Explicitly import Meshes.Geometry

[mikeingold]

Changes

• Explicitly import Meshes.Geometry. This is a frequently-used reference with a fairly obvious namespace source, and being able to reference it implicitly will reduce function signature line lengths.
• Update method signatures, eliminating unnecessary where clauses. I'm told that there's no functional/performance difference between a function argument being f(x::SuperType) vs f(x::ST) where {ST <: SuperType}.

[mikeingold]

This partially conflicts with #114 (as-is at the moment). I wanted to shrink the line length of _units to fit it into a single-line definition and had this idea, but implementing this for the whole package felt out-of-scope for that PR. If this PR is merged, I'll update the #114 branch to merge it with these changes.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Please upload report for BASE (main@93a9b1d). Learn more about missing BASE report.

Additional details and impacted files

@@           Coverage Diff           @@
##             main     #118   +/-   ##
=======================================
  Coverage        ?   95.75%           
=======================================
  Files           ?       17           
  Lines           ?      259           
  Branches        ?        0           
=======================================
  Hits            ?      248           
  Misses          ?       11           
  Partials        ?        0           

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

[mikeingold]

I'm continuing to consider condensing all of our function definitions to use single-line headers, but this currently would still require splitting somewhere into multiple lines to respect the 92-column soft limit. I'm hoping that this change would also make that change more practical.