output.txt

{["test.rb", 9, :T_STRING]=>[9, 0, 0, 0, 0, 0],
 ["test.rb", 9, :T_ARRAY]=>[1, 0, 0, 0, 0, 0],
 ["test.rb", 9, :T_REGEXP]=>[1, 0, 0, 0, 0, 0]}

-------------------------------------------------

{["test.rb", 13, :T_STRING]=>[5, 0, 0, 0, 0, 0],
 ["test.rb", 13, :T_REGEXP]=>[1, 0, 0, 0, 0, 0]}

Calculating -------------------------------------
             Regular    18.013k i/100ms
          Regexp new    21.582k i/100ms
-------------------------------------------------
             Regular    198.856k (± 3.2%) i/s -      1.009M
          Regexp new    240.396k (± 3.7%) i/s -      1.209M

Comparison:
          Regexp new:   240395.9 i/s
             Regular:   198855.5 i/s - 1.21x slower

test.rb

require 'objspace'
require 'allocation_tracer'
require 'pp'
require 'benchmark/ips'

ObjectSpace::AllocationTracer.setup(%i{path line type})
a = :content
result = ObjectSpace::AllocationTracer.trace do
  /\A#{Regexp.escape(a)}\(\d+[if]?\)\z/
end

result2 = ObjectSpace::AllocationTracer.trace do
  Regexp.new("\\A#{Regexp.escape(a)}\\(\\d+[if]?\\)\\z")
end

pp result
puts "\n"
puts "-------------------------------------------------"
puts "\n"
pp result2

puts "\n"

Benchmark.ips do |x|
  x.report('Regular') do
    /\A#{Regexp.escape(a)}\(\d+[if]?\)\z/
  end

  x.report('Regexp new') do
    Regexp.new("\\A#{Regexp.escape(a)}\\(\\d+[if]?\\)\\z")
  end

  x.compare!
end