Do some stack work to speed up Rhino hotspots

[adamhathcock]

Part of https://linear.app/speckle/issue/CNX-1324/improving-send-and-receive-perf

Tasked to do some profiling.

First change is to a converter hotspot that seems to speed up by doing some work on the stack then allocating for arrays instead of doing it upfront:

before:

after:

I didn't see a real difference in allocates but probably saves some capacity setting. It wouldn't expect a big difference because it is ultimately allocating for these values.

BrepEncoding is a big allocator:

before:

after:

It can probably be reduced by only doing the byte array then the serializer knows what to do with it (base64 later) to reduce memory usage

More to come...

[codecov]

Codecov Report

Attention: Patch coverage is 0% with 25 lines in your changes missing coverage. Please review.

Project coverage is 8.71%. Comparing base (fe5b77e) to head (094ecc3).
Report is 1 commits behind head on dev.

Files with missing lines	Patch %	Lines
...hinoShared/ToSpeckle/Raw/MeshToSpeckleConverter.cs	0.00%	25 Missing ⚠️

Additional details and impacted files

@@           Coverage Diff            @@
##             dev    #590      +/-   ##
========================================
- Coverage   8.71%   8.71%   -0.01%     
========================================
  Files        227     227              
  Lines       4370    4374       +4     
  Branches     545     546       +1     
========================================
  Hits         381     381              
- Misses      3974    3978       +4     
  Partials      15      15              

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[linear]

CNX-1324 Improving Send and Receive perf