ACSetSpec PEG.jl-based Implemenation

[cscaff]

The following PR contains two main files: src/parsers.jl as well as test/parsers.jl. This contains a module for constructing a parse expression grammar for creating an ADT representation from an ACSet specification. The use of PEG.jl allows us to build ASTs that are not constrained to the way Julia treats code as data. We can define our expressions in our tree the way we see fit. I have also included a testing module to ensure that the parser functions properly. The new parser can be called by a string macro: "acsetspec_str". For instance, we can call:

acsetspec"""
    LabeledGraph{Symbol}
    begin
      V(label=a)
      V(label=b)
      V(label=c)
      E(src=1,tgt=3)
      E(src=2,tgt=3)
    end"""

[lukem12345]

I noticed that the function parse_identifier happened to always be passed Chars, so I devised a test for the case in which an identifier is more than one character.

This test fails:

@test args("foo, b ,c")[1] == ([Value(:foo), Value(:b), Value(:c)])

The LHS evaluates to ACSets.ADTs.Value{Symbol}[f, b, c].

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 91.72%. Comparing base (0c5b066) to head (aaaed94).
Report is 21 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #153      +/-   ##
==========================================
- Coverage   92.06%   91.72%   -0.35%     
==========================================
  Files          23       24       +1     
  Lines        2118     2139      +21     
==========================================
+ Hits         1950     1962      +12     
- Misses        168      177       +9     

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

[lukem12345]

@jpfairbanks It looks like @quffaro and I both approve. Merge?

[jpfairbanks]

Merge