Turn a Rails controller into a web-proxy (web.archive.org-style)

proxy_controller.rb

require 'pp'

class ProxyController < ApplicationController

  # Example headers:
  #   {"GATEWAY_INTERFACE"=>"CGI/1.1",
  #    "PATH_INFO"=>"/proxy/http:/google.com",
  #    "QUERY_STRING"=>"",
  #    "REMOTE_ADDR"=>"127.0.0.1",
  #    "REMOTE_HOST"=>"127.0.0.1",
  #    "REQUEST_METHOD"=>"GET",
  #    "REQUEST_URI"=>"http://localhost:3000/proxy/http://google.com/",
  #    "SCRIPT_NAME"=>"",
  #    "SERVER_NAME"=>"localhost",
  #    "SERVER_PORT"=>"3000",
  #    "SERVER_PROTOCOL"=>"HTTP/1.1",
  #    "SERVER_SOFTWARE"=>"WEBrick/1.3.1 (Ruby/2.0.0/2013-06-27)",
  #    "HTTP_USER_AGENT"=>
  #     "Opera/9.80 (X11; Linux x86_64) Presto/2.12.388 Version/12.16",
  #    "HTTP_HOST"=>"localhost:3000",
  #    "HTTP_ACCEPT"=>
  #     "text/html, application/xml;q=0.9, application/xhtml+xml, image/png, image/webp, image/jpeg, image/gif, image/x-xbitmap, */*;q=0.1",
  #    "HTTP_ACCEPT_LANGUAGE"=>"en-US,en;q=0.9",
  #    "HTTP_ACCEPT_ENCODING"=>"gzip, deflate",
  #    "HTTP_COOKIE"=>
  #     "remember_user_token=BAhbB1sGaQZJIiIkMmEkMTAkU2E2NjJYTDIyVUxYYlVXbnhEM3dWTwY6BkVU--66f2f6480b5e1a7e1293959ba4653ee9f1a48311; _hibar_session=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%3D--579c8e19bafc03e0f6e8f92639161fe22901ca76",
  #    "HTTP_CACHE_CONTROL"=>"no-cache",
  #    "HTTP_CONNECTION"=>"Keep-Alive",
  #    "HTTP_VERSION"=>"HTTP/1.1",
  #    "REQUEST_PATH"=>"/proxy/http://google.com/",
  #    "ORIGINAL_FULLPATH"=>"/proxy/http://google.com/",
  #    "ORIGINAL_SCRIPT_NAME"=>"",
  #    "ROUTES_25203060_SCRIPT_NAME"=>""}

  def index
    # request variables:
    #   referer
    #   base_url
    #   original_fullpath
    #   host
    #   port
    #   fullpath
    #   url
    #   host_with_port
    #   original_url

    http = ::HTTPClient.new(
      cookie_file:      current_user.cookie_file, 
      verbose:          true,
      follow_redirects: false,
      raise_exceptions: false,
      logger:           Rails.logger
    )

    logger.info "HTTP Referer: #{request.referer}"

    # binding.pry

    # logger.info "Rack environment:\n" + request.env.select{|k,v| v =~ /proxy/ }.pretty_inspect

    if request.original_fullpath =~ %r{^/proxy/(.+)}
      url = $1
      url += "?#{request.query_string}" unless request.query_string.blank?
    else
      raise "original_fullpath is broken: #{request.original_fullpath}"
    end

    # loop do
    #   puts "* Downloading: #{url}"
    #   page = http.get(url)
    #   puts "* Info: #{page.link_info}"

    #   if page.redirect
    #     puts "* Following redirect: #{page.redirect}"
    #     url = page.redirect
    #   else
    #     break
    #   end
    # end

    puts "* Downloading: #{url}"
    page = http.get(url)

    puts "* Info: #{page.link_info}"

    if page.mime_type[/^text\/html/]
      # # set base url
      # # head.add_child("<base href=\"#{url}\">")

      rewrite_urls!(page, url)
      body = page.parser.to_html
    else
      body = page.body
    end

    if page.redirect
      logger.info "redirecting_to: #{page.redirect}"
      redirect_to proxify_url(url, page.redirect)
    else
      render text: body, content_type: page.mime_type, status: page.code
    end
  end

protected
  
  def fix_url(url)
    url.gsub(" ", "%20")
  end

  def proxify_url(current_url, link)
    current_url = URI.parse(fix_url(current_url)) unless current_url.is_a? URI
    "/proxy/#{current_url + fix_url(link)}"
  end
  
  #
  # Rewrite all urls on the page so that they're absolute urls (relative to "base"),
  # and prefixed by "/proxy"
  #
  def rewrite_urls!(page, current_url)
    # Add a <base href=""> tag
    # page.at("head").children.before("<base href=\"#{base_url}\">")

    current_url = URI.parse fix_url(current_url)

    # Replace other tags...

    tag_map = {
      "a"      => "href",
      "img"    => "src",
      "form"   => "action",
      "script" => "src",
      "link"   => "href"
    }

    for tagname, attrname in tag_map

      page.search(tagname).each do |node|
        if tag_url = node[attrname]
          tag_url = fix_url(tag_url)
          begin
            new_url = proxify_url(current_url, tag_url)
            node[attrname] = new_url
          rescue URI::InvalidURIError => e
            logger.error e.inspect
            node[attrname] = "javascript:alert('Unable to parse original url: #{tag_url}');"
          end
        end
      end

    end

    if meta = page.at("meta") and meta["http-equiv"] == "refresh"
      #<meta http-equiv="refresh" content="0; url=http://tinyurl.com/pemjju3">
      meta["content"] = meta["content"].gsub(/(?<=\burl=)(.+)(?=\s|$)/) do |m|
        proxify_url current_url, $1
      end
    end

  end

end