W3C home > Mailing lists > Public > public-rdf-wg@w3.org > March 2013

Re: N-Triples test suite

From: Gregg Kellogg <gregg@greggkellogg.net>
Date: Wed, 13 Mar 2013 08:07:06 -0700
Cc: public-rdf-wg@w3.org
Message-Id: <77FF332F-8733-4A86-A196-FCEFF5C88F80@greggkellogg.net>
To: Andy Seaborne <andy.seaborne@epimorphics.com>
Thanks, these are all oversights, which I'll correct.

Gregg

On Mar 13, 2013, at 3:40 AM, Andy Seaborne <andy.seaborne@epimorphics.com> wrote:

> 
> 
> On 13/03/13 00:28, Gregg Kellogg wrote:
>> Spurred on by a suggestion from Gavin, I've created a test suite for
>> N-Triples. This is mostly based on the Turtle syntax tests and the
>> original RDF Tests version from Dave Beckett. I've added some
>> additional negative syntax tests, but I'm sure there could be more. I
>> didn't think that evaluation tests were particularly important, but
>> we could have some to verify the equivalence of UTF-8 escaped and
>> un-escaped.
>> 
>> The format is the same as for Turtle, and the Manifest is in Turtle,
>> but it uses rdft:TestNTriplesPositiveSyntax and
>> rdft:TestNTriplesNegativeSyntax types.
> 
> As the format is so regular, I don't think it is much to worry about because you could use perl to generate the information about each test without needing to understand the RDF that the Turtle serializes.
> 
>> I didn't look at the Turtle "coverage" suite, and there are probably
>> more specific things that can be added based on that. as well as more
>> things that are legal in Turtle, but illegal in N-Triples.
>> 
>> Tests are at
>> <https://dvcs.w3.org/hg/rdf/raw-file/default/rdf-turtle/tests-nt/manifest.ttl>.
>> 
>> Gregg Kellogg gregg@greggkellogg.net
>> 
>> 
> 
> Gregg,
> 
> Thanks for doing this.
> 
> nt-syntax-bad-num-02.nt
> nt-syntax-bad-num-03.nt
> 
> The files are empty - zero bytes
> hg sync issue?
> 
> These tests
> 
> nt-syntax-datatypes-01
> nt-syntax-datatypes-02
> 
> have relative URIs.  Could we make them absolute?
> 
> It does not affect the intent of the test and many N-Triples parsers aren't going to do IRI resolution for location base.
> 
> Spec comment - I think the spec should say that in the syntax, unlike Turtle, all IRIREFs MUST be absolute, i.e. no resolution required even against the location base.
> 
> 	Andy
> 
Received on Wednesday, 13 March 2013 15:07:42 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 17:04:26 UTC