plomlombot.py

   1 #!/usr/bin/python3
   2
   3 import argparse
   4 import socket
   5 import datetime
   6 import select
   7 import time
   8 import re
   9 import requests
  10 import bs4
  11 import random
  12 import hashlib
  13 import os
  14 import plomsearch
  15
  16 # Defaults, may be overwritten by command line arguments.
  17 SERVER = "irc.freenode.net"
  18 PORT = 6667
  19 TIMEOUT = 240
  20 USERNAME = "plomlombot"
  21 NICKNAME = USERNAME
  22 TWTFILE = ""
  23
  24
  25 class ExceptionForRestart(Exception):
  26     pass
  27
  28
  29 class IO:
  30
  31     def __init__(self, server, port, timeout):
  32         self.timeout = timeout
  33         self.socket = socket.socket()
  34         self.socket.connect((server, port))
  35         self.socket.setblocking(0)
  36         self.line_buffer = []
  37         self.rune_buffer = ""
  38         self.last_pong = time.time()
  39         self.servername = self.recv_line(send_ping=False).split(" ")[0][1:]
  40
  41     def _pingtest(self, send_ping=True):
  42         if self.last_pong + self.timeout < time.time():
  43             print("SERVER NOT ANSWERING")
  44             raise ExceptionForRestart
  45         if send_ping:
  46             self.send_line("PING " + self.servername)
  47
  48     def send_line(self, msg):
  49         msg = msg.replace("\r", " ")
  50         msg = msg.replace("\n", " ")
  51         if len(msg.encode("utf-8")) > 510:
  52             print("NOT SENT LINE TO SERVER (too long): " + msg)
  53         print("LINE TO SERVER: "
  54               + str(datetime.datetime.now()) + ": " + msg)
  55         msg = msg + "\r\n"
  56         msg_len = len(msg)
  57         total_sent_len = 0
  58         while total_sent_len < msg_len:
  59             sent_len = self.socket.send(bytes(msg[total_sent_len:], "UTF-8"))
  60             if sent_len == 0:
  61                 print("SOCKET CONNECTION BROKEN")
  62                 raise ExceptionForRestart
  63             total_sent_len += sent_len
  64
  65     def _recv_line_wrapped(self, send_ping=True):
  66         if len(self.line_buffer) > 0:
  67             return self.line_buffer.pop(0)
  68         while True:
  69             ready = select.select([self.socket], [], [], int(self.timeout / 2))
  70             if not ready[0]:
  71                 self._pingtest(send_ping)
  72                 return None
  73             self.last_pong = time.time()
  74             received_bytes = self.socket.recv(1024)
  75             try:
  76                 received_runes = received_bytes.decode("UTF-8")
  77             except UnicodeDecodeError:
  78                 received_runes = received_bytes.decode("latin1")
  79             if len(received_runes) == 0:
  80                 print("SOCKET CONNECTION BROKEN")
  81                 raise ExceptionForRestart
  82             self.rune_buffer += received_runes
  83             lines_split = str.split(self.rune_buffer, "\r\n")
  84             self.line_buffer += lines_split[:-1]
  85             self.rune_buffer = lines_split[-1]
  86             if len(self.line_buffer) > 0:
  87                 return self.line_buffer.pop(0)
  88
  89     def recv_line(self, send_ping=True):
  90         line = self._recv_line_wrapped(send_ping)
  91         if line:
  92             print("LINE FROM SERVER " + str(datetime.datetime.now()) + ": " +
  93                   line)
  94         return line
  95
  96
  97 def handle_command(command, argument, notice, target, session):
  98     hash_string = hashlib.md5(target.encode("utf-8")).hexdigest()
  99     quotesfile_name = "quotes_" + hash_string
 100
 101     def addquote():
 102         if not os.access(quotesfile_name, os.F_OK):
 103             quotesfile = open(quotesfile_name, "w")
 104             quotesfile.write("QUOTES FOR " + target + ":\n")
 105             quotesfile.close()
 106         quotesfile = open(quotesfile_name, "a")
 107         quotesfile.write(argument + "\n")
 108         quotesfile.close()
 109         quotesfile = open(quotesfile_name, "r")
 110         lines = quotesfile.readlines()
 111         quotesfile.close()
 112         notice("ADDED QUOTE #" + str(len(lines) - 1))
 113
 114     def quote():
 115
 116         def help():
 117             notice("SYNTAX: !quote [int] OR !quote search QUERY")
 118             notice("QUERY may be a boolean grouping of quoted or unquoted " +
 119                    "search terms, examples:")
 120             notice("!quote search foo")
 121             notice("!quote search foo AND (bar OR NOT baz)")
 122             notice("!quote search \"foo\\\"bar\" AND ('NOT\"' AND \"'foo'\"" +
 123                    " OR 'bar\\'baz')")
 124
 125         if "" == argument:
 126             tokens = []
 127         else:
 128             tokens = argument.split(" ")
 129         if (len(tokens) > 1 and tokens[0] != "search") or \
 130             (len(tokens) == 1 and
 131                 (tokens[0] == "search" or not tokens[0].isdigit())):
 132             help()
 133             return
 134         if not os.access(quotesfile_name, os.F_OK):
 135             notice("NO QUOTES AVAILABLE")
 136             return
 137         quotesfile = open(quotesfile_name, "r")
 138         lines = quotesfile.readlines()
 139         quotesfile.close()
 140         lines = lines[1:]
 141         if len(tokens) == 1:
 142             i = int(tokens[0])
 143             if i == 0 or i > len(lines):
 144                 notice("THERE'S NO QUOTE OF THAT INDEX")
 145                 return
 146             i = i - 1
 147         elif len(tokens) > 1:
 148             query = str.join(" ", tokens[1:])
 149             try:
 150                 results = plomsearch.search(query, lines)
 151             except plomsearch.LogicParserError as err:
 152                 notice("FAILED QUERY PARSING: " + str(err))
 153                 return
 154             if len(results) == 0:
 155                 notice("NO QUOTES MATCHING QUERY")
 156             else:
 157                 for result in results:
 158                     notice("QUOTE #" + str(result[0] + 1) + " : " + result[1])
 159             return
 160         else:
 161             i = random.randrange(len(lines))
 162         notice("QUOTE #" + str(i + 1) + ": " + lines[i])
 163
 164     def markov():
 165         from random import choice
 166         select_length = 2
 167         selections = []
 168
 169         def markov(snippet):
 170             usable_selections = []
 171             for i in range(select_length, 0, -1):
 172                 for selection in selections:
 173                     add = True
 174                     for j in range(i):
 175                         j += 1
 176                         if snippet[-j] != selection[-(j+1)]:
 177                             add = False
 178                             break
 179                     if add:
 180                         usable_selections += [selection]
 181                 if [] != usable_selections:
 182                     break
 183             if [] == usable_selections:
 184                 usable_selections = selections
 185             selection = choice(usable_selections)
 186             return selection[select_length]
 187
 188         hash_string = hashlib.md5(target.encode("utf-8")).hexdigest()
 189         markovfeed_name = "markovfeed_" + hash_string
 190         if not os.access(markovfeed_name, os.F_OK):
 191             notice("NOT ENOUGH TEXT TO MARKOV.")
 192             return
 193
 194         # Lowercase incoming lines, ensure they end in a sentence end mark.
 195         file = open(markovfeed_name, "r")
 196         lines = file.readlines()
 197         file.close()
 198         tokens = []
 199         for line in lines:
 200             line = line.lower().replace("\n", "")
 201             if line[-1] not in ".!?":
 202                 line += "."
 203             tokens += line.split()
 204         if len(tokens) <= select_length:
 205             notice("NOT ENOUGH TEXT TO MARKOV.")
 206             return
 207
 208         # Replace URLs with escape string for now, so that the Markov selector
 209         # won't see them as different strings. Stash replaced URLs in urls.
 210         urls = []
 211         url_escape = "\nURL"
 212         url_starts = ["http://", "https://", "<http://", "<https://"]
 213         for i in range(len(tokens)):
 214             for url_start in url_starts:
 215                 if tokens[i][:len(url_start)] == url_start:
 216                     length = len(tokens[i])
 217                     if url_start[0] == "<":
 218                         try:
 219                             length = tokens[i].index(">") + 1
 220                         except ValueError:
 221                             pass
 222                     urls += [tokens[i][:length]]
 223                     tokens[i] = url_escape + tokens[i][length:]
 224                     break
 225
 226         # For each snippet of select_length, use markov() to find continuation
 227         # token from selections. Replace present users' names with malkovich.
 228         for i in range(len(tokens) - select_length):
 229             token_list = []
 230             for j in range(select_length + 1):
 231                 token_list += [tokens[i + j]]
 232             selections += [token_list]
 233         snippet = []
 234         for i in range(select_length):
 235             snippet += [""]
 236         msg = ""
 237         malkovich = "malkovich"
 238         while 1:
 239             new_end = markov(snippet)
 240             for name in session.users_in_chan:
 241                 if new_end[:len(name)] == name.lower():
 242                     new_end = malkovich + new_end[len(name):]
 243                     break
 244             if len(msg) + len(new_end) > 200:
 245                 break
 246             msg += new_end + " "
 247             for i in range(select_length - 1):
 248                 snippet[i] = snippet[i + 1]
 249             snippet[select_length - 1] = new_end
 250
 251         # Replace occurences of url escape string with random choice from urls.
 252         while True:
 253             index = msg.find(url_escape)
 254             if index < 0:
 255                 break
 256             msg = msg.replace(url_escape, choice(urls), 1)
 257
 258         # More meaningful ways to randomly end sentences.
 259         notice(msg + malkovich + ".")
 260
 261     def twt():
 262         def try_open(mode):
 263             try:
 264                 twtfile = open(session.twtfile, mode)
 265             except (PermissionError, FileNotFoundError) as err:
 266                 notice("CAN'T ACCESS OR CREATE TWT FILE: " + str(err))
 267                 return None
 268             return twtfile
 269
 270         from datetime import datetime
 271         if not os.access(session.twtfile, os.F_OK):
 272             twtfile = try_open("w")
 273             if None == twtfile:
 274                 return
 275             twtfile.close()
 276         twtfile = try_open("a")
 277         if None == twtfile:
 278             return
 279         twtfile.write(datetime.utcnow().isoformat() + "\t" + argument + "\n")
 280         twtfile.close()
 281         notice("WROTE TWT.")
 282
 283     if "addquote" == command:
 284         addquote()
 285     elif "quote" == command:
 286         quote()
 287     elif "markov" == command:
 288         markov()
 289     elif "twt" == command:
 290         twt()
 291
 292
 293 def handle_url(url, notice, show_url=False):
 294
 295     def mobile_twitter_hack(url):
 296         re1 = 'https?://(mobile.twitter.com/)[^/]+(/status/)'
 297         re2 = 'https?://mobile.twitter.com/([^/]+)/status/([^\?/]+)'
 298         m = re.search(re1, url)
 299         if m and m.group(1) == 'mobile.twitter.com/' \
 300                 and m.group(2) == '/status/':
 301             m = re.search(re2, url)
 302             url = 'https://twitter.com/' + m.group(1) + '/status/' + m.group(2)
 303             handle_url(url, notice, True)
 304             return True
 305
 306     try:
 307         r = requests.get(url, timeout=15)
 308     except (requests.exceptions.TooManyRedirects,
 309             requests.exceptions.ConnectionError,
 310             requests.exceptions.InvalidURL,
 311             UnicodeError,
 312             requests.exceptions.InvalidSchema) as error:
 313         notice("TROUBLE FOLLOWING URL: " + str(error))
 314         return
 315     if mobile_twitter_hack(url):
 316         return
 317     title = bs4.BeautifulSoup(r.text, "html.parser").title
 318     if title:
 319         prefix = "PAGE TITLE: "
 320         if show_url:
 321             prefix = "PAGE TITLE FOR <" + url + ">: "
 322         notice(prefix + title.string.strip())
 323     else:
 324         notice("PAGE HAS NO TITLE TAG")
 325
 326
 327 class Session:
 328
 329     def __init__(self, io, username, nickname, channel, twtfile):
 330         self.io = io
 331         self.nickname = nickname
 332         self.channel = channel
 333         self.users_in_chan = []
 334         self.twtfile = twtfile
 335         self.io.send_line("NICK " + self.nickname)
 336         self.io.send_line("USER " + username + " 0 * : ")
 337         self.io.send_line("JOIN " + self.channel)
 338
 339     def loop(self):
 340
 341         def handle_privmsg(tokens):
 342
 343             def handle_input(msg, target):
 344
 345                 def notice(msg):
 346                     self.io.send_line("NOTICE " + target + " :" + msg)
 347
 348                 matches = re.findall("(https?://[^\s>]+)", msg)
 349                 for i in range(len(matches)):
 350                     handle_url(matches[i], notice)
 351                 if "!" == msg[0]:
 352                     tokens = msg[1:].split()
 353                     argument = str.join(" ", tokens[1:])
 354                     handle_command(tokens[0], argument, notice, target, self)
 355                     return
 356                 hash_string = hashlib.md5(target.encode("utf-8")).hexdigest()
 357                 markovfeed_name = "markovfeed_" + hash_string
 358                 file = open(markovfeed_name, "a")
 359                 file.write(msg + "\n")
 360                 file.close()
 361
 362             sender = ""
 363             for rune in tokens[0]:
 364                 if rune == "!":
 365                     break
 366                 if rune != ":":
 367                     sender += rune
 368             receiver = ""
 369             for rune in tokens[2]:
 370                 if rune == "!":
 371                     break
 372                 if rune != ":":
 373                     receiver += rune
 374             target = sender
 375             if receiver != self.nickname:
 376                 target = receiver
 377             msg = str.join(" ", tokens[3:])[1:]
 378             handle_input(msg, target)
 379
 380         def name_from_join_or_part(tokens):
 381             token = tokens[0][1:]
 382             index_cut = token.find("@")
 383             index_ex = token.find("!")
 384             if index_ex > 0 and index_ex < index_cut:
 385                 index_cut = index_ex
 386             return token[:index_cut]
 387
 388         while True:
 389             line = self.io.recv_line()
 390             if not line:
 391                 continue
 392             tokens = line.split(" ")
 393             if len(tokens) > 1:
 394                 if tokens[0] == "PING":
 395                     self.io.send_line("PONG " + tokens[1])
 396                 elif tokens[1] == "PRIVMSG":
 397                     handle_privmsg(tokens)
 398                 elif tokens[1] == "353":
 399                     names = tokens[5:]
 400                     names[0] = names[0][1:]
 401                     self.users_in_chan += names
 402                 elif tokens[1] == "JOIN":
 403                     name = name_from_join_or_part(tokens)
 404                     if name != self.nickname:
 405                         self.users_in_chan += [name]
 406                 elif tokens[1] == "PART":
 407                     name = name_from_join_or_part(tokens)
 408                     del(self.users_in_chan[self.users_in_chan.index(name)])
 409
 410 def parse_command_line_arguments():
 411     parser = argparse.ArgumentParser()
 412     parser.add_argument("-s, --server", action="store", dest="server",
 413                         default=SERVER,
 414                         help="server or server net to connect to (default: "
 415                         + SERVER + ")")
 416     parser.add_argument("-p, --port", action="store", dest="port", type=int,
 417                         default=PORT, help="port to connect to (default : "
 418                         + str(PORT) + ")")
 419     parser.add_argument("-w, --wait", action="store", dest="timeout",
 420                         type=int, default=TIMEOUT,
 421                         help="timeout in seconds after which to attempt " +
 422                         "reconnect (default: " + str(TIMEOUT) + ")")
 423     parser.add_argument("-u, --username", action="store", dest="username",
 424                         default=USERNAME, help="username to use (default: "
 425                         + USERNAME + ")")
 426     parser.add_argument("-n, --nickname", action="store", dest="nickname",
 427                         default=NICKNAME, help="nickname to use (default: "
 428                         + NICKNAME + ")")
 429     parser.add_argument("-t, --twtfile", action="store", dest="twtfile",
 430                         default=TWTFILE, help="twtfile to use (default: "
 431                         + TWTFILE + ")")
 432     parser.add_argument("CHANNEL", action="store", help="channel to join")
 433     opts, unknown = parser.parse_known_args()
 434     return opts
 435
 436
 437 opts = parse_command_line_arguments()
 438 while True:
 439     try:
 440         io = IO(opts.server, opts.port, opts.timeout)
 441         session = Session(io, opts.username, opts.nickname, opts.CHANNEL,
 442             opts.twtfile)
 443         session.loop()
 444     except ExceptionForRestart:
 445         io.socket.close()
 446         continue