Question

I m using Ruby+Watir to request pages through Firefox.

I would like to record the headers and content of every http request made through the browser.

Would it be possible to configure a proxy solution to store this information, either in a file or pipe it straight into an application? Could I use something such as squid or nginx to record header/content information?

PS: Running Ubuntu x64.

Answer 1

If you don t want a proxy, you could just use tcpdump e.g. tcpdump -i en0 -n -s 0 -w output.pcap

You can then retrospectively look at all traffic in wireshark etc.

The BrowserMob proxy might be a good fit for you (runs as a jar file)

You could also roll your own in Ruby (but only works with HTTP, not HTTPS) e.g.

  require  rubygems 
  require  webrick/httpproxy 

  @proxy_port    = ARGV[0] || 9090

  # Optional flags
  @print_headers  = false
  @print_body     = true

  server = WEBrick::HTTPProxyServer.new(
    :Port => @proxy_port,
    :AccessLog => [],
    :ProxyContentHandler => Proc.new do |req,res|
        puts "-"*75
        puts ">>> #{req.request_line.chomp}
"
        req.header.keys.each { |key| puts "#{key.capitalize}: #{req.header[key]}" if @print_headers }
        puts "<<<" if @print_headers
        puts res.status_line if @print_headers
        res.header.keys.each { |key| puts "#{key.capitalize}: #{res.header[key]}" if @print_headers }
        puts res.body unless res.body.nil? or !@print_body
    end
  )
  trap("INT") { server.shutdown }
  server.start

Answer 2

For Windows there is a program called Fiddler that does exactly what you need, so I did a Google search for "Fiddler for Linux" and came up with Charles. Looks pretty strong.

I didn t notice the price tag for the non-trial version of Charles. Another app worth looking into is Poster, an add-on for Firefox. It is not clear to me whether it captures all traffic or only returns response for directly input requests, but still could help you with your project.

友情链接