Follow redirect links

redirect.rb

require 'net/https'
require 'uri'

=begin
This simple module provides two methods to follow, and return, the redirects of a URL
It goes to a depth of 10 unless specified otherwise.
Redirect.redirect_urls(<url>) returns a dictionary with the following keys:
:completed : true if reached final direct before hitting limit
:uris : a list of URI structures. The first will be the final destination
:urls: a string version of the above
:hosts: the unique hosts of the URIs

Redirect.resolve(<url>) returns the last resolved URL.

Redirect.redirect_urls("entish.org") => dictionary as above 
Redirect.resolve("entish.org") => "http://entish.org"

By design, this doesn't do much error checking, though it does try
to add the Scheme ("http") and reuse URL information on relocations.

It also uses HEAD method, and sets a User-agent of "Ruby redirect script"
=end


module Redirect
  def self.redirect_urls(url, options = {})
    redirect_lookup_depth = options[:depth].to_i > 0 ? options[:depth].to_i : 10
    current_uri = URI.parse(url)
    current_uri = URI.parse('http://' + url) if current_uri.scheme.nil?
    redirs = get_redirects(current_uri, [current_uri], redirect_lookup_depth, redirect_lookup_depth)
    redirs[:urls] = redirs[:uris].map(&:to_s)
    redirs[:hosts] = redirs[:uris].map(&:host).uniq
    redirs
  end

  def self.resolve(url)
    self.redirect_urls(url)[:urls][0]
  end

  private

  def self.get_redirects(current_uri, uris, limit, limit_count)
    return { completed: false, uris: uris } if limit_count < 1
    http = Net::HTTP.new(current_uri.host, current_uri.port)
    http.use_ssl = true if current_uri.scheme == 'https'
    request = Net::HTTP::Head.new(current_uri.request_uri)
    request.initialize_http_header('User-Agent' => 'Ruby redirect script')
    response = http.request(request)
    case response
    when Net::HTTPSuccess then
      return { completed: true, uris: uris }
    when Net::HTTPRedirection then
      redirect_location = response['location']
      location_uri = URI.parse(redirect_location)
      if location_uri.host.nil?
        location_uri = URI.parse(uri.scheme + '://' + uri.host + redirect_location)
      end
      # puts("Redirecting from #{current_uri} to #{location_uri}")
      get_redirects(location_uri, [location_uri] + uris, limit, limit - 1)
    else
      raise 'Non-success/redirect response: ' + response.inspect
    end
  end
end

>> Redirect.redirect_urls("https://t.co/TSoXOTDBYd")
=> {:completed=>true,
       :uris=>[#<URI::HTTP http://twistedsifter.com/2017/05/subway-maps-compared-to-their-actual-geography/>, #<URI::HTTP http://bit.ly/2qZ9oFw>, #<URI::HTTPS https://t.co/TSoXOTDBYd>], 
       :urls=>["http://twistedsifter.com/2017/05/subway-maps-compared-to-their-actual-geography/", "http://bit.ly/2qZ9oFw", "https://t.co/TSoXOTDBYd"], 
      :hosts=>["twistedsifter.com", "bit.ly", "t.co"]}

>> Redirect.resolve("https://t.co/TSoXOTDBYd")
=> "http://twistedsifter.com/2017/05/subway-maps-compared-to-their-actual-geography/"

Oh, this is sooooo excellent.