WRITING_TEST_CASES.md

Writing test cases

Test cases are expressed as either SPARQL ASK or SELECT queries.

• ASK queries are considered to have failed if they evaluate to true so should be written to find invalid statements.
• SELECT queries are considered to have failed if they return any matching solutions.

We recommend that you prefer using SELECT queries as then the results provide an indication as to the cause of the problem.

For example this query is a port of IC-1 from the RDF Data Cube specification into SELECT style. It will return any qb:Observations that are not also in a qb:dataSet:

PREFIX qb: <http://purl.org/linked-data/cube#>

SELECT (?obs AS ?obsWithNoDataset)
WHERE {
  {
    # Check observation has a data set
    ?obs a qb:Observation .
    FILTER NOT EXISTS { ?obs qb:dataSet ?dataset1 . }
  }
}

Some more example SELECT queries for validating RDF Data cubes can be found here.

Query templates

You can write SPARQL validations with handlebars-style templates that will be pre-processed with selmer. This let's you include variables like {{ myvariable }} in your query templates. To set the values, pass a map of bindings in edn format using the --variables CLI option.

For example, if you have a followers-have-names.sparql template including the variable person:

PREFIX schema: <https://schema.org/>

SELECT ?follower WHERE {
  ?follower schema:follows <{{ person }}> .

  FILTER NOT EXISTS { ?follower scheme:name ?name }
}

And a person.edn data file include the bindings in a map from the variable name (as a keyword) to a string value that will be interpolated in it's place:

{:person "http://example.net/id/person/Ada"}

Then you can call it like:

$ clojure -M:rdf-validator --suite followers-have-names.sparql --variables person.edn --endpoint http://some.domain/sparql/query