Now that we have the methods in our own stack to do this, implement in Cashtab. This allows us to stop using bitcoinjs-message.
Details
Details
- Reviewers
- None
- Group Reviewers
Restricted Project
npm test
Diff Detail
Diff Detail
- Repository
- rABC Bitcoin ABC
- Branch
- ecash-lib-sign-msgs
- Lint
Lint Errors Severity Location Code Message Error cashtab/src/components/SignVerifyMsg/SignVerifyMsg.js:20 ESLINT @typescript-eslint/no-unused-vars - Unit
No Test Coverage - Build Status
Event Timeline
Comment Actions
implement a de facto spec implementation of sign and verify at the lib level and apply in Cashtab
| cashtab/src/components/SignVerifyMsg/__tests__/SignVerifyMsg.test.js | ||
|---|---|---|
| 81 ↗ | (On Diff #53033) | the signature changing is ... potentially not expected behavior. I am not sure why it changes. The prefixes are identical. There are some conventions in bitcoinjs-message that could lead to distinct values, i.e. adjusting the recovery byte https://github.com/bitcoinjs/bitcoinjs-message/blob/master/index.js#L34 does electrum / the node do this? |
| 81 ↗ | (On Diff #53032) | the signature changing is ... potentially not expected behavior. I am not sure why it changes. The prefixes are identical. There are some conventions in bitcoinjs-message that could lead to distinct values, i.e. adjusting the recovery byte https://github.com/bitcoinjs/bitcoinjs-message/blob/master/index.js#L34 does electrum / the node do this? |
| cashtab/src/components/SignVerifyMsg/__tests__/SignVerifyMsg.test.js | ||
|---|---|---|
| 81 ↗ | (On Diff #53033) |
relevant file looks like electrumabc/ecc.py def sign_message(
self,
message: bytes,
is_compressed,
*,
sigtype: SignatureType = SignatureType.ECASH,
) -> bytes:
def bruteforce_recid(sig_string):
for recid in range(4):
sig65 = construct_sig65(sig_string, recid, is_compressed)
if not self.verify_message(sig65, message, sigtype=sigtype):
continue
return sig65, recid
else:
raise Exception("error: cannot sign message. no recid fits..")
message = to_bytes(message, "utf8")
msg_hash = Hash(msg_magic(message, sigtype))
sig_string = self.sign(
msg_hash,
sigencode=sig_string_from_r_and_s,
)
sig65, recid = bruteforce_recid(sig_string)
return sig65ref def construct_sig65(sig_string, recid, is_compressed):
comp = 4 if is_compressed else 0
return bytes([27 + recid + comp]) + sig_stringso, it looks like there is a standard as such about signing with / without a compressed key ... and the standard exists bc we need to know which key to recover. that said ... I'm not sure we need to have this in ecash-lib. the way ecash-lib is contained, we would always be signing with 32-byte sk, and always be verifying against the 33-byte compressed public key. So question -- do we need to support uncompressed public keys? Cashtab does not need this. I can see the value in having deterministic signatures for any given private key tho. |