Benchmark interpret vs compiled code evaluation

[ReubenJ]

As I was working on HerbBenchmarks.jl I realized we might be spending a lot of our time interpreting built expressions (candidate programs). This PR is a start at benchmarking our default interpreter so that we have some concrete numbers to talk about when making changes to it.

So far, the PR makes the package compliant with PackageBenchmark.jl's structure it expects, making the following possible:

using Pkg
Pkg.activate(; temp=true)
Pkg.add("PkgBenchmark")
Pkg.add("BenchmarkTools")
Pkg.add(PackageSpec(name="HerbInterpret", rev="benchmark-interpret-vs-compiled"))
using PkgBenchmark

import HerbInterpret

benchmarkpkg(HerbInterpret)

TODO:

• Add more expressions to benchmark
• Add GH action to automatically run benchmarks on PRs and comment the results (see either the GitHub action for continuous benchmarking or BenchmarkCI.jl for possible ways to move forward)

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 28.57%. Comparing base (584ea42) to head (c3d295a).

Additional details and impacted files

@@           Coverage Diff           @@
##           master      #14   +/-   ##
=======================================
  Coverage   28.57%   28.57%           
=======================================
  Files           2        2           
  Lines         105      105           
=======================================
  Hits           30       30           
  Misses         75       75           

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