In Files

  • webrick/cgi.rb

Parent

WEBrick::CGI

A CGI library using WEBrick requests and responses.

Example:

class MyCGI < WEBrick::CGI
  def do_GET req, res
    res.body = 'it worked!'
    res.status = 200
  end
end

MyCGI.new.start

Constants

CGIError

The CGI error exception class

Attributes

config[R]

The CGI configuration. This is based on WEBrick::Config::HTTP

logger[R]

The CGI logger

Public Class Methods

new(*args) click to toggle source

Creates a new CGI interface.

The first argument in args is a configuration hash which would update WEBrick::Config::HTTP.

Any remaining arguments are stored in the @options instance variable for use by a subclass.

 
               # File webrick/cgi.rb, line 55
def initialize(*args)
  if defined?(MOD_RUBY)
    unless ENV.has_key?("GATEWAY_INTERFACE")
      Apache.request.setup_cgi_env
    end
  end
  if %r{HTTP/(\d+\.\d+)} =~ ENV["SERVER_PROTOCOL"]
    httpv = $1
  end
  @config = WEBrick::Config::HTTP.dup.update(
    :ServerSoftware => ENV["SERVER_SOFTWARE"] || "null",
    :HTTPVersion    => HTTPVersion.new(httpv || "1.0"),
    :RunOnCGI       => true,   # to detect if it runs on CGI.
    :NPH            => false   # set true to run as NPH script.
  )
  if config = args.shift
    @config.update(config)
  end
  @config[:Logger] ||= WEBrick::BasicLog.new($stderr)
  @logger = @config[:Logger]
  @options = args
end
            

Public Instance Methods

[](key) click to toggle source

Reads key from the configuration

 
               # File webrick/cgi.rb, line 81
def [](key)
  @config[key]
end
            
service(req, res) click to toggle source

Services the request req which will fill in the response res. See WEBrick::HTTPServlet::AbstractServlet#service for details.

 
               # File webrick/cgi.rb, line 156
def service(req, res)
  method_name = "do_" + req.request_method.gsub(/-/, "_")
  if respond_to?(method_name)
    __send__(method_name, req, res)
  else
    raise HTTPStatus::MethodNotAllowed,
          "unsupported method `#{req.request_method}'."
  end
end
            
start(env=ENV, stdin=$stdin, stdout=$stdout) click to toggle source

Starts the CGI process with the given environment env and standard input and output stdin and stdout.

 
               # File webrick/cgi.rb, line 89
def start(env=ENV, stdin=$stdin, stdout=$stdout)
  sock = WEBrick::CGI::Socket.new(@config, env, stdin, stdout)
  req = HTTPRequest.new(@config)
  res = HTTPResponse.new(@config)
  unless @config[:NPH] or defined?(MOD_RUBY)
    def res.setup_header
      unless @header["status"]
        phrase = HTTPStatus::reason_phrase(@status)
        @header["status"] = "#{@status} #{phrase}"
      end
      super
    end
    def res.status_line
      ""
    end
  end

  begin
    req.parse(sock)
    req.script_name = (env["SCRIPT_NAME"] || File.expand_path($0)).dup
    req.path_info = (env["PATH_INFO"] || "").dup
    req.query_string = env["QUERY_STRING"]
    req.user = env["REMOTE_USER"]
    res.request_method = req.request_method
    res.request_uri = req.request_uri
    res.request_http_version = req.http_version
    res.keep_alive = req.keep_alive?
    self.service(req, res)
  rescue HTTPStatus::Error => ex
    res.set_error(ex)
  rescue HTTPStatus::Status => ex
    res.status = ex.code
  rescue Exception => ex
    @logger.error(ex)
    res.set_error(ex, true)
  ensure
    req.fixup
    if defined?(MOD_RUBY)
      res.setup_header
      Apache.request.status_line = "#{res.status} #{res.reason_phrase}"
      Apache.request.status = res.status
      table = Apache.request.headers_out
      res.header.each{|key, val|
        case key
        when /^content-encoding$/i
          Apache::request.content_encoding = val
        when /^content-type$/i
          Apache::request.content_type = val
        else
          table[key] = val.to_s
        end
      }
      res.cookies.each{|cookie|
        table.add("Set-Cookie", cookie.to_s)
      }
      Apache.request.send_http_header
      res.send_body(sock)
    else
      res.send_response(sock)
    end
  end
end
            

Commenting is here to help enhance the documentation. For example, code samples, or clarification of the documentation.

If you have questions about Ruby or the documentation, please post to one of the Ruby mailing lists. You will get better, faster, help that way.

If you wish to post a correction of the docs, please do so, but also file bug report so that it can be corrected for the next release. Thank you.

If you want to help improve the Ruby documentation, please visit Documenting-ruby.org.

blog comments powered by Disqus