Test case for $ref resolution in case of URN $id

[vearutop]

It is not apparent how reference resolution should happen in case base $id is having urn:... value and $ref is subschema.json.

[awwright]

+1 me for this. RFC 3986 doesn't describe this much, it's not very intuitive if I recall correctly, but it's not an error, and I can imagine implementations are likely to get it wrong.

…

On Mon, Jan 28, 2019, 15:59 Viacheslav Poturaev ***@***.*** wrote: It is not apparent how reference resolution should happen in case base $id is having urn:... value and $ref is subschema.json. — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#250>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AAatDZE2LZU6yQWWXVsdkCD3CT6Bo08Uks5vH4C2gaJpZM4aW8uw> .

[vearutop]

RFC 3986:

When presented with a URI that violates one or more scheme-specific
restrictions, the scheme-specific resolution process should flag the
reference as an error rather than ignore the unused parts; doing so
reduces the number of equivalent URIs and helps detect abuses of the
generic syntax, which might indicate that the URI has been
constructed to mislead the user (Section 7.6).

According to RFC it seems such case should be an invalid schema, so it might not belong to JSON Schema Test Suite as suite expects valid schema with valid/invalid test data.

[handrews]

@vearutop let's be careful about "URI scheme-specific error" and "invalid schema". schemE and schemA are quite different here.

However, I agree that checking for an error here would be something that a reference resolver that understands the urn: scheme should handle, and not part of JSON Schema's test suite.

We can test that an implementation can handle urn: references (because in that case we are testing that the implementation provides the library support at all). But we probably shouldn't get into testing how well the URN implementation chosen by the JSON Schema implementation handles weird cases.

scheme-specific reference resolution information for urn: can be found in https://tools.ietf.org/html/rfc8141#section-4.3

[Julian]

Think this is a duplicate of #179. Going to close this in favor of that one, though yeah definitely seems like we need some of these.