Optimize value coercion

[JoviDeCroock]

Optimize Argument Coercion Performance

This PR attempts to optimize the argument coercion process in GraphQL-Ruby by adding fast paths for common scenarios. I thought I'd create it as a draft so folks could give input and the CI would run.

Looks like I was a bit too eager, looking into the tests etc

Changes

1. List Type Coercion (schema/list.rb)
   • Added fast path to avoid unnecessary array creation when input is already an array
   • Special handling for scalar inputs that need conversion to arrays
2. Arguments Cache (execution/interpreter/arguments_cache.rb)
   • Added optimization to skip argument materialization for empty arguments
   • Fast path when argument owner has no arguments defined
3. Argument Handling (schema/member/has_arguments.rb)
   • Added fast path for simple cases with small number of arguments (0-2)
   • Simplified execution flow for cases without lazy values
   • Added method to detect if arguments have default values
4. Argument Coercion (schema/argument.rb)
   • Optimized coerce_into_values with a fast path for statically coercible arguments
   • Direct value handling for simple cases without loading or preparation

Benchmarks

Performance improvements measured with isolated benchmarks, these were custom made as I couldn't get the benchmarks of the repository to run well 😅 :

• List Coercion:

  • Array input: ~6% improvement
  • Scalar input: ~48% improvement

• Argument Coercion:

  • Empty arguments: ~74% improvement
  • Simple arguments (1-2 args): ~8% improvement
  • Complex arguments (5+ args): ~5% improvement

[rmosolgo]

Hey, thanks for looking into these! If you're interested in pursuing them further, I'd be happy to help debug test failures -- but it'd be a lot easier if you could open PRs for the one-by-one, just so there wouldn't be quite as many moving parts involved.