back-ports privateKey validation of the 2.x Accounts module to 1.x #3323
Conversation
nivida
added
Enhancement
nivida
changed the title
…se we do not left pad the private key to 32 bytes in web3.js
|
@cgewecke I've removed the test vectors with 30 or 31bytes long private keys stored in the keystore v3 JSON file because web3.js does not left-pad the PK to 32bytes as go-ethereum is doing. I think the path of keeping the PK untouched is the right one. WDYT? |
There was a problem hiding this comment.
What if you have encrypted with 1.2.4 using a non-standard length, stored the result somewhere and want to decrypt with a later version? The test case suggests the non-standard length should work for this purpose.
Not completely certain what's correct here - have opened #3328 as a possible alternative.
…dded to eth.accounts.encrypt-decrypt test file
|
@cgewecke I've applied the requested changes and re-added the 30/31 byte test vectors. |
There was a problem hiding this comment.
This looks great and improves on the suggestion in #3328.
The test file changes should really come out though, sorry. I think those files should only be changed for logic so the implications of the new code are crystal clear. They don't require "helpful reformatting" - unless done as a separate initiative.
|
@nivida This was merged over a requested change? |
Description
Thanks to @DalderupMaurice to give me a hint about the more advanced solution in the 2.x branch.
Type of change
Checklist:
npm run dtslintwith success and extended the tests and types if necessary.npm run test:unitwith success and extended the tests if necessary.npm run build-alland tested the resulting file/'s fromdistfolder in a browser.CHANGELOG.mdfile in the root folder.