X-Git-Url: https://plomlompom.com/repos/foo.html?a=blobdiff_plain;ds=inline;f=plomlombot.py;h=9d04830ce92156b028dd809c349a33e112fe7015;hb=81f01ae26ee8082302e635693f59701911bbfb45;hp=786b072bf483d1c69cfd19e18c18895214bbbf6b;hpb=1961e9f73924c2b27625bdb31545e93ab5e49f89;p=plomlombot-irc.git
diff --git a/plomlombot.py b/plomlombot.py
index 786b072..9d04830 100755
--- a/plomlombot.py
+++ b/plomlombot.py
@@ -20,6 +20,7 @@ TIMEOUT = 240
USERNAME = "plomlombot"
NICKNAME = USERNAME
TWTFILE = ""
+DBDIR = os.path.expanduser("~/plomlombot_db")
class ExceptionForRestart(Exception):
@@ -96,7 +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):
@@ -162,7 +163,7 @@ def handle_command(command, argument, notice, target, session):
notice("QUOTE #" + str(i + 1) + ": " + lines[i])
def markov():
- from random import choice
+ from random import choice, shuffle
select_length = 2
selections = []
@@ -186,20 +187,28 @@ def handle_command(command, argument, notice, target, session):
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://", " 200:
break
@@ -236,17 +256,21 @@ def handle_command(command, argument, notice, target, session):
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):
try:
- twtfile = open(session.twtfile, "w")
+ twtfile = open(session.twtfile, mode)
except (PermissionError, FileNotFoundError) as err:
notice("CAN'T ACCESS OR CREATE TWT FILE: " + str(err))
return None
@@ -293,13 +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 + ">: "
@@ -310,12 +335,13 @@ def handle_url(url, notice, show_url=False):
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.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)
@@ -338,7 +364,7 @@ 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()
@@ -382,14 +408,14 @@ class Session:
elif tokens[1] == "353":
names = tokens[5:]
names[0] = names[0][1:]
- self.uses_in_chan += names
+ self.users_in_chan += names
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]
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)])
def parse_command_line_arguments():
parser = argparse.ArgumentParser()
@@ -410,9 +436,11 @@ 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, --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
@@ -423,7 +451,7 @@ while True:
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()