USERNAME = "plomlombot"
NICKNAME = USERNAME
TWTFILE = ""
+DBDIR = os.path.expanduser("~/plomlombot_db")
class ExceptionForRestart(Exception):
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):
notice("QUOTE #" + str(i + 1) + ": " + lines[i])
def markov():
- from random import choice
+ from random import choice, shuffle
select_length = 2
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()
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://", "<http://", "<https://"]
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):
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):]
+ new_end = malkovich + new_end[len(name):]
break
if len(msg) + len(new_end) > 200:
break
for i in range(select_length - 1):
snippet[i] = snippet[i + 1]
snippet[select_length - 1] = new_end
+
+ # 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)
- notice(msg + "malkovich.")
+
+ # More meaningful ways to randomly end sentences.
+ notice(msg + malkovich + ".")
def twt():
def try_open(mode):
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 + ">: "
class Session:
- def __init__(self, io, username, nickname, channel, twtfile):
+ def __init__(self, io, username, nickname, channel, twtfile, dbdir):
self.io = io
self.nickname = nickname
self.channel = channel
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):
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()
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):
elif tokens[1] == "353":
names = tokens[5:]
names[0] = names[0][1:]
+ 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.users_in_chan += [name]
+ log(line)
elif tokens[1] == "PART":
name = name_from_join_or_part(tokens)
del(self.users_in_chan[self.users_in_chan.index(name)])
+ log(line)
+ else:
+ log(line)
+
def parse_command_line_arguments():
parser = argparse.ArgumentParser()
parser.add_argument("-n, --nickname", action="store", dest="nickname",
default=NICKNAME, help="nickname to use (default: "
+ NICKNAME + ")")
- parser.add_argument("-t, --twtfile", action="store", dest="twtfile",
- default=TWTFILE, help="twtfile to use (default: "
+ 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
try:
io = IO(opts.server, opts.port, opts.timeout)
session = Session(io, opts.username, opts.nickname, opts.CHANNEL,
- opts.twtfile)
+ opts.twtfile, opts.dbdir)
session.loop()
except ExceptionForRestart:
io.socket.close()
projects / plomlombot-irc.git / blobdiff