Program for reading and writing linked data in various formats.
To install,
npm install -g rabel
Commands look like unix options are executed in order from left to right. They include:
-base=rrrr Set the current base URI (relative URI, default is file:///$PWD)
-clear Clear the current store
-dump Serialize the current store in current content type
-format=cccc Set the current content-type
-help This message
-in=uri Load a web resource or file
-out=filename Output in the current content type
-report=file set the report file destination for future validation
-size Give the current store
-spray=base Write out linked data to lots of different linked files CAREFUL!
-test=manifest Run tests as described in the test manifest
-validate=shapeFile Run a SHACL validator on the data loaded by previous in=x
-version Give the version of this program
Formats cccc are given as MIME types. These can be used for input or output:
- text/turtle (default)
- application/rdf+xml
whereas these can only input:
- application/rdfa
- application/xml
rabel -format=application/xml -in=foo.xml -format=text/turtle -out=foo.ttl
rabel part*.ttl -out=whole.ttl
Currently rabel can read from the web or files, and write only to files. Filenames are deemed to be relative URIs just taken relative to file:///{pwd}/ where {pwd} is the current working directory.
One use case is testing all the parsers. Another is providing a stable serialization. The output serialization is designed to be stable under small changes of the the data, to allow data files to be checked into source code control systems.
The name comes from RDF and Babel.
When loading XML, elements are mapped to arcs, and text content to trimmed RDF strings. For the XML namespace used for IANA registry documents, custom mapping is done, both of properties and datatypes, and local identifier generation. (See the source for details!)