Class: EventMachine::Protocols::HttpClient

Inherits:
Connection
  • Object
show all
Includes:
Deferrable
Defined in:
lib/em/protocols/httpclient.rb

Overview

Usage

EventMachine.run {
  http = EventMachine::Protocols::HttpClient.request(
    :host => server,
    :port => 80,
    :request => "/index.html",
    :query_string => "parm1=value1&parm2=value2"
  )
  http.callback {|response|
    puts response[:status]
    puts response[:headers]
    puts response[:content]
  }
}

– TODO: Add streaming so we can support enormous POSTs. Current max is 20meg. Timeout for connections that run too long or hang somewhere in the middle. Persistent connections (HTTP/1.1), may need a associated delegate object. DNS: Some way to cache DNS lookups for hostnames we connect to. Ruby’s DNS lookups are unbelievably slow. HEAD requests. Chunked transfer encoding. Convenience methods for requests. get, post, url, etc. SSL. Handle status codes like 304, 100, etc. Refactor this code so that protocol errors all get handled one way (an exception?), instead of sprinkling set_deferred_status :failed calls everywhere.

Constant Summary collapse

MaxPostContentLength = 
20 * 1024 * 1024

Instance Attribute Summary

Attributes inherited from Connection

#signature

Class Method Summary collapse

Instance Method Summary collapse

Methods included from Deferrable

#callback, #cancel_timeout, #errback, #fail, future, #set_deferred_status, #succeed, #timeout

Methods inherited from Connection

#associate_callback_target, #close_connection, #close_connection_after_writing, #comm_inactivity_timeout, #comm_inactivity_timeout=, #detach, #error?, #get_outbound_data_size, #get_peer_cert, #get_peername, #get_pid, #get_sock_opt, #get_sockname, #get_status, #initialize, new, #notify_readable=, #notify_readable?, #notify_writable=, #notify_writable?, #pause, #paused?, #pending_connect_timeout, #pending_connect_timeout=, #proxy_incoming_to, #proxy_target_unbound, #reconnect, #resume, #send_data, #send_datagram, #send_file_data, #set_comm_inactivity_timeout, #set_pending_connect_timeout, #ssl_handshake_completed, #ssl_verify_peer, #start_tls, #stop_proxying, #stream_file_data

Constructor Details

This class inherits a constructor from EventMachine::Connection

Class Method Details

.request(args = {}) ⇒ Object

Arg list

:host => ‘ip/dns’, :port => fixnum, :verb => ‘GET’, :request => ‘path’, :basic_auth => => ”, :password => ”, :content => ‘content’, :contenttype => ‘text/plain’, :query_string => ”, :host_header => ”, :cookie => ”



70
71
72
73
74
75
76
 
# File 'lib/em/protocols/httpclient.rb', line 70

def self.request( args = {} )
  args[:port] ||= 80
  EventMachine.connect( args[:host], args[:port], self ) {|c|
    # According to the docs, we will get here AFTER post_init is called.
    c.instance_eval {@args = args}
  }
end

Instance Method Details

#connection_completed ⇒ Object

We send the request when we get a connection. AND, we set an instance variable to indicate we passed through here. That allows #unbind to know whether there was a successful connection. NB: This naive technique won’t work when we have to support multiple requests on a single connection.



89
90
91
92
 
# File 'lib/em/protocols/httpclient.rb', line 89

def connection_completed
  @connected = true
  send_request @args
end

#dispatch_response ⇒ Object 



242
243
244
245
246
247
248
249
250
251
 
# File 'lib/em/protocols/httpclient.rb', line 242

def dispatch_response
  @read_state = :base
  set_deferred_status :succeeded, {
    :content => @content,
    :headers => @headers,
    :status => @status
  }
  # TODO, we close the connection for now, but this is wrong for persistent clients.
  close_connection
end

#post_init ⇒ Object 



78
79
80
81
82
 
# File 'lib/em/protocols/httpclient.rb', line 78

def post_init
  @start_time = Time.now
  @data = ""
  @read_state = :base
end

#receive_data(data) ⇒ Object 



161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
 
# File 'lib/em/protocols/httpclient.rb', line 161

def receive_data data
  while data and data.length > 0
    case @read_state
    when :base
      # Perform any per-request initialization here and don't consume any data.
      @data = ""
      @headers = []
      @content_length = nil # not zero
      @content = ""
      @status = nil
      @read_state = :header
      @connection_close = nil
    when :header
      ary = data.split( /\r?\n/m, 2 )
      if ary.length == 2
        data = ary.last
        if ary.first == ""
          if (@content_length and @content_length > 0) || @connection_close
            @read_state = :content
          else
            dispatch_response
            @read_state = :base
          end
        else
          @headers << ary.first
          if @headers.length == 1
            parse_response_line
          elsif ary.first =~ /\Acontent-length:\s*/i
            # Only take the FIRST content-length header that appears,
            # which we can distinguish because @content_length is nil.
            # TODO, it's actually a fatal error if there is more than one
            # content-length header, because the caller is presumptively
            # a bad guy. (There is an exploit that depends on multiple
            # content-length headers.)
            @content_length ||= $'.to_i
          elsif ary.first =~ /\Aconnection:\s*close/i
            @connection_close = true
          end
        end
      else
        @data << data
        data = ""
      end
    when :content
      # If there was no content-length header, we have to wait until the connection
      # closes. Everything we get until that point is content.
      # TODO: Must impose a content-size limit, and also must implement chunking.
      # Also, must support either temporary files for large content, or calling
      # a content-consumer block supplied by the user.
      if @content_length
        bytes_needed = @content_length - @content.length
        @content += data[0, bytes_needed]
        data = data[bytes_needed..-1] || ""
        if @content_length == @content.length
          dispatch_response
          @read_state = :base
        end
      else
        @content << data
        data = ""
      end
    end
  end
end

#send_request(args) ⇒ Object 



94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
 
# File 'lib/em/protocols/httpclient.rb', line 94

def send_request args
  args[:verb] ||= args[:method] # Support :method as an alternative to :verb.
  args[:verb] ||= :get # IS THIS A GOOD IDEA, to default to GET if nothing was specified?

  verb = args[:verb].to_s.upcase
  unless ["GET", "POST", "PUT", "DELETE", "HEAD"].include?(verb)
    set_deferred_status :failed, {:status => 0} # TODO, not signalling the error type
    return # NOTE THE EARLY RETURN, we're not sending any data.
  end

  request = args[:request] || "/"
  unless request[0,1] == "/"
    request = "/" + request
  end

  qs = args[:query_string] || ""
  if qs.length > 0 and qs[0,1] != '?'
    qs = "?" + qs
  end

  version = args[:version] || "1.1"

  # Allow an override for the host header if it's not the connect-string.
  host = args[:host_header] || args[:host] || "_"
  # For now, ALWAYS tuck in the port string, although we may want to omit it if it's the default.
  port = args[:port]

  # POST items.
  postcontenttype = args[:contenttype] || "application/octet-stream"
  postcontent = args[:content] || ""
  raise "oversized content in HTTP POST" if postcontent.length > MaxPostContentLength

  # ESSENTIAL for the request's line-endings to be CRLF, not LF. Some servers misbehave otherwise.
  # TODO: We ASSUME the caller wants to send a 1.1 request. May not be a good assumption.
  req = [
    "#{verb} #{request}#{qs} HTTP/#{version}",
    "Host: #{host}:#{port}",
    "User-agent: Ruby EventMachine",
  ]

    if verb == "POST" || verb == "PUT"
      req << "Content-type: #{postcontenttype}"
      req << "Content-length: #{postcontent.length}"
    end

    # TODO, this cookie handler assumes it's getting a single, semicolon-delimited string.
    # Eventually we will want to deal intelligently with arrays and hashes.
    if args[:cookie]
      req << "Cookie: #{args[:cookie]}"
    end

    # Basic-auth stanza contributed by Matt Murphy.
    if args[:basic_auth]
      basic_auth_string = ["#{args[:basic_auth][:username]}:#{args[:basic_auth][:password]}"].pack('m').strip.gsub(/\n/,'')
      req << "Authorization: Basic #{basic_auth_string}"
    end

    req << ""
    reqstring = req.map {|l| "#{l}\r\n"}.join
    send_data reqstring

    if verb == "POST" || verb == "PUT"
      send_data postcontent
    end
end

#unbind ⇒ Object 



253
254
255
256
257
258
259
 
# File 'lib/em/protocols/httpclient.rb', line 253

def unbind
  if !@connected
    set_deferred_status :failed, {:status => 0} # YECCCCH. Find a better way to signal no-connect/network error.
  elsif (@read_state == :content and @content_length == nil)
    dispatch_response
  end
end