- From: Gregory Williams <greg@evilfunhouse.com>
- Date: Fri, 24 May 2024 15:15:29 -0700
- To: Martynas Jusevičius <martynas@atomgraph.com>
- Cc: public-sparql-dev@w3.org
On May 24, 2024, at 1:46 PM, Martynas Jusevičius <martynas@atomgraph.com> wrote: > > Hi, > > I was trying to figure out how to run the SPARQL test suite which lead > me to links > https://github.com/w3c/rdf-tests/issues/47 > https://github.com/kasei/sparql11-protocolvalidator > and several dead links too. > > I don't have the answer yet. It doesn't look good for a standards suite IMO. Those links are specifically about testing SPARQL Protocol implementations. There's probably similar variation in how people run the tests for query, update, etc. The protocol validator you linked to is mine, and was developed during the SPARQL 1.1 standardization process to support the working group’s need to evaluate the implementation support for the SPARQL Protocol. I’m not opposed to it being used on an ongoing basis, but it’s not actively being maintained. I’d be happy to accept pull requests for adding a Dockerfile if anyone wants to attempt that. > Moreover, I found a quote "everyone has their own mechanism for > running the tests" (I suppose meaning every implementation), which is > weird given that SPARQL provides a standard protocol for running > queries. I think mostly the “everyone has their own mechanism” situation is because to do the testing using just standard APIs, you’d have to assume that part of the system under test already worked (e.g. assuming SPARQL Update or GSP worked so that you could set up the dataset to run a query test). The Protocol tests are a bit different in this respect, because they build on top of the other specs, and the GitHub repo states right up front in the README that full support of SPARQL 1.1 Query and Update is assumed. .greg
Received on Friday, 24 May 2024 22:15:47 UTC