get_directory_names_under_app.rb

require 'csv'
require 'uri'
require 'net/http'
require 'json'

CSV_REPOS_DATA = <<EOS
discourse,discourse,https://github.com/discourse/discourse
gitlabhq,gitlabhq,https://github.com/gitlabhq/gitlabhq
huginn,huginn,https://github.com/huginn/huginn
diaspora,diaspora,https://github.com/diaspora/diaspora
tootsuite,mastodon,https://github.com/tootsuite/mastodon
errbit,errbit,https://github.com/errbit/errbit
fatfreecrm,fat_free_crm,https://github.com/fatfreecrm/fat_free_crm
edavis10,redmine,https://github.com/edavis10/redmine
Netflix,Scumblr,https://github.com/Netflix/Scumblr
locomotivecms,engine,https://github.com/locomotivecms/engine
rubygems,rubygems.org,https://github.com/rubygems/rubygems.org
opf,openproject,https://github.com/opf/openproject
houndci,hound,https://github.com/houndci/hound
loomio,loomio,https://github.com/loomio/loomio
jcs,lobsters,https://github.com/jcs/lobsters
EOS

repositories = CSV.parse(CSV_REPOS_DATA)

class ApiClient
  def initialize(access_token)
    @access_token = access_token
  end
  
  def get(url)
    url = URI.parse(url)
    req = Net::HTTP::Get.new(url)
    req.add_field("Authorization", "token #{@access_token}")
    
    Net::HTTP.new(url.host, url.port).tap { |http| http.use_ssl = true}.start do |http|
      http.request(req)
    end
  end
end
api_client = ApiClient.new(ENV['GITHUB_API_ACCESS_TOKEN'])

result = repositories.map do |owner, repo, _|
  response = api_client.get("https://api.github.com/repos/#{owner}/#{repo}/contents/app")
  directories = JSON.parse(response.body).select { |item| item['type'] == 'dir' }
  directory_names = directories.map { |d| d['name'] }
  ["#{owner}/#{repo}", directory_names]
end

summary =  Hash.new { |h, k| h[k] = [] }
result.each do |repo, dirs|
  dirs.each do |dir|
    summary[dir] << repo
  end
end

summary.each do |key, dirs|
  puts "#{key} #{dirs.length} #{dirs.join(',')}"
end