Chore: Remove Boxing of SymbolicExpressions

[ASuciuX]

Description

SymbolicExpressions and PreSymbolicExpressions are unnecessarily Boxed in a few places in the clarity crate when used in Vecs. As Vecs are already on the heap, a Box is not necessary.

Closes #4445

[codecov]

Codecov Report

Attention: Patch coverage is 93.40659% with 6 lines in your changes are missing coverage. Please review.

Project coverage is 83.07%. Comparing base (5150acf) to head (462356c).
Report is 41 commits behind head on next.

Additional details and impacted files

@@            Coverage Diff             @@
##             next    #4530      +/-   ##
==========================================
- Coverage   83.29%   83.07%   -0.22%     
==========================================
  Files         452      452              
  Lines      326071   326068       -3     
  Branches      323      323              
==========================================
- Hits       271589   270876     -713     
- Misses      54474    55184     +710     
  Partials        8        8              

Files	Coverage Δ
clarity/src/vm/ast/parser/v1.rs	90.93% <100.00%> (-0.02%)	⬇️
clarity/src/vm/ast/sugar_expander/mod.rs	94.20% <100.00%> (ø)
clarity/src/vm/costs/mod.rs	81.22% <100.00%> (-0.54%)	⬇️
clarity/src/vm/mod.rs	88.86% <100.00%> (-0.26%)	⬇️
clarity/src/vm/representations.rs	82.28% <100.00%> (ø)
clarity/src/vm/ast/parser/v2/mod.rs	56.32% <33.33%> (+0.14%)	⬆️

... and 33 files with indirect coverage changes

---

Continue to review full report in Codecov by Sentry.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 5150acf...462356c. Read the comment docs.

[cylewitruk]

Lgtm :) This suggestion stemmed from playing with different binary serialization formats, which aren't as forgiving as JSON.

[jcnelson]

@kantai Is there a reason we were Box-ing these before?

[kantai]

Classic case of early optimization, I think. These structs never need to be resized, so boxed slices were maybe a closer match to their usage than vecs. No real reason that they need to be boxed slices instead of vecs, though, and I doubt that there's a performance benefit to keeping them as boxed slices (I wouldn't be surprised if it was the opposite).

[jcnelson]

If it's fine with @kantai it's fine with me 👍

[jbencin]

I doubt that there's a performance benefit to keeping them as boxed slices (I wouldn't be surprised if it was the opposite).

They compile to pretty much the same thing. A Boxed slice is represented on the stack with 2 elements: ptr and len. Vec has an additional element: capacity. So the cost of using a Vec is just an extra usize, which is trivial

[cylewitruk]

Boxed slices also had issues with several binary encodings I tried for these types, where Vecs were handled otob -- that's essentially the birth place of the issue, preparing for that kind of change.

[blockstack-devops]

This pull request has been automatically locked since there has not been any recent activity after it was closed. Please open a new issue for related bugs.