Bugfix with repeated names

Previous versions allowed the production of schemas with duplicate shape names. That's fixed with this release

Bugfix for RDF-Config generation and bugfix for SPARQL exploration

• RDF-Config: forcing lowercase for some names that the standard reject when containing capital letters.
• SPARQL: some endpoints retireve empty values for variables even when no OPTIONAL is specified, which raises errors. That has been fixed.

Release for first sync with Zenodo

2.6.3.0

Merge branch 'enhancement-#190'

Bugfix for RDF-Config generation and annotations

• Bugfix: double quotes are scaped when appearing inside a string, both in rdf-config and shex annotations.
• Bugfix: multiline annotations / examples in RDF-Config are tranformed into single line strings scaped newlines.
• Bugfix: RDF-Config generation was preventing ShEx/SHACL generation with some inputs. Now you can generate both in every case in a single execution.

Work on RDF-Config generation

• Improvement: variable names do not include numbers (except in extreme situations). More meaningful names are produced by using contextual information.

Fix a bug related to multiline strings in TURTLE_ITER format

Multiline strings are now supported by the parser used when the input format TURTLE_ITER is chosen

Make shape maps and SHACL generation compatible

Even if shape maps are a concept related to ShEx, in this context can be used to sprcify custom groupings of nodes beyond usual class-instance relations for SHACL as well.

For each selector associated to a shape label in a shape map, sheXer will produce as many sh:targetNode declarations associated to a shape as nodes are targeted by the selector.

Fix TypeError when generating some results

The previous release introduced some bug when asked to save results to file under certain circunstances. This releases fixes that.

Generation of rdf-config files

Besides ShEx, SHACL and UML, now you can also generate a representation of the schema using rdf-config files which comply with the rdf-config specification.

WARNING: there is a bug saving results to file in some context this version. Release 2.5.12 fixes it

Support for shapes for federated queries

You can configure sheXer to look for synonyms of entities in the target source in one or several SPARQL endpoints. To do that, you must use the parameter federated_sources in Shaper, and set it to a list of shexer.model.FederatedSource objects. Each object will contain the information required to lacate those synonyms: endpoint URL, property to expresss equivalence (owl:sameAs, schema:sameAs, etc.), expected location of the triple linking the sources (target source or endpoint), position of the nodes whose synonyms we are looking for in that triple (subject or object), etc.

WARNING: these feature may not still be fully integrated with every posible input in sheXer. In case you find any problem, please, write an issue in sheXer's repository.