rdflib.term.fromn3 method

[ghost]

gromgull, 2012-01-25T10:46:53.000Z

RDFLib Terms have a .n3() method that return the n3 representation of the term.

For round-tripping, it would be nice to have a method that does the opposite, i.e. something that creates a term object from the n3 representation.

Comment 1 by [email protected]

see:

from rdflib.util import from_n3

Comment 2 by joernhees

thanks. would still be cool to have that in term.
even cooler would be if the n3 parser used it in a similar way as the n3 serializer uses .n3()

[gweis]

As far as I can tell, the TSV Parser could use it as well.
However from_n3 doesn't to much validation. I guess it would be good to add validation to from_n3, so that it could be re-used in all parsers.

[gromgull]

I would think it should be possible to have a common code base for parsing rdf1.1 ntriples (utf8!) and turtle/n3 nodes.

[joernhees]

related: #502, #503

as mentioned i also think it would be useful to only have one place where we parse n3... #503 fixes util.from_n3()'s parsing of literals with datatypes, and allows parsing PrefixedNames by introducing a NamespaceManager.

Still, as util.from_n3() is a "single term parser", it for example doesn't really seem to do well with parsing quoted graphs.

I think some of the points to decide here are:

• should we unify the parsing of n3?
  • should we use parser/n3.py code from util.from_n3() or
  • vice versa?
• should we be able to parse all rdflib.terms from n3, e.g. Variable, QuotedGraph?