Closed clehner closed 3 years ago
Merge conflicts.
I'd also like a few people to run these tests against their implementations to make sure they agree with the changes.
Rebased to fix merge conflicts. I'm also open to other ways to make this work.
Rebased to fix merge conflicts. I'm also open to other ways to make this work.
Conflicts are back :(
Rebased again. The conflicts were in implementations/index.html
; fixed using node implementations/generate.js
@brentzundel -- can you please check the changes in this PR... it modifies the ZKP tests in a way that I think is correct, but want to double check with you to make sure I'm not missing something.
This addresses #112 to allow a
VerifiablePresentation
without theverifiableCredential
property, for the use case of proving control over a DID as in vp-request-spec. I also removed the requirement for theproof
property, as that property is similarly mentioned in the VC Data Model with "If present," implying optionality (https://w3c.github.io/vc-data-model/#presentations-0) - except for ZKPs where it is required (https://w3c.github.io/vc-data-model/#zero-knowledge-proofs).I renamed the positive test case for a VP containing
verifiableCredential
andproof
to say "MAY" instead of "MUST". "MAY" isn't used elsewhere in the tests, but I thought it would be useful to express that the credential may have this property (which the implementation must support). Since this test case is the same but only renamed, I considered updating the implementation reports JSON with the renaming, so that their info would still appear in the generated implementation report HTML file. But I left them as-is so they remain true to what the implementors submitted. This particular row in the generated implementation report then disappears until implementation reports are updated.The negative case for missing
verifiableCredential
is changed to a positive case ("MAY omitverifiableCredential
"). This means previous implementations passing this test will now fail it, and vice versa. But it will be useful to know which ones support this for the use cases in question.A possible use case for omitting the proof property is in https://github.com/w3c-ccg/vc-http-api/issues/88, but I did not add any positive test case for omitting the proof property.
To replace the negative test case for missing proof property, I added a negative test case for a proof with missing
type
property, as that is a required property: https://w3c.github.io/vc-data-model/#proofs-signaturesFixes #112