feat: Allow ENS resolver to answer more queries

[gskril]

Motivation

Accompanies farcasterxyz/fname-registry#505 to make nice ENS profiles from fnames. I also added additional checks on the CCIP Read callback function for extra safety.

I wasn't sure what the version should be, so left that untouched for now.

Change Summary

Adds support for addr(node, cointype) and text(node, key) lookups by changing the EIP-712 type.

Merge Checklist

• The PR title adheres to the conventional commits standard
• The PR has been tagged with change type label(s) (i.e. documentation, feature, bugfix, or chore)
• The PR's changes adhere to all the requirements in the contribution guidelines
• All commits have been signed

---

PR-Codex overview

This PR focuses on enhancing the FnameResolver contract and its related tests by introducing new parameters, modifying function signatures, and improving error handling for resolver functions.

Detailed summary

• Updated pre-commit hook in .rusty-hook.toml.
• Added FNAME_RESOLVER_DNS_ENCODED_NAME and FNAME_RESOLVER_PASSTHROUGH_ADDRESS environment variables.
• Modified DeploymentParams struct and runDeploy function in script/DeployL1.s.sol.
• Enhanced the FnameResolver contract with new mappings for allowed functions and text records.
• Updated resolve and resolveWithProof functions to improve validation and error handling.
• Added a MockResolver contract for testing purposes.
• Revised test cases in FnameResolverTest to cover new functionality and edge cases.

✨ Ask PR-Codex anything about this PR by commenting with /codex {your question}

[gskril]

Adding note to self that I can implement in the morning: if the name arg in resolve() is farcaster.eth, we should pass that request to the ENS public resolver. This allows the farcaster.eth name itself to have onchain records while all subnames read from the fname registry.

Edit: implemented this. The owner of farcaster.eth can set records on the ENS public resolver (resolver.eth) even if the name uses a different resolver, and requests will pass through seamlessly. Figured this was cleaner than adding all the relevant code to the FnameResolver directly.