fix(sema): reject rational expressions in abi.encode* arguments

[aryanbaranwal001]

This PR fixes a compiler panic when a rational (fractional/decimal) expression is passed as an argument to abi.encode*.

• Fixes Compiler panic in codegen: unreachable!("Type should not exist in codegen") when abi-encoding a rational expression #1864

The root cause is that sema's abi.encode* argument resolution loop at sema/builtin.rs:1403 only guards against mapping and recursive types. It has no guard for Type::Rational, which is a sema level placeholder for expressions involving decimal literals, 24.24, 1 - 24.24, 1.5 * 2 that haven't been coerced to a concrete integer type, which only happens when a target type is available (e.g. uint256 x = 32). The rational expression is stored in AST as Type::Rational without any error which then passes into codegen.

codegen/expression.rs:1048 matches ast::Builtin::AbiEncodePacked and calls abi_encode_packed() at codegen/expression.rs:1853.

abi_encode_packed maps expression() over each arg and passes the result to abi_encode() at codegen/expression.rs:1858.

abi_encode calls calculate_size_args() at codegen/encoding/mod.rs:48, which calls get_expr_size() at codegen/encoding/mod.rs:176.

get_expr_size matches on the expression's type. Type::Rational hits the unreachable arm at codegen/encoding/mod.rs:1454–1456:

Type::UserType(_) | Type::Unresolved | Type::Rational => unreachable!("Type should not exist in codegen")

Fix

Add a Type::Rational guard in the arg resolution loop at sema/builtin.rs:1407, immediately after the existing mapping/recursive check:

if matches!(ty, Type::Rational) {
    diagnostics.push(Diagnostic::error(
        arg.loc(),
        "rational expression has no concrete type; cast to an integer type before encoding".to_string(),
    ));
    return Err(());
}

This makes sema reject unresolved rational expressions before they reach codegen. The fix covers all abi.encode* variants of abi.encode, abi.encodePacked, abi.encodeWithSelector, and abi.encodeWithSignature, since they all share the same loop at sema/builtin.rs:1403.

Previously panicking inputs:

• abi.encodePacked(1 - 24.24);
• abi.encode(1.7);

Reproduces on polkadot, solana, and evm targets.