chore(fuzz): Test AST print-and-parse roundtrip

[aakoshh]

Description

Problem*

Resolves #9058

Summary*

Adds a new prop test to the AST fuzzer to ensure that if we render the monomorphized AST created by the fuzzer and parse it back, it renders the same way a second time, indicating that there are no significant differences between the fuzzer and the compiler, no hidden assumptions that the generator does not respect.

cargo test -p noir_ast_fuzzer --test mono

Additional Context

The can be subtle differences, for example rendering Unary { op: Minus, rhs: Literal(1) } could be parsed back as Literal(-1). These don't cause a test failure, because we only compare the printed Noir source, and I made some changes to the printer to ensure they appear identical. Other instances that cause no harm such as --1 being parsed as 1 are handled in the test itself.

The test did highlight some actual semantic differences, such as a misunderstanding in the return_type of a unary reference, which resulted in an extra &mut in the printed output. Similarly it uncovered that the separation of array indexes into a temporary variable had a different definition of "side effect" than the fuzzer.

After investigation there remains a discrepancy in how these array indexes are handled for multi-dimensional arrays: the compiler creates temporary variables for all dimensions in a single block, ie. a[0 + 1][2 - 1] = 3; is turned into { let i_1 = 0 - 1; let i_2 = 2 - 1; a[i][j] = 3;. The generator only considers a single index. This doesn't cause an issue at the moment because the generator doesn't assign to multiple dimensions, only the top one. I'll create a follow-up PR to add multi-dimensional assignments.

Documentation*

Check one:

• No documentation needed.
• Documentation included in this PR.
• [For Experimental Features] Documentation to be submitted in a separate PR.

PR Checklist*

• I have tested the changes locally.
• I have formatted the changes with Prettier and/or cargo fmt on default settings.

[github-actions]

⚠️ Performance Alert ⚠️

Possible performance regression was detected for benchmark 'Test Suite Duration'.
Benchmark result of this commit is worse than the previous benchmark result exceeding threshold 1.20.

Benchmark suite	Current: 8450e51	Previous: 6a68ca1	Ratio
test_report_noir-lang_noir_bigcurve_	314 s	240 s	1.31

This comment was automatically generated by workflow using github-action-benchmark.

CC: @TomAFrench

[asterite]

Looks great!