Add support for parsing N-Quads #47
Conversation
Exposes Oxigraph parsers (oxigraph#45)
tests/data/test.nq
Outdated
| <http://example.org/stuff/1.0/document> <http://xmlns.com/foaf/0.1/creator> <http://example.org/stuff/1.0/creator> <http://example.com/graph1> . | ||
| <http://example.org/stuff/1.0/creator> <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> <http://xmlns.com/foaf/0.1/Person> <http://example.com/graph1> . | ||
| <http://example.org/stuff/1.0/creator> <http://xmlns.com/foaf/0.1/name> "John Doe" <http://example.com/graph1> . | ||
| <http://example.org/stuff/1.0/creator> <http://xmlns.com/foaf/0.1/mbox> <mailto:[email protected]> <http://example.com/graph1> . |
There was a problem hiding this comment.
Would be great to have a few quads in the default graph to check that they are properly loaded into the default graph
| <http://example.org/stuff/1.0/document> <http://xmlns.com/foaf/0.1/title> "Example Document" <http://example.com/graph2> . | ||
| <http://example.org/stuff/1.0/creator> <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> <http://xmlns.com/foaf/0.1/Person> <http://example.com/graph3> . | ||
| <http://example.org/stuff/1.0/creator> <http://xmlns.com/foaf/0.1/name> "John Doe" <urn:x-rdflib:default> . | ||
| <http://example.org/stuff/1.0/creator> <http://xmlns.com/foaf/0.1/mbox> <mailto:[email protected]> <urn:x-rdflib:default> . |
There was a problem hiding this comment.
It would be great to test with the actual NQuads default graph like <http://example.org/stuff/1.0/creator> <http://xmlns.com/foaf/0.1/mbox> <mailto:[email protected]> ..
I believe proper support for the default graph would require moving the quads in a named graph (likely urn:x-rdflib:default). Sadly pyoxigraph 0.3.x does not support that but pyoxigraph 0.4 alphas does.
oxrdflib/parser.py
Outdated
| @@ -47,6 +47,7 @@ def parse( | |||
|
|
|||
| else: | |||
| base_iri = sink.absolutize(source.getPublicId() or source.getSystemId() or "") | |||
| graph_identifier = to_ox(sink.identifier) if format not in ("ox-nq", "ox-nquads") else None | |||
There was a problem hiding this comment.
We should map the default graph for dataset formats too.
| graph_identifier = to_ox(sink.identifier) if format not in ("ox-nq", "ox-nquads") else None | |
| graph_identifier = to_ox(sink.identifier) |
|
@nikokaoja I have merged pyoxigraph 0.4 support in the main branch. Do you want to rebase your MR? If not, may I build on top of it? |
|
@Tpt great ! Will do update of this PR by end of this week. |
|
@nikokaoja Amazing! I already implemented serializers in #51 and found a few ways to simplify the code a bit. |
|
@Tpt still same issue when rebasing main, |
|
Thank you for pushing this! I wrote an untested draft that should properly insert in the correct graph and supports other store: #53 Feel free to reuse the code. I also found a nice way to automate tests for serializers, it might be handy for the parsers too |
@Tpt still the same issue, |
This is weird, I have updated #53 with some tests that are passing. I also fixed the loading into not-Oxigraph stores that seems now to work. Edit: tests are passing with Python 3.12 but not with Python 3.8. It sounds like something sensitive to the Python version. |
I am running it with Python 3.11 and have the above reported behaviour. So it seems it works only with 3.12 |
|
I understood the error, it's because of changes in rdflib 7.1 that fixes the support of default graph. I am afraid it will be hard to make things work with older versions. I have gated the tests in my PR on v7.1 and also fixed a small bug in OxigraphStore in #54 that was making iteration on the dataset fail. If it's fine with you, I can maybe just merge my MR. WDYT? |
Sure, np , remove my PR then. |
|
This PR has been abandoned |
No description provided.