chore(fuzz): Capture comptime print output

[aakoshh]

Description

Problem*

Resolves #8606
Fixes #8658

Needs AztecProtocol/aztec-packages#14507, #8660 and #8665 to be merged to pass CI.

Summary*

Adds an interpreter_output to Context which is accessed by the comptime Interpreter through the Elaborator to do the printing. The field holds a trait object which implements std::io::Write; we use dynamic dispatch to access it for writing.

In the fuzz test tests the content goes in a byte vector and is used to compare the output from the interpreter with the real execution.

Fixes the comptime interpreter to display Field with hexadecimal format.

Adds the ability to the AST printer to show the type in let bindings to combat the ambiguity between the generated AST, which carries type information, and the printed source, where the frontend has to infer it from literals. This manifested itself as a value being printed as a number in the test and a field in nargo, because the test knew the type was u32, but that did not appear in the printed source, which was parsed for the comptime snippet.

Additional Context

I started off by adding a generic W: std::io::Write, and borrowing a mutable reference to it through the Elaborator, but the number of places I had to add it was huge relative to how little I expect comptime printing to be used. I thought the performance impact of dynamic dispatch should be negligible, as this is most likely used for debugging, and this way the code churn is kept to a minimum.

I ended up using Rc<RefCell<dyn std::io::Write>> so that the we can share the ownership over something writeable between the test and the Context without losing type information, which is what would happen if we used Box and passed it over to Context, then took it back.

Discrepancies

Run the test with the following command:

cargo test -p noir_ast_fuzzer_fuzz comptime_vs_brillig

The first test failure shows our formatting is different:

thread 'targets::comptime_vs_brillig::tests::fuzz_with_arbtest' panicked at tooling/ast_fuzzer/fuzz/src/targets/mod.rs:46:18:
called `Result::unwrap()` on an `Err` value: programs disagree on printed output:
---
[278759899407742179878585973846349345507, 172403831841634215884590204055802967033]
106748601

--- != ---
[0xd1b7342960b48d8b50bde6d8d02a72e3, 0x81b3c52d9a442b048766f07cd2d173f9]
106748601

---

This was fixed by changing the formatting of `Field` in the interpreter, and then by including the type information in `let`.

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 'Execution Time'.
Benchmark result of this commit is worse than the previous benchmark result exceeding threshold 1.20.

Benchmark suite	Current: 4647a8a	Previous: 176cfd4	Ratio
rollup-merge	0.004 s	0.003 s	1.33

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

CC: @TomAFrench

[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: d66b8d3	Previous: d3cb2b2	Ratio
test_report_AztecProtocol_aztec-packages_noir-projects_noir-protocol-circuits_crates_blob	2 s	1 s	2
test_report_AztecProtocol_aztec-packages_noir-projects_noir-protocol-circuits_crates_rollup-lib	2 s	1 s	2

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

CC: @TomAFrench

[TomAFrench]

LGTM