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 import irclog
  16
  17 # Defaults, may be overwritten by command line arguments.
  18 SERVER = "irc.freenode.net"
  19 PORT = 6667
  20 TIMEOUT = 240
  21 USERNAME = "plomlombot"
  22 NICKNAME = USERNAME
  23 TWTFILE = ""
  24 DBDIR = os.path.expanduser("~/plomlombot_db")
  25
  26
  27 class ExceptionForRestart(Exception):
  28     pass
  29
  30
  31 class Line:
  32
  33     def __init__(self, line):
  34         self.line = line
  35         self.tokens = line.split(" ")
  36         self.sender = ""
  37         if self.tokens[0][0] == ":":
  38             for rune in self.tokens[0][1:]:
  39                 if rune in {"!", "@"}:
  40                     break
  41                 self.sender += rune
  42         self.receiver = ""
  43         if len(self.tokens) > 2:
  44             for rune in self.tokens[2]:
  45                 if rune in {"!", "@"}:
  46                     break
  47                 if rune != ":":
  48                     self.receiver += rune
  49
  50
  51 class IO:
  52
  53     def __init__(self, server, port, timeout):
  54         self.timeout = timeout
  55         self.socket = socket.socket()
  56         try:
  57             self.socket.connect((server, port))
  58         except TimeoutError:
  59             raise ExceptionForRestart
  60         self.socket.setblocking(0)
  61         self.line_buffer = []
  62         self.rune_buffer = ""
  63         self.last_pong = time.time()
  64         self.servername = self.recv_line(send_ping=False).split(" ")[0][1:]
  65
  66     def _pingtest(self, send_ping=True):
  67         if self.last_pong + self.timeout < time.time():
  68             print("SERVER NOT ANSWERING")
  69             raise ExceptionForRestart
  70         if send_ping:
  71             self.send_line("PING " + self.servername)
  72
  73     def send_line(self, msg):
  74         msg = msg.replace("\r", " ")
  75         msg = msg.replace("\n", " ")
  76         if len(msg.encode("utf-8")) > 510:
  77             print("NOT SENT LINE TO SERVER (too long): " + msg)
  78         print("LINE TO SERVER: "
  79               + str(datetime.datetime.now()) + ": " + msg)
  80         msg = msg + "\r\n"
  81         msg_len = len(msg)
  82         total_sent_len = 0
  83         while total_sent_len < msg_len:
  84             sent_len = self.socket.send(bytes(msg[total_sent_len:], "UTF-8"))
  85             if sent_len == 0:
  86                 print("SOCKET CONNECTION BROKEN")
  87                 raise ExceptionForRestart
  88             total_sent_len += sent_len
  89
  90     def _recv_line_wrapped(self, send_ping=True):
  91         if len(self.line_buffer) > 0:
  92             return self.line_buffer.pop(0)
  93         while True:
  94             ready = select.select([self.socket], [], [], int(self.timeout / 2))
  95             if not ready[0]:
  96                 self._pingtest(send_ping)
  97                 return None
  98             self.last_pong = time.time()
  99             received_bytes = self.socket.recv(1024)
 100             try:
 101                 received_runes = received_bytes.decode("UTF-8")
 102             except UnicodeDecodeError:
 103                 received_runes = received_bytes.decode("latin1")
 104             if len(received_runes) == 0:
 105                 print("SOCKET CONNECTION BROKEN")
 106                 raise ExceptionForRestart
 107             self.rune_buffer += received_runes
 108             lines_split = str.split(self.rune_buffer, "\r\n")
 109             self.line_buffer += lines_split[:-1]
 110             self.rune_buffer = lines_split[-1]
 111             if len(self.line_buffer) > 0:
 112                 return self.line_buffer.pop(0)
 113
 114     def recv_line(self, send_ping=True):
 115         line = self._recv_line_wrapped(send_ping)
 116         if line:
 117             print("LINE FROM SERVER " + str(datetime.datetime.now()) + ": " +
 118                   line)
 119         return line
 120
 121
 122 def handle_command(command, argument, notice, target, session):
 123
 124     def addquote():
 125         if not os.access(session.quotesfile, os.F_OK):
 126             quotesfile = open(session.quotesfile, "w")
 127             quotesfile.write("QUOTES FOR " + target + ":\n")
 128             quotesfile.close()
 129         quotesfile = open(session.quotesfile, "a")
 130         quotesfile.write(argument + "\n")
 131         quotesfile.close()
 132         quotesfile = open(session.quotesfile, "r")
 133         lines = quotesfile.readlines()
 134         quotesfile.close()
 135         notice("ADDED QUOTE #" + str(len(lines) - 1))
 136
 137     def quote():
 138
 139         def help():
 140             notice("SYNTAX: !quote [int] OR !quote search QUERY")
 141             notice("QUERY may be a boolean grouping of quoted or unquoted " +
 142                    "search terms, examples:")
 143             notice("!quote search foo")
 144             notice("!quote search foo AND (bar OR NOT baz)")
 145             notice("!quote search \"foo\\\"bar\" AND ('NOT\"' AND \"'foo'\"" +
 146                    " OR 'bar\\'baz')")
 147
 148         if "" == argument:
 149             tokens = []
 150         else:
 151             tokens = argument.split(" ")
 152         if (len(tokens) > 1 and tokens[0] != "search") or \
 153             (len(tokens) == 1 and
 154                 (tokens[0] == "search" or not tokens[0].isdigit())):
 155             help()
 156             return
 157         if not os.access(session.quotesfile, os.F_OK):
 158             notice("NO QUOTES AVAILABLE")
 159             return
 160         quotesfile = open(session.quotesfile, "r")
 161         lines = quotesfile.readlines()
 162         quotesfile.close()
 163         lines = lines[1:]
 164         if len(tokens) == 1:
 165             i = int(tokens[0])
 166             if i == 0 or i > len(lines):
 167                 notice("THERE'S NO QUOTE OF THAT INDEX")
 168                 return
 169             i = i - 1
 170         elif len(tokens) > 1:
 171             query = str.join(" ", tokens[1:])
 172             try:
 173                 results = plomsearch.search(query, lines)
 174             except plomsearch.LogicParserError as err:
 175                 notice("FAILED QUERY PARSING: " + str(err))
 176                 return
 177             if len(results) == 0:
 178                 notice("NO QUOTES MATCHING QUERY")
 179             else:
 180                 for result in results:
 181                     notice("QUOTE #" + str(result[0] + 1) + " : "
 182                            + result[1][-1])
 183             return
 184         else:
 185             i = random.randrange(len(lines))
 186         notice("QUOTE #" + str(i + 1) + ": " + lines[i][:-1])
 187
 188     def markov():
 189         from random import choice, shuffle
 190         select_length = 2
 191         selections = []
 192
 193         def markov(snippet):
 194             usable_selections = []
 195             for i in range(select_length, 0, -1):
 196                 for selection in selections:
 197                     add = True
 198                     for j in range(i):
 199                         j += 1
 200                         if snippet[-j] != selection[-(j+1)]:
 201                             add = False
 202                             break
 203                     if add:
 204                         usable_selections += [selection]
 205                 if [] != usable_selections:
 206                     break
 207             if [] == usable_selections:
 208                 usable_selections = selections
 209             selection = choice(usable_selections)
 210             return selection[select_length]
 211
 212         if not os.access(session.markovfile, os.F_OK):
 213             notice("NOT ENOUGH TEXT TO MARKOV.")
 214             return
 215
 216         # Lowercase incoming lines, ensure they end in a sentence end mark.
 217         file = open(session.markovfile, "r")
 218         lines = file.readlines()
 219         file.close()
 220         tokens = []
 221         sentence_end_markers = ".!?)("
 222         for line in lines:
 223             line = line.lower().replace("\n", "")
 224             if line[-1] not in sentence_end_markers:
 225                 line += "."
 226             tokens += line.split()
 227         if len(tokens) <= select_length:
 228             notice("NOT ENOUGH TEXT TO MARKOV.")
 229             return
 230
 231         # Replace URLs with escape string for now, so that the Markov selector
 232         # won't see them as different strings. Stash replaced URLs in urls.
 233         urls = []
 234         url_escape = "\nURL"
 235         url_starts = ["http://", "https://", "<http://", "<https://"]
 236         for i in range(len(tokens)):
 237             for url_start in url_starts:
 238                 if tokens[i][:len(url_start)] == url_start:
 239                     length = len(tokens[i])
 240                     if url_start[0] == "<":
 241                         try:
 242                             length = tokens[i].index(">") + 1
 243                         except ValueError:
 244                             pass
 245                     urls += [tokens[i][:length]]
 246                     tokens[i] = url_escape + tokens[i][length:]
 247                     break
 248
 249         # For each snippet of select_length, use markov() to find continuation
 250         # token from selections. Replace present users' names with malkovich.
 251         # Start snippets with the beginning of a sentence, if possible.
 252         for i in range(len(tokens) - select_length):
 253             token_list = []
 254             for j in range(select_length + 1):
 255                 token_list += [tokens[i + j]]
 256             selections += [token_list]
 257         snippet = []
 258         for i in range(select_length):
 259             snippet += [""]
 260         shuffle(selections)
 261         for i in range(len(selections)):
 262             if selections[i][0][-1] in sentence_end_markers:
 263                 for i in range(select_length):
 264                     snippet[i] = selections[i][i + 1]
 265                 break
 266         msg = ""
 267         malkovich = "malkovich"
 268         while 1:
 269             new_end = markov(snippet)
 270             for name in session.users_in_chan:
 271                 if new_end[:len(name)] == name.lower():
 272                     new_end = malkovich + new_end[len(name):]
 273                     break
 274             if len(msg) + len(new_end) > 200:
 275                 break
 276             msg += new_end + " "
 277             for i in range(select_length - 1):
 278                 snippet[i] = snippet[i + 1]
 279             snippet[select_length - 1] = new_end
 280
 281         # Replace occurences of url escape string with random choice from urls.
 282         while True:
 283             index = msg.find(url_escape)
 284             if index < 0:
 285                 break
 286             msg = msg.replace(url_escape, choice(urls), 1)
 287
 288         # More meaningful ways to randomly end sentences.
 289         notice(msg + malkovich + ".")
 290
 291     def twt():
 292         def try_open(mode):
 293             try:
 294                 twtfile = open(session.twtfile, mode)
 295             except (PermissionError, FileNotFoundError) as err:
 296                 notice("CAN'T ACCESS OR CREATE TWT FILE: " + str(err))
 297                 return None
 298             return twtfile
 299
 300         from datetime import datetime
 301         if not os.access(session.twtfile, os.F_OK):
 302             twtfile = try_open("w")
 303             if None == twtfile:
 304                 return
 305             twtfile.close()
 306         twtfile = try_open("a")
 307         if None == twtfile:
 308             return
 309         twtfile.write(datetime.utcnow().isoformat() + "\t" + argument + "\n")
 310         twtfile.close()
 311         notice("WROTE TWT.")
 312
 313     if "addquote" == command:
 314         addquote()
 315     elif "quote" == command:
 316         quote()
 317     elif "markov" == command:
 318         markov()
 319     elif "twt" == command:
 320         twt()
 321
 322
 323 def handle_url(url, notice, show_url=False):
 324
 325     def mobile_twitter_hack(url):
 326         re1 = 'https?://(mobile.twitter.com/)[^/]+(/status/)'
 327         re2 = 'https?://mobile.twitter.com/([^/]+)/status/([^\?/]+)'
 328         m = re.search(re1, url)
 329         if m and m.group(1) == 'mobile.twitter.com/' \
 330                 and m.group(2) == '/status/':
 331             m = re.search(re2, url)
 332             url = 'https://twitter.com/' + m.group(1) + '/status/' + m.group(2)
 333             handle_url(url, notice, True)
 334             return True
 335
 336     try:
 337         r = requests.get(url, timeout=15)
 338     except (requests.exceptions.TooManyRedirects,
 339             requests.exceptions.ConnectionError,
 340             requests.exceptions.InvalidURL,
 341             UnicodeError,
 342             requests.exceptions.InvalidSchema) as error:
 343         notice("TROUBLE FOLLOWING URL: " + str(error))
 344         return
 345     if mobile_twitter_hack(url):
 346         return
 347     title = bs4.BeautifulSoup(r.text, "html5lib").title
 348     if title and title.string:
 349         prefix = "PAGE TITLE: "
 350         if show_url:
 351             prefix = "PAGE TITLE FOR <" + url + ">: "
 352         notice(prefix + title.string.strip())
 353     else:
 354         notice("PAGE HAS NO TITLE TAG")
 355
 356
 357 class Session:
 358
 359     def __init__(self, io, username, nickname, channel, twtfile, dbdir, rmlogs):
 360         self.io = io
 361         self.nickname = nickname
 362         self.username = username
 363         self.channel = channel
 364         self.users_in_chan = []
 365         self.twtfile = twtfile
 366         self.dbdir = dbdir
 367         self.rmlogs = rmlogs
 368         self.io.send_line("NICK " + self.nickname)
 369         self.io.send_line("USER " + self.username + " 0 * : ")
 370         self.io.send_line("JOIN " + self.channel)
 371         hash_channel = hashlib.md5(self.channel.encode("utf-8")).hexdigest()
 372         self.chandir = self.dbdir + "/" + hash_channel + "/"
 373         self.rawlogdir = self.chandir + "raw_logs/"
 374         self.logdir = self.chandir + "logs/"
 375         if not os.path.exists(self.logdir):
 376             os.makedirs(self.logdir)
 377         if not os.path.exists(self.rawlogdir):
 378             os.makedirs(self.rawlogdir)
 379         self.markovfile = self.chandir + "markovfeed"
 380         self.quotesfile = self.chandir + "quotes"
 381
 382     def loop(self):
 383
 384         def log(line):
 385             if type(line) == str:
 386                 line = Line(":" + self.nickname + "!~" + self.username +
 387                             "@localhost" + " " + line)
 388             now = datetime.datetime.utcnow()
 389             logfile = open(self.rawlogdir + now.strftime("%Y-%m-%d") + ".txt", "a")
 390             form = "%Y-%m-%d %H:%M:%S UTC\t"
 391             logfile.write(now.strftime(form) + " " + line.line + "\n")
 392             logfile.close()
 393             to_log = irclog.format_logline(line, self.channel)
 394             if to_log != None:
 395                 logfile = open(self.logdir + now.strftime("%Y-%m-%d") + ".txt", "a")
 396                 logfile.write(now.strftime(form) + " " + to_log + "\n")
 397                 logfile.close()
 398
 399         def handle_privmsg(line):
 400
 401             def notice(msg):
 402                 line = "NOTICE " + target + " :" + msg
 403                 self.io.send_line(line)
 404                 log(line)
 405
 406             target = line.sender
 407             if line.receiver != self.nickname:
 408                 target = line.receiver
 409             msg = str.join(" ", line.tokens[3:])[1:]
 410             matches = re.findall("(https?://[^\s>]+)", msg)
 411             for i in range(len(matches)):
 412                 handle_url(matches[i], notice)
 413             if "!" == msg[0]:
 414                 tokens = msg[1:].split()
 415                 argument = str.join(" ", tokens[1:])
 416                 handle_command(tokens[0], argument, notice, target, self)
 417                 return
 418             file = open(self.markovfile, "a")
 419             file.write(msg + "\n")
 420             file.close()
 421
 422         while True:
 423             if self.rmlogs > 0:
 424                 for f in os.listdir(self.logdir):
 425                     f = os.path.join(self.logdir, f)
 426                     if os.path.isfile(f) and \
 427                             os.stat(f).st_mtime < time.time() - self.rmlogs:
 428                         os.remove(f)
 429             line = self.io.recv_line()
 430             if not line:
 431                 continue
 432             line = Line(line)
 433             log(line)
 434             if len(line.tokens) > 1:
 435                 if line.tokens[0] == "PING":
 436                     self.io.send_line("PONG " + line.tokens[1])
 437                 elif line.tokens[1] == "PRIVMSG":
 438                     handle_privmsg(line)
 439                 elif line.tokens[1] == "353":
 440                     names = line.tokens[5:]
 441                     names[0] = names[0][1:]
 442                     for i in range(len(names)):
 443                         names[i] = names[i].replace("@", "").replace("+", "")
 444                     self.users_in_chan += names
 445                 elif line.tokens[1] == "JOIN" and line.sender != self.nickname:
 446                     self.users_in_chan += [line.sender]
 447                 elif line.tokens[1] == "PART":
 448                     del(self.users_in_chan[self.users_in_chan.index(line.sender)])
 449                 elif line.tokens[1] == "NICK":
 450                     del(self.users_in_chan[self.users_in_chan.index(line.sender)])
 451                     self.users_in_chan += [line.receiver]
 452
 453
 454 def parse_command_line_arguments():
 455     parser = argparse.ArgumentParser()
 456     parser.add_argument("-s, --server", action="store", dest="server",
 457                         default=SERVER,
 458                         help="server or server net to connect to (default: "
 459                         + SERVER + ")")
 460     parser.add_argument("-p, --port", action="store", dest="port", type=int,
 461                         default=PORT, help="port to connect to (default : "
 462                         + str(PORT) + ")")
 463     parser.add_argument("-w, --wait", action="store", dest="timeout",
 464                         type=int, default=TIMEOUT,
 465                         help="timeout in seconds after which to attempt "
 466                         "reconnect (default: " + str(TIMEOUT) + ")")
 467     parser.add_argument("-u, --username", action="store", dest="username",
 468                         default=USERNAME, help="username to use (default: "
 469                         + USERNAME + ")")
 470     parser.add_argument("-n, --nickname", action="store", dest="nickname",
 471                         default=NICKNAME, help="nickname to use (default: "
 472                         + NICKNAME + ")")
 473     parser.add_argument("-t, --twtxtfile", action="store", dest="twtfile",
 474                         default=TWTFILE, help="twtxt file to use (default: "
 475                         + TWTFILE + ")")
 476     parser.add_argument("-d, --dbdir", action="store", dest="dbdir",
 477                         default=DBDIR, help="directory to store DB files in")
 478     parser.add_argument("-r, --rmlogs", action="store", dest="rmlogs",
 479                         type=int, default=0,
 480                         help="maximum age in seconds for logfiles in logs/ "
 481                         "(0 means: never delete, and is default)")
 482     parser.add_argument("CHANNEL", action="store", help="channel to join")
 483     opts, unknown = parser.parse_known_args()
 484     return opts
 485
 486
 487 opts = parse_command_line_arguments()
 488 while True:
 489     try:
 490         io = IO(opts.server, opts.port, opts.timeout)
 491         hash_server = hashlib.md5(opts.server.encode("utf-8")).hexdigest()
 492         dbdir = opts.dbdir + "/" + hash_server
 493         session = Session(io, opts.username, opts.nickname, opts.CHANNEL,
 494             opts.twtfile, dbdir, opts.rmlogs)
 495         session.loop()
 496     except ExceptionForRestart:
 497         io.socket.close()
 498         continue