- From: Alex Nelson via GitHub <noreply@w3.org>
- Date: Fri, 03 Apr 2026 22:33:38 +0000
- To: public-shacl@w3.org
ajnelson-nist has just created a new issue for https://github.com/w3c/data-shapes: == [PROF] Add Profile of SHACL for a certain desired validation result quality == This came out of a discussion from last week's profiling task force call. A certain profile of SHACL can be created to enable "Pretty validation." (Term c/o someone else on the call. It lines up conceptually well with "pretty printing.") What the profile requires is: 1. All property shapes must be IRI-identified. 2. All SPARQL constraints must be IRI-identified. 3. `sh:targetObjectsOf` is forbidden. Those might not seem related, but they have a few compacting effects on validation reports. What is sought, and possibly achieved (there might be some foiling cases from complex `sh:path` values---some experimenting is needed), is every validation result gets a compact IRI---perhaps even URL!---a user can see for the full shape definition causing the result; AND, a user gains enough information from focus node, value node, and path to find the triple (or at least, path connecting first subject and last object) triggering the result. Anti-use cases this profile helps mitigate: * A shape written using targetOjectsOf against a datatype can lead to a truly strange-looking validation result of "the focus node triggering this shape is the number 2." * A detailed property shape spelled out in a blank node can produce a whelming `sh:sourceShape` line. * Most SPARQL constraints produce a whelming `sh:sourceConstraint` line. Please view or discuss this issue at https://github.com/w3c/data-shapes/issues/850 using your GitHub account -- Sent via github-notify-ml as configured in https://github.com/w3c/github-notify-ml-config
Received on Friday, 3 April 2026 22:33:39 UTC