Alloy binding failed: unable to parse token from invalid AST

[alxiong]

Component

Forge

Have you ensured that all of these are up to date?

• Foundry
• Foundryup

What version of Foundry are you on?

1.0.0-nightly

What version of Foundryup are you on?

No response

What command(s) is the bug in?

just gen-bindings

Operating System

MacOS

Describe the bug

On 6e86d3c of EspressoSystems/espresso-network#2706, just run gen-bindings which internally run

forge bind --skip test --skip script --libraries contracts/src/libraries/PlonkVerifier.sol:PlonkVerifier:0x5fbdb2315678afecb367f032d93f642f64180aa3 --alloy --contracts ./contracts/src/ --crate-name contract-bindings-alloy --bindings-path contract-bindings-alloy --select "^LightClient(V\d+)?$|^LightClientArbitrum(V\d+)?$|^FeeContract$|^PlonkVerifier(V\d+)?$|^ERC1967Proxy$|^LightClient(V\d+)?Mock$|^PermissionedStakeTable$" --overwrite --force

and got:

[⠊] Compiling 44 files with Solc 0.8.23
[⠒] Solc 0.8.23 finished in 8.32s
Compiler run successful!
Generating bindings for 11 contracts
Error: failed to parse generated tokens as an AST for contract-bindings-alloy/src/lightclientarbitrumv2.rs:lightclientarbitrumv2;
this is likely a bug

Context:
- expected identifier, found keyword `self`

Before this PR, our main has no problem generating bindings (for both ethers and alloys) using the same just command.

p.s. the problem with contract-bindings-alloy/src/lightclientarbitrumv2.rs:lightclientarbitrumv2; also exists for contract-bindings-alloy/src/lightclientv2.rs:lightclientv2; as I have observed that before as well.

cc @sveitser

[grandizzy]

@alxiong mind to check if problem persists in latest stable? thank you! (be aware that ethers was removed and we're it defaults to alloy)

[alxiong]

just updated to the latest, still got this issue:

┌─  espresso-network on  ax/lc-epoch (MERGING) [=$✘»!+] via 🐳 desktop-linux is 📦 v0.1.0 via  v20.18.0 via 🦀 v1.83.0 
└─○ forge bind --skip test --skip script --libraries contracts/src/libraries/PlonkVerifier.sol:PlonkVerifier:0x5fbdb2315678afecb367f032d93f642f64180aa3 --alloy --contracts ./contracts/src/ --crate-name contract-bindings-alloy --bindings-path contract-bindings-alloy --select "^LightClient(V\\d+)?$|^LightClientArbitrum(V\\d+)?$|^FeeContract$|^PlonkVerifier(V\\d+)?$|^ERC1967Proxy$|^LightClient(V\\d+)?Mock$|^PermissionedStakeTable$" --overwrite --force
Warning: This is a nightly build of Foundry. It is recommended to use the latest stable version. Visit https://book.getfoundry.sh/announcements for more information. 
To mute this warning set `FOUNDRY_DISABLE_NIGHTLY_WARNING` in your environment. 

[⠊] Compiling...
[⠰] Compiling 44 files with Solc 0.8.23
[⠔] Solc 0.8.23 finished in 4.50s
Compiler run successful!
Generating bindings for 11 contracts
Error: failed to parse generated tokens as an AST for contract-bindings-alloy/src/lightclientarbitrumv2.rs:lightclientarbitrumv2;
this is likely a bug

Context:
- expected identifier, found keyword `self`

┌─  espresso-network on  ax/lc-epoch (MERGING) [=$✘»!+] via 🐳 desktop-linux is 📦 v0.1.0 via  v20.18.0 via 🦀 v1.83.0 took 5s 
└─○ forge --version
forge Version: 1.0.0-nightly
Commit SHA: 3c048be05726218c405d5d4deec4d3f1ab515f6f
Build Timestamp: 2025-03-11T06:02:02.800485000Z (1741672922)
Build Profile: maxperf

(and yes, I'm only generating alloy bindings)

[alxiong]

are you able to reproduce @grandizzy ? (you don't need to enter nix shell, just clone the repo, and run the forge bind above.)

(this is a blocking bug on a somewhat time-sensitive feature I'm working on, so would appreciate if its gets prioritized, I'm happy to hop on call or explaining more about our contracts, if necessary, to help you debug and locate the bug)

[grandizzy]

yep, going to check why it happens

[grandizzy]

@alxiong self is a reserved keyword, please rename all EvalDomain memory self within contracts/src/libraries/PolynomialEvalV2.sol like this one into something like EvalDomain memory evalDomain (contract PlonkVerifierV2 is PolynomialEvalV2 so bindings fail with self)

Please reopen if still issues, thank you

[alxiong]

I see, 😂 didn't realize that.

Using the name "self" in library is not uncommon as it appears in the solidity official example: https://docs.soliditylang.org/en/v0.8.29/contracts.html#libraries

While this ticket can be closed, may I suggest adding a compiler warning or hint during binding generations?

[grandizzy]

the error says

Context:
- expected identifier, found keyword `self`

agree could be misleading

[yash-atreya]

Reopening as we have a fix here: alloy-rs/core#903 which will be applied in the next alloy core bump

@alxiong we still can't generate bindings for self due to rust limitations: https://internals.rust-lang.org/t/raw-identifiers-dont-work-for-all-identifiers/9094/4

Having said that, we've opted for mapping self to this in the codegen.

function something(address self) returns (uint256); 

will translate to:

pub struct somethingCall { this: Address };

[alxiong]

yes, that's lovely. self or _this are both good choices IMO. This is helpful, thanks for the quick fix.