- From: Dave Beckett <dave.beckett@bristol.ac.uk>
- Date: Tue, 05 Jun 2001 15:10:30 +0100
- To: w3c-rdfcore-wg@w3.org
RDF/XML Parser Tests
http://ilrt.org/discovery/swsw/pt
An online comparison demo of 5 parsers.
RDF Parser Tests Schema
http://ilrt.org/discovery/2001/03/parser-tests/
Document about the RDF schema used to describe the parser, the
test case and the crude results of running a test with a parser.
RDF schema file
http://ilrt.org/discovery/2001/03/parser-tests/schema.rdfs
So what is all this? Well Redland is used to parse various online
tests from http://www.w3.org/2000/10/rdf-tests/ with all the 5
parsers that it supports. The results of these are crudely done by
counting triples and some RDF/XML files describing the tests and the
results are emitted:
http://ilrt.org/discovery/swsw/pt/inputs/parsers/
e.g. http://ilrt.org/discovery/swsw/pt/inputs/parsers/rapier.rdf
describes the results for my parser Rapier.
The 5 results are then taken along with three RDF Schema files - RDF,
DC and the RDF Parser Test schema above and an expected results file
http://ilrt.org/discovery/swsw/pt/inputs/expected.rdf
generated by me with some likely answers and all of that is merged
into a big RDF graph.
The result is the online demo but you can browse the graph itself from
http://www.redland.opensource.ac.uk/demo?db=pt;command=print
This is the other part of the parser testing, and probably can be
extended to read the n-triple answers from test cases and check that
they really are emitted, do the graph comparison etc. when some of
that software gets written.
Dave
Received on Tuesday, 5 June 2001 10:10:36 UTC