Class: RaptorIO::Protocol::HTTP::Client

Inherits:
Object
  • Object
show all
Defined in:
lib/raptor-io/protocol/http/client.rb

Overview

HTTP Client class.

Author:

Constant Summary collapse

DEFAULT_OPTIONS =

Default client options.

{
    concurrency:      20,
    user_agent:       "RaptorIO::HTTP/#{RaptorIO::VERSION}",
    timeout:          10,
    manipulators:     {},
    ssl_version:      :TLSv1,
    ssl_verify_mode:  OpenSSL::SSL::VERIFY_NONE,
    ssl_context:      nil
}.freeze

Instance Attribute Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(options = {}) ⇒ Client

Returns a new instance of Client.

Parameters:

  • options (Hash) (defaults to: {}) —

    Request options.

Options Hash (options):

  • :concurrency (Integer) — default: 20 —

    Amount of open sockets at any given time.

  • :user_agent (String) — default: 'RaptorIO::HTTP/<RaptorIO::VERSION>' —

    User-agent string to include in the requests.

  • :timeout (Integer, Float) — default: 10 —

    Timeout in seconds.

  • :manipulators (Hash{Symbol=>Hash}) —

    Request manipulators and their options.

  • :ssl_version (Symbol) — default: :TLSv1
  • :ssl_verify_mode (Constant) — default: OpenSSL::SSL::VERIFY_NONE —

    Peer verification mode.

  • :ssl_context (OpenSSL::SSL::SSLContext) — default: nil —

    SSL context to use.

  • :switch_board (Socket::SwitchBoard) —

    The switch board from which we can create new connections.

Raises:

  • RaptorIO::Protocol::HTTP::Request::Manipulator::Error::InvalidOptions On invalid manipulator options.



74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
 
# File 'lib/raptor-io/protocol/http/client.rb', line 74

def initialize( options = {} )
  @switch_board = options.delete(:switch_board)
  unless @switch_board.respond_to?(:create_tcp)
    raise ArgumentError, 'Must provide a :switch_board'
  end

  DEFAULT_OPTIONS.merge( options ).each do |k, v|
    begin
      send( "#{k}=", try_dup( v ) )
    rescue NoMethodError
      instance_variable_set( "@#{k}".to_sym, try_dup( v ) )
    end
  end

  validate_manipulators!( manipulators )

  # Holds Request objects.
  @queue = []

  # Persistent storage for request manipulators.
  @datastore = Hash.new { |h, k| h[k] = {} }

  reset_sockets
  reset_pending_responses
end

Instance Attribute Details

#concurrency ⇒ Integer

Returns Maximum open sockets.

Returns:

  • (Integer) —

    Maximum open sockets.



18
19
20
 
# File 'lib/raptor-io/protocol/http/client.rb', line 18

def concurrency
  @concurrency
end

#datastore ⇒ Hash

Returns Persistent storage for the manipulators..

Returns:

  • (Hash) —

    Persistent storage for the manipulators..



29
30
31
 
# File 'lib/raptor-io/protocol/http/client.rb', line 29

def datastore
  @datastore
end

#manipulators ⇒ Hash{Symbol=>Hash}

Returns Request manipulators, and their options, to be run against each queued request.

Returns:

  • (Hash{Symbol=>Hash}) —

    Request manipulators, and their options, to be run against each queued request.



26
27
28
 
# File 'lib/raptor-io/protocol/http/client.rb', line 26

def manipulators
  @manipulators
end

#ssl_context ⇒ OpenSSL::SSL::SSLContext

Returns SSL context to use.

Returns:

  • (OpenSSL::SSL::SSLContext) —

    SSL context to use.



38
39
40
 
# File 'lib/raptor-io/protocol/http/client.rb', line 38

def ssl_context
  @ssl_context
end

#ssl_verify_mode ⇒ Constant

Returns Peer verification mode.

Returns:

  • (Constant) —

    Peer verification mode.



35
36
37
 
# File 'lib/raptor-io/protocol/http/client.rb', line 35

def ssl_verify_mode
  @ssl_verify_mode
end

#ssl_version ⇒ Symbol

Returns SSL version.

Returns:

  • (Symbol) —

    SSL version.



32
33
34
 
# File 'lib/raptor-io/protocol/http/client.rb', line 32

def ssl_version
  @ssl_version
end

#switch_board ⇒ SwitchBoard (readonly)

Returns The routing table from which this RaptorIO::Protocol::HTTP::Client will make new TCP connections.

Returns:



42
43
44
 
# File 'lib/raptor-io/protocol/http/client.rb', line 42

def switch_board
  @switch_board
end

#timeout ⇒ Integer, Float

Returns Timeout in seconds.

Returns:

  • (Integer, Float) —

    Timeout in seconds.



15
16
17
 
# File 'lib/raptor-io/protocol/http/client.rb', line 15

def timeout
  @timeout
end

#user_agent ⇒ String

Returns User-agent string to use.

Returns:

  • (String) —

    User-agent string to use.



21
22
23
 
# File 'lib/raptor-io/protocol/http/client.rb', line 21

def user_agent
  @user_agent
end

Instance Method Details

#get(url, options = {}, &block) ⇒ Request, Response

Creates and queues a GET Request.

Parameters:

  • url (String) —

    URL of the remote resource.

  • options (Hash) (defaults to: {}) —

    Request options with the following extras:

  • block (Block) —

    Callback to be passed the Response.

Returns:



171
172
173
 
# File 'lib/raptor-io/protocol/http/client.rb', line 171

def get( url, options = {}, &block )
  request( url, options.merge( http_method: :get ), &block )
end

#open_socket_count ⇒ Integer

Returns Amount of open sockets.

Returns:

  • (Integer) —

    Amount of open sockets.



304
305
306
 
# File 'lib/raptor-io/protocol/http/client.rb', line 304

def open_socket_count
  open_sockets.size
end

#post(url, options = {}, &block) ⇒ Request, Response

Creates and queues a POST Request.

Parameters:

  • url (String) —

    URL of the remote resource.

  • options (Hash) (defaults to: {}) —

    Request options with the following extras:

  • block (Block) —

    Callback to be passed the Response.

Returns:



181
182
183
 
# File 'lib/raptor-io/protocol/http/client.rb', line 181

def post( url, options = {}, &block )
  request( url, options.merge( http_method: :post ), &block )
end

#queue(request, manipulators = {}) ⇒ Request Also known as: <<

Queues a Request.

Parameters:

  • request (Request)
  • manipulators (Hash{Symbol=>Hash}) (defaults to: {}) —

    Manipulator names for keys and their options as their values.

Returns:



198
199
200
201
202
203
204
205
206
207
208
209
 
# File 'lib/raptor-io/protocol/http/client.rb', line 198

def queue( request, manipulators = {} )
  validate_manipulators!( manipulators )

  request.timeout ||= timeout

  @manipulators.merge( manipulators ).each do |manipulator, options|
    Request::Manipulators.process( manipulator, self, request, options )
  end

  @queue << request
  request
end

#queue_size ⇒ Integer

Returns The amount of queued requests.

Returns:

  • (Integer) —

    The amount of queued requests.



186
187
188
 
# File 'lib/raptor-io/protocol/http/client.rb', line 186

def queue_size
  @queue.size
end

#request(url, options = {}, &block) ⇒ Request, Response

Creates and queues a Request.

Parameters:

  • url (String) —

    URL of the remote resource.

  • options (Hash) (defaults to: {}) —

    Request options with the following extras:

  • block (Block) —

    Callback to be passed the Response.

Options Hash (options):

  • :mode (Symbol) — default: :async —

    Mode to use for the request, available options are:

    • ‘:async` – Adds the request to the queue.

    • ‘:sync` – Performs the request in a blocking manner and returns the Response.

  • :cookies (Hash, String) —

    Cookies as name=>pair values – should already be escaped.

  • :manipulators (Hash{Symbol=>Hash}) —

    Manipulator names for keys and their options as their values.

Returns:

See Also:



141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
 
# File 'lib/raptor-io/protocol/http/client.rb', line 141

def request( url, options = {}, &block )
  options = options.dup
  options[:timeout] ||= @timeout

  req = Request.new( options.merge( url: url ) )

  req.headers['User-Agent'] = @user_agent if !@user_agent.to_s.empty?

  case options[:cookies]
    when Hash
      req.headers['Cookie'] =
          options[:cookies].map { |k, v| "#{k}=#{v}" }.join( ';' )
    when String
      req.headers['Cookie'] = options[:cookies]
  end

  return sync_request( req, options[:manipulators] || {} ) if options[:mode] == :sync

  req.on_complete( &block ) if block_given?

  queue( req, options[:manipulators] || {} )
  req
end

#run ⇒ Object

Runs the queued Request.



213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
 
# File 'lib/raptor-io/protocol/http/client.rb', line 213

def run
  while @queue.any?
    # Get us some seeds.
    consume_requests

    while @sockets[:done].size != @sockets[:lookup_request].size

      if @sockets[:reads].any?
        # Use the lowest available timeout for #select.
        lowest_timeout =
            @sockets[:reads].map { |socket| @pending_responses[socket][:timeout] }.sort.first

        clock = Time.now
        res = select( @sockets[:reads], nil, @sockets[:reads], lowest_timeout )
        waiting_time = Time.now - clock

        # Adjust the timeouts for *all* sockets since they all benefited from
        # the #select waiting period which just elapsed.
        #
        # And this is the whole reason for keeping track of timeouts externally.
        @pending_responses.each do |_, pending_response|
          pending_response[:timeout] -= waiting_time
          pending_response[:timeout]  = 0 if pending_response[:timeout] < 0
        end

        # #select timed out, go digging.
        if !res
          # Find and handle the sockets which timed out.
          @sockets[:reads].each do |socket|
            if waiting_time >= @pending_responses[socket][:timeout]
              error = RaptorIO::Error::Timeout.new( 'Request timed-out.' )
              error.set_backtrace( caller )
              handle_error( @sockets[:lookup_request][socket], error, socket )
            end

            # Fill the available pool space.
            consume_requests
          end

        # #select didn't time out, yay!
        else

          # Handle sockets with errors -- like reset connections.
          if res[2].any?
            res[2].each do |socket|
              handle_error( @sockets[:lookup_request][socket], nil, socket )

              # Fill the available pool space.
              consume_requests
            end
          end

          # Handle sockets which are ready to be read.
          if res[0].any?
            res[0].each do |socket|
              # Buffer/handle the response for the given socket.
              read( socket )

              # Fill the available pool space.
              consume_requests
            end
          end
        end
      end

      next if @sockets[:writes].empty?

      _, writes, errors = select( nil, @sockets[:writes], @sockets[:writes] )

      errors.each do |socket|
        handle_error( @sockets[:lookup_request][socket], nil, socket )
      end

      writes.each do |socket|
        # Send the request for the given socket.
        write( socket )

        # Fully utilize our socket allowance.
        consume_requests
      end

    end
  end

  reset_sockets
  reset_pending_responses

  nil
end

#update_manipulators(manipulators) ⇒ Object

Updates the client #manipulators and perform and sanity check on their options.

Parameters:

  • manipulators (Hash{String=>Hash}) —

    Manipulators and their options.

Raises:

  • RaptorIO::Protocol::HTTP::Request::Manipulator::Error::InvalidOptions



108
109
110
111
 
# File 'lib/raptor-io/protocol/http/client.rb', line 108

def update_manipulators( manipulators )
  validate_manipulators!( manipulators )
  @manipulators.merge!( manipulators )
end