The following notes for metadata TG requirements were taken down during the workshop in Ispra 2nd & 3rd Dec 2014.
Check that the hierarchyLevel element is there. Combine Req 1 with TG Req 2. Value must be one of the options in Req 2.
Note to the TG development: add a new explicit requirement for using the XML encoding defined in ISO 19139
Maybe combine Req 3 & 4 (data & service). Check if the values of all the Resource Locators structure are a valid URLs. Each URL must resolve to return a document: return "ok" and length > 0, do not try to parse the returned document Authentication/authorization/redirection? Break down into two tests: parse the URL (validity) and resolve it, but require both to pass
Combine Reqs 5 & 6.
Two options:
- if the type if MD_identifier it must have the code.
- if the type is RS_Identifier it must have the code. If the codeSpace is also given, it must not be empty.
Requirement wording feedback to TG?
Test only applied to services.
If it operatesOn element is there check that
- the href attribute is there,
- the URL is resolvable,
- the returned (resolved) fragment should(must?) resolve to the MD_Metadata element
Thijs: Maybe this is more easily tested from the service side?
This test should be part of the NS interoperability tests.
Not possible to automatically check if the given keywords originate from a Controlled Vocabulary. However, it's a reasonable requirement for the data providers to check that they include a valid reference to the thesaurus or ontology if the keyword they use originates from such a vocabulary.
Maybe a manual tests?
Feedback to TG editors: Use resolvable URIs/URLs for thesauri references instead of the (translated) document titles. This would require entries in the INSPIRE registry?
If the vocabulary is mentioned, the date must be given.
Validation for titles (and other "free text"?): text content must include at least one alpha-numeric letter