Include fragment in schemaLocation

[marksparkza]

https://github.com/json-schema-org/json-schema-spec/blob/70af7dcf147759f9eaecae5c574f0b36cc86a276/jsonschema-core.xml#L2851

[gregsdennis]

The fragment isn't necessary when you're looking at the root. Given that they're URIs, they should be considered equivalent. That line you linked to it's more informative (no MUST retirement).

I realize that for the test, however, it's checking a string value against a string value. Not sure what to do here.

[marksparkza]

I should probably have referenced this clause, which does say MUST.

The fragment isn't necessary to identify/resolve the root, but I don't think there's any reason not to mandate that it must be present in "schemaLocation". It ensures consistency across implementations, and it's also simpler to specify - "there must be a fragment" as opposed to "there must be a fragment unless it's the root in which case there may be a fragment"...

[Julian]

Merging since the spec seems clear this is correct (and wasn't previously correct) but if someone has some reason to disagree, follow up after.