Feat: Added private type leak validation to structs

[mkhan45]

Also exposed struct field visibility in the HIR and created a struct validation module

[codecov]

Codecov Report

Merging #344 (078c93a) into main (69594b0) will increase coverage by 0.01%.
The diff coverage is 91.83%.

@@            Coverage Diff             @@
##             main     #344      +/-   ##
==========================================
+ Coverage   80.27%   80.29%   +0.01%     
==========================================
  Files         259      262       +3     
  Lines       15636    15671      +35     
==========================================
+ Hits        12552    12583      +31     
- Misses       3084     3088       +4     

Impacted Files	Coverage Δ
crates/mun_hir/src/code_model.rs	11.11% <ø> (ø)
crates/mun_hir/src/expr/validator/tests.rs	100.00% <ø> (ø)
crates/mun_hir/src/code_model/struct/validator.rs	84.00% <84.00%> (ø)
crates/mun_hir/src/code_model/struct.rs	89.36% <100.00%> (+0.47%)	⬆️
...s/mun_hir/src/code_model/struct/validator/tests.rs	100.00% <100.00%> (ø)
crates/mun_hir/src/utils.rs	100.00% <100.00%> (ø)
crates/mun_hir/src/ty/infer.rs	70.04% <0.00%> (-0.14%)	⬇️
... and 1 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 69594b0...078c93a. Read the comment docs.

[mkhan45]

Both are required to be pub(crate) since they're used by mun_hir::src::utils::tests, which I extracted out of expr::tests to make common to the struct validator and expression validator tests.

The first force-push was because I'd removed the (crate) part of pub(crate) from both of them instead of the whole pub(crate), and then the second force-push is from when I turned my brain on and realized that that's not what we want.