'escaped pointer ref' percent encoding

[gavinwahl]

In ref.json, the 'escaped pointer ref' tests test that the json pointer '#/percent%25field' looks up a field called 'percent%field'. Is this a valid test? I can't find any mention of percent-encoding in https://tools.ietf.org/html/draft-ietf-appsawg-json-pointer-07#section-4.

[handrews]

@gavinwahl the percent encoding comes from section 6 of the JSON Pointer RFC 6901. So while it's technically not part of the JSON Schema suite of specifications, it's something that would be easy to get wrong if you're doing your own URI encoding of JSON Pointers.

Does that address your concerns about the validity of the test?

[gavinwahl]

"A JSON Pointer can be represented in a URI fragment identifier by encoding it into octets using UTF-8 [RFC3629], while percent-encoding those characters not allowed by the fragment rule in [RFC3986]."

Are refs JSON Pointers, or JSON Pointers encoded as URI fragment identifiers? I don't think the linked section can apply, because refs are not encoded into octets. (There's no way to represent octets in JSON, only unicode code points)

[handrews]

@gavinwahl percent-encoded octets are being used here the same way as they are throughout URIs as described in RFC3986. The value of a "$ref" is a URI reference. There is nothing special about it. It's encoded into UTF-8 (or UTF-16 or UTF-32) like any other string data for JSON. URI-encoded JSON Pointers are one type of fragment supported by application/schema+json. So the pointer gets encoded into a URI fragment per RFCs 6901 and 3986, and the resulting URI is represented in JSON the same way any URI would be.

[epoberezkin]

close?

[handrews]

Yeah, no reply for two months.