Update the OpenAPI pipeline to run SHACL validations

[rduyme]

At the moment there is automatic validation of OpenAPI examples with the open api yaml.
https://github.com/skg-if/api/tree/main/openapi/ver/current/sample_data

ex:

{
    "@context": [
        "https://w3id.org/skg-if/context/1.1.0/skg-if.json", => data model context
        "https://w3id.org/skg-if/context/1.0.0/skg-if-api.json",
        {
            "@base": "https://w3id.org/skg-if/sandbox/my-skg-acronym/"
        }
    ],
    "@graph": [
        {
            "local_identifier": "http://example.com/skg-if/api/grants/6f368a3a-b1cf-498f-b2de-27135d1e0011",
            "identifiers": [
                {
                    "scheme": "doi",
                    "value": "https://doi.org/10.3030/101095129"
                }
            ],

This validation is done in github actions whenever we update a sample_data
or the yaml specifications (https://github.com/skg-if/api/blob/main/openapi/ver/current/skg-if-openapi.yaml)

• https://github.com/skg-if/api/blob/main/.github/scripts/validate_files.py (see prism docker container used)
• https://github.com/skg-if/api/blob/main/.github/workflows/validate_specs_and_jsons.yml

@essepuntato Silvio requested we add SHACL validation for these sample_data examples

SHACL rules are defined here. https://github.com/skg-if/data-model/blob/main/README.md, we have a directory with rules for each data model version

This issue is about updating the https://github.com/skg-if/api github action/workflow to perform validation of openapi/ver/xxx/sample_data with version xxx SHACL rules.

• No need to change when this validation is triggered we can keep it in openapi repo
• we want to rely on official openapi examples in https://github.com/skg-if/api/tree/main/openapi/ver/xxx/sample_data (we validate these against both openapi yaml + SHACL )

[rduyme]

@essepuntato could you take this one ? :-)

[arcangelo7]

Hey @rduyme, I took a look at this and opened a PR with a possible implementation: #59.

In a few words, I added data-model as a submodule (to get the SHACL shapes), then a Python script that parses each sample JSON-LD file into RDF and validates it against the version-matched shapes using pyshacl. Runs as a new CI job alongside the existing OpenAPI validation.

I run it locally and found two issues in the current sample data:

• In grant_1.json, funded_amount is set to 50000, which is an integer while a float is expected
• In journal_article_full.json, collected dates are years (2013, 2014), while they should be xsd:dateTime

[rduyme]

@arcangelo7 great thanks ! We will take time to merge your PR.