Re: shapes-ISSUE-75 (error vs violation): How to distinguish constraint violations from errors [SHACL Spec]

On Mon, Jul 13, 2015 at 9:43 PM, RDF Data Shapes Working Group Issue
Tracker <sysbot+tracker@w3.org> wrote:

> shapes-ISSUE-75 (error vs violation): How to distinguish constraint
> violations from errors [SHACL Spec]
>
> http://www.w3.org/2014/data-shapes/track/issues/75
>
> Raised by: Peter Patel-Schneider
> On product: SHACL Spec
>
> There are two kinds of things that can happen when running SHACL:
> 1/ A constraint violation can happen, e.g., the name of a person is a
> number.
> 2/ An error can occur, perhaps because an illegal recursion is discovered,
> or some SPARQL macro tries to do an operation an unbound variable.
>
> Currently constraint violations produce sh:Error and documents talk about
> these violations being errors.  This can easily lead to a confusion between
> constraint violations and the other kinds of errors.  This did not use to
> be much of a problem because it was possible to treat errors locally, but
> the handling of illegal recursion may require treating errors very
> differently from constraint violations.
>

This is something that is covered by the RDFUnit vocabulary (see slide 5 in
http://www.slideshare.net/jimkont/rdfunitmethodologywww2014research-track#5)
Of course RDFUnit does not support recursion but there are other kinds or
errors that can be encountered such as timeout, a network error or even a
syntax error if the SPARQL engine does not support some constructs (sparql
specific)

In RDFUnit there are two types of report: 1) a SHACL like where only
violations are written (ignoring other errors) and 2) a full one where for
every test the user gets a status (success, fail, timeout, error) and a
level (error, warning, etc).

One way to resolve this is to approve the (already under consideration)
requirement "Constraint Violations Reporting Details"
https://www.w3.org/2014/data-shapes/wiki/Requirements#Constraint_Violations_Reporting_Details
and create a similar to the RDFUnit full report.

Best,
Dimitris



-- 
Dimitris Kontokostas
Department of Computer Science, University of Leipzig & DBpedia Association
Projects: http://dbpedia.org, http://http://aligned-project.eu,
http://rdfunit.aksw.org
Homepage:http://aksw.org/DimitrisKontokostas
Research Group: http://aksw.org