Re: RDF's relative IRI resolution is ambiguous

Dear all,

Here is a proposed test case for IRI resolution:
https://gist.github.com/RubenVerborgh/39f0e8d63e33e435371a

For a correct Turtle parser, iri-resolution.ttl and iri-resolution.nt
should contain equal graphs.

These tests:
a) assume that base IRIs are not normalized
b) are based on the RFC3986 tests, with two changes:
– I included several variations on the base IRI
– the base IRI includes path parts with multiple characters,
   to also capture off-by-one string offset bugs that could occur

Feedback is most welcome;
I still plan to include more variations
(paths with mixed ../ and ./ sequences).

Best,

Ruben