Lib/asynchat.py

   1 # -*- Mode: Python; tab-width: 4 -*-
   2 #       Id: asynchat.py,v 2.26 2000/09/07 22:29:26 rushing Exp
   3 #       Author: Sam Rushing <rushing@nightmare.com>
   4
   5 # ======================================================================
   6 # Copyright 1996 by Sam Rushing
   7 #
   8 #                         All Rights Reserved
   9 #
  10 # Permission to use, copy, modify, and distribute this software and
  11 # its documentation for any purpose and without fee is hereby
  12 # granted, provided that the above copyright notice appear in all
  13 # copies and that both that copyright notice and this permission
  14 # notice appear in supporting documentation, and that the name of Sam
  15 # Rushing not be used in advertising or publicity pertaining to
  16 # distribution of the software without specific, written prior
  17 # permission.
  18 #
  19 # SAM RUSHING DISCLAIMS ALL WARRANTIES WITH REGARD TO THIS SOFTWARE,
  20 # INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS, IN
  21 # NO EVENT SHALL SAM RUSHING BE LIABLE FOR ANY SPECIAL, INDIRECT OR
  22 # CONSEQUENTIAL DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS
  23 # OF USE, DATA OR PROFITS, WHETHER IN AN ACTION OF CONTRACT,
  24 # NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF OR IN
  25 # CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
  26 # ======================================================================
  27
  28 r"""A class supporting chat-style (command/response) protocols.
  29
  30 This class adds support for 'chat' style protocols - where one side
  31 sends a 'command', and the other sends a response (examples would be
  32 the common internet protocols - smtp, nntp, ftp, etc..).
  33
  34 The handle_read() method looks at the input stream for the current
  35 'terminator' (usually '\r\n' for single-line responses, '\r\n.\r\n'
  36 for multi-line output), calling self.found_terminator() on its
  37 receipt.
  38
  39 for example:
  40 Say you build an async nntp client using this class.  At the start
  41 of the connection, you'll have self.terminator set to '\r\n', in
  42 order to process the single-line greeting.  Just before issuing a
  43 'LIST' command you'll set it to '\r\n.\r\n'.  The output of the LIST
  44 command will be accumulated (using your own 'collect_incoming_data'
  45 method) up to the terminator, and then control will be returned to
  46 you - by calling your self.found_terminator() method.
  47 """
  48
  49 import socket
  50 import asyncore
  51
  52 class async_chat (asyncore.dispatcher):
  53     """This is an abstract class.  You must derive from this class, and add
  54     the two methods collect_incoming_data() and found_terminator()"""
  55
  56     # these are overridable defaults
  57
  58     ac_in_buffer_size       = 4096
  59     ac_out_buffer_size      = 4096
  60
  61     def __init__ (self, conn=None):
  62         self.ac_in_buffer = ''
  63         self.ac_out_buffer = ''
  64         self.producer_fifo = fifo()
  65         asyncore.dispatcher.__init__ (self, conn)
  66
  67     def collect_incoming_data(self, data):
  68         raise NotImplementedError, "must be implemented in subclass"
  69
  70     def found_terminator(self):
  71         raise NotImplementedError, "must be implemented in subclass"
  72
  73     def set_terminator (self, term):
  74         "Set the input delimiter.  Can be a fixed string of any length, an integer, or None"
  75         self.terminator = term
  76
  77     def get_terminator (self):
  78         return self.terminator
  79
  80     # grab some more data from the socket,
  81     # throw it to the collector method,
  82     # check for the terminator,
  83     # if found, transition to the next state.
  84
  85     def handle_read (self):
  86
  87         try:
  88             data = self.recv (self.ac_in_buffer_size)
  89         except socket.error, why:
  90             self.handle_error()
  91             return
  92
  93         self.ac_in_buffer = self.ac_in_buffer + data
  94
  95         # Continue to search for self.terminator in self.ac_in_buffer,
  96         # while calling self.collect_incoming_data.  The while loop
  97         # is necessary because we might read several data+terminator
  98         # combos with a single recv(1024).
  99
 100         while self.ac_in_buffer:
 101             lb = len(self.ac_in_buffer)
 102             terminator = self.get_terminator()
 103             if terminator is None:
 104                 # no terminator, collect it all
 105                 self.collect_incoming_data (self.ac_in_buffer)
 106                 self.ac_in_buffer = ''
 107             elif type(terminator) == type(0):
 108                 # numeric terminator
 109                 n = terminator
 110                 if lb < n:
 111                     self.collect_incoming_data (self.ac_in_buffer)
 112                     self.ac_in_buffer = ''
 113                     self.terminator = self.terminator - lb
 114                 else:
 115                     self.collect_incoming_data (self.ac_in_buffer[:n])
 116                     self.ac_in_buffer = self.ac_in_buffer[n:]
 117                     self.terminator = 0
 118                     self.found_terminator()
 119             else:
 120                 # 3 cases:
 121                 # 1) end of buffer matches terminator exactly:
 122                 #    collect data, transition
 123                 # 2) end of buffer matches some prefix:
 124                 #    collect data to the prefix
 125                 # 3) end of buffer does not match any prefix:
 126                 #    collect data
 127                 terminator_len = len(terminator)
 128                 index = self.ac_in_buffer.find(terminator)
 129                 if index != -1:
 130                     # we found the terminator
 131                     if index > 0:
 132                         # don't bother reporting the empty string (source of subtle bugs)
 133                         self.collect_incoming_data (self.ac_in_buffer[:index])
 134                     self.ac_in_buffer = self.ac_in_buffer[index+terminator_len:]
 135                     # This does the Right Thing if the terminator is changed here.
 136                     self.found_terminator()
 137                 else:
 138                     # check for a prefix of the terminator
 139                     index = find_prefix_at_end (self.ac_in_buffer, terminator)
 140                     if index:
 141                         if index != lb:
 142                             # we found a prefix, collect up to the prefix
 143                             self.collect_incoming_data (self.ac_in_buffer[:-index])
 144                             self.ac_in_buffer = self.ac_in_buffer[-index:]
 145                         break
 146                     else:
 147                         # no prefix, collect it all
 148                         self.collect_incoming_data (self.ac_in_buffer)
 149                         self.ac_in_buffer = ''
 150
 151     def handle_write (self):
 152         self.initiate_send ()
 153
 154     def handle_close (self):
 155         self.close()
 156
 157     def push (self, data):
 158         self.producer_fifo.push (simple_producer (data))
 159         self.initiate_send()
 160
 161     def push_with_producer (self, producer):
 162         self.producer_fifo.push (producer)
 163         self.initiate_send()
 164
 165     def readable (self):
 166         "predicate for inclusion in the readable for select()"
 167         return (len(self.ac_in_buffer) <= self.ac_in_buffer_size)
 168
 169     def writable (self):
 170         "predicate for inclusion in the writable for select()"
 171         # return len(self.ac_out_buffer) or len(self.producer_fifo) or (not self.connected)
 172         # this is about twice as fast, though not as clear.
 173         return not (
 174                 (self.ac_out_buffer == '') and
 175                 self.producer_fifo.is_empty() and
 176                 self.connected
 177                 )
 178
 179     def close_when_done (self):
 180         "automatically close this channel once the outgoing queue is empty"
 181         self.producer_fifo.push (None)
 182
 183     # refill the outgoing buffer by calling the more() method
 184     # of the first producer in the queue
 185     def refill_buffer (self):
 186         _string_type = type('')
 187         while 1:
 188             if len(self.producer_fifo):
 189                 p = self.producer_fifo.first()
 190                 # a 'None' in the producer fifo is a sentinel,
 191                 # telling us to close the channel.
 192                 if p is None:
 193                     if not self.ac_out_buffer:
 194                         self.producer_fifo.pop()
 195                         self.close()
 196                     return
 197                 elif type(p) is _string_type:
 198                     self.producer_fifo.pop()
 199                     self.ac_out_buffer = self.ac_out_buffer + p
 200                     return
 201                 data = p.more()
 202                 if data:
 203                     self.ac_out_buffer = self.ac_out_buffer + data
 204                     return
 205                 else:
 206                     self.producer_fifo.pop()
 207             else:
 208                 return
 209
 210     def initiate_send (self):
 211         obs = self.ac_out_buffer_size
 212         # try to refill the buffer
 213         if (len (self.ac_out_buffer) < obs):
 214             self.refill_buffer()
 215
 216         if self.ac_out_buffer and self.connected:
 217             # try to send the buffer
 218             try:
 219                 num_sent = self.send (self.ac_out_buffer[:obs])
 220                 if num_sent:
 221                     self.ac_out_buffer = self.ac_out_buffer[num_sent:]
 222
 223             except socket.error, why:
 224                 self.handle_error()
 225                 return
 226
 227     def discard_buffers (self):
 228         # Emergencies only!
 229         self.ac_in_buffer = ''
 230         self.ac_out_buffer = ''
 231         while self.producer_fifo:
 232             self.producer_fifo.pop()
 233
 234
 235 class simple_producer:
 236
 237     def __init__ (self, data, buffer_size=512):
 238         self.data = data
 239         self.buffer_size = buffer_size
 240
 241     def more (self):
 242         if len (self.data) > self.buffer_size:
 243             result = self.data[:self.buffer_size]
 244             self.data = self.data[self.buffer_size:]
 245             return result
 246         else:
 247             result = self.data
 248             self.data = ''
 249             return result
 250
 251 class fifo:
 252     def __init__ (self, list=None):
 253         if not list:
 254             self.list = []
 255         else:
 256             self.list = list
 257
 258     def __len__ (self):
 259         return len(self.list)
 260
 261     def is_empty (self):
 262         return self.list == []
 263
 264     def first (self):
 265         return self.list[0]
 266
 267     def push (self, data):
 268         self.list.append (data)
 269
 270     def pop (self):
 271         if self.list:
 272             result = self.list[0]
 273             del self.list[0]
 274             return (1, result)
 275         else:
 276             return (0, None)
 277
 278 # Given 'haystack', see if any prefix of 'needle' is at its end.  This
 279 # assumes an exact match has already been checked.  Return the number of
 280 # characters matched.
 281 # for example:
 282 # f_p_a_e ("qwerty\r", "\r\n") => 1
 283 # f_p_a_e ("qwertydkjf", "\r\n") => 0
 284 # f_p_a_e ("qwerty\r\n", "\r\n") => <undefined>
 285
 286 # this could maybe be made faster with a computed regex?
 287 # [answer: no; circa Python-2.0, Jan 2001]
 288 # new python:   28961/s
 289 # old python:   18307/s
 290 # re:        12820/s
 291 # regex:     14035/s
 292
 293 def find_prefix_at_end (haystack, needle):
 294     l = len(needle) - 1
 295     while l and not haystack.endswith(needle[:l]):
 296         l -= 1
 297     return l