Added a unit type

[g-r-a-n-t]

What was wrong?

closes #293
closes #407

We had two separate representations of empty tuples in the Yul output. One being 0x0 and the other being a function call with no return. This led to the issue discovered in #293.

How was it fixed?

1.) Enabled empty struct definitions using pass (also did events and contracts).
2.) Lowered empty tuples to a struct type named empty_tuple.
3.) Removed all tuple mappings in the Yul codegen stage.

To-Do

• OPTIONAL: Update Spec if applicable

• Add entry to the release notes (may forgo for trivial changes)

• Clean up commit history

[g-r-a-n-t]

decided Im going to hold off on this and wait for zero-sized structs so we can full implement tuple lowering

[codecov-commenter]

Codecov Report

Merging #406 (0ab2a1f) into master (934ffb5) will increase coverage by 0.12%.
The diff coverage is 89.58%.

❗ Current head 0ab2a1f differs from pull request most recent head 862d6f1. Consider uploading reports for the commit 862d6f1 to get more accurate results

@@            Coverage Diff             @@
##           master     #406      +/-   ##
==========================================
+ Coverage   86.07%   86.20%   +0.12%     
==========================================
  Files          67       67              
  Lines        4008     4110     +102     
==========================================
+ Hits         3450     3543      +93     
- Misses        558      567       +9     

Impacted Files	Coverage Δ
analyzer/src/operations.rs	83.58% <0.00%> (-1.27%)	⬇️
compiler/src/yul/mappers/expressions.rs	88.67% <50.00%> (+0.21%)	⬆️
analyzer/src/namespace/types.rs	85.05% <62.50%> (-2.59%)	⬇️
analyzer/src/context.rs	92.30% <66.66%> (+0.32%)	⬆️
compiler/src/lowering/mappers/functions.rs	98.91% <96.15%> (-1.09%)	⬇️
analyzer/src/traversal/expressions.rs	92.71% <100.00%> (+0.40%)	⬆️
analyzer/src/traversal/functions.rs	99.15% <100.00%> (+0.37%)	⬆️
compiler/src/abi/elements.rs	84.61% <100.00%> (ø)
compiler/src/lowering/mappers/expressions.rs	100.00% <100.00%> (ø)
compiler/src/lowering/mappers/types.rs	100.00% <100.00%> (ø)
... and 14 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 934ffb5...862d6f1. Read the comment docs.

[sbillig]

@g-r-a-n-t It's maybe a bit late, but what if we introduce a Unit type (https://doc.rust-lang.org/std/primitive.unit.html) which just happens to resemble a tuple with no elements, and disallow empty tuples? We could use https://crates.io/crates/nonempty or https://crates.io/crates/vec1 to enforce the non-emptiness of the tuple elements vec.

That seems like a more aesthetically pleasing solution than lowering to an empty struct with a special name.

[g-r-a-n-t]

@sbillig Yeah, I think that makes sense. The is_empty_tuple functions are also something that I'm not a big fan of.

I'd be happy to try it out in this PR.

[g-r-a-n-t]

Added a unit type @sbillig, this is ready for a review.

[sbillig]

👍