feat: COSE Key Thumbprint (RFC 9679) feature and its test

[kentakayama]

Calculates canonical (deterministic) thumbprint from fields of Key.
https://www.rfc-editor.org/info/rfc9679

[kentakayama]

As noted in Pull Request #223, the behavior of Key.validate() has changed, which also affects the error codes received in the Thumbprint() test cases.
If #223 is merged, I will add another commit to ensure Thumbprint() produces the intended error codes.

[OR13]

Looks good to me, I think we should add this

[thomas-fossati]

As noted in Pull Request #223, the behavior of Key.validate() has changed, which also affects the error codes received in the Thumbprint() test cases. If #223 is merged, I will add another commit to ensure Thumbprint() produces the intended error codes.

Is that the reason why TestKey_Thumbprint fails at present?

[thomas-fossati]

Hey @kentakayama , thanks for the PR!

You requested reviews, but it's still marked as a draft.
Could you please clarify its status?

cheers!

[kentakayama]

Hey @kentakayama , thanks for the PR!

You requested reviews, but it's still marked as a draft. Could you please clarify its status?

cheers!

@thomas-fossati
Sorry for the confusion. At first I thought this Pull Request was “perfect!” and submitted it in a non‑Draft state.
However, I realized that many of the checks for computing the Thumbprint overlap with the validate() function, which isn’t great from a maintainability perspective, so I switched it back to Draft.
As a result, it’s now in the state of “review requested but Draft.”

While looking more closely at the validate() function, I revisited RFC 9053 in light of the COSE Key Thumbprint functionality, and I started to think that perhaps it should be handled as in #223.
I believe it might be better to discuss that first.
In fact, because the error checking in #223 was made stricter, the expected error codes in the Thumbprint test cases have also changed, so I think #223 should be resolved first.

[thomas-fossati]

@thomas-fossati Sorry for the confusion. At first I thought this Pull Request was “perfect!” and submitted it in a non‑Draft state. However, I realized that many of the checks for computing the Thumbprint overlap with the validate() function, which isn’t great from a maintainability perspective, so I switched it back to Draft. As a result, it’s now in the state of “review requested but Draft.”

While looking more closely at the validate() function, I revisited RFC 9053 in light of the COSE Key Thumbprint functionality, and I started to think that perhaps it should be handled as in #223. I believe it might be better to discuss that first. In fact, because the error checking in #223 was made stricter, the expected error codes in the Thumbprint test cases have also changed, so I think #223 should be resolved first.

Thank you for taking the time to explain it to me so clearly.