fixed up test case rdfs-entailment-2

• From: Peter F. Patel-Schneider <pfpschneider@gmail.com>
• Date: Fri, 18 Oct 2013 10:25:12 -0400
• To: 'RDF WG' <public-rdf-wg@w3.org>
• Message-ID: <526144C8.7070603@gmail.com>

The test case rdfs-entailment-2 appears to have been messed up from the 
beginning of testing.

It was supposed to test that a random satisfiable graph followed from an 
unsatisfiable graph in RDFS.  However, the unsatisfiable "graph" in the 
original test suite was FALSE, which isn't a real graph.  I don't know how 
this was tested.  (Presumably testers put in some unsatisfiable graph.)   The 
test somehow mutated in the hg repository to something different.

I modified the test to have a real unsatisfiable graph (in RDFS, where 
rdf:langString and xsd:string are available).  Although I could have just used 
an unsatisfiable literal, I wanted this to remain an RDFS test.  I also 
modified the result graph to be just an unexceptional graph (instead of 
rdf:type rdf:type rdf:type .).


I left rdfs-entailment-1 alone, although it uses rdf:XMLLiteral, which isn't 
mandatory anymore.

peter

Received on Friday, 18 October 2013 14:25:40 UTC