add sign and verify test vectors #292
Conversation
nitro-neal
requested review from
KendallWeihe,
kirahsapong,
shamilovtim,
diehuxx,
decentralgabe,
frankhinek,
leordev and
ALRubinger
as code owners
|
this pr goes along with this one - decentralized-identity/web5-spec#172 |
| private val rustCoreVerifier: RustCoreEd25519Verifier | ||
|
|
||
| constructor(privateKey: Jwk) { | ||
| this.rustCoreVerifier = RustCoreEd25519Verifier(privateKey) |
There was a problem hiding this comment.
I think with the latest on main you'll need to rebase and then this'll be...
| this.rustCoreVerifier = RustCoreEd25519Verifier(privateKey) | |
| this.rustCoreVerifier = RustCoreEd25519Verifier(privateKey.rustCoreJwkData) |
| val key: TestVectorJwk, | ||
| ) | ||
|
|
||
| data class TestVectorJwk( |
There was a problem hiding this comment.
I know we're not testing the JWK in this test, but ideally we could use the real Jwk, but to your point on the other PR, we may need to add #297 first?
There was a problem hiding this comment.
yup after 297 is in we can use it
| let signature = result.expect("Signing should not fail"); | ||
|
|
||
| // Convert the signature to a hex string | ||
| let signature_hex = byte_array_to_hex_string(&signature); |
There was a problem hiding this comment.
I'm surprised this works. ed25519 signatures are different every time because they use a random nonce to prevent computing the private key if the same private key is used for multiple signatures
There was a problem hiding this comment.
oh no..
This works in every sdk
Are we doing something wrong lol, let me check on this..
There was a problem hiding this comment.
I think we are good - https://cryptobook.nakov.com/digital-signatures/eddsa-and-ed25519
There was a problem hiding this comment.
see Section 4.2.5 on Deterministic Nonce Generation
Ed25519 uses deterministic signing which removes the need for fresh random numbers during the signing process. This does not lead to any particular consequences for our security analysis since we model the key derivation function as a random oracle. However, it is well known not to reduce security
|
|
||
| let result = verifier.verify(&data, &signature); | ||
|
|
||
| let is_valid = result.expect("Verification should not fail"); |
There was a problem hiding this comment.
what's the difference between Err() and Ok(false) for verifier.verify?
There was a problem hiding this comment.
chad says:
Ok(false): The operation itself succeeded, but the verification result is false. This means that the signature is valid in terms of format and process, but it does not match the data or is otherwise invalid according to the verification logic.
and yea then err is an exception basically if I'm understnading rust correctly
There was a problem hiding this comment.
oh and as far as the implementation yea just verified false is invalid signature vs catastrophic failure
Updating web5-spec and adding ed25519 test vectors