Python scandir

results.md

This is on OSX:

$ time ruby /tmp/scan.rb .
21953
29899

real	0m3.448s
user	0m0.713s
sys	0m2.712s
$ time ruby /tmp/scan.rb .
21953
29899

real	0m4.728s
user	0m0.747s
sys	0m3.948s

Interestingly Python find a few more files and dirs, I suspect the semantic of Dir.Entry.is_dir() doesn't quite match File.directory?

$ time python3 /tmp/scan.py .
22655
29905

real	0m1.118s
user	0m0.367s
sys	0m0.729s
$ time python3 /tmp/scan.py .
22655
29905

real	0m1.123s
user	0m0.371s
sys	0m0.737s

scan.py

import os, sys
import timeit

def call(path):
    path = os.path.abspath(path)
    dirs = []
    requirables = []

    for entry in walk(path):
        if entry.is_dir():
            dirs.append(entry.path)
        elif entry.name.endswith('.rb'):
            requirables.append(entry.path)
    return [dirs, requirables]

def walk(absolute_dir_path):
    for entry in os.scandir(absolute_dir_path):
        if entry.name.startswith('.'):
            next
        yield entry
        if entry.is_dir():
            yield from walk(entry.path)


dirs, files = call(sys.argv[1])
print(len(dirs))
print(len(files))

scan.rb

def call(path)
  path = File.expand_path(path.to_s).freeze
  dirs = []
  requirables = []
  walk(path) do |absolute_path, is_directory|
    if is_directory
      dirs << absolute_path
    elsif absolute_path.end_with?('.rb')
      requirables << absolute_path
    end
  end
  [requirables, dirs]
end

def walk(absolute_dir_path, &block)
  Dir.each_child(absolute_dir_path) do |name|
    next if name.start_with?('.')
    absolute_path = "#{absolute_dir_path}/#{name}"
    if File.directory?(absolute_path)
      yield absolute_path, true
      walk(absolute_path, &block)
    else
      yield absolute_path, false
    end
  end
end

files, dirs = call(ARGV.first)
p dirs.size
p files.size