bug: uint and uint256 parsed different in EIP-712 Typed Data event signatures

[krzkaczor]

Is there an existing issue for this?

• I have searched the existing issues

Package Version

0.1.7

Current Behavior

This results in two different types ie. uint doesn't get coerced to uint256.

// works
  const logs = await client.getLogs({
    address: erc20Address,
    event: parseAbiItem('event Transfer(address indexed from, address indexed to, uint256 value)'),
    fromBlock: 16790000n,
  })

//doesnt work (no results)
  const logs2 = await client.getLogs({
    address: erc20Address,
    event: parseAbiItem('event Transfer(address indexed from, address indexed to, uint value)'), // uint instead of uint256 here
    fromBlock: 16790000n,
  })

NOTE: Maybe this issue should be moved to abitype, i am not sure which library is responsible for ABI normalization.

Expected Behavior

Both queries works.

Steps To Reproduce

Just run the code stored above.

Link to Minimal Reproducible Example (StackBlitz, CodeSandbox, GitHub repo etc.)

https://stackblitz.com/edit/viem-getting-started-vwxiwl?file=index.ts

Anything else?

Sorry I couldn't create repro on stackblitz due to some random error. I think the problem is also that template new.viev.sh uses old viem version in package.json.

Anyway, great job on the library! I can't wait to use it more.

[greenlucid]

Isn't this intended? uint is a solidity alias, is it ABI conforming?

[krzkaczor]

@greenlucid you are right that EIP-712 doesn't support uint alias (i copied the whole selector from the sourcecode -- that's why i hit that).

However, I think that there should be some kind of error instead of a silent failure.

[tmm]

Thanks for reporting! Think we already have some code in PR that will solve this.

[Raiden1411]

@tmm not sure if the current PRs have a fix for this. Unless I am missing something on your side.

If you don't have anything I can work on a PR that fix this.

[tmm]

Based on my interpretation of this (also below):

uint, int: synonyms for uint256, int256 respectively. For computing the function selector, uint256 and int256 have to be used.

"type": "uint" and "type": "int" are both allowed. Not sure if any compilers output these as-is or convert to the fixed versions. Either way, converting things in #110

---

Separately, uint and int are already not allowed in EIP-712 Typed Data (see TypedDataType).