IdentityEvidenceType should be an open enumeration

[Former user]

IdentityEvidenceType is made final, which makes it impossible to extend the framework for new evidence types (non arguably there are more ways to confirm one's identity than by utility bill, QES and ID document).

Also, the parsing machinery in the parse method of IdentityEvidence allows for no pluggability to parse new evidence types description. For example, if the evidence was a DNA fingerprint of a person, there is no way this could be parsed.

Thus, if the id token contains any non standard evidence, it becomes unparsable at any higher level of abstraction than raw JWT.

[Yavor Vasilev]

Hi,

The IdentityEvidenceType class is for declaring identifiers, that fact that it is made final doesn’t prevent users from creating instances like new Identity"videnceType(“DNA-FINGERPRINT”) .

https://www.javadoc.io/doc/com.nimbusds/oauth2-oidc-sdk/latest/com/nimbusds/openid/connect/sdk/assurance/evidences/IdentityEvidenceType.html

As for the parsing machinery, you’re right that at present it doesn’t have an interface for plugging factories for new types. Do you have ideas or suggestions for a PR to handle this?

[Yavor Vasilev]

In release 9.21 the evidence types were updates for draft 12 of the eKYC spec:

https://www.javadoc.io/doc/com.nimbusds/oauth2-oidc-sdk/9.21/com/nimbusds/openid/connect/sdk/assurance/evidences/IdentityEvidenceType.html

[Yavor Vasilev]

Draft 12 defines those 5 evidence types, see if you can fit into them:

https://openid.bitbucket.io/eKYC-IDA/openid-connect-4-identity-assurance-1_0-master.html#name-evidence-element

If you there is a concrete need for a custom evidence type, class and parser reopen the ticket with more details and an example.