class Mongrel::Rails::RailsHandler

Implements a handler that can run Rails and serve files out of the Rails application's public directory. This lets you run your Rails application with Mongrel during development and testing, then use it also in production behind a server that's better at serving the static files.

The RailsHandler takes a mime_map parameter which is a simple suffix=mimetype mapping that it should add to the list of valid mime types.

It also supports page caching directly and will try to resolve a request in the following order:

This means that if you are using page caching it will actually work with Mongrel and you should see a decent speed boost (but not as fast as if you use a static server like Apache or Litespeed).

Attributes

files[R]
guard[R]

Public Class Methods

new(dir, mime_map = {}) click to toggle source
# File lib/mongrel/rails.rb, line 37
def initialize(dir, mime_map = {})
  @files = Mongrel::DirHandler.new(dir,false)
  @guard = Mutex.new

  # Register the requested MIME types
  mime_map.each {|k,v| Mongrel::DirHandler::add_mime_type(k,v) }
end

Public Instance Methods

process(request, response) click to toggle source

Attempts to resolve the request as follows:

  • If the requested exact PATH_INFO exists as a file then serve it.

  • If it exists at PATH_INFO+“.html” exists then serve that.

  • Finally, construct a Mongrel::CGIWrapper and run Dispatcher.dispatch to have Rails go.

# File lib/mongrel/rails.rb, line 50
def process(request, response)
  return if response.socket.closed?
  
  path_info = request.params[Mongrel::Const::PATH_INFO]
  rest_operator = request.params[Mongrel::Const::REQUEST_URI][/^#{Regexp.escape path_info}(;[^\?]+)/, 1].to_s
  path_info.chomp!("/")
  
  page_cached = path_info + rest_operator + ActionController::Base.page_cache_extension
  get_or_head = @@file_only_methods.include? request.params[Mongrel::Const::REQUEST_METHOD]

  if get_or_head and @files.can_serve(path_info)
    # File exists as-is so serve it up
    @files.process(request,response)
  elsif get_or_head and @files.can_serve(page_cached)
    # Possible cached page, serve it up
    request.params[Mongrel::Const::PATH_INFO] = page_cached
    @files.process(request,response)
  else
    begin
      cgi = Mongrel::CGIWrapper.new(request, response)
      cgi.handler = self
      # We don't want the output to be really final until we're out of the lock
      cgi.default_really_final = false

      @guard.synchronize {
        @active_request_path = request.params[Mongrel::Const::PATH_INFO] 
        Dispatcher.dispatch(cgi, ActionController::CgiRequest::DEFAULT_SESSION_OPTIONS, response.body)
        @active_request_path = nil
      }

      # This finalizes the output using the proper HttpResponse way
      cgi.out("text/html",true) {""}
    rescue Errno::EPIPE
      response.socket.close
    rescue Object => rails_error
      STDERR.puts "#{Time.now}: Error calling Dispatcher.dispatch #{rails_error.inspect}"
      STDERR.puts rails_error.backtrace.join("\n")
    end
  end
end
reload!() click to toggle source

Does the internal reload for Rails. It might work for most cases, but sometimes you get exceptions. In that case just do a real restart.

# File lib/mongrel/rails.rb, line 93
def reload!
  begin
    @guard.synchronize {
      $".replace $orig_dollar_quote
      GC.start
      Dispatcher.reset_application!
      ActionController::Routing::Routes.reload
    }
  end
end