the maintenance folder under the main mediawiki folder.

php <wikidir>/maintenance/importDump.php semantic_mediawiki_pages_templates.xml
php <wikidir>/maintenance/importDump.php semantic_mediawiki_pages_properties.xml
php <wikidir>/maintenance/importDump.php semantic_mediawiki_pages.xml

Note that the order above is highly recommended (templates, then properties, then the rest), so as to avoid unnecessary re-computing of semantic data after the import is done.

Find below a schematic illustration of the flow-based programming process graph of the rdf2smw program:

Illustration created with drawfbp

Only N-triples is supported as input format right now. We plan to add more formats shortly.

rdf2smw is based on the FlowBase flow-based programming micro-framework.

rdf2smw makes heavy use of Petter Goksøyr Åsen's awesome RDF parsing library.