Trig serialiser writing empty named graph name for default graph

[dschallis]

When parsing then serialising TriG, the serialiser fails to correctly detect when it's dealing with the default graph, so attempts to serialise it as a named graph (which ends up with 'None' instead of a URI or qname).

E.g.:

import rdflib                                                                   

data = """                                                                      
@prefix : <http://example.com/> .                                               

<:a> <:b> <:c> .                                                                
"""                                                                             

g = rdflib.ConjunctiveGraph()                                                   
g.parse(data=data, format='trig')                                               
print g.serialize(format='trig')

Produces the output:

@prefix : <http://example.com/> .
@prefix ns1: <:> .
@prefix rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#> .
@prefix rdfs: <http://www.w3.org/2000/01/rdf-schema#> .
@prefix xml: <http://www.w3.org/XML/1998/namespace> .
@prefix xsd: <http://www.w3.org/2001/XMLSchema#> .

<None> {
    ns1:a ns1:b ns1:c .
}

whereas for latter part of the output, it should produce either:

ns1:a ns1:b ns1:c .

or

{ ns1:a ns1:b ns1:c . }

I've added a unit test for this to #431 .

[niklasl]

This has been partially fixed. The erroneous graph identifier syntax was fixed in bd57db7 and f167ecf.

The expected result can be achieved by parsing datasets into the default_context of the conjunctive graph (or dataset), like:

g = rdflib.ConjunctiveGraph()
g.default_context.parse(data=data, format='trig')
print g.serialize(format='trig')

Should we close this (and leave the remaining details to the eventual improvements outlined in #436)?

[gromgull]

Should we close this (and leave the remaining details to the eventual improvements outlined in #436)?

yes :)