X-Git-Url: https://plomlompom.com/repos/?p=plomlombot-irc.git;a=blobdiff_plain;f=plomlombot.py;h=7f412ea63d5505e11f68b90893c515cbdbe129e4;hp=82f96a21e4c93b7371ef1669f749fd73c96afde0;hb=e7fab400e3f2bfa5ea5320a7ce8eebf8a01afcdd;hpb=30c070d0602b36ed7ebee10046b9f5331a1cb412 diff --git a/plomlombot.py b/plomlombot.py index 82f96a2..7f412ea 100755 --- a/plomlombot.py +++ b/plomlombot.py @@ -13,14 +13,14 @@ import hashlib import os import plomsearch -URLREGEX = "(https?://[^\s>]+)" - # Defaults, may be overwritten by command line arguments. SERVER = "irc.freenode.net" PORT = 6667 TIMEOUT = 240 USERNAME = "plomlombot" NICKNAME = USERNAME +TWTFILE = "" +DBDIR = os.path.expanduser("~/plomlombot_db") class ExceptionForRestart(Exception): @@ -97,7 +97,7 @@ class IO: def handle_command(command, argument, notice, target, session): hash_string = hashlib.md5(target.encode("utf-8")).hexdigest() - quotesfile_name = "quotes_" + hash_string + quotesfile_name = session.dbdir + "/quotes_" + hash_string def addquote(): if not os.access(quotesfile_name, os.F_OK): @@ -163,7 +163,7 @@ def handle_command(command, argument, notice, target, session): notice("QUOTE #" + str(i + 1) + ": " + lines[i]) def markov(): - from random import shuffle + from random import choice, shuffle select_length = 2 selections = [] @@ -173,7 +173,8 @@ def handle_command(command, argument, notice, target, session): for selection in selections: add = True for j in range(i): - if snippet[j] != selection[j]: + j += 1 + if snippet[-j] != selection[-(j+1)]: add = False break if add: @@ -182,37 +183,51 @@ def handle_command(command, argument, notice, target, session): break if [] == usable_selections: usable_selections = selections - shuffle(usable_selections) - return usable_selections[0][select_length] - - def purge_undesired(tokens): - for token in tokens: - if None != re.match("^" + URLREGEX, token): - del(tokens[tokens.index(token)]) - for name in session.uses_in_chan: - while True: - try: - del(tokens[tokens.index(name.lower())]) - except ValueError: - break - return tokens + selection = choice(usable_selections) + return selection[select_length] hash_string = hashlib.md5(target.encode("utf-8")).hexdigest() - markovfeed_name = "markovfeed_" + hash_string + markovfeed_name = session.dbdir + "/markovfeed_" + hash_string if not os.access(markovfeed_name, os.F_OK): notice("NOT ENOUGH TEXT TO MARKOV.") return + + # Lowercase incoming lines, ensure they end in a sentence end mark. file = open(markovfeed_name, "r") lines = file.readlines() file.close() tokens = [] + sentence_end_markers = ".!?)(" for line in lines: - line = line.replace("\n", "").lower() + line = line.lower().replace("\n", "") + if line[-1] not in sentence_end_markers: + line += "." tokens += line.split() - tokens = purge_undesired(tokens) if len(tokens) <= select_length: notice("NOT ENOUGH TEXT TO MARKOV.") return + + # Replace URLs with escape string for now, so that the Markov selector + # won't see them as different strings. Stash replaced URLs in urls. + urls = [] + url_escape = "\nURL" + url_starts = ["http://", "https://", "") + 1 + except ValueError: + pass + urls += [tokens[i][:length]] + tokens[i] = url_escape + tokens[i][length:] + break + + # For each snippet of select_length, use markov() to find continuation + # token from selections. Replace present users' names with malkovich. + # Start snippets with the beginning of a sentence, if possible. for i in range(len(tokens) - select_length): token_list = [] for j in range(select_length + 1): @@ -221,16 +236,58 @@ def handle_command(command, argument, notice, target, session): snippet = [] for i in range(select_length): snippet += [""] + shuffle(selections) + for i in range(len(selections)): + if selections[i][0][-1] in sentence_end_markers: + for i in range(select_length): + snippet[i] = selections[i][i + 1] + break msg = "" + malkovich = "malkovich" while 1: new_end = markov(snippet) + for name in session.users_in_chan: + if new_end[:len(name)] == name.lower(): + new_end = malkovich + new_end[len(name):] + break if len(msg) + len(new_end) > 200: break msg += new_end + " " for i in range(select_length - 1): snippet[i] = snippet[i + 1] snippet[select_length - 1] = new_end - notice(msg + "malkovich.") + + # Replace occurences of url escape string with random choice from urls. + while True: + index = msg.find(url_escape) + if index < 0: + break + msg = msg.replace(url_escape, choice(urls), 1) + + # More meaningful ways to randomly end sentences. + notice(msg + malkovich + ".") + + def twt(): + def try_open(mode): + try: + twtfile = open(session.twtfile, mode) + except (PermissionError, FileNotFoundError) as err: + notice("CAN'T ACCESS OR CREATE TWT FILE: " + str(err)) + return None + return twtfile + + from datetime import datetime + if not os.access(session.twtfile, os.F_OK): + twtfile = try_open("w") + if None == twtfile: + return + twtfile.close() + twtfile = try_open("a") + if None == twtfile: + return + twtfile.write(datetime.utcnow().isoformat() + "\t" + argument + "\n") + twtfile.close() + notice("WROTE TWT.") if "addquote" == command: addquote() @@ -238,6 +295,8 @@ def handle_command(command, argument, notice, target, session): quote() elif "markov" == command: markov() + elif "twt" == command: + twt() def handle_url(url, notice, show_url=False): @@ -258,13 +317,14 @@ def handle_url(url, notice, show_url=False): except (requests.exceptions.TooManyRedirects, requests.exceptions.ConnectionError, requests.exceptions.InvalidURL, + UnicodeError, requests.exceptions.InvalidSchema) as error: notice("TROUBLE FOLLOWING URL: " + str(error)) return if mobile_twitter_hack(url): return - title = bs4.BeautifulSoup(r.text, "html.parser").title - if title: + title = bs4.BeautifulSoup(r.text, "html5lib").title + if title and title.string: prefix = "PAGE TITLE: " if show_url: prefix = "PAGE TITLE FOR <" + url + ">: " @@ -275,17 +335,30 @@ def handle_url(url, notice, show_url=False): class Session: - def __init__(self, io, username, nickname, channel): + def __init__(self, io, username, nickname, channel, twtfile, dbdir): self.io = io self.nickname = nickname self.channel = channel - self.uses_in_chan = [] + self.users_in_chan = [] + self.twtfile = twtfile + self.dbdir = dbdir self.io.send_line("NICK " + self.nickname) self.io.send_line("USER " + username + " 0 * : ") self.io.send_line("JOIN " + self.channel) + hash_string = hashlib.md5(self.channel.encode("utf-8")).hexdigest() + self.logdir = self.dbdir + "/irclogs_" + hash_string + "/" + if not os.path.exists(self.logdir): + os.makedirs(self.logdir) def loop(self): + def log(line): + now = datetime.datetime.utcnow() + logfile = open(self.logdir + now.strftime("%Y-%m-%d") + ".txt", "a") + form = "%Y-%m-%d %H:%M:%S UTC\t" + logfile.write(now.strftime(form) + " " + line + "\n") + logfile.close() + def handle_privmsg(tokens): def handle_input(msg, target): @@ -293,7 +366,7 @@ class Session: def notice(msg): self.io.send_line("NOTICE " + target + " :" + msg) - matches = re.findall(URLREGEX, msg) + matches = re.findall("(https?://[^\s>]+)", msg) for i in range(len(matches)): handle_url(matches[i], notice) if "!" == msg[0]: @@ -302,7 +375,7 @@ class Session: handle_command(tokens[0], argument, notice, target, self) return hash_string = hashlib.md5(target.encode("utf-8")).hexdigest() - markovfeed_name = "markovfeed_" + hash_string + markovfeed_name = self.dbdir + "/markovfeed_" + hash_string file = open(markovfeed_name, "a") file.write(msg + "\n") file.close() @@ -323,6 +396,8 @@ class Session: if receiver != self.nickname: target = receiver msg = str.join(" ", tokens[3:])[1:] + if target == self.channel: + log("<" + sender + "> " + msg) handle_input(msg, target) def name_from_join_or_part(tokens): @@ -346,14 +421,22 @@ class Session: elif tokens[1] == "353": names = tokens[5:] names[0] = names[0][1:] - self.uses_in_chan += names + for i in range(len(names)): + names[i] = names[i].replace("@", "").replace("+", "") + self.users_in_chan += names + log(line) elif tokens[1] == "JOIN": name = name_from_join_or_part(tokens) if name != self.nickname: - self.uses_in_chan += [name] + self.users_in_chan += [name] + log(line) elif tokens[1] == "PART": name = name_from_join_or_part(tokens) - del(self.uses_in_chan[self.uses_in_chan.index(name)]) + del(self.users_in_chan[self.users_in_chan.index(name)]) + log(line) + else: + log(line) + def parse_command_line_arguments(): parser = argparse.ArgumentParser() @@ -364,7 +447,7 @@ def parse_command_line_arguments(): parser.add_argument("-p, --port", action="store", dest="port", type=int, default=PORT, help="port to connect to (default : " + str(PORT) + ")") - parser.add_argument("-t, --timeout", action="store", dest="timeout", + parser.add_argument("-w, --wait", action="store", dest="timeout", type=int, default=TIMEOUT, help="timeout in seconds after which to attempt " + "reconnect (default: " + str(TIMEOUT) + ")") @@ -374,6 +457,11 @@ def parse_command_line_arguments(): parser.add_argument("-n, --nickname", action="store", dest="nickname", default=NICKNAME, help="nickname to use (default: " + NICKNAME + ")") + parser.add_argument("-t, --twtxtfile", action="store", dest="twtfile", + default=TWTFILE, help="twtxt file to use (default: " + + TWTFILE + ")") + parser.add_argument("-d, --dbdir", action="store", dest="dbdir", + default=DBDIR, help="directory to store DB files in") parser.add_argument("CHANNEL", action="store", help="channel to join") opts, unknown = parser.parse_known_args() return opts @@ -383,7 +471,8 @@ opts = parse_command_line_arguments() while True: try: io = IO(opts.server, opts.port, opts.timeout) - session = Session(io, opts.username, opts.nickname, opts.CHANNEL) + session = Session(io, opts.username, opts.nickname, opts.CHANNEL, + opts.twtfile, opts.dbdir) session.loop() except ExceptionForRestart: io.socket.close()