Speed up lookup of possible types for Fragments

[amomchilov]

Really this is just a readability improvement, but there's a nice perf side effect, too :)

This change speeds up rake bench:profile_large_result benchmark by about 1.8%.

Baseline (master at cbdae76)

Warming up --------------------------------------
Querying for 1000 objects
                         1.000  i/100ms
Calculating -------------------------------------
Querying for 1000 objects
                          8.149  (± 0.0%) i/s -    815.000  in 100.061220s

After this change:

Warming up --------------------------------------
Querying for 1000 objects
                         1.000  i/100ms
Calculating -------------------------------------
Querying for 1000 objects
                          8.295  (± 0.0%) i/s -    830.000  in 100.093047s

---

This logic used to be more complicated (t.metadata[:type_class] == owner_type), but has since been simplified to doing a search for an exact-match (t == owner_type).

graphql-ruby/lib/graphql/execution/interpreter/runtime.rb

Lines 83 to 89 in aec347f

	# Faster than .map{}.include?()
	query.warden.possible_types(type_defn).each do |t|
	if t.metadata[:type_class] == owner_type
	gather_selections(owner_type, node.selections, selections_by_name)
	break
	end
	end

include? can do this more expressively than each + break, and has the added benefit of being way faster (since it can run the loop body in C code, rather than needing to call into a Ruby block). Here's a dumb little synthetic benchmark to demonstrate that:

Benchmark

#!/usr/bin/ruby

require "benchmark/ips"

a = Array(1...100)

def found_it!; end

Benchmark.ips do |x|
  x.time = 30
  x.warmup = 10
  x.time = 1
  x.warmup = 1
  
  x.report("each+break") do |times|
    i = 0
    while i < times
      i += 1
    
      a.each do |x|
        if x == 90
          found_it!
          break
        end
      end
    
    end
  end
  
  x.report("include?") do |times|
    i = 0
    while i < times
      i += 1
      
      if a.include?(90)
        found_it!
      end
      
    end
  end
  
  x.compare!
end

Warming up --------------------------------------
          each+break    32.413k i/100ms
            include?   264.562k i/100ms
Calculating -------------------------------------
          each+break    322.606k (± 2.6%) i/s -    324.130k in   1.005439s
            include?      2.631M (± 1.1%) i/s -      2.646M in   1.005860s

Comparison:
            include?:  2630515.3 i/s
          each+break:   322606.5 i/s - 8.15x  slower

[rmosolgo]

Nice find! Thanks for sharing this improvement.