class Bot: """ Main class for the twitch bot """ version = '1.32' date_fmt = '%H:%M' admin = None url_regex_str = r'\(?(?:(http|https):\/\/)?(?:((?:[^\W\s]|\.|-|[:]{1})+)@{1})?((?:www.)?(?:[^\W\s]|\.|-)+[\.][^\W\s]{2,4}|localhost(?=\/)|\d{1,3}\.\d{1,3}\.\d{1,3}\.\d{1,3})(?::(\d*))?([\/]?[^\s\?]*[\/]{1})*(?:\/?([^\s\n\?\[\]\{\}\#]*(?:(?=\.)){1}|[^\s\n\?\[\]\{\}\.\#]*)?([\.]{1}[^\s\?\#]*)?)?(?:\?{1}([^\s\n\#\[\]]*))?([\#][^\s\n]*)?\)?' last_ping = datetime.datetime.now() last_pong = datetime.datetime.now() def parse_args(): parser = argparse.ArgumentParser() parser.add_argument('--config', '-c', default='config.ini', help='Specify which config file to use ' '(default: config.ini)') parser.add_argument('--silent', action='count', help='Decides whether the bot should be ' 'silent or not') # TODO: Add a log level argument. return parser.parse_args() def load_config(self, config): self.config = config pajbot.models.user.Config.se_sync_token = config['main'].get('se_sync_token', None) pajbot.models.user.Config.se_channel = config['main'].get('se_channel', None) self.domain = config['web'].get('domain', 'localhost') self.nickname = config['main'].get('nickname', 'pajbot') self.password = config['main'].get('password', 'abcdef') self.timezone = config['main'].get('timezone', 'UTC') os.environ['TZ'] = self.timezone if config['main'].getboolean('verified', False): TMI.promote_to_verified() self.trusted_mods = config.getboolean('main', 'trusted_mods') TimeManager.init_timezone(self.timezone) if 'streamer' in config['main']: self.streamer = config['main']['streamer'] self.channel = '#' + self.streamer elif 'target' in config['main']: self.channel = config['main']['target'] self.streamer = self.channel[1:] self.silent = False self.dev = False if 'flags' in config: self.silent = True if 'silent' in config['flags'] and config['flags']['silent'] == '1' else self.silent self.dev = True if 'dev' in config['flags'] and config['flags']['dev'] == '1' else self.dev DBManager.init(self.config['main']['db']) redis_options = {} if 'redis' in config: redis_options = config._sections['redis'] RedisManager.init(**redis_options) def __init__(self, config, args=None): # Load various configuration variables from the given config object # The config object that should be passed through should # come from pajbot.utils.load_config self.load_config(config) # Update the database scheme if necessary using alembic # In case of errors, i.e. if the database is out of sync or the alembic # binary can't be called, we will shut down the bot. pajbot.utils.alembic_upgrade() # Actions in this queue are run in a separate thread. # This means actions should NOT access any database-related stuff. self.action_queue = ActionQueue() self.action_queue.start() self.reactor = irc.client.Reactor(self.on_connect) self.start_time = datetime.datetime.now() ActionParser.bot = self HandlerManager.init_handlers() self.socket_manager = SocketManager(self) self.stream_manager = StreamManager(self) StreamHelper.init_bot(self, self.stream_manager) ScheduleManager.init() self.users = UserManager() self.decks = DeckManager() self.module_manager = ModuleManager(self.socket_manager, bot=self).load() self.commands = CommandManager( socket_manager=self.socket_manager, module_manager=self.module_manager, bot=self).load() self.filters = FilterManager().reload() self.banphrase_manager = BanphraseManager(self).load() self.timer_manager = TimerManager(self).load() self.kvi = KVIManager() self.emotes = EmoteManager(self) self.twitter_manager = TwitterManager(self) HandlerManager.trigger('on_managers_loaded') # Reloadable managers self.reloadable = { 'filters': self.filters, } # Commitable managers self.commitable = { 'commands': self.commands, 'filters': self.filters, 'banphrases': self.banphrase_manager, } self.execute_every(10 * 60, self.commit_all) self.execute_every(1, self.do_tick) try: self.admin = self.config['main']['admin'] except KeyError: log.warning('No admin user specified. See the [main] section in config.example.ini for its usage.') if self.admin: with self.users.get_user_context(self.admin) as user: pass # user.level = 2000 self.parse_version() relay_host = self.config['main'].get('relay_host', None) relay_password = self.config['main'].get('relay_password', None) if relay_host is None or relay_password is None: self.irc = MultiIRCManager(self) else: self.irc = SingleIRCManager(self) self.reactor.add_global_handler('all_events', self.irc._dispatcher, -10) twitch_client_id = None twitch_oauth = None if 'twitchapi' in self.config: twitch_client_id = self.config['twitchapi'].get('client_id', None) twitch_oauth = self.config['twitchapi'].get('oauth', None) # A client ID is required for the bot to work properly now, give an error for now if twitch_client_id is None: log.error('MISSING CLIENT ID, SET "client_id" VALUE UNDER [twitchapi] SECTION IN CONFIG FILE') self.twitchapi = TwitchAPI(twitch_client_id, twitch_oauth) self.data = {} self.data_cb = {} self.url_regex = re.compile(self.url_regex_str, re.IGNORECASE) self.data['broadcaster'] = self.streamer self.data['version'] = self.version self.data['version_brief'] = self.version_brief self.data['bot_name'] = self.nickname self.data_cb['status_length'] = self.c_status_length self.data_cb['stream_status'] = self.c_stream_status self.data_cb['bot_uptime'] = self.c_uptime self.data_cb['current_time'] = self.c_current_time self.silent = True if args.silent else self.silent if self.silent: log.info('Silent mode enabled') """ For actions that need to access the main thread, we can use the mainthread_queue. """ self.mainthread_queue = ActionQueue() self.execute_every(1, self.mainthread_queue.parse_action) self.websocket_manager = WebSocketManager(self) try: if self.config['twitchapi']['update_subscribers'] == '1': self.execute_every(30 * 60, self.action_queue.add, (self.update_subscribers_stage1, )) except: pass # XXX: TEMPORARY UGLY CODE HandlerManager.add_handler('on_user_gain_tokens', self.on_user_gain_tokens) HandlerManager.add_handler('send_whisper', self.whisper) def on_connect(self, sock): return self.irc.on_connect(sock) def on_user_gain_tokens(self, user, tokens_gained): self.whisper(user.username, 'You finished todays quest! You have been awarded with {} tokens.'.format(tokens_gained)) def update_subscribers_stage1(self): limit = 25 offset = 0 subscribers = [] log.info('Starting stage1 subscribers update') try: retry_same = 0 while True: log.debug('Beginning sub request {0} {1}'.format(limit, offset)) subs, retry_same, error = self.twitchapi.get_subscribers(self.streamer, 0, offset, 0 if retry_same is False else retry_same) log.debug('got em!') if error is True: log.error('Too many attempts, aborting') return False if retry_same is False: offset += limit if len(subs) == 0: # We don't need to retry, and the last query finished propery # Break out of the loop and start fiddling with the subs! log.debug('Done!') break else: log.debug('Fetched {0} subs'.format(len(subs))) subscribers.extend(subs) if retry_same is not False: # In case the next attempt is a retry, wait for 3 seconds log.debug('waiting for 3 seconds...') time.sleep(3) log.debug('waited enough!') log.debug('Finished with the while True loop!') except: log.exception('Caught an exception while trying to get subscribers') return log.info('Ended stage1 subscribers update') if len(subscribers) > 0: log.info('Got some subscribers, so we are pushing them to stage 2!') self.mainthread_queue.add(self.update_subscribers_stage2, args=[subscribers]) log.info('Pushed them now.') def update_subscribers_stage2(self, subscribers): self.kvi['active_subs'].set(len(subscribers) - 1) self.users.reset_subs() self.users.update_subs(subscribers) def start(self): """Start the IRC client.""" self.reactor.process_forever() def get_kvi_value(self, key, extra={}): return self.kvi[key].get() def get_last_tweet(self, key, extra={}): return self.twitter_manager.get_last_tweet(key) def get_emote_tm(self, key, extra={}): val = self.emotes.get_emote_epm(key) if not val: return None return '{0:,d}'.format(val) def get_emote_count(self, key, extra={}): val = self.emotes.get_emote_count(key) if not val: return None return '{0:,d}'.format(val) def get_emote_tm_record(self, key, extra={}): val = self.emotes.get_emote_epmrecord(key) if not val: return None return '{0:,d}'.format(val) def get_source_value(self, key, extra={}): try: return getattr(extra['source'], key) except: log.exception('Caught exception in get_source_value') return None def get_user_value(self, key, extra={}): try: user = self.users.find(extra['argument']) if user: return getattr(user, key) except: log.exception('Caught exception in get_source_value') return None def get_command_value(self, key, extra={}): try: return getattr(extra['command'].data, key) except: log.exception('Caught exception in get_source_value') return None def get_usersource_value(self, key, extra={}): try: user = self.users.find(extra['argument']) if user: return getattr(user, key) else: return getattr(extra['source'], key) except: log.exception('Caught exception in get_source_value') return None def get_time_value(self, key, extra={}): try: tz = timezone(key) return datetime.datetime.now(tz).strftime(self.date_fmt) except: log.exception('Unhandled exception in get_time_value') return None def get_current_song_value(self, key, extra={}): if self.stream_manager.online: current_song = PleblistManager.get_current_song(self.stream_manager.current_stream.id) inner_keys = key.split('.') val = current_song for inner_key in inner_keys: val = getattr(val, inner_key, None) if val is None: return None if val is not None: return val return None def get_strictargs_value(self, key, extra={}): ret = self.get_args_value(key, extra) if len(ret) == 0: return None return ret def get_args_value(self, key, extra={}): range = None try: msg_parts = extra['message'].split(' ') except (KeyError, AttributeError): msg_parts = [''] try: if '-' in key: range_str = key.split('-') if len(range_str) == 2: range = (int(range_str[0]), int(range_str[1])) if range is None: range = (int(key), len(msg_parts)) except (TypeError, ValueError): range = (0, len(msg_parts)) try: return ' '.join(msg_parts[range[0]:range[1]]) except AttributeError: return '' except: log.exception('UNHANDLED ERROR IN get_args_value') return '' def get_notify_value(self, key, extra={}): payload = { 'message': extra['message'] or '', 'trigger': extra['trigger'], 'user': extra['source'].username_raw, } self.websocket_manager.emit('notify', payload) return '' def get_value(self, key, extra={}): if key in extra: return extra[key] elif key in self.data: return self.data[key] elif key in self.data_cb: return self.data_cb[key]() log.warning('Unknown key passed to get_value: {0}'.format(key)) return None def privmsg_arr(self, arr): for msg in arr: self.privmsg(msg) def privmsg_from_file(self, url, per_chunk=35, chunk_delay=30): try: r = requests.get(url) lines = r.text.split('\n') i = 0 while len(lines) > 0: if i == 0: self.privmsg_arr(lines[:per_chunk]) else: self.execute_delayed(chunk_delay * i, self.privmsg_arr, (lines[:per_chunk],)) del lines[:per_chunk] i = i + 1 except: log.exception('error in privmsg_from_file') # event is an event to clone and change the text from. def eval_from_file(self, event, url): try: r = requests.get(url) lines = r.text.splitlines() import copy for msg in lines: cloned_event = copy.deepcopy(event) cloned_event.arguments = [msg] # omit the source connectino as None (since its not used) self.on_pubmsg(None, cloned_event) except: log.exception('BabyRage') self.say('Exception') self.say('All lines done') def privmsg(self, message, channel=None, increase_message=True): if channel is None: channel = self.channel return self.irc.privmsg(message, channel, increase_message=increase_message) def c_uptime(self): return time_since(datetime.datetime.now().timestamp(), self.start_time.timestamp()) def c_current_time(self): return datetime.datetime.now() @property def is_online(self): return self.stream_manager.online def c_stream_status(self): return 'online' if self.stream_manager.online else 'offline' def c_status_length(self): if self.stream_manager.online: return time_since(time.time(), self.stream_manager.current_stream.stream_start.timestamp()) else: if self.stream_manager.last_stream is not None: return time_since(time.time(), self.stream_manager.last_stream.stream_end.timestamp()) else: return 'No recorded stream FeelsBadMan ' def execute_at(self, at, function, arguments=()): self.reactor.scheduler.execute_at(at, lambda: function(*arguments)) def execute_delayed(self, delay, function, arguments=()): self.reactor.scheduler.execute_after(delay, lambda: function(*arguments)) def execute_every(self, period, function, arguments=()): self.reactor.scheduler.execute_every(period, lambda: function(*arguments)) def _ban(self, username, reason=''): self.privmsg('.ban {0} {1}'.format(username, reason), increase_message=False) def ban(self, username, reason=''): log.debug('Banning {}'.format(username)) self._timeout(username, 30, reason) self.execute_delayed(1, self._ban, (username, reason)) def ban_user(self, user, reason=''): self._timeout(user.username, 30, reason) self.execute_delayed(1, self._ban, (user.username, reason)) def unban(self, username): self.privmsg('.unban {0}'.format(username), increase_message=False) def _timeout(self, username, duration, reason=''): self.privmsg('.timeout {0} {1} {2}'.format(username, duration, reason), increase_message=False) def timeout(self, username, duration, reason=''): log.debug('Timing out {} for {} seconds'.format(username, duration)) self._timeout(username, duration, reason) # self.execute_delayed(1, self._timeout, (username, duration, reason)) def timeout_warn(self, user, duration, reason=''): duration, punishment = user.timeout(duration, warning_module=self.module_manager['warning']) self.timeout(user.username, duration, reason) return (duration, punishment) def timeout_user(self, user, duration, reason=''): self._timeout(user.username, duration, reason) self.execute_delayed(1, self._timeout, (user.username, duration, reason)) def _timeout_user(self, user, duration, reason=''): self._timeout(user.username, duration, reason) def whisper(self, username, *messages, separator='. '): """ Takes a sequence of strings and concatenates them with separator. Then sends that string as a whisper to username """ if len(messages) < 0: return False message = separator.join(messages) return self.irc.whisper(username, message) def send_message_to_user(self, user, message, event, separator='. ', method='say'): if method == 'say': self.say(user.username + ', ' + lowercase_first_letter(message), separator=separator) elif method == 'whisper': self.whisper(user.username, message, separator=separator) elif method == 'me': self.me(message, separator=separator) elif method == 'reply': if event.type in ['action', 'pubmsg']: self.say(message, separator=separator) elif event.type == 'whisper': self.whisper(user.username, message, separator=separator) else: log.warning('Unknown send_message method: {}'.format(method)) def safe_privmsg(self, message, channel=None, increase_message=True): # Check for banphrases res = self.banphrase_manager.check_message(message, None) if res is not False: self.privmsg('filtered message ({})'.format(res.id), channel, increase_message) return self.privmsg(message, channel, increase_message) def say(self, *messages, channel=None, separator='. '): """ Takes a sequence of strings and concatenates them with separator. Then sends that string to the given channel. """ if len(messages) < 0: return False if not self.silent: message = separator.join(messages).strip() message = clean_up_message(message) if not message: return False # log.info('Sending message: {0}'.format(message)) self.privmsg(message[:510], channel) def is_bad_message(self, message): return self.banphrase_manager.check_message(message, None) is not False def safe_me(self, message, channel=None): if not self.is_bad_message(message): self.me(message, channel) def me(self, message, channel=None): if not self.silent: self.say('.me ' + message[:500], channel=channel) def parse_version(self): self.version = self.version self.version_brief = self.version if self.dev: try: current_branch = subprocess.check_output(['git', 'rev-parse', '--abbrev-ref', 'HEAD']).decode('utf8').strip() latest_commit = subprocess.check_output(['git', 'rev-parse', 'HEAD']).decode('utf8').strip()[:8] commit_number = subprocess.check_output(['git', 'rev-list', 'HEAD', '--count']).decode('utf8').strip() self.version = '{0} DEV ({1}, {2}, commit {3})'.format(self.version, current_branch, latest_commit, commit_number) except: pass def on_welcome(self, chatconn, event): return self.irc.on_welcome(chatconn, event) def connect(self): return self.irc.start() def on_disconnect(self, chatconn, event): self.irc.on_disconnect(chatconn, event) def parse_message(self, msg_raw, source, event, tags={}, whisper=False): msg_lower = msg_raw.lower() emote_tag = None for tag in tags: if tag['key'] == 'subscriber' and event.target == self.channel: source.subscriber = tag['value'] == '1' elif tag['key'] == 'emotes' and tag['value']: emote_tag = tag['value'] elif tag['key'] == 'display-name' and tag['value']: source.username_raw = tag['value'] elif tag['key'] == 'user-type': source.moderator = tag['value'] == 'mod' or source.username == self.streamer # source.num_lines += 1 if source is None: log.error('No valid user passed to parse_message') return False if source.banned: self.ban(source.username) return False # If a user types when timed out, we assume he's been unbanned for a good reason and remove his flag. if source.timed_out is True: source.timed_out = False # Parse emotes in the message message_emotes = self.emotes.parse_message_twitch_emotes(source, msg_raw, emote_tag, whisper) urls = self.find_unique_urls(msg_raw) if whisper: self.whisper('datguy1', '{} said: {}'.format(source.username, msg_raw)) # log.debug('{2}{0}: {1}'.format(source.username, msg_raw, '<w>' if whisper else '')) res = HandlerManager.trigger('on_message', source, msg_raw, message_emotes, whisper, urls, event, stop_on_false=True) if res is False: return False source.last_seen = datetime.datetime.now() source.last_active = datetime.datetime.now() if source.ignored: return False if msg_lower[:1] == '!': msg_lower_parts = msg_lower.split(' ') trigger = msg_lower_parts[0][1:] msg_raw_parts = msg_raw.split(' ') remaining_message = ' '.join(msg_raw_parts[1:]) if len(msg_raw_parts) > 1 else None if trigger in self.commands: command = self.commands[trigger] extra_args = { 'emotes': message_emotes, 'trigger': trigger, } command.run(self, source, remaining_message, event=event, args=extra_args, whisper=whisper) @time_method def redis_test(self, username): try: pipeline = RedisManager.get().pipeline() pipeline.hget('pajlada:users:points', username) pipeline.hget('pajlada:users:ignored', username) pipeline.hget('pajlada:users:banned', username) pipeline.hget('pajlada:users:last_active', username) except: pipeline.reset() finally: b = pipeline.execute() log.info(b) @time_method def redis_test2(self, username): redis = RedisManager.get() values = [ redis.hget('pajlada:users:points', username), redis.hget('pajlada:users:ignored', username), redis.hget('pajlada:users:banned', username), redis.hget('pajlada:users:last_active', username), ] return values def on_whisper(self, chatconn, event): # We use .lower() in case twitch ever starts sending non-lowercased usernames username = event.source.user.lower() with self.users.get_user_context(username) as source: self.parse_message(event.arguments[0], source, event, whisper=True, tags=event.tags) def on_ping(self, chatconn, event): # self.say('Received a ping. Last ping received {} ago'.format(time_since(datetime.datetime.now().timestamp(), self.last_ping.timestamp()))) # log.info('Received a ping. Last ping received {} ago'.format(time_since(datetime.datetime.now().timestamp(), self.last_ping.timestamp()))) self.last_ping = datetime.datetime.now() def on_pong(self, chatconn, event): # self.say('Received a pong. Last pong received {} ago'.format(time_since(datetime.datetime.now().timestamp(), self.last_pong.timestamp()))) # log.info('Received a pong. Last pong received {} ago'.format(time_since(datetime.datetime.now().timestamp(), self.last_pong.timestamp()))) self.last_pong = datetime.datetime.now() def on_pubnotice(self, chatconn, event): return type = 'whisper' if chatconn in self.whisper_manager else 'normal' log.debug('NOTICE {}@{}: {}'.format(type, event.target, event.arguments)) def on_usernotice(self, chatconn, event): # We use .lower() in case twitch ever starts sending non-lowercased usernames tags = {} for d in event.tags: tags[d['key']] = d['value'] if 'login' not in tags: return username = tags['login'] with self.users.get_user_context(username) as source: msg = '' if len(event.arguments) > 0: msg = event.arguments[0] HandlerManager.trigger('on_usernotice', source, msg, tags) def on_action(self, chatconn, event): self.on_pubmsg(chatconn, event) def on_pubmsg(self, chatconn, event): if event.source.user == self.nickname: return False username = event.source.user.lower() # We use .lower() in case twitch ever starts sending non-lowercased usernames with self.users.get_user_context(username) as source: res = HandlerManager.trigger('on_pubmsg', source, event.arguments[0], stop_on_false=True) if res is False: return False self.parse_message(event.arguments[0], source, event, tags=event.tags) @time_method def reload_all(self): log.info('Reloading all...') for key, manager in self.reloadable.items(): log.debug('Reloading {0}'.format(key)) manager.reload() log.debug('Done with {0}'.format(key)) log.info('ok!') @time_method def commit_all(self): log.info('Commiting all...') for key, manager in self.commitable.items(): log.info('Commiting {0}'.format(key)) manager.commit() log.info('Done with {0}'.format(key)) log.info('ok!') HandlerManager.trigger('on_commit', stop_on_false=False) def do_tick(self): HandlerManager.trigger('on_tick') def quit(self, message, event, **options): quit_chub = self.config['main'].get('control_hub', None) quit_delay = 0 if quit_chub is not None and event.target == ('#{}'.format(quit_chub)): quit_delay_random = 300 try: if message is not None and int(message.split()[0]) >= 1: quit_delay_random = int(message.split()[0]) except (IndexError, ValueError, TypeError): pass quit_delay = random.randint(0, quit_delay_random) log.info('{} is restarting in {} seconds.'.format(self.nickname, quit_delay)) self.execute_delayed(quit_delay, self.quit_bot) def quit_bot(self, **options): self.commit_all() quit = '{nickname} {version} shutting down...' phrase_data = { 'nickname': self.nickname, 'version': self.version, } try: ScheduleManager.base_scheduler.print_jobs() ScheduleManager.base_scheduler.shutdown(wait=False) except: log.exception('Error while shutting down the apscheduler') try: self.say('I have to leave PepeHands') except Exception: log.exception('Exception caught while trying to say quit phrase') self.twitter_manager.quit() self.socket_manager.quit() self.irc.quit() sys.exit(0) def apply_filter(self, resp, filter): available_filters = { 'strftime': _filter_strftime, 'lower': lambda var, args: var.lower(), 'upper': lambda var, args: var.upper(), 'time_since_minutes': lambda var, args: 'no time' if var == 0 else time_since(var * 60, 0, format='long'), 'time_since': lambda var, args: 'no time' if var == 0 else time_since(var, 0, format='long'), 'time_since_dt': _filter_time_since_dt, 'urlencode': _filter_urlencode, 'join': _filter_join, 'number_format': _filter_number_format, 'add': _filter_add, } if filter.name in available_filters: return available_filters[filter.name](resp, filter.arguments) return resp def find_unique_urls(self, message): from pajbot.modules.linkchecker import find_unique_urls return find_unique_urls(self.url_regex, message)
class FollowAgeModule(BaseModule): ID = __name__.split(".")[-1] NAME = "Follow age" DESCRIPTION = "Makes two commands available: !followage and !followsince" CATEGORY = "Feature" SETTINGS = [ ModuleSetting( key="action_followage", label="MessageAction for !followage", type="options", required=True, default="say", options=["say", "whisper", "reply"], ), ModuleSetting( key="action_followsince", label="MessageAction for !followsince", type="options", required=True, default="say", options=["say", "whisper", "reply"], ), ModuleSetting( key="global_cd", label="Global cooldown (seconds)", type="number", required=True, placeholder="", default=4, constraints={"min_value": 0, "max_value": 120}, ), ModuleSetting( key="user_cd", label="Per-user cooldown (seconds)", type="number", required=True, placeholder="", default=8, constraints={"min_value": 0, "max_value": 240}, ), ] def __init__(self, bot): super().__init__(bot) self.action_queue = ActionQueue() self.action_queue.start() def load_commands(self, **options): # TODO: Have delay modifiable in settings self.commands["followage"] = Command.raw_command( self.follow_age, delay_all=self.settings["global_cd"], delay_user=self.settings["user_cd"], description="Check your or someone elses follow age for a channel", can_execute_with_whisper=True, examples=[ CommandExample( None, "Check your own follow age", chat="user:!followage\n" "bot:pajlada, you have been following Karl_Kons for 4 months and 24 days", description="Check how long you have been following the current streamer (Karl_Kons in this case)", ).parse(), CommandExample( None, "Check someone elses follow age", chat="user:!followage NightNacht\n" "bot:pajlada, NightNacht has been following Karl_Kons for 5 months and 4 days", description="Check how long any user has been following the current streamer (Karl_Kons in this case)", ).parse(), CommandExample( None, "Check someones follow age for a certain streamer", chat="user:!followage NightNacht forsenlol\n" "bot:pajlada, NightNacht has been following forsenlol for 1 year and 4 months", description="Check how long NightNacht has been following forsenlol", ).parse(), CommandExample( None, "Check your own follow age for a certain streamer", chat="user:!followage pajlada forsenlol\n" "bot:pajlada, you have been following forsenlol for 1 year and 3 months", description="Check how long you have been following forsenlol", ).parse(), ], ) self.commands["followsince"] = Command.raw_command( self.follow_since, delay_all=self.settings["global_cd"], delay_user=self.settings["user_cd"], description="Check from when you or someone else first followed a channel", can_execute_with_whisper=True, examples=[ CommandExample( None, "Check your own follow since", chat="user:!followsince\n" "bot:pajlada, you have been following Karl_Kons since 04 March 2015, 07:02:01 UTC", description="Check when you first followed the current streamer (Karl_Kons in this case)", ).parse(), CommandExample( None, "Check someone elses follow since", chat="user:!followsince NightNacht\n" "bot:pajlada, NightNacht has been following Karl_Kons since 03 July 2014, 04:12:42 UTC", description="Check when NightNacht first followed the current streamer (Karl_Kons in this case)", ).parse(), CommandExample( None, "Check someone elses follow since for another streamer", chat="user:!followsince NightNacht forsenlol\n" "bot:pajlada, NightNacht has been following forsenlol since 13 June 2013, 13:10:51 UTC", description="Check when NightNacht first followed the given streamer (forsenlol)", ).parse(), CommandExample( None, "Check your follow since for another streamer", chat="user:!followsince pajlada forsenlol\n" "bot:pajlada, you have been following forsenlol since 16 December 1990, 03:06:51 UTC", description="Check when you first followed the given streamer (forsenlol)", ).parse(), ], ) def check_follow_age(self, bot, source, username, streamer, event): streamer = bot.streamer if streamer is None else streamer.lower() age = bot.twitchapi.get_follow_relationship(username, streamer) is_self = source.username == username message = "" streamer = streamer.replace('admiralbulldog', 'Buldog') if age: # Following human_age = time_since(utils.now().timestamp() - age.timestamp(), 0) message = "{} has been following {} for {}".format(username, streamer, human_age) else: # Not following message = "{} is not following {}".format(username, streamer) bot.send_message_to_user(source, message, event, method=self.settings["action_followage"]) def check_follow_since(self, bot, source, username, streamer, event): streamer = bot.streamer if streamer is None else streamer.lower() follow_since = bot.twitchapi.get_follow_relationship(username, streamer) is_self = source.username == username message = "" streamer = streamer.replace('admiralbulldog', 'Buldog') if follow_since: # Following human_age = follow_since.strftime("%d %B %Y, %X") message = "{} has been following {} since {} UTC".format(username, streamer, human_age) else: # Not following message = "{} is not following {}".format(username, streamer) bot.send_message_to_user(source, message, event, method=self.settings["action_followsince"]) def follow_age(self, **options): source = options["source"] message = options["message"] bot = options["bot"] event = options["event"] username, streamer = self.parse_message(bot, source, message) self.action_queue.add(self.check_follow_age, args=[bot, source, username, streamer, event]) def follow_since(self, **options): bot = options["bot"] source = options["source"] message = options["message"] event = options["event"] username, streamer = self.parse_message(bot, source, message) self.action_queue.add(self.check_follow_since, args=[bot, source, username, streamer, event]) @staticmethod def parse_message(bot, source, message): username = source.username streamer = None if message is not None and len(message) > 0: message_split = message.split(" ") if len(message_split[0]) and message_split[0].replace("_", "").isalnum(): username = message_split[0].lower() if len(message_split) > 1 and message_split[1].replace("_", "").isalnum(): streamer = message_split[1] return username, streamer
class Bot: """ Main class for the twitch bot """ version = '1.30' date_fmt = '%H:%M' admin = None url_regex_str = r'\(?(?:(http|https):\/\/)?(?:((?:[^\W\s]|\.|-|[:]{1})+)@{1})?((?:www.)?(?:[^\W\s]|\.|-)+[\.][^\W\s]{2,4}|localhost(?=\/)|\d{1,3}\.\d{1,3}\.\d{1,3}\.\d{1,3})(?::(\d*))?([\/]?[^\s\?]*[\/]{1})*(?:\/?([^\s\n\?\[\]\{\}\#]*(?:(?=\.)){1}|[^\s\n\?\[\]\{\}\.\#]*)?([\.]{1}[^\s\?\#]*)?)?(?:\?{1}([^\s\n\#\[\]]*))?([\#][^\s\n]*)?\)?' last_ping = datetime.datetime.now() last_pong = datetime.datetime.now() def parse_args(): parser = argparse.ArgumentParser() parser.add_argument('--config', '-c', default='config.ini', help='Specify which config file to use ' '(default: config.ini)') parser.add_argument('--silent', action='count', help='Decides whether the bot should be ' 'silent or not') # TODO: Add a log level argument. return parser.parse_args() def load_config(self, config): self.config = config self.domain = config['web'].get('domain', 'localhost') self.nickname = config['main'].get('nickname', 'pajbot') self.password = config['main'].get('password', 'abcdef') self.timezone = config['main'].get('timezone', 'UTC') self.trusted_mods = config.getboolean('main', 'trusted_mods') TimeManager.init_timezone(self.timezone) if 'streamer' in config['main']: self.streamer = config['main']['streamer'] self.channel = '#' + self.streamer elif 'target' in config['main']: self.channel = config['main']['target'] self.streamer = self.channel[1:] self.wolfram = None if 'wolfram' in config['main']: try: import wolframalpha self.wolfram = wolframalpha.Client(config['main']['wolfram']) except ImportError: pass self.silent = False self.dev = False if 'flags' in config: self.silent = True if 'silent' in config['flags'] and config['flags']['silent'] == '1' else self.silent self.dev = True if 'dev' in config['flags'] and config['flags']['dev'] == '1' else self.dev DBManager.init(self.config['main']['db']) redis_options = {} if 'redis' in config: redis_options = config._sections['redis'] RedisManager.init(**redis_options) def __init__(self, config, args=None): # Load various configuration variables from the given config object # The config object that should be passed through should # come from pajbot.utils.load_config self.load_config(config) # Update the database scheme if necessary using alembic # In case of errors, i.e. if the database is out of sync or the alembic # binary can't be called, we will shut down the bot. pajbot.utils.alembic_upgrade() # Actions in this queue are run in a separate thread. # This means actions should NOT access any database-related stuff. self.action_queue = ActionQueue() self.action_queue.start() self.reactor = irc.client.Reactor(self.on_connect) self.start_time = datetime.datetime.now() ActionParser.bot = self HandlerManager.init_handlers() self.socket_manager = SocketManager(self) self.stream_manager = StreamManager(self) StreamHelper.init_bot(self, self.stream_manager) ScheduleManager.init() self.users = UserManager() self.decks = DeckManager() self.module_manager = ModuleManager(self.socket_manager, bot=self).load() self.commands = CommandManager( socket_manager=self.socket_manager, module_manager=self.module_manager, bot=self).load() self.filters = FilterManager().reload() self.banphrase_manager = BanphraseManager(self).load() self.timer_manager = TimerManager(self).load() self.kvi = KVIManager() self.emotes = EmoteManager(self) self.twitter_manager = TwitterManager(self) HandlerManager.trigger('on_managers_loaded') # Reloadable managers self.reloadable = { 'filters': self.filters, } # Commitable managers self.commitable = { 'commands': self.commands, 'filters': self.filters, 'banphrases': self.banphrase_manager, } self.execute_every(10 * 60, self.commit_all) self.execute_every(1, self.do_tick) try: self.admin = self.config['main']['admin'] except KeyError: log.warning('No admin user specified. See the [main] section in config.example.ini for its usage.') if self.admin: with self.users.get_user_context(self.admin) as user: user.level = 2000 self.parse_version() relay_host = self.config['main'].get('relay_host', None) relay_password = self.config['main'].get('relay_password', None) if relay_host is None or relay_password is None: self.irc = MultiIRCManager(self) else: self.irc = SingleIRCManager(self) self.reactor.add_global_handler('all_events', self.irc._dispatcher, -10) twitch_client_id = None twitch_oauth = None if 'twitchapi' in self.config: twitch_client_id = self.config['twitchapi'].get('client_id', None) twitch_oauth = self.config['twitchapi'].get('oauth', None) # A client ID is required for the bot to work properly now, give an error for now if twitch_client_id is None: log.error('MISSING CLIENT ID, SET "client_id" VALUE UNDER [twitchapi] SECTION IN CONFIG FILE') self.twitchapi = TwitchAPI(twitch_client_id, twitch_oauth) self.data = {} self.data_cb = {} self.url_regex = re.compile(self.url_regex_str, re.IGNORECASE) self.data['broadcaster'] = self.streamer self.data['version'] = self.version self.data['version_brief'] = self.version_brief self.data['bot_name'] = self.nickname self.data_cb['status_length'] = self.c_status_length self.data_cb['stream_status'] = self.c_stream_status self.data_cb['bot_uptime'] = self.c_uptime self.data_cb['current_time'] = self.c_current_time self.silent = True if args.silent else self.silent if self.silent: log.info('Silent mode enabled') """ For actions that need to access the main thread, we can use the mainthread_queue. """ self.mainthread_queue = ActionQueue() self.execute_every(1, self.mainthread_queue.parse_action) self.websocket_manager = WebSocketManager(self) try: if self.config['twitchapi']['update_subscribers'] == '1': self.execute_every(30 * 60, self.action_queue.add, (self.update_subscribers_stage1, )) except: pass # XXX: TEMPORARY UGLY CODE HandlerManager.add_handler('on_user_gain_tokens', self.on_user_gain_tokens) HandlerManager.add_handler('send_whisper', self.whisper) def on_connect(self, sock): return self.irc.on_connect(sock) def on_user_gain_tokens(self, user, tokens_gained): self.whisper(user.username, 'You finished todays quest! You have been awarded with {} tokens.'.format(tokens_gained)) def update_subscribers_stage1(self): limit = 100 offset = 0 subscribers = [] log.info('Starting stage1 subscribers update') try: retry_same = 0 while True: log.debug('Beginning sub request {0} {1}'.format(limit, offset)) subs, retry_same, error = self.twitchapi.get_subscribers(self.streamer, limit, offset, 0 if retry_same is False else retry_same) log.debug('got em!') if error is True: log.error('Too many attempts, aborting') return False if retry_same is False: offset += limit if len(subs) == 0: # We don't need to retry, and the last query finished propery # Break out of the loop and start fiddling with the subs! log.debug('Done!') break else: log.debug('Fetched {0} subs'.format(len(subs))) subscribers.extend(subs) if retry_same is not False: # In case the next attempt is a retry, wait for 3 seconds log.debug('waiting for 3 seconds...') time.sleep(3) log.debug('waited enough!') log.debug('Finished with the while True loop!') except: log.exception('Caught an exception while trying to get subscribers') return log.info('Ended stage1 subscribers update') if len(subscribers) > 0: log.info('Got some subscribers, so we are pushing them to stage 2!') self.mainthread_queue.add(self.update_subscribers_stage2, args=[subscribers]) log.info('Pushed them now.') def update_subscribers_stage2(self, subscribers): self.kvi['active_subs'].set(len(subscribers) - 1) self.users.reset_subs() self.users.update_subs(subscribers) def start(self): """Start the IRC client.""" self.reactor.process_forever() def get_kvi_value(self, key, extra={}): return self.kvi[key].get() def get_last_tweet(self, key, extra={}): return self.twitter_manager.get_last_tweet(key) def get_emote_tm(self, key, extra={}): val = self.emotes.get_emote_epm(key) if not val: return None return '{0:,d}'.format(val) def get_emote_count(self, key, extra={}): val = self.emotes.get_emote_count(key) if not val: return None return '{0:,d}'.format(val) def get_emote_tm_record(self, key, extra={}): val = self.emotes.get_emote_epmrecord(key) if not val: return None return '{0:,d}'.format(val) def get_source_value(self, key, extra={}): try: return getattr(extra['source'], key) except: log.exception('Caught exception in get_source_value') return None def get_user_value(self, key, extra={}): try: user = self.users.find(extra['argument']) if user: return getattr(user, key) except: log.exception('Caught exception in get_source_value') return None def get_command_value(self, key, extra={}): try: return getattr(extra['command'].data, key) except: log.exception('Caught exception in get_source_value') return None def get_usersource_value(self, key, extra={}): try: user = self.users.find(extra['argument']) if user: return getattr(user, key) else: return getattr(extra['source'], key) except: log.exception('Caught exception in get_source_value') return None def get_time_value(self, key, extra={}): try: tz = timezone(key) return datetime.datetime.now(tz).strftime(self.date_fmt) except: log.exception('Unhandled exception in get_time_value') return None def get_current_song_value(self, key, extra={}): if self.stream_manager.online: current_song = PleblistManager.get_current_song(self.stream_manager.current_stream.id) inner_keys = key.split('.') val = current_song for inner_key in inner_keys: val = getattr(val, inner_key, None) if val is None: return None if val is not None: return val return None def get_strictargs_value(self, key, extra={}): ret = self.get_args_value(key, extra) if len(ret) == 0: return None return ret def get_args_value(self, key, extra={}): range = None try: msg_parts = extra['message'].split(' ') except (KeyError, AttributeError): msg_parts = [''] try: if '-' in key: range_str = key.split('-') if len(range_str) == 2: range = (int(range_str[0]), int(range_str[1])) if range is None: range = (int(key), len(msg_parts)) except (TypeError, ValueError): range = (0, len(msg_parts)) try: return ' '.join(msg_parts[range[0]:range[1]]) except AttributeError: return '' except: log.exception('UNHANDLED ERROR IN get_args_value') return '' def get_notify_value(self, key, extra={}): payload = { 'message': extra['message'] or '', 'trigger': extra['trigger'], 'user': extra['source'].username_raw, } self.websocket_manager.emit('notify', payload) return '' def get_value(self, key, extra={}): if key in extra: return extra[key] elif key in self.data: return self.data[key] elif key in self.data_cb: return self.data_cb[key]() log.warning('Unknown key passed to get_value: {0}'.format(key)) return None def privmsg_arr(self, arr): for msg in arr: self.privmsg(msg) def privmsg_from_file(self, url, per_chunk=35, chunk_delay=30): try: r = requests.get(url) lines = r.text.split('\n') i = 0 while len(lines) > 0: if i == 0: self.privmsg_arr(lines[:per_chunk]) else: self.execute_delayed(chunk_delay * i, self.privmsg_arr, (lines[:per_chunk],)) del lines[:per_chunk] i = i + 1 except: log.exception('error in privmsg_from_file') def privmsg(self, message, channel=None, increase_message=True): if channel is None: channel = self.channel return self.irc.privmsg(message, channel, increase_message=increase_message) def c_uptime(self): return time_since(datetime.datetime.now().timestamp(), self.start_time.timestamp()) def c_current_time(self): return datetime.datetime.now() @property def is_online(self): return self.stream_manager.online def c_stream_status(self): return 'online' if self.stream_manager.online else 'offline' def c_status_length(self): if self.stream_manager.online: return time_since(time.time(), self.stream_manager.current_stream.stream_start.timestamp()) else: if self.stream_manager.last_stream is not None: return time_since(time.time(), self.stream_manager.last_stream.stream_end.timestamp()) else: return 'No recorded stream FeelsBadMan ' def execute_at(self, at, function, arguments=()): self.reactor.scheduler.execute_at(at, lambda: function(*arguments)) def execute_delayed(self, delay, function, arguments=()): self.reactor.scheduler.execute_after(delay, lambda: function(*arguments)) def execute_every(self, period, function, arguments=()): self.reactor.scheduler.execute_every(period, lambda: function(*arguments)) def _ban(self, username, reason=''): self.privmsg('.ban {0} {1}'.format(username, reason), increase_message=False) def ban(self, username, reason=''): log.debug('Banning {}'.format(username)) self._timeout(username, 30, reason) self.execute_delayed(1, self._ban, (username, reason)) def ban_user(self, user, reason=''): self._timeout(user.username, 30, reason) self.execute_delayed(1, self._ban, (user.username, reason)) def unban(self, username): self.privmsg('.unban {0}'.format(username), increase_message=False) def _timeout(self, username, duration, reason=''): self.privmsg('.timeout {0} {1} {2}'.format(username, duration, reason), increase_message=False) def timeout(self, username, duration, reason=''): log.debug('Timing out {} for {} seconds'.format(username, duration)) self._timeout(username, duration, reason) self.execute_delayed(1, self._timeout, (username, duration, reason)) def timeout_warn(self, user, duration, reason=''): duration, punishment = user.timeout(duration, warning_module=self.module_manager['warning']) self.timeout(user.username, duration, reason) return (duration, punishment) def timeout_user(self, user, duration, reason=''): self._timeout(user.username, duration, reason) self.execute_delayed(1, self._timeout, (user.username, duration, reason)) def whisper(self, username, *messages, separator='. '): """ Takes a sequence of strings and concatenates them with separator. Then sends that string as a whisper to username """ if len(messages) < 0: return False message = separator.join(messages) return self.irc.whisper(username, message) def send_message_to_user(self, user, message, separator='. ', method='say'): if method == 'say': self.say(user.username + ', ' + lowercase_first_letter(message), separator=separator) elif method == 'whisper': self.whisper(user.username, message, separator=separator) else: log.warning('Unknown send_message method: {}'.format(method)) def say(self, *messages, channel=None, separator='. '): """ Takes a sequence of strings and concatenates them with separator. Then sends that string to the given channel. """ if len(messages) < 0: return False if not self.silent: message = separator.join(messages).strip() if len(message) >= 1: if (message[0] == '.' or message[0] == '/') and not message[1:3] == 'me': log.warning('Message we attempted to send started with . or /, skipping.') return log.info('Sending message: {0}'.format(message)) self.privmsg(message[:510], channel) def me(self, message, channel=None): if not self.silent: message = message.strip() if len(message) >= 1: if message[0] == '.' or message[0] == '/': log.warning('Message we attempted to send started with . or /, skipping.') return log.info('Sending message: {0}'.format(message)) self.privmsg('.me ' + message[:500], channel) def parse_version(self): self.version = self.version self.version_brief = self.version if self.dev: try: current_branch = subprocess.check_output(['git', 'rev-parse', '--abbrev-ref', 'HEAD']).decode('utf8').strip() latest_commit = subprocess.check_output(['git', 'rev-parse', 'HEAD']).decode('utf8').strip()[:8] commit_number = subprocess.check_output(['git', 'rev-list', 'HEAD', '--count']).decode('utf8').strip() self.version = '{0} DEV ({1}, {2}, commit {3})'.format(self.version, current_branch, latest_commit, commit_number) except: pass def on_welcome(self, chatconn, event): return self.irc.on_welcome(chatconn, event) def connect(self): return self.irc.start() def on_disconnect(self, chatconn, event): self.irc.on_disconnect(chatconn, event) def parse_message(self, msg_raw, source, event, tags={}, whisper=False): msg_lower = msg_raw.lower() emote_tag = None for tag in tags: if tag['key'] == 'subscriber' and event.target == self.channel: source.subscriber = tag['value'] == '1' elif tag['key'] == 'emotes' and tag['value']: emote_tag = tag['value'] elif tag['key'] == 'display-name' and tag['value']: source.username_raw = tag['value'] elif tag['key'] == 'user-type': source.moderator = tag['value'] == 'mod' or source.username == self.streamer # source.num_lines += 1 if source is None: log.error('No valid user passed to parse_message') return False if source.banned: self.ban(source.username) return False # If a user types when timed out, we assume he's been unbanned for a good reason and remove his flag. if source.timed_out is True: source.timed_out = False # Parse emotes in the message message_emotes = self.emotes.parse_message_twitch_emotes(source, msg_raw, emote_tag, whisper) urls = self.find_unique_urls(msg_raw) log.debug('{2}{0}: {1}'.format(source.username, msg_raw, '<w>' if whisper else '')) res = HandlerManager.trigger('on_message', source, msg_raw, message_emotes, whisper, urls, event, stop_on_false=True) if res is False: return False source.last_seen = datetime.datetime.now() source.last_active = datetime.datetime.now() """ if self.streamer == 'forsenlol' and whisper is False: follow_time = self.twitchapi.get_follow_relationship2(source.username, self.streamer) if follow_time is False: self._timeout(source.username, 600, '2 years follow mode (or api is down?)') return follow_age = datetime.datetime.now() - follow_time log.debug(follow_age) if follow_age.days < 730: log.debug('followed less than 730 days LUL') self._timeout(source.username, 600, '2 years follow mode') """ if source.ignored: return False if msg_lower[:1] == '!': msg_lower_parts = msg_lower.split(' ') trigger = msg_lower_parts[0][1:] msg_raw_parts = msg_raw.split(' ') remaining_message = ' '.join(msg_raw_parts[1:]) if len(msg_raw_parts) > 1 else None if trigger in self.commands: command = self.commands[trigger] extra_args = { 'emotes': message_emotes, 'trigger': trigger, } command.run(self, source, remaining_message, event=event, args=extra_args, whisper=whisper) @time_method def redis_test(self, username): try: pipeline = RedisManager.get().pipeline() pipeline.hget('pajlada:users:points', username) pipeline.hget('pajlada:users:ignored', username) pipeline.hget('pajlada:users:banned', username) pipeline.hget('pajlada:users:last_active', username) except: pipeline.reset() finally: b = pipeline.execute() log.info(b) @time_method def redis_test2(self, username): redis = RedisManager.get() values = [ redis.hget('pajlada:users:points', username), redis.hget('pajlada:users:ignored', username), redis.hget('pajlada:users:banned', username), redis.hget('pajlada:users:last_active', username), ] return values def on_whisper(self, chatconn, event): # We use .lower() in case twitch ever starts sending non-lowercased usernames username = event.source.user.lower() with self.users.get_user_context(username) as source: self.parse_message(event.arguments[0], source, event, whisper=True, tags=event.tags) def on_ping(self, chatconn, event): # self.say('Received a ping. Last ping received {} ago'.format(time_since(datetime.datetime.now().timestamp(), self.last_ping.timestamp()))) log.info('Received a ping. Last ping received {} ago'.format(time_since(datetime.datetime.now().timestamp(), self.last_ping.timestamp()))) self.last_ping = datetime.datetime.now() def on_pong(self, chatconn, event): # self.say('Received a pong. Last pong received {} ago'.format(time_since(datetime.datetime.now().timestamp(), self.last_pong.timestamp()))) log.info('Received a pong. Last pong received {} ago'.format(time_since(datetime.datetime.now().timestamp(), self.last_pong.timestamp()))) self.last_pong = datetime.datetime.now() def on_pubnotice(self, chatconn, event): return type = 'whisper' if chatconn in self.whisper_manager else 'normal' log.debug('NOTICE {}@{}: {}'.format(type, event.target, event.arguments)) def on_usernotice(self, chatconn, event): # We use .lower() in case twitch ever starts sending non-lowercased usernames tags = {} for d in event.tags: tags[d['key']] = d['value'] if 'login' not in tags: return username = tags['login'] with self.users.get_user_context(username) as source: msg = '' if len(event.arguments) > 0: msg = event.arguments[0] HandlerManager.trigger('on_usernotice', source, msg, tags) def on_action(self, chatconn, event): self.on_pubmsg(chatconn, event) def on_pubmsg(self, chatconn, event): if event.source.user == self.nickname: return False username = event.source.user.lower() # We use .lower() in case twitch ever starts sending non-lowercased usernames with self.users.get_user_context(username) as source: res = HandlerManager.trigger('on_pubmsg', source, event.arguments[0], stop_on_false=True) if res is False: return False self.parse_message(event.arguments[0], source, event, tags=event.tags) @time_method def reload_all(self): log.info('Reloading all...') for key, manager in self.reloadable.items(): log.debug('Reloading {0}'.format(key)) manager.reload() log.debug('Done with {0}'.format(key)) log.info('ok!') @time_method def commit_all(self): log.info('Commiting all...') for key, manager in self.commitable.items(): log.info('Commiting {0}'.format(key)) manager.commit() log.info('Done with {0}'.format(key)) log.info('ok!') HandlerManager.trigger('on_commit', stop_on_false=False) def do_tick(self): HandlerManager.trigger('on_tick') def quit(self, message, event, **options): quit_chub = self.config['main'].get('control_hub', None) quit_delay = 0 if quit_chub is not None and event.target == ('#{}'.format(quit_chub)): quit_delay_random = 300 try: if message is not None and int(message.split()[0]) >= 1: quit_delay_random = int(message.split()[0]) except (IndexError, ValueError, TypeError): pass quit_delay = random.randint(0, quit_delay_random) log.info('{} is restarting in {} seconds.'.format(self.nickname, quit_delay)) self.execute_delayed(quit_delay, self.quit_bot) def quit_bot(self, **options): self.commit_all() quit = '{nickname} {version} shutting down...' phrase_data = { 'nickname': self.nickname, 'version': self.version, } try: ScheduleManager.base_scheduler.print_jobs() ScheduleManager.base_scheduler.shutdown(wait=False) except: log.exception('Error while shutting down the apscheduler') try: self.say(quit.format(**phrase_data)) except Exception: log.exception('Exception caught while trying to say quit phrase') self.twitter_manager.quit() self.socket_manager.quit() self.irc.quit() sys.exit(0) def apply_filter(self, resp, filter): available_filters = { 'strftime': _filter_strftime, 'lower': lambda var, args: var.lower(), 'upper': lambda var, args: var.upper(), 'time_since_minutes': lambda var, args: 'no time' if var == 0 else time_since(var * 60, 0, format='long'), 'time_since': lambda var, args: 'no time' if var == 0 else time_since(var, 0, format='long'), 'time_since_dt': _filter_time_since_dt, 'urlencode': _filter_urlencode, 'join': _filter_join, 'number_format': _filter_number_format, } if filter.name in available_filters: return available_filters[filter.name](resp, filter.arguments) return resp def find_unique_urls(self, message): from pajbot.modules.linkchecker import find_unique_urls return find_unique_urls(self.url_regex, message)
class LinkCheckerModule(BaseModule): ID = __name__.split('.')[-1] NAME = 'Link Checker' DESCRIPTION = 'Checks links if they\'re bad' ENABLED_DEFAULT = True CATEGORY = 'Filter' SETTINGS = [ ModuleSetting(key='ban_pleb_links', label='Disallow links from non-subscribers', type='boolean', required=True, default=False) ] def __init__(self): super().__init__() self.db_session = None self.links = {} self.blacklisted_links = [] self.whitelisted_links = [] self.cache = LinkCheckerCache( ) # cache[url] = True means url is safe, False means the link is bad self.action_queue = ActionQueue() self.action_queue.start() def enable(self, bot): self.bot = bot pajbot.managers.handler.HandlerManager.add_handler('on_message', self.on_message, priority=100) pajbot.managers.handler.HandlerManager.add_handler( 'on_commit', self.on_commit) if bot: self.run_later = bot.execute_delayed if 'safebrowsingapi' in bot.config['main']: # XXX: This should be loaded as a setting instead. # There needs to be a setting for settings to have them as "passwords" # so they're not displayed openly self.safeBrowsingAPI = SafeBrowsingAPI( bot.config['main']['safebrowsingapi'], bot.nickname, bot.version) else: self.safeBrowsingAPI = None if self.db_session is not None: self.db_session.commit() self.db_session.close() self.db_session = None self.db_session = DBManager.create_session() self.blacklisted_links = [] for link in self.db_session.query(BlacklistedLink): self.blacklisted_links.append(link) self.whitelisted_links = [] for link in self.db_session.query(WhitelistedLink): self.whitelisted_links.append(link) def disable(self, bot): pajbot.managers.handler.HandlerManager.remove_handler( 'on_message', self.on_message) pajbot.managers.handler.HandlerManager.remove_handler( 'on_commit', self.on_commit) if self.db_session is not None: self.db_session.commit() self.db_session.close() self.db_session = None self.blacklisted_links = [] self.whitelisted_links = [] def reload(self): log.info('Loaded {0} bad links and {1} good links'.format( len(self.blacklisted_links), len(self.whitelisted_links))) return self super_whitelist = ['pajlada.se', 'pajlada.com', 'forsen.tv', 'pajbot.com'] def on_message(self, source, message, emotes, whisper, urls, event): if not whisper and source.level < 500 and source.moderator is False: if self.settings[ 'ban_pleb_links'] is True and source.subscriber is False and len( urls) > 0: # Check if the links are in our super-whitelist. i.e. on the pajlada.se domain o forsen.tv for url in urls: parsed_url = Url(url) if len(parsed_url.parsed.netloc.split('.')) < 2: continue whitelisted = False for whitelist in self.super_whitelist: if is_subdomain(parsed_url.parsed.netloc, whitelist): whitelisted = True break if whitelisted is False: self.bot.timeout(source.username, 30, reason='Non-subs cannot post links') if source.minutes_in_chat_online > 60: self.bot.whisper( source.username, 'You cannot post non-verified links in chat if you\'re not a subscriber.' ) return False for url in urls: # Action which will be taken when a bad link is found action = Action(self.bot.timeout, args=[source.username, 20], kwargs={'reason': 'Banned link'}) # First we perform a basic check if self.simple_check(url, action) == self.RET_FURTHER_ANALYSIS: # If the basic check returns no relevant data, we queue up a proper check on the URL self.action_queue.add(self.check_url, args=[url, action]) def on_commit(self): if self.db_session is not None: self.db_session.commit() def delete_from_cache(self, url): if url in self.cache: log.debug('LinkChecker: Removing url {0} from cache'.format(url)) del self.cache[url] def cache_url(self, url, safe): if url in self.cache and self.cache[url] == safe: return log.debug('LinkChecker: Caching url {0} as {1}'.format( url, 'SAFE' if safe is True else 'UNSAFE')) self.cache[url] = safe self.run_later(20, self.delete_from_cache, (url, )) def counteract_bad_url(self, url, action=None, want_to_cache=True, want_to_blacklist=False): log.debug('LinkChecker: BAD URL FOUND {0}'.format(url.url)) if action: action.run() if want_to_cache: self.cache_url(url.url, False) if want_to_blacklist: self.blacklist_url(url.url, url.parsed) return True def blacklist_url(self, url, parsed_url=None, level=0): if not (url.lower().startswith('http://') or url.lower().startswith('https://')): url = 'http://' + url if parsed_url is None: parsed_url = urllib.parse.urlparse(url) if self.is_blacklisted(url, parsed_url): return False domain = parsed_url.netloc.lower() path = parsed_url.path.lower() if domain.startswith('www.'): domain = domain[4:] if path.endswith('/'): path = path[:-1] if path == '': path = '/' link = BlacklistedLink(domain, path, level) self.db_session.add(link) self.blacklisted_links.append(link) self.db_session.commit() def whitelist_url(self, url, parsed_url=None): if not (url.lower().startswith('http://') or url.lower().startswith('https://')): url = 'http://' + url if parsed_url is None: parsed_url = urllib.parse.urlparse(url) if self.is_whitelisted(url, parsed_url): return domain = parsed_url.netloc.lower() path = parsed_url.path.lower() if domain.startswith('www.'): domain = domain[4:] if path.endswith('/'): path = path[:-1] if path == '': path = '/' link = WhitelistedLink(domain, path) self.db_session.add(link) self.whitelisted_links.append(link) self.db_session.commit() def is_blacklisted(self, url, parsed_url=None, sublink=False): if parsed_url is None: parsed_url = urllib.parse.urlparse(url) domain = parsed_url.netloc.lower() path = parsed_url.path.lower() if path == '': path = '/' domain_split = domain.split('.') if len(domain_split) < 2: return False for link in self.blacklisted_links: if link.is_subdomain(domain): if link.is_subpath(path): if not sublink: return True elif link.level >= 1: # if it's a sublink, but the blacklisting level is 0, we don't consider it blacklisted return True return False def is_whitelisted(self, url, parsed_url=None): if parsed_url is None: parsed_url = urllib.parse.urlparse(url) domain = parsed_url.netloc.lower() path = parsed_url.path.lower() if path == '': path = '/' domain_split = domain.split('.') if len(domain_split) < 2: return False for link in self.whitelisted_links: if link.is_subdomain(domain): if link.is_subpath(path): return True return False RET_BAD_LINK = -1 RET_FURTHER_ANALYSIS = 0 RET_GOOD_LINK = 1 def basic_check(self, url, action, sublink=False): """ Check if the url is in the cache, or if it's Return values: 1 = Link is OK -1 = Link is bad 0 = Link needs further analysis """ if url.url in self.cache: log.debug('LinkChecker: Url {0} found in cache'.format(url.url)) if not self.cache[url.url]: # link is bad self.counteract_bad_url(url, action, False, False) return self.RET_BAD_LINK return self.RET_GOOD_LINK log.info('Checking if link is blacklisted...') if self.is_blacklisted(url.url, url.parsed, sublink): log.debug('LinkChecker: Url {0} is blacklisted'.format(url.url)) self.counteract_bad_url(url, action, want_to_blacklist=False) return self.RET_BAD_LINK log.info('Checking if link is whitelisted...') if self.is_whitelisted(url.url, url.parsed): log.debug('LinkChecker: Url {0} allowed by the whitelist'.format( url.url)) self.cache_url(url.url, True) return self.RET_GOOD_LINK return self.RET_FURTHER_ANALYSIS def simple_check(self, url, action): url = Url(url) if len(url.parsed.netloc.split('.')) < 2: # The URL is broken, ignore it return self.RET_FURTHER_ANALYSIS return self.basic_check(url, action) def check_url(self, url, action): url = Url(url) if len(url.parsed.netloc.split('.')) < 2: # The URL is broken, ignore it return try: self._check_url(url, action) except: log.exception('LinkChecker unhanled exception while _check_url') def _check_url(self, url, action): log.debug('LinkChecker: Checking url {0}'.format(url.url)) # XXX: The basic check is currently performed twice on links found in messages. Solve res = self.basic_check(url, action) if res == self.RET_GOOD_LINK: return elif res == self.RET_BAD_LINK: return connection_timeout = 2 read_timeout = 1 try: r = requests.head(url.url, allow_redirects=True, timeout=connection_timeout) except: self.cache_url(url.url, True) return checkcontenttype = ('content-type' in r.headers and r.headers['content-type'] == 'application/octet-stream') checkdispotype = ('disposition-type' in r.headers and r.headers['disposition-type'] == 'attachment') if checkcontenttype or checkdispotype: # triggering a download not allowed self.counteract_bad_url(url, action) return redirected_url = Url(r.url) if is_same_url(url, redirected_url) is False: res = self.basic_check(redirected_url, action) if res == self.RET_GOOD_LINK: return elif res == self.RET_BAD_LINK: return if self.safeBrowsingAPI: if self.safeBrowsingAPI.check_url( redirected_url.url): # harmful url detected log.debug('Bad url because google api') self.counteract_bad_url(url, action, want_to_blacklist=False) self.counteract_bad_url(redirected_url, want_to_blacklist=False) return if 'content-type' not in r.headers or not r.headers[ 'content-type'].startswith('text/html'): return # can't analyze non-html content maximum_size = 1024 * 1024 * 10 # 10 MB receive_timeout = 3 html = '' try: response = requests.get(url=url.url, stream=True, timeout=(connection_timeout, read_timeout)) content_length = response.headers.get('Content-Length') if content_length and int( response.headers.get('Content-Length')) > maximum_size: log.error('This file is too big!') return size = 0 start = time.time() for chunk in response.iter_content(1024): if time.time() - start > receive_timeout: log.error('The site took too long to load') return size += len(chunk) if size > maximum_size: log.error('This file is too big! (fake header)') return html += str(chunk) except requests.exceptions.ConnectTimeout: log.warning('Connection timed out while checking {0}'.format( url.url)) self.cache_url(url.url, True) return except requests.exceptions.ReadTimeout: log.warning('Reading timed out while checking {0}'.format(url.url)) self.cache_url(url.url, True) return except: log.exception('Unhandled exception') return try: soup = BeautifulSoup(html, 'html.parser') except: return original_url = url original_redirected_url = redirected_url urls = [] for link in soup.find_all( 'a'): # get a list of links to external sites url = link.get('href') if url is None: continue if url.startswith('//'): urls.append('http:' + url) elif url.startswith('http://') or url.startswith('https://'): urls.append(url) for url in urls: # check if the site links to anything dangerous url = Url(url) if is_subdomain(url.parsed.netloc, original_url.parsed.netloc): # log.debug('Skipping because internal link') continue log.debug('Checking sublink {0}'.format(url.url)) res = self.basic_check(url, action, sublink=True) if res == self.RET_BAD_LINK: self.counteract_bad_url(url) self.counteract_bad_url(original_url, want_to_blacklist=False) self.counteract_bad_url(original_redirected_url, want_to_blacklist=False) return elif res == self.RET_GOOD_LINK: continue try: r = requests.head(url.url, allow_redirects=True, timeout=connection_timeout) except: continue redirected_url = Url(r.url) if not is_same_url(url, redirected_url): res = self.basic_check(redirected_url, action, sublink=True) if res == self.RET_BAD_LINK: self.counteract_bad_url(url) self.counteract_bad_url(original_url, want_to_blacklist=False) self.counteract_bad_url(original_redirected_url, want_to_blacklist=False) return elif res == self.RET_GOOD_LINK: continue if self.safeBrowsingAPI: if self.safeBrowsingAPI.check_url( redirected_url.url): # harmful url detected log.debug('Evil sublink {0} by google API'.format(url)) self.counteract_bad_url(original_url, action) self.counteract_bad_url(original_redirected_url) self.counteract_bad_url(url) self.counteract_bad_url(redirected_url) return # if we got here, the site is clean for our standards self.cache_url(original_url.url, True) self.cache_url(original_redirected_url.url, True) return def load_commands(self, **options): self.commands['add'] = pajbot.models.command.Command.multiaction_command( level=100, delay_all=0, delay_user=0, default=None, command='add', commands={ 'link': pajbot.models.command.Command.multiaction_command( level=500, delay_all=0, delay_user=0, default=None, commands={ 'blacklist': pajbot.models. command.Command.raw_command( self .add_link_blacklist, level=500, delay_all=0, delay_user=0, description= 'Blacklist a link', examples=[ pajbot. models.command.CommandExample( None, 'Add a link to the blacklist for a shallow search', chat= 'user:!add link blacklist --shallow scamlink.lonk/\n' 'bot>user:Successfully added your links', description= 'Added the link scamlink.lonk/ to the blacklist for a shallow search' ).parse(), pajbot.models.command.CommandExample( None, 'Add a link to the blacklist for a deep search', chat= 'user:!add link blacklist --deep scamlink.lonk/\n' 'bot>user:Successfully added your links', description= 'Added the link scamlink.lonk/ to the blacklist for a deep search' ).parse(), ]), 'whitelist': pajbot.models.command.Command.raw_command( self.add_link_whitelist, level=500, delay_all=0, delay_user=0, description='Whitelist a link', examples=[ pajbot.models.command.CommandExample( None, 'Add a link to the whitelist', chat= 'user:!add link whitelink safelink.lonk/\n' 'bot>user:Successfully added your links', description= 'Added the link safelink.lonk/ to the whitelist' ).parse(), ]), }) }) self.commands['remove'] = pajbot.models.command.Command.multiaction_command( level=100, delay_all=0, delay_user=0, default=None, command='remove', commands={ 'link': pajbot.models.command.Command.multiaction_command( level=500, delay_all=0, delay_user=0, default=None, commands={ 'blacklist': pajbot.models. command.Command.raw_command( self .remove_link_blacklist, level=500, delay_all=0, delay_user=0, description= 'Remove a link from the blacklist.', examples=[ pajbot. models.command.CommandExample( None, 'Remove a link from the blacklist.', chat ='user:!remove link blacklist 20\n' 'bot>user:Successfully removed blacklisted link with id 20', description= 'Remove a link from the blacklist with an ID' ).parse(), ]), 'whitelist': pajbot.models.command.Command.raw_command( self .remove_link_whitelist, level=500, delay_all=0, delay_user=0, description='Remove a link from the whitelist.', examples=[ pajbot.models.command.CommandExample( None, 'Remove a link from the whitelist.', chat='user:!remove link whitelist 12\n' 'bot>user:Successfully removed blacklisted link with id 12', description= 'Remove a link from the whitelist with an ID' ).parse(), ]), }), }) def add_link_blacklist(self, **options): bot = options['bot'] message = options['message'] source = options['source'] options, new_links = self.parse_link_blacklist_arguments(message) if new_links: parts = new_links.split(' ') try: for link in parts: if len(link) > 1: self.blacklist_url(link, **options) AdminLogManager.post('Blacklist link added', source, link) bot.whisper(source.username, 'Successfully added your links') return True except: log.exception('Unhandled exception in add_link_blacklist') bot.whisper(source.username, 'Some error occurred while adding your links') return False else: bot.whisper(source.username, 'Usage: !add link blacklist LINK') return False def add_link_whitelist(self, **options): bot = options['bot'] message = options['message'] source = options['source'] parts = message.split(' ') try: for link in parts: self.whitelist_url(link) AdminLogManager.post('Whitelist link added', source, link) except: log.exception('Unhandled exception in add_link') bot.whisper(source.username, 'Some error occurred white adding your links') return False bot.whisper(source.username, 'Successfully added your links') def remove_link_blacklist(self, **options): message = options['message'] bot = options['bot'] source = options['source'] if message: id = None try: id = int(message) except ValueError: pass link = self.db_session.query(BlacklistedLink).filter_by( id=id).one_or_none() if link: self.blacklisted_links.remove(link) self.db_session.delete(link) self.db_session.commit() else: bot.whisper(source.username, 'No link with the given id found') return False AdminLogManager.post('Blacklist link removed', source, link.domain) bot.whisper( source.username, 'Successfully removed blacklisted link with id {0}'.format( link.id)) else: bot.whisper(source.username, 'Usage: !remove link blacklist ID') return False def remove_link_whitelist(self, **options): message = options['message'] bot = options['bot'] source = options['source'] if message: id = None try: id = int(message) except ValueError: pass link = self.db_session.query(WhitelistedLink).filter_by( id=id).one_or_none() if link: self.whitelisted_links.remove(link) self.db_session.delete(link) self.db_session.commit() else: bot.whisper(source.username, 'No link with the given id found') return False AdminLogManager.post('Whitelist link removed', source, link.domain) bot.whisper( source.username, 'Successfully removed whitelisted link with id {0}'.format( link.id)) else: bot.whisper(source.username, 'Usage: !remove link whitelist ID') return False def parse_link_blacklist_arguments(self, message): parser = argparse.ArgumentParser() parser.add_argument('--deep', dest='level', action='store_true') parser.add_argument('--shallow', dest='level', action='store_false') parser.set_defaults(level=False) try: args, unknown = parser.parse_known_args(message.split()) except SystemExit: return False, False except: log.exception('Unhandled exception in add_link_blacklist') return False, False # Strip options of any values that are set as None options = {k: v for k, v in vars(args).items() if v is not None} response = ' '.join(unknown) return options, response
class LinkCheckerModule(BaseModule): ID = __name__.split('.')[-1] NAME = 'Link Checker' DESCRIPTION = 'Checks links if they\'re bad' ENABLED_DEFAULT = True SETTINGS = [] def __init__(self): super().__init__() self.db_session = None self.links = {} self.blacklisted_links = [] self.whitelisted_links = [] self.cache = LinkCheckerCache() # cache[url] = True means url is safe, False means the link is bad self.action_queue = ActionQueue() self.action_queue.start() def enable(self, bot): self.bot = bot if bot: bot.add_handler('on_message', self.on_message, priority=100) bot.add_handler('on_commit', self.on_commit) self.run_later = bot.execute_delayed if 'safebrowsingapi' in bot.config['main']: # XXX: This should be loaded as a setting instead. # There needs to be a setting for settings to have them as "passwords" # so they're not displayed openly self.safeBrowsingAPI = SafeBrowsingAPI(bot.config['main']['safebrowsingapi'], bot.nickname, bot.version) else: self.safeBrowsingAPI = None if self.db_session is not None: self.db_session.commit() self.db_session.close() self.db_session = None self.db_session = DBManager.create_session() self.blacklisted_links = [] for link in self.db_session.query(BlacklistedLink): self.blacklisted_links.append(link) self.whitelisted_links = [] for link in self.db_session.query(WhitelistedLink): self.whitelisted_links.append(link) def disable(self, bot): if bot: bot.remove_handler('on_message', self.on_message) bot.remove_handler('on_commit', self.on_commit) if self.db_session is not None: self.db_session.commit() self.db_session.close() self.db_session = None self.blacklisted_links = [] self.whitelisted_links = [] def reload(self): log.info('Loaded {0} bad links and {1} good links'.format(len(self.blacklisted_links), len(self.whitelisted_links))) return self def on_message(self, source, message, emotes, whisper, urls): if not whisper and source.level < 500 and source.moderator is False: for url in urls: # Action which will be taken when a bad link is found action = Action(self.bot.timeout, args=[source.username, 20]) # First we perform a basic check if self.simple_check(url, action) == self.RET_FURTHER_ANALYSIS: # If the basic check returns no relevant data, we queue up a proper check on the URL self.action_queue.add(self.check_url, args=[url, action]) def on_commit(self): if self.db_session is not None: self.db_session.commit() def delete_from_cache(self, url): if url in self.cache: log.debug("LinkChecker: Removing url {0} from cache".format(url)) del self.cache[url] def cache_url(self, url, safe): if url in self.cache and self.cache[url] == safe: return log.debug("LinkChecker: Caching url {0} as {1}".format(url, 'SAFE' if safe is True else 'UNSAFE')) self.cache[url] = safe self.run_later(20, self.delete_from_cache, (url, )) def counteract_bad_url(self, url, action=None, want_to_cache=True, want_to_blacklist=True): log.debug("LinkChecker: BAD URL FOUND {0}".format(url.url)) if action: action.run() if want_to_cache: self.cache_url(url.url, False) if want_to_blacklist: self.blacklist_url(url.url, url.parsed) def unlist_url(self, url, list_type, parsed_url=None): """ list_type is either 'blacklist' or 'whitelist' """ if not (url.startswith('http://') or url.startswith('https://')): url = 'http://' + url if parsed_url is None: parsed_url = urllib.parse.urlparse(url) domain = parsed_url.netloc path = parsed_url.path if domain.startswith('www.'): domain = domain[4:] if path.endswith('/'): path = path[:-1] if path == '': path = '/' if list_type == 'blacklist': link = self.db_session.query(BlacklistedLink).filter_by(domain=domain, path=path).one_or_none() if link: self.blacklisted_links.remove(link) self.db_session.delete(link) else: log.warning('Unable to unlist {0}{1}'.format(domain, path)) elif list_type == 'whitelist': link = self.db_session.query(WhitelistedLink).filter_by(domain=domain, path=path).one_or_none() if link: self.whitelisted_links.remove(link) self.db_session.delete(link) else: log.warning('Unable to unlist {0}{1}'.format(domain, path)) def blacklist_url(self, url, parsed_url=None, level=1): if not (url.lower().startswith('http://') or url.lower().startswith('https://')): url = 'http://' + url if parsed_url is None: parsed_url = urllib.parse.urlparse(url) if self.is_blacklisted(url, parsed_url): return False domain = parsed_url.netloc.lower() path = parsed_url.path.lower() if domain.startswith('www.'): domain = domain[4:] if path.endswith('/'): path = path[:-1] if path == '': path = '/' link = BlacklistedLink(domain, path, level) self.db_session.add(link) self.blacklisted_links.append(link) return True def whitelist_url(self, url, parsed_url=None): if not (url.lower().startswith('http://') or url.lower().startswith('https://')): url = 'http://' + url if parsed_url is None: parsed_url = urllib.parse.urlparse(url) if self.is_whitelisted(url, parsed_url): return domain = parsed_url.netloc.lower() path = parsed_url.path.lower() if domain.startswith('www.'): domain = domain[4:] if path.endswith('/'): path = path[:-1] if path == '': path = '/' link = WhitelistedLink(domain, path) self.db_session.add(link) self.whitelisted_links.append(link) def is_blacklisted(self, url, parsed_url=None, sublink=False): if parsed_url is None: parsed_url = urllib.parse.urlparse(url) domain = parsed_url.netloc.lower() path = parsed_url.path.lower() if path == '': path = '/' domain_split = domain.split('.') if len(domain_split) < 2: return False for link in self.blacklisted_links: if link.is_subdomain(domain): if link.is_subpath(path): if not sublink: return True elif link.level >= 1: # if it's a sublink, but the blacklisting level is 0, we don't consider it blacklisted return True return False def is_whitelisted(self, url, parsed_url=None): if parsed_url is None: parsed_url = urllib.parse.urlparse(url) domain = parsed_url.netloc.lower() path = parsed_url.path.lower() if path == '': path = '/' domain_split = domain.split('.') if len(domain_split) < 2: return False for link in self.whitelisted_links: if link.is_subdomain(domain): if link.is_subpath(path): return True return False RET_BAD_LINK = -1 RET_FURTHER_ANALYSIS = 0 RET_GOOD_LINK = 1 def basic_check(self, url, action, sublink=False): """ Check if the url is in the cache, or if it's Return values: 1 = Link is OK -1 = Link is bad 0 = Link needs further analysis """ if url.url in self.cache: log.debug("LinkChecker: Url {0} found in cache".format(url.url)) if not self.cache[url.url]: # link is bad self.counteract_bad_url(url, action, False, False) return self.RET_BAD_LINK return self.RET_GOOD_LINK log.info('Checking if link is blacklisted...') if self.is_blacklisted(url.url, url.parsed, sublink): log.debug("LinkChecker: Url {0} is blacklisted".format(url.url)) self.counteract_bad_url(url, action, want_to_blacklist=False) return self.RET_BAD_LINK log.info('Checking if link is whitelisted...') if self.is_whitelisted(url.url, url.parsed): log.debug("LinkChecker: Url {0} allowed by the whitelist".format(url.url)) self.cache_url(url.url, True) return self.RET_GOOD_LINK return self.RET_FURTHER_ANALYSIS def simple_check(self, url, action): url = Url(url) if len(url.parsed.netloc.split('.')) < 2: # The URL is broken, ignore it return self.RET_FURTHER_ANALYSIS return self.basic_check(url, action) def check_url(self, url, action): url = Url(url) if len(url.parsed.netloc.split('.')) < 2: # The URL is broken, ignore it return try: self._check_url(url, action) except: log.exception("LinkChecker unhanled exception while _check_url") def _check_url(self, url, action): log.debug("LinkChecker: Checking url {0}".format(url.url)) # XXX: The basic check is currently performed twice on links found in messages. Solve res = self.basic_check(url, action) if res == self.RET_GOOD_LINK: return elif res == self.RET_BAD_LINK: return connection_timeout = 2 read_timeout = 1 try: r = requests.head(url.url, allow_redirects=True, timeout=connection_timeout) except: self.cache_url(url.url, True) return checkcontenttype = ('content-type' in r.headers and r.headers['content-type'] == 'application/octet-stream') checkdispotype = ('disposition-type' in r.headers and r.headers['disposition-type'] == 'attachment') if checkcontenttype or checkdispotype: # triggering a download not allowed self.counteract_bad_url(url, action) return redirected_url = Url(r.url) if is_same_url(url, redirected_url) is False: res = self.basic_check(redirected_url, action) if res == self.RET_GOOD_LINK: return elif res == self.RET_BAD_LINK: return if self.safeBrowsingAPI: if self.safeBrowsingAPI.check_url(redirected_url.url): # harmful url detected log.debug("Bad url because google api") self.counteract_bad_url(url, action) self.counteract_bad_url(redirected_url) return if 'content-type' not in r.headers or not r.headers['content-type'].startswith('text/html'): return # can't analyze non-html content maximum_size = 1024 * 1024 * 10 # 10 MB receive_timeout = 3 html = '' try: response = requests.get(url=url.url, stream=True, timeout=(connection_timeout, read_timeout)) content_length = response.headers.get('Content-Length') if content_length and int(response.headers.get('Content-Length')) > maximum_size: log.error('This file is too big!') return size = 0 start = time.time() for chunk in response.iter_content(1024): if time.time() - start > receive_timeout: log.error('The site took too long to load') return size += len(chunk) if size > maximum_size: log.error('This file is too big! (fake header)') return html += str(chunk) except requests.exceptions.ConnectTimeout: log.warning('Connection timed out while checking {0}'.format(url.url)) self.cache_url(url.url, True) return except requests.exceptions.ReadTimeout: log.warning('Reading timed out while checking {0}'.format(url.url)) self.cache_url(url.url, True) return except: log.exception('Unhandled exception') return try: soup = BeautifulSoup(html, 'html.parser') except: return original_url = url original_redirected_url = redirected_url urls = [] for link in soup.find_all('a'): # get a list of links to external sites url = link.get('href') if url is None: continue if url.startswith('//'): urls.append('http:' + url) elif url.startswith('http://') or url.startswith('https://'): urls.append(url) for url in urls: # check if the site links to anything dangerous url = Url(url) if is_subdomain(url.parsed.netloc, original_url.parsed.netloc): # log.debug("Skipping because internal link") continue log.debug("Checking sublink {0}".format(url.url)) res = self.basic_check(url, action, sublink=True) if res == self.RET_BAD_LINK: self.counteract_bad_url(url) self.counteract_bad_url(original_url, want_to_blacklist=False) self.counteract_bad_url(original_redirected_url, want_to_blacklist=False) return elif res == self.RET_GOOD_LINK: continue try: r = requests.head(url.url, allow_redirects=True, timeout=connection_timeout) except: continue redirected_url = Url(r.url) if not is_same_url(url, redirected_url): res = self.basic_check(redirected_url, action, sublink=True) if res == self.RET_BAD_LINK: self.counteract_bad_url(url) self.counteract_bad_url(original_url, want_to_blacklist=False) self.counteract_bad_url(original_redirected_url, want_to_blacklist=False) return elif res == self.RET_GOOD_LINK: continue if self.safeBrowsingAPI: if self.safeBrowsingAPI.check_url(redirected_url.url): # harmful url detected log.debug("Evil sublink {0} by google API".format(url)) self.counteract_bad_url(original_url, action) self.counteract_bad_url(original_redirected_url) self.counteract_bad_url(url) self.counteract_bad_url(redirected_url) return # if we got here, the site is clean for our standards self.cache_url(original_url.url, True) self.cache_url(original_redirected_url.url, True) return def load_commands(self, **options): self.commands['add'] = Command.multiaction_command( level=100, delay_all=0, delay_user=0, default=None, command='add', commands={ 'link': Command.multiaction_command( level=500, delay_all=0, delay_user=0, default=None, commands={ 'blacklist': Command.raw_command(self.add_link_blacklist, level=500, description='Blacklist a link', examples=[ CommandExample(None, 'Add a link to the blacklist for shallow search', chat='user:!add link blacklist 0 scamlink.lonk/\n' 'bot>user:Successfully added your links', description='Added the link scamlink.lonk/ to the blacklist for a shallow search').parse(), CommandExample(None, 'Add a link to the blacklist for deep search', chat='user:!add link blacklist 1 scamlink.lonk/\n' 'bot>user:Successfully added your links', description='Added the link scamlink.lonk/ to the blacklist for a deep search').parse(), ]), 'whitelist': Command.raw_command(self.add_link_whitelist, level=500, description='Whitelist a link', examples=[ CommandExample(None, 'Add a link to the whitelist', chat='user:!add link whitelink safelink.lonk/\n' 'bot>user:Successfully added your links', description='Added the link safelink.lonk/ to the whitelist').parse(), ]), } ) } ) self.commands['remove'] = Command.multiaction_command( level=100, delay_all=0, delay_user=0, default=None, command='remove', commands={ 'link': Command.multiaction_command( level=500, delay_all=0, delay_user=0, default=None, commands={ 'blacklist': Command.raw_command(self.remove_link_blacklist, level=500, description='Unblacklist a link', examples=[ CommandExample(None, 'Remove a blacklist link', chat='user:!remove link blacklist scamtwitch.scam\n' 'bot>user:Successfully removed your links', description='Removes scamtwitch.scam as a blacklisted link').parse(), ]), 'whitelist': Command.raw_command(self.remove_link_whitelist, level=500, description='Unwhitelist a link', examples=[ CommandExample(None, 'Remove a whitelist link', chat='user:!remove link whitelist twitch.safe\n' 'bot>user:Successfully removed your links', description='Removes twitch.safe as a whitelisted link').parse(), ]), } ), } ) def add_link_blacklist(self, **options): bot = options['bot'] message = options['message'] source = options['source'] parts = message.split(' ') try: if not parts[0].isnumeric(): for link in parts: self.blacklist_url(link) else: for link in parts[1:]: self.blacklist_url(link, level=int(parts[0])) except: log.exception('Unhandled exception in add_link') bot.whisper(source.username, 'Some error occurred white adding your links') return False bot.whisper(source.username, 'Successfully added your links') def add_link_whitelist(self, **options): bot = options['bot'] message = options['message'] source = options['source'] parts = message.split(' ') try: for link in parts: self.whitelist_url(link) except: log.exception('Unhandled exception in add_link') bot.whisper(source.username, 'Some error occurred white adding your links') return False bot.whisper(source.username, 'Successfully added your links') def remove_link_blacklist(self, **options): bot = options['bot'] message = options['message'] source = options['source'] parts = message.split(' ') try: for link in parts: self.unlist_url(link, 'blacklist') except: log.exception('Unhandled exception in add_link') bot.whisper(source.username, 'Some error occurred white adding your links') return False bot.whisper(source.username, 'Successfully removed your links') def remove_link_whitelist(self, **options): bot = options['bot'] message = options['message'] source = options['source'] parts = message.split(' ') try: for link in parts: self.unlist_url(link, 'whitelist') except: log.exception('Unhandled exception in add_link') bot.whisper(source.username, 'Some error occurred white adding your links') return False bot.whisper(source.username, 'Successfully removed your links')
class LinkCheckerModule(BaseModule): ID = __name__.split('.')[-1] NAME = 'Link Checker' DESCRIPTION = 'Checks links if they\'re bad' ENABLED_DEFAULT = True CATEGORY = 'Filter' SETTINGS = [ ModuleSetting( key='ban_pleb_links', label='Disallow links from non-subscribers', type='boolean', required=True, default=False) ] def __init__(self): super().__init__() self.db_session = None self.links = {} self.blacklisted_links = [] self.whitelisted_links = [] self.cache = LinkCheckerCache() # cache[url] = True means url is safe, False means the link is bad self.action_queue = ActionQueue() self.action_queue.start() def enable(self, bot): self.bot = bot pajbot.managers.handler.HandlerManager.add_handler('on_message', self.on_message, priority=100) pajbot.managers.handler.HandlerManager.add_handler('on_commit', self.on_commit) if bot: self.run_later = bot.execute_delayed if 'safebrowsingapi' in bot.config['main']: # XXX: This should be loaded as a setting instead. # There needs to be a setting for settings to have them as "passwords" # so they're not displayed openly self.safeBrowsingAPI = SafeBrowsingAPI(bot.config['main']['safebrowsingapi'], bot.nickname, bot.version) else: self.safeBrowsingAPI = None if self.db_session is not None: self.db_session.commit() self.db_session.close() self.db_session = None self.db_session = DBManager.create_session() self.blacklisted_links = [] for link in self.db_session.query(BlacklistedLink): self.blacklisted_links.append(link) self.whitelisted_links = [] for link in self.db_session.query(WhitelistedLink): self.whitelisted_links.append(link) def disable(self, bot): pajbot.managers.handler.HandlerManager.remove_handler('on_message', self.on_message) pajbot.managers.handler.HandlerManager.remove_handler('on_commit', self.on_commit) if self.db_session is not None: self.db_session.commit() self.db_session.close() self.db_session = None self.blacklisted_links = [] self.whitelisted_links = [] def reload(self): log.info('Loaded {0} bad links and {1} good links'.format(len(self.blacklisted_links), len(self.whitelisted_links))) return self super_whitelist = ['pajlada.se', 'pajlada.com', 'forsen.tv', 'pajbot.com'] def on_message(self, source, message, emotes, whisper, urls, event): if not whisper and source.level < 500 and source.moderator is False: if self.settings['ban_pleb_links'] is True and source.subscriber is False and len(urls) > 0: # Check if the links are in our super-whitelist. i.e. on the pajlada.se domain o forsen.tv for url in urls: parsed_url = Url(url) if len(parsed_url.parsed.netloc.split('.')) < 2: continue whitelisted = False for whitelist in self.super_whitelist: if is_subdomain(parsed_url.parsed.netloc, whitelist): whitelisted = True break if whitelisted is False: self.bot.timeout(source.username, 30, reason='Non-subs cannot post links') if source.minutes_in_chat_online > 60: self.bot.whisper(source.username, 'You cannot post non-verified links in chat if you\'re not a subscriber.') return False for url in urls: # Action which will be taken when a bad link is found action = Action(self.bot.timeout, args=[source.username, 20], kwargs={'reason': 'Banned link'}) # First we perform a basic check if self.simple_check(url, action) == self.RET_FURTHER_ANALYSIS: # If the basic check returns no relevant data, we queue up a proper check on the URL self.action_queue.add(self.check_url, args=[url, action]) def on_commit(self): if self.db_session is not None: self.db_session.commit() def delete_from_cache(self, url): if url in self.cache: log.debug('LinkChecker: Removing url {0} from cache'.format(url)) del self.cache[url] def cache_url(self, url, safe): if url in self.cache and self.cache[url] == safe: return log.debug('LinkChecker: Caching url {0} as {1}'.format(url, 'SAFE' if safe is True else 'UNSAFE')) self.cache[url] = safe self.run_later(20, self.delete_from_cache, (url, )) def counteract_bad_url(self, url, action=None, want_to_cache=True, want_to_blacklist=False): log.debug('LinkChecker: BAD URL FOUND {0}'.format(url.url)) if action: action.run() if want_to_cache: self.cache_url(url.url, False) if want_to_blacklist: self.blacklist_url(url.url, url.parsed) return True def blacklist_url(self, url, parsed_url=None, level=0): if not (url.lower().startswith('http://') or url.lower().startswith('https://')): url = 'http://' + url if parsed_url is None: parsed_url = urllib.parse.urlparse(url) if self.is_blacklisted(url, parsed_url): return False domain = parsed_url.netloc.lower() path = parsed_url.path.lower() if domain.startswith('www.'): domain = domain[4:] if path.endswith('/'): path = path[:-1] if path == '': path = '/' link = BlacklistedLink(domain, path, level) self.db_session.add(link) self.blacklisted_links.append(link) self.db_session.commit() def whitelist_url(self, url, parsed_url=None): if not (url.lower().startswith('http://') or url.lower().startswith('https://')): url = 'http://' + url if parsed_url is None: parsed_url = urllib.parse.urlparse(url) if self.is_whitelisted(url, parsed_url): return domain = parsed_url.netloc.lower() path = parsed_url.path.lower() if domain.startswith('www.'): domain = domain[4:] if path.endswith('/'): path = path[:-1] if path == '': path = '/' link = WhitelistedLink(domain, path) self.db_session.add(link) self.whitelisted_links.append(link) self.db_session.commit() def is_blacklisted(self, url, parsed_url=None, sublink=False): if parsed_url is None: parsed_url = urllib.parse.urlparse(url) domain = parsed_url.netloc.lower() path = parsed_url.path.lower() if path == '': path = '/' domain_split = domain.split('.') if len(domain_split) < 2: return False for link in self.blacklisted_links: if link.is_subdomain(domain): if link.is_subpath(path): if not sublink: return True elif link.level >= 1: # if it's a sublink, but the blacklisting level is 0, we don't consider it blacklisted return True return False def is_whitelisted(self, url, parsed_url=None): if parsed_url is None: parsed_url = urllib.parse.urlparse(url) domain = parsed_url.netloc.lower() path = parsed_url.path.lower() if path == '': path = '/' domain_split = domain.split('.') if len(domain_split) < 2: return False for link in self.whitelisted_links: if link.is_subdomain(domain): if link.is_subpath(path): return True return False RET_BAD_LINK = -1 RET_FURTHER_ANALYSIS = 0 RET_GOOD_LINK = 1 def basic_check(self, url, action, sublink=False): """ Check if the url is in the cache, or if it's Return values: 1 = Link is OK -1 = Link is bad 0 = Link needs further analysis """ if url.url in self.cache: log.debug('LinkChecker: Url {0} found in cache'.format(url.url)) if not self.cache[url.url]: # link is bad self.counteract_bad_url(url, action, False, False) return self.RET_BAD_LINK return self.RET_GOOD_LINK log.info('Checking if link is blacklisted...') if self.is_blacklisted(url.url, url.parsed, sublink): log.debug('LinkChecker: Url {0} is blacklisted'.format(url.url)) self.counteract_bad_url(url, action, want_to_blacklist=False) return self.RET_BAD_LINK log.info('Checking if link is whitelisted...') if self.is_whitelisted(url.url, url.parsed): log.debug('LinkChecker: Url {0} allowed by the whitelist'.format(url.url)) self.cache_url(url.url, True) return self.RET_GOOD_LINK return self.RET_FURTHER_ANALYSIS def simple_check(self, url, action): url = Url(url) if len(url.parsed.netloc.split('.')) < 2: # The URL is broken, ignore it return self.RET_FURTHER_ANALYSIS return self.basic_check(url, action) def check_url(self, url, action): url = Url(url) if len(url.parsed.netloc.split('.')) < 2: # The URL is broken, ignore it return try: self._check_url(url, action) except: log.exception('LinkChecker unhanled exception while _check_url') def _check_url(self, url, action): log.debug('LinkChecker: Checking url {0}'.format(url.url)) # XXX: The basic check is currently performed twice on links found in messages. Solve res = self.basic_check(url, action) if res == self.RET_GOOD_LINK: return elif res == self.RET_BAD_LINK: return connection_timeout = 2 read_timeout = 1 try: r = requests.head(url.url, allow_redirects=True, timeout=connection_timeout) except: self.cache_url(url.url, True) return checkcontenttype = ('content-type' in r.headers and r.headers['content-type'] == 'application/octet-stream') checkdispotype = ('disposition-type' in r.headers and r.headers['disposition-type'] == 'attachment') if checkcontenttype or checkdispotype: # triggering a download not allowed self.counteract_bad_url(url, action) return redirected_url = Url(r.url) if is_same_url(url, redirected_url) is False: res = self.basic_check(redirected_url, action) if res == self.RET_GOOD_LINK: return elif res == self.RET_BAD_LINK: return if self.safeBrowsingAPI: if self.safeBrowsingAPI.check_url(redirected_url.url): # harmful url detected log.debug('Bad url because google api') self.counteract_bad_url(url, action, want_to_blacklist=False) self.counteract_bad_url(redirected_url, want_to_blacklist=False) return if 'content-type' not in r.headers or not r.headers['content-type'].startswith('text/html'): return # can't analyze non-html content maximum_size = 1024 * 1024 * 10 # 10 MB receive_timeout = 3 html = '' try: response = requests.get(url=url.url, stream=True, timeout=(connection_timeout, read_timeout)) content_length = response.headers.get('Content-Length') if content_length and int(response.headers.get('Content-Length')) > maximum_size: log.error('This file is too big!') return size = 0 start = time.time() for chunk in response.iter_content(1024): if time.time() - start > receive_timeout: log.error('The site took too long to load') return size += len(chunk) if size > maximum_size: log.error('This file is too big! (fake header)') return html += str(chunk) except requests.exceptions.ConnectTimeout: log.warning('Connection timed out while checking {0}'.format(url.url)) self.cache_url(url.url, True) return except requests.exceptions.ReadTimeout: log.warning('Reading timed out while checking {0}'.format(url.url)) self.cache_url(url.url, True) return except: log.exception('Unhandled exception') return try: soup = BeautifulSoup(html, 'html.parser') except: return original_url = url original_redirected_url = redirected_url urls = [] for link in soup.find_all('a'): # get a list of links to external sites url = link.get('href') if url is None: continue if url.startswith('//'): urls.append('http:' + url) elif url.startswith('http://') or url.startswith('https://'): urls.append(url) for url in urls: # check if the site links to anything dangerous url = Url(url) if is_subdomain(url.parsed.netloc, original_url.parsed.netloc): # log.debug('Skipping because internal link') continue log.debug('Checking sublink {0}'.format(url.url)) res = self.basic_check(url, action, sublink=True) if res == self.RET_BAD_LINK: self.counteract_bad_url(url) self.counteract_bad_url(original_url, want_to_blacklist=False) self.counteract_bad_url(original_redirected_url, want_to_blacklist=False) return elif res == self.RET_GOOD_LINK: continue try: r = requests.head(url.url, allow_redirects=True, timeout=connection_timeout) except: continue redirected_url = Url(r.url) if not is_same_url(url, redirected_url): res = self.basic_check(redirected_url, action, sublink=True) if res == self.RET_BAD_LINK: self.counteract_bad_url(url) self.counteract_bad_url(original_url, want_to_blacklist=False) self.counteract_bad_url(original_redirected_url, want_to_blacklist=False) return elif res == self.RET_GOOD_LINK: continue if self.safeBrowsingAPI: if self.safeBrowsingAPI.check_url(redirected_url.url): # harmful url detected log.debug('Evil sublink {0} by google API'.format(url)) self.counteract_bad_url(original_url, action) self.counteract_bad_url(original_redirected_url) self.counteract_bad_url(url) self.counteract_bad_url(redirected_url) return # if we got here, the site is clean for our standards self.cache_url(original_url.url, True) self.cache_url(original_redirected_url.url, True) return def load_commands(self, **options): self.commands['add'] = pajbot.models.command.Command.multiaction_command( level=100, delay_all=0, delay_user=0, default=None, command='add', commands={ 'link': pajbot.models.command.Command.multiaction_command( level=500, delay_all=0, delay_user=0, default=None, commands={ 'blacklist': pajbot.models.command.Command.raw_command(self.add_link_blacklist, level=500, delay_all=0, delay_user=0, description='Blacklist a link', examples=[ pajbot.models.command.CommandExample(None, 'Add a link to the blacklist for a shallow search', chat='user:!add link blacklist --shallow scamlink.lonk/\n' 'bot>user:Successfully added your links', description='Added the link scamlink.lonk/ to the blacklist for a shallow search').parse(), pajbot.models.command.CommandExample(None, 'Add a link to the blacklist for a deep search', chat='user:!add link blacklist --deep scamlink.lonk/\n' 'bot>user:Successfully added your links', description='Added the link scamlink.lonk/ to the blacklist for a deep search').parse(), ]), 'whitelist': pajbot.models.command.Command.raw_command(self.add_link_whitelist, level=500, delay_all=0, delay_user=0, description='Whitelist a link', examples=[ pajbot.models.command.CommandExample(None, 'Add a link to the whitelist', chat='user:!add link whitelink safelink.lonk/\n' 'bot>user:Successfully added your links', description='Added the link safelink.lonk/ to the whitelist').parse(), ]), } ) } ) self.commands['remove'] = pajbot.models.command.Command.multiaction_command( level=100, delay_all=0, delay_user=0, default=None, command='remove', commands={ 'link': pajbot.models.command.Command.multiaction_command( level=500, delay_all=0, delay_user=0, default=None, commands={ 'blacklist': pajbot.models.command.Command.raw_command(self.remove_link_blacklist, level=500, delay_all=0, delay_user=0, description='Remove a link from the blacklist.', examples=[ pajbot.models.command.CommandExample(None, 'Remove a link from the blacklist.', chat='user:!remove link blacklist 20\n' 'bot>user:Successfully removed blacklisted link with id 20', description='Remove a link from the blacklist with an ID').parse(), ]), 'whitelist': pajbot.models.command.Command.raw_command(self.remove_link_whitelist, level=500, delay_all=0, delay_user=0, description='Remove a link from the whitelist.', examples=[ pajbot.models.command.CommandExample(None, 'Remove a link from the whitelist.', chat='user:!remove link whitelist 12\n' 'bot>user:Successfully removed blacklisted link with id 12', description='Remove a link from the whitelist with an ID').parse(), ]), } ), } ) def add_link_blacklist(self, **options): bot = options['bot'] message = options['message'] source = options['source'] options, new_links = self.parse_link_blacklist_arguments(message) if new_links: parts = new_links.split(' ') try: for link in parts: if len(link) > 1: self.blacklist_url(link, **options) AdminLogManager.post('Blacklist link added', source, link) bot.whisper(source.username, 'Successfully added your links') return True except: log.exception('Unhandled exception in add_link_blacklist') bot.whisper(source.username, 'Some error occurred while adding your links') return False else: bot.whisper(source.username, 'Usage: !add link blacklist LINK') return False def add_link_whitelist(self, **options): bot = options['bot'] message = options['message'] source = options['source'] parts = message.split(' ') try: for link in parts: self.whitelist_url(link) AdminLogManager.post('Whitelist link added', source, link) except: log.exception('Unhandled exception in add_link') bot.whisper(source.username, 'Some error occurred white adding your links') return False bot.whisper(source.username, 'Successfully added your links') def remove_link_blacklist(self, **options): message = options['message'] bot = options['bot'] source = options['source'] if message: id = None try: id = int(message) except ValueError: pass link = self.db_session.query(BlacklistedLink).filter_by(id=id).one_or_none() if link: self.blacklisted_links.remove(link) self.db_session.delete(link) self.db_session.commit() else: bot.whisper(source.username, 'No link with the given id found') return False AdminLogManager.post('Blacklist link removed', source, link.domain) bot.whisper(source.username, 'Successfully removed blacklisted link with id {0}'.format(link.id)) else: bot.whisper(source.username, 'Usage: !remove link blacklist ID') return False def remove_link_whitelist(self, **options): message = options['message'] bot = options['bot'] source = options['source'] if message: id = None try: id = int(message) except ValueError: pass link = self.db_session.query(WhitelistedLink).filter_by(id=id).one_or_none() if link: self.whitelisted_links.remove(link) self.db_session.delete(link) self.db_session.commit() else: bot.whisper(source.username, 'No link with the given id found') return False AdminLogManager.post('Whitelist link removed', source, link.domain) bot.whisper(source.username, 'Successfully removed whitelisted link with id {0}'.format(link.id)) else: bot.whisper(source.username, 'Usage: !remove link whitelist ID') return False def parse_link_blacklist_arguments(self, message): parser = argparse.ArgumentParser() parser.add_argument('--deep', dest='level', action='store_true') parser.add_argument('--shallow', dest='level', action='store_false') parser.set_defaults(level=False) try: args, unknown = parser.parse_known_args(message.split()) except SystemExit: return False, False except: log.exception('Unhandled exception in add_link_blacklist') return False, False # Strip options of any values that are set as None options = {k: v for k, v in vars(args).items() if v is not None} response = ' '.join(unknown) return options, response
class LinkCheckerModule(BaseModule): ID = __name__.split('.')[-1] NAME = 'Link Checker' DESCRIPTION = 'Checks links if they\'re bad' ENABLED_DEFAULT = True SETTINGS = [] def __init__(self): super().__init__() self.db_session = None self.links = {} self.blacklisted_links = [] self.whitelisted_links = [] self.cache = LinkCheckerCache( ) # cache[url] = True means url is safe, False means the link is bad self.action_queue = ActionQueue() self.action_queue.start() def enable(self, bot): self.bot = bot if bot: bot.add_handler('on_message', self.on_message, priority=100) bot.add_handler('on_commit', self.on_commit) self.run_later = bot.execute_delayed if 'safebrowsingapi' in bot.config['main']: # XXX: This should be loaded as a setting instead. # There needs to be a setting for settings to have them as "passwords" # so they're not displayed openly self.safeBrowsingAPI = SafeBrowsingAPI( bot.config['main']['safebrowsingapi'], bot.nickname, bot.version) else: self.safeBrowsingAPI = None if self.db_session is not None: self.db_session.commit() self.db_session.close() self.db_session = None self.db_session = DBManager.create_session() self.blacklisted_links = [] for link in self.db_session.query(BlacklistedLink): self.blacklisted_links.append(link) self.whitelisted_links = [] for link in self.db_session.query(WhitelistedLink): self.whitelisted_links.append(link) def disable(self, bot): if bot: bot.remove_handler('on_message', self.on_message) bot.remove_handler('on_commit', self.on_commit) if self.db_session is not None: self.db_session.commit() self.db_session.close() self.db_session = None self.blacklisted_links = [] self.whitelisted_links = [] def reload(self): log.info('Loaded {0} bad links and {1} good links'.format( len(self.blacklisted_links), len(self.whitelisted_links))) return self def on_message(self, source, message, emotes, whisper, urls): if not whisper and source.level < 500 and source.moderator is False: for url in urls: # Action which will be taken when a bad link is found action = Action(self.bot.timeout, args=[source.username, 20]) # First we perform a basic check if self.simple_check(url, action) == self.RET_FURTHER_ANALYSIS: # If the basic check returns no relevant data, we queue up a proper check on the URL self.action_queue.add(self.check_url, args=[url, action]) def on_commit(self): if self.db_session is not None: self.db_session.commit() def delete_from_cache(self, url): if url in self.cache: log.debug("LinkChecker: Removing url {0} from cache".format(url)) del self.cache[url] def cache_url(self, url, safe): if url in self.cache and self.cache[url] == safe: return log.debug("LinkChecker: Caching url {0} as {1}".format( url, 'SAFE' if safe is True else 'UNSAFE')) self.cache[url] = safe self.run_later(20, self.delete_from_cache, (url, )) def counteract_bad_url(self, url, action=None, want_to_cache=True, want_to_blacklist=True): log.debug("LinkChecker: BAD URL FOUND {0}".format(url.url)) if action: action.run() if want_to_cache: self.cache_url(url.url, False) if want_to_blacklist: self.blacklist_url(url.url, url.parsed) def unlist_url(self, url, list_type, parsed_url=None): """ list_type is either 'blacklist' or 'whitelist' """ if not (url.startswith('http://') or url.startswith('https://')): url = 'http://' + url if parsed_url is None: parsed_url = urllib.parse.urlparse(url) domain = parsed_url.netloc path = parsed_url.path if domain.startswith('www.'): domain = domain[4:] if path.endswith('/'): path = path[:-1] if path == '': path = '/' if list_type == 'blacklist': link = self.db_session.query(BlacklistedLink).filter_by( domain=domain, path=path).one_or_none() if link: self.blacklisted_links.remove(link) self.db_session.delete(link) else: log.warning('Unable to unlist {0}{1}'.format(domain, path)) elif list_type == 'whitelist': link = self.db_session.query(WhitelistedLink).filter_by( domain=domain, path=path).one_or_none() if link: self.whitelisted_links.remove(link) self.db_session.delete(link) else: log.warning('Unable to unlist {0}{1}'.format(domain, path)) def blacklist_url(self, url, parsed_url=None, level=1): if not (url.lower().startswith('http://') or url.lower().startswith('https://')): url = 'http://' + url if parsed_url is None: parsed_url = urllib.parse.urlparse(url) if self.is_blacklisted(url, parsed_url): return False domain = parsed_url.netloc.lower() path = parsed_url.path.lower() if domain.startswith('www.'): domain = domain[4:] if path.endswith('/'): path = path[:-1] if path == '': path = '/' link = BlacklistedLink(domain, path, level) self.db_session.add(link) self.blacklisted_links.append(link) return True def whitelist_url(self, url, parsed_url=None): if not (url.lower().startswith('http://') or url.lower().startswith('https://')): url = 'http://' + url if parsed_url is None: parsed_url = urllib.parse.urlparse(url) if self.is_whitelisted(url, parsed_url): return domain = parsed_url.netloc.lower() path = parsed_url.path.lower() if domain.startswith('www.'): domain = domain[4:] if path.endswith('/'): path = path[:-1] if path == '': path = '/' link = WhitelistedLink(domain, path) self.db_session.add(link) self.whitelisted_links.append(link) def is_blacklisted(self, url, parsed_url=None, sublink=False): if parsed_url is None: parsed_url = urllib.parse.urlparse(url) domain = parsed_url.netloc.lower() path = parsed_url.path.lower() if path == '': path = '/' domain_split = domain.split('.') if len(domain_split) < 2: return False for link in self.blacklisted_links: if link.is_subdomain(domain): if link.is_subpath(path): if not sublink: return True elif link.level >= 1: # if it's a sublink, but the blacklisting level is 0, we don't consider it blacklisted return True return False def is_whitelisted(self, url, parsed_url=None): if parsed_url is None: parsed_url = urllib.parse.urlparse(url) domain = parsed_url.netloc.lower() path = parsed_url.path.lower() if path == '': path = '/' domain_split = domain.split('.') if len(domain_split) < 2: return False for link in self.whitelisted_links: if link.is_subdomain(domain): if link.is_subpath(path): return True return False RET_BAD_LINK = -1 RET_FURTHER_ANALYSIS = 0 RET_GOOD_LINK = 1 def basic_check(self, url, action, sublink=False): """ Check if the url is in the cache, or if it's Return values: 1 = Link is OK -1 = Link is bad 0 = Link needs further analysis """ if url.url in self.cache: log.debug("LinkChecker: Url {0} found in cache".format(url.url)) if not self.cache[url.url]: # link is bad self.counteract_bad_url(url, action, False, False) return self.RET_BAD_LINK return self.RET_GOOD_LINK log.info('Checking if link is blacklisted...') if self.is_blacklisted(url.url, url.parsed, sublink): log.debug("LinkChecker: Url {0} is blacklisted".format(url.url)) self.counteract_bad_url(url, action, want_to_blacklist=False) return self.RET_BAD_LINK log.info('Checking if link is whitelisted...') if self.is_whitelisted(url.url, url.parsed): log.debug("LinkChecker: Url {0} allowed by the whitelist".format( url.url)) self.cache_url(url.url, True) return self.RET_GOOD_LINK return self.RET_FURTHER_ANALYSIS def simple_check(self, url, action): url = Url(url) if len(url.parsed.netloc.split('.')) < 2: # The URL is broken, ignore it return self.RET_FURTHER_ANALYSIS return self.basic_check(url, action) def check_url(self, url, action): url = Url(url) if len(url.parsed.netloc.split('.')) < 2: # The URL is broken, ignore it return try: self._check_url(url, action) except: log.exception("LinkChecker unhanled exception while _check_url") def _check_url(self, url, action): log.debug("LinkChecker: Checking url {0}".format(url.url)) # XXX: The basic check is currently performed twice on links found in messages. Solve res = self.basic_check(url, action) if res == self.RET_GOOD_LINK: return elif res == self.RET_BAD_LINK: return connection_timeout = 2 read_timeout = 1 try: r = requests.head(url.url, allow_redirects=True, timeout=connection_timeout) except: self.cache_url(url.url, True) return checkcontenttype = ('content-type' in r.headers and r.headers['content-type'] == 'application/octet-stream') checkdispotype = ('disposition-type' in r.headers and r.headers['disposition-type'] == 'attachment') if checkcontenttype or checkdispotype: # triggering a download not allowed self.counteract_bad_url(url, action) return redirected_url = Url(r.url) if is_same_url(url, redirected_url) is False: res = self.basic_check(redirected_url, action) if res == self.RET_GOOD_LINK: return elif res == self.RET_BAD_LINK: return if self.safeBrowsingAPI: if self.safeBrowsingAPI.check_url( redirected_url.url): # harmful url detected log.debug("Bad url because google api") self.counteract_bad_url(url, action) self.counteract_bad_url(redirected_url) return if 'content-type' not in r.headers or not r.headers[ 'content-type'].startswith('text/html'): return # can't analyze non-html content maximum_size = 1024 * 1024 * 10 # 10 MB receive_timeout = 3 html = '' try: response = requests.get(url=url.url, stream=True, timeout=(connection_timeout, read_timeout)) content_length = response.headers.get('Content-Length') if content_length and int( response.headers.get('Content-Length')) > maximum_size: log.error('This file is too big!') return size = 0 start = time.time() for chunk in response.iter_content(1024): if time.time() - start > receive_timeout: log.error('The site took too long to load') return size += len(chunk) if size > maximum_size: log.error('This file is too big! (fake header)') return html += str(chunk) except requests.exceptions.ConnectTimeout: log.warning('Connection timed out while checking {0}'.format( url.url)) self.cache_url(url.url, True) return except requests.exceptions.ReadTimeout: log.warning('Reading timed out while checking {0}'.format(url.url)) self.cache_url(url.url, True) return except: log.exception('Unhandled exception') return try: soup = BeautifulSoup(html, 'html.parser') except: return original_url = url original_redirected_url = redirected_url urls = [] for link in soup.find_all( 'a'): # get a list of links to external sites url = link.get('href') if url is None: continue if url.startswith('//'): urls.append('http:' + url) elif url.startswith('http://') or url.startswith('https://'): urls.append(url) for url in urls: # check if the site links to anything dangerous url = Url(url) if is_subdomain(url.parsed.netloc, original_url.parsed.netloc): # log.debug("Skipping because internal link") continue log.debug("Checking sublink {0}".format(url.url)) res = self.basic_check(url, action, sublink=True) if res == self.RET_BAD_LINK: self.counteract_bad_url(url) self.counteract_bad_url(original_url, want_to_blacklist=False) self.counteract_bad_url(original_redirected_url, want_to_blacklist=False) return elif res == self.RET_GOOD_LINK: continue try: r = requests.head(url.url, allow_redirects=True, timeout=connection_timeout) except: continue redirected_url = Url(r.url) if not is_same_url(url, redirected_url): res = self.basic_check(redirected_url, action, sublink=True) if res == self.RET_BAD_LINK: self.counteract_bad_url(url) self.counteract_bad_url(original_url, want_to_blacklist=False) self.counteract_bad_url(original_redirected_url, want_to_blacklist=False) return elif res == self.RET_GOOD_LINK: continue if self.safeBrowsingAPI: if self.safeBrowsingAPI.check_url( redirected_url.url): # harmful url detected log.debug("Evil sublink {0} by google API".format(url)) self.counteract_bad_url(original_url, action) self.counteract_bad_url(original_redirected_url) self.counteract_bad_url(url) self.counteract_bad_url(redirected_url) return # if we got here, the site is clean for our standards self.cache_url(original_url.url, True) self.cache_url(original_redirected_url.url, True) return def load_commands(self, **options): self.commands['add'] = Command.multiaction_command( level=100, delay_all=0, delay_user=0, default=None, command='add', commands={ 'link': Command.multiaction_command( level=500, delay_all=0, delay_user=0, default=None, commands={ 'blacklist': Command.raw_command( self.add_link_blacklist, level=500, description='Blacklist a link', examples=[ CommandExample( None, 'Add a link to the blacklist for shallow search', chat= 'user:!add link blacklist 0 scamlink.lonk/\n' 'bot>user:Successfully added your links', description= 'Added the link scamlink.lonk/ to the blacklist for a shallow search' ).parse(), CommandExample( None, 'Add a link to the blacklist for deep search', chat= 'user:!add link blacklist 1 scamlink.lonk/\n' 'bot>user:Successfully added your links', description= 'Added the link scamlink.lonk/ to the blacklist for a deep search' ).parse(), ]), 'whitelist': Command.raw_command( self.add_link_whitelist, level=500, description='Whitelist a link', examples=[ CommandExample( None, 'Add a link to the whitelist', chat= 'user:!add link whitelink safelink.lonk/\n' 'bot>user:Successfully added your links', description= 'Added the link safelink.lonk/ to the whitelist' ).parse(), ]), }) }) self.commands['remove'] = Command.multiaction_command( level=100, delay_all=0, delay_user=0, default=None, command='remove', commands={ 'link': Command.multiaction_command( level=500, delay_all=0, delay_user=0, default=None, commands={ 'blacklist': Command.raw_command( self.remove_link_blacklist, level=500, description='Unblacklist a link', examples=[ CommandExample( None, 'Remove a blacklist link', chat= 'user:!remove link blacklist scamtwitch.scam\n' 'bot>user:Successfully removed your links', description= 'Removes scamtwitch.scam as a blacklisted link' ).parse(), ]), 'whitelist': Command.raw_command( self.remove_link_whitelist, level=500, description='Unwhitelist a link', examples=[ CommandExample( None, 'Remove a whitelist link', chat= 'user:!remove link whitelist twitch.safe\n' 'bot>user:Successfully removed your links', description= 'Removes twitch.safe as a whitelisted link' ).parse(), ]), }), }) def add_link_blacklist(self, **options): bot = options['bot'] message = options['message'] source = options['source'] parts = message.split(' ') try: if not parts[0].isnumeric(): for link in parts: self.blacklist_url(link) else: for link in parts[1:]: self.blacklist_url(link, level=int(parts[0])) except: log.exception('Unhandled exception in add_link') bot.whisper(source.username, 'Some error occurred white adding your links') return False bot.whisper(source.username, 'Successfully added your links') def add_link_whitelist(self, **options): bot = options['bot'] message = options['message'] source = options['source'] parts = message.split(' ') try: for link in parts: self.whitelist_url(link) except: log.exception('Unhandled exception in add_link') bot.whisper(source.username, 'Some error occurred white adding your links') return False bot.whisper(source.username, 'Successfully added your links') def remove_link_blacklist(self, **options): bot = options['bot'] message = options['message'] source = options['source'] parts = message.split(' ') try: for link in parts: self.unlist_url(link, 'blacklist') except: log.exception('Unhandled exception in add_link') bot.whisper(source.username, 'Some error occurred white adding your links') return False bot.whisper(source.username, 'Successfully removed your links') def remove_link_whitelist(self, **options): bot = options['bot'] message = options['message'] source = options['source'] parts = message.split(' ') try: for link in parts: self.unlist_url(link, 'whitelist') except: log.exception('Unhandled exception in add_link') bot.whisper(source.username, 'Some error occurred white adding your links') return False bot.whisper(source.username, 'Successfully removed your links')
class FollowAgeModule(BaseModule): ID = __name__.split('.')[-1] NAME = 'Follow age' DESCRIPTION = 'Makes two commands available: !followage and !followsince' CATEGORY = 'Feature' def __init__(self): super().__init__() self.action_queue = ActionQueue() self.action_queue.start() def load_commands(self, **options): # TODO: Have delay modifiable in settings self.commands['followage'] = pajbot.models.command.Command.raw_command(self.follow_age, delay_all=4, delay_user=8, description='Check your or someone elses follow age for a channel', examples=[ pajbot.models.command.CommandExample(None, 'Check your own follow age', chat='user:!followage\n' 'bot:pajlada, you have been following Karl_Kons for 4 months and 24 days', description='Check how long you have been following the current streamer (Karl_Kons in this case)').parse(), pajbot.models.command.CommandExample(None, 'Check someone elses follow age', chat='user:!followage NightNacht\n' 'bot:pajlada, NightNacht has been following Karl_Kons for 5 months and 4 days', description='Check how long any user has been following the current streamer (Karl_Kons in this case)').parse(), pajbot.models.command.CommandExample(None, 'Check someones follow age for a certain streamer', chat='user:!followage NightNacht forsenlol\n' 'bot:pajlada, NightNacht has been following forsenlol for 1 year and 4 months', description='Check how long NightNacht has been following forsenlol').parse(), pajbot.models.command.CommandExample(None, 'Check your own follow age for a certain streamer', chat='user:!followage pajlada forsenlol\n' 'bot:pajlada, you have been following forsenlol for 1 year and 3 months', description='Check how long you have been following forsenlol').parse(), ], ) self.commands['followsince'] = pajbot.models.command.Command.raw_command(self.follow_since, delay_all=4, delay_user=8, description='Check from when you or someone else first followed a channel', examples=[ pajbot.models.command.CommandExample(None, 'Check your own follow since', chat='user:!followsince\n' 'bot:pajlada, you have been following Karl_Kons since 04 March 2015, 07:02:01 UTC', description='Check when you first followed the current streamer (Karl_Kons in this case)').parse(), pajbot.models.command.CommandExample(None, 'Check someone elses follow since', chat='user:!followsince NightNacht\n' 'bot:pajlada, NightNacht has been following Karl_Kons since 03 July 2014, 04:12:42 UTC', description='Check when NightNacht first followed the current streamer (Karl_Kons in this case)').parse(), pajbot.models.command.CommandExample(None, 'Check someone elses follow since for another streamer', chat='user:!followsince NightNacht forsenlol\n' 'bot:pajlada, NightNacht has been following forsenlol since 13 June 2013, 13:10:51 UTC', description='Check when NightNacht first followed the given streamer (forsenlol)').parse(), pajbot.models.command.CommandExample(None, 'Check your follow since for another streamer', chat='user:!followsince pajlada forsenlol\n' 'bot:pajlada, you have been following forsenlol since 16 December 1990, 03:06:51 UTC', description='Check when you first followed the given streamer (forsenlol)').parse(), ], ) def check_follow_age(self, bot, source, username, streamer): streamer = bot.streamer if streamer is None else streamer.lower() age = bot.twitchapi.get_follow_relationship(username, streamer) if source.username == username: if age is False: bot.say('{}, you are not following {}'.format(source.username_raw, streamer)) else: bot.say('{}, you have been following {} for {}'.format(source.username_raw, streamer, time_since(datetime.datetime.now().timestamp() - age.timestamp(), 0))) else: if age is False: bot.say('{}, {} is not following {}'.format(source.username_raw, username, streamer)) else: bot.say('{}, {} has been following {} for {}'.format( source.username_raw, username, streamer, time_since(datetime.datetime.now().timestamp() - age.timestamp(), 0))) def check_follow_since(self, bot, source, username, streamer): streamer = bot.streamer if streamer is None else streamer.lower() follow_since = bot.twitchapi.get_follow_relationship(username, streamer) if source.username == username: if follow_since is False: bot.say('{}, you are not following {}'.format(source.username_raw, streamer)) else: bot.say('{}, you have been following {} since {} UTC'.format( source.username_raw, streamer, follow_since.strftime('%d %B %Y, %X'))) else: if follow_since is False: bot.say('{}, {} is not following {}'.format(source.username_raw, username, streamer)) else: bot.say('{}, {} has been following {} since {} UTC'.format( source.username_raw, username, streamer, follow_since.strftime('%d %B %Y, %X'))) def follow_age(self, **options): source = options['source'] message = options['message'] bot = options['bot'] username, streamer = self.parse_message(bot, source, message) self.action_queue.add(self.check_follow_age, args=[bot, source, username, streamer]) def follow_since(self, **options): bot = options['bot'] source = options['source'] message = options['message'] username, streamer = self.parse_message(bot, source, message) self.action_queue.add(self.check_follow_since, args=[bot, source, username, streamer]) def parse_message(self, bot, source, message): username = source.username streamer = None if message is not None and len(message) > 0: message_split = message.split(' ') if len(message_split[0]) and message_split[0].replace('_', '').isalnum(): username = message_split[0].lower() if len(message_split) > 1 and message_split[1].replace('_', '').isalnum(): streamer = message_split[1] return username, streamer
class LinkCheckerModule(BaseModule): ID = __name__.split(".")[-1] NAME = "Link Checker" DESCRIPTION = "Checks links if they're bad" ENABLED_DEFAULT = True CATEGORY = "Filter" SETTINGS = [ ModuleSetting( key="ban_pleb_links", label="Disallow links from non-subscribers", type="boolean", required=True, default=False, ), ModuleSetting(key="ban_sub_links", label="Disallow links from subscribers", type="boolean", required=True, default=False), ModuleSetting( key="timeout_length", label="Timeout length", type="number", required=True, placeholder="Timeout length in seconds", default=60, constraints={ "min_value": 1, "max_value": 3600 }, ), ] def __init__(self, bot): super().__init__(bot) self.db_session = None self.links = {} self.blacklisted_links = [] self.whitelisted_links = [] self.cache = LinkCheckerCache( ) # cache[url] = True means url is safe, False means the link is bad self.action_queue = ActionQueue() self.action_queue.start() def enable(self, bot): HandlerManager.add_handler("on_message", self.on_message, priority=100) HandlerManager.add_handler("on_commit", self.on_commit) if bot: self.run_later = bot.execute_delayed if "safebrowsingapi" in bot.config["main"]: # XXX: This should be loaded as a setting instead. # There needs to be a setting for settings to have them as "passwords" # so they're not displayed openly self.safeBrowsingAPI = SafeBrowsingAPI( bot.config["main"]["safebrowsingapi"], bot.nickname, bot.version) else: self.safeBrowsingAPI = None if self.db_session is not None: self.db_session.commit() self.db_session.close() self.db_session = None self.db_session = DBManager.create_session() self.blacklisted_links = [] for link in self.db_session.query(BlacklistedLink): self.blacklisted_links.append(link) self.whitelisted_links = [] for link in self.db_session.query(WhitelistedLink): self.whitelisted_links.append(link) def disable(self, bot): pajbot.managers.handler.HandlerManager.remove_handler( "on_message", self.on_message) pajbot.managers.handler.HandlerManager.remove_handler( "on_commit", self.on_commit) if self.db_session is not None: self.db_session.commit() self.db_session.close() self.db_session = None self.blacklisted_links = [] self.whitelisted_links = [] def reload(self): log.info("Loaded {0} bad links and {1} good links".format( len(self.blacklisted_links), len(self.whitelisted_links))) return self super_whitelist = ["pajlada.se", "pajlada.com", "forsen.tv", "pajbot.com"] def on_message(self, source, whisper, urls, **rest): if whisper: return if source.level >= 500 or source.moderator is True: return if len(urls) > 0: do_timeout = False ban_reason = "You are not allowed to post links in chat" whisper_reason = "??? KKona" if self.settings[ "ban_pleb_links"] is True and source.subscriber is False: do_timeout = True whisper_reason = "You cannot post non-verified links in chat if you're not a subscriber." elif self.settings[ "ban_sub_links"] is True and source.subscriber is True: do_timeout = True whisper_reason = "You cannot post non-verified links in chat." if do_timeout is True: # Check if the links are in our super-whitelist. i.e. on the pajlada.se domain o forsen.tv for url in urls: parsed_url = Url(url) if len(parsed_url.parsed.netloc.split(".")) < 2: continue whitelisted = False for whitelist in self.super_whitelist: if is_subdomain(parsed_url.parsed.netloc, whitelist): whitelisted = True break if whitelisted is False: self.bot.timeout(source.username, 30, reason=ban_reason) if source.minutes_in_chat_online > 60: self.bot.whisper(source.username, whisper_reason) return False for url in urls: # Action which will be taken when a bad link is found action = Action( self.bot.timeout, args=[source.username, self.settings["timeout_length"]], kwargs={"reason": "Banned link"}, ) # First we perform a basic check if self.simple_check(url, action) == self.RET_FURTHER_ANALYSIS: # If the basic check returns no relevant data, we queue up a proper check on the URL self.action_queue.add(self.check_url, args=[url, action]) def on_commit(self, **rest): if self.db_session is not None: self.db_session.commit() def delete_from_cache(self, url): if url in self.cache: log.debug("LinkChecker: Removing url {0} from cache".format(url)) del self.cache[url] def cache_url(self, url, safe): if url in self.cache and self.cache[url] == safe: return log.debug("LinkChecker: Caching url {0} as {1}".format( url, "SAFE" if safe is True else "UNSAFE")) self.cache[url] = safe self.run_later(20, self.delete_from_cache, (url, )) def counteract_bad_url(self, url, action=None, want_to_cache=True, want_to_blacklist=False): log.debug("LinkChecker: BAD URL FOUND {0}".format(url.url)) if action: action.run() if want_to_cache: self.cache_url(url.url, False) if want_to_blacklist: self.blacklist_url(url.url, url.parsed) return True def blacklist_url(self, url, parsed_url=None, level=0): if not (url.lower().startswith("http://") or url.lower().startswith("https://")): url = "http://" + url if parsed_url is None: parsed_url = urllib.parse.urlparse(url) if self.is_blacklisted(url, parsed_url): return False domain = parsed_url.netloc.lower() path = parsed_url.path.lower() if domain.startswith("www."): domain = domain[4:] if path.endswith("/"): path = path[:-1] if path == "": path = "/" link = BlacklistedLink(domain, path, level) self.db_session.add(link) self.blacklisted_links.append(link) self.db_session.commit() def whitelist_url(self, url, parsed_url=None): if not (url.lower().startswith("http://") or url.lower().startswith("https://")): url = "http://" + url if parsed_url is None: parsed_url = urllib.parse.urlparse(url) if self.is_whitelisted(url, parsed_url): return domain = parsed_url.netloc.lower() path = parsed_url.path.lower() if domain.startswith("www."): domain = domain[4:] if path.endswith("/"): path = path[:-1] if path == "": path = "/" link = WhitelistedLink(domain, path) self.db_session.add(link) self.whitelisted_links.append(link) self.db_session.commit() def is_blacklisted(self, url, parsed_url=None, sublink=False): if parsed_url is None: parsed_url = urllib.parse.urlparse(url) domain = parsed_url.netloc.lower() path = parsed_url.path.lower() if path == "": path = "/" domain_split = domain.split(".") if len(domain_split) < 2: return False for link in self.blacklisted_links: if link.is_subdomain(domain): if link.is_subpath(path): if not sublink: return True elif ( link.level >= 1 ): # if it's a sublink, but the blacklisting level is 0, we don't consider it blacklisted return True return False def is_whitelisted(self, url, parsed_url=None): if parsed_url is None: parsed_url = urllib.parse.urlparse(url) domain = parsed_url.netloc.lower() path = parsed_url.path.lower() if path == "": path = "/" domain_split = domain.split(".") if len(domain_split) < 2: return False for link in self.whitelisted_links: if link.is_subdomain(domain): if link.is_subpath(path): return True return False RET_BAD_LINK = -1 RET_FURTHER_ANALYSIS = 0 RET_GOOD_LINK = 1 def basic_check(self, url, action, sublink=False): """ Check if the url is in the cache, or if it's Return values: 1 = Link is OK -1 = Link is bad 0 = Link needs further analysis """ if url.url in self.cache: log.debug("LinkChecker: Url {0} found in cache".format(url.url)) if not self.cache[url.url]: # link is bad self.counteract_bad_url(url, action, False, False) return self.RET_BAD_LINK return self.RET_GOOD_LINK log.info("Checking if link is blacklisted...") if self.is_blacklisted(url.url, url.parsed, sublink): log.debug("LinkChecker: Url {0} is blacklisted".format(url.url)) self.counteract_bad_url(url, action, want_to_blacklist=False) return self.RET_BAD_LINK log.info("Checking if link is whitelisted...") if self.is_whitelisted(url.url, url.parsed): log.debug("LinkChecker: Url {0} allowed by the whitelist".format( url.url)) self.cache_url(url.url, True) return self.RET_GOOD_LINK return self.RET_FURTHER_ANALYSIS def simple_check(self, url, action): url = Url(url) if len(url.parsed.netloc.split(".")) < 2: # The URL is broken, ignore it return self.RET_FURTHER_ANALYSIS return self.basic_check(url, action) def check_url(self, url, action): url = Url(url) if len(url.parsed.netloc.split(".")) < 2: # The URL is broken, ignore it return try: self._check_url(url, action) except: log.exception("LinkChecker unhanled exception while _check_url") def _check_url(self, url, action): log.debug("LinkChecker: Checking url {0}".format(url.url)) # XXX: The basic check is currently performed twice on links found in messages. Solve res = self.basic_check(url, action) if res == self.RET_GOOD_LINK: return elif res == self.RET_BAD_LINK: return connection_timeout = 2 read_timeout = 1 try: r = requests.head(url.url, allow_redirects=True, timeout=connection_timeout) except: self.cache_url(url.url, True) return checkcontenttype = "content-type" in r.headers and r.headers[ "content-type"] == "application/octet-stream" checkdispotype = "disposition-type" in r.headers and r.headers[ "disposition-type"] == "attachment" if checkcontenttype or checkdispotype: # triggering a download not allowed self.counteract_bad_url(url, action) return redirected_url = Url(r.url) if is_same_url(url, redirected_url) is False: res = self.basic_check(redirected_url, action) if res == self.RET_GOOD_LINK: return elif res == self.RET_BAD_LINK: return if self.safeBrowsingAPI: if self.safeBrowsingAPI.check_url( redirected_url.url): # harmful url detected log.debug("Bad url because google api") self.counteract_bad_url(url, action, want_to_blacklist=False) self.counteract_bad_url(redirected_url, want_to_blacklist=False) return if "content-type" not in r.headers or not r.headers[ "content-type"].startswith("text/html"): return # can't analyze non-html content maximum_size = 1024 * 1024 * 10 # 10 MB receive_timeout = 3 html = "" try: response = requests.get(url=url.url, stream=True, timeout=(connection_timeout, read_timeout)) content_length = response.headers.get("Content-Length") if content_length and int( response.headers.get("Content-Length")) > maximum_size: log.error("This file is too big!") return size = 0 start = pajbot.utils.now().timestamp() for chunk in response.iter_content(1024): if pajbot.utils.now().timestamp() - start > receive_timeout: log.error("The site took too long to load") return size += len(chunk) if size > maximum_size: log.error("This file is too big! (fake header)") return html += str(chunk) except requests.exceptions.ConnectTimeout: log.warning("Connection timed out while checking {0}".format( url.url)) self.cache_url(url.url, True) return except requests.exceptions.ReadTimeout: log.warning("Reading timed out while checking {0}".format(url.url)) self.cache_url(url.url, True) return except: log.exception("Unhandled exception") return try: soup = BeautifulSoup(html, "html.parser") except: return original_url = url original_redirected_url = redirected_url urls = [] for link in soup.find_all( "a"): # get a list of links to external sites url = link.get("href") if url is None: continue if url.startswith("//"): urls.append("http:" + url) elif url.startswith("http://") or url.startswith("https://"): urls.append(url) for url in urls: # check if the site links to anything dangerous url = Url(url) if is_subdomain(url.parsed.netloc, original_url.parsed.netloc): # log.debug('Skipping because internal link') continue log.debug("Checking sublink {0}".format(url.url)) res = self.basic_check(url, action, sublink=True) if res == self.RET_BAD_LINK: self.counteract_bad_url(url) self.counteract_bad_url(original_url, want_to_blacklist=False) self.counteract_bad_url(original_redirected_url, want_to_blacklist=False) return elif res == self.RET_GOOD_LINK: continue try: r = requests.head(url.url, allow_redirects=True, timeout=connection_timeout) except: continue redirected_url = Url(r.url) if not is_same_url(url, redirected_url): res = self.basic_check(redirected_url, action, sublink=True) if res == self.RET_BAD_LINK: self.counteract_bad_url(url) self.counteract_bad_url(original_url, want_to_blacklist=False) self.counteract_bad_url(original_redirected_url, want_to_blacklist=False) return elif res == self.RET_GOOD_LINK: continue if self.safeBrowsingAPI: if self.safeBrowsingAPI.check_url( redirected_url.url): # harmful url detected log.debug("Evil sublink {0} by google API".format(url)) self.counteract_bad_url(original_url, action) self.counteract_bad_url(original_redirected_url) self.counteract_bad_url(url) self.counteract_bad_url(redirected_url) return # if we got here, the site is clean for our standards self.cache_url(original_url.url, True) self.cache_url(original_redirected_url.url, True) return def load_commands(self, **options): self.commands["add"] = Command.multiaction_command( level=100, delay_all=0, delay_user=0, default=None, command="add", commands={ "link": Command.multiaction_command( level=500, delay_all=0, delay_user=0, default=None, commands={ "blacklist": Command.raw_command( self.add_link_blacklist, level=500, delay_all=0, delay_user=0, description="Blacklist a link", examples=[ CommandExample( None, "Add a link to the blacklist for a shallow search", chat= "user:!add link blacklist --shallow scamlink.lonk/\n" "bot>user:Successfully added your links", description= "Added the link scamlink.lonk/ to the blacklist for a shallow search", ).parse(), CommandExample( None, "Add a link to the blacklist for a deep search", chat= "user:!add link blacklist --deep scamlink.lonk/\n" "bot>user:Successfully added your links", description= "Added the link scamlink.lonk/ to the blacklist for a deep search", ).parse(), ], ), "whitelist": Command.raw_command( self.add_link_whitelist, level=500, delay_all=0, delay_user=0, description="Whitelist a link", examples=[ CommandExample( None, "Add a link to the whitelist", chat= "user:!add link whitelink safelink.lonk/\n" "bot>user:Successfully added your links", description= "Added the link safelink.lonk/ to the whitelist", ).parse() ], ), }, ) }, ) self.commands["remove"] = Command.multiaction_command( level=100, delay_all=0, delay_user=0, default=None, command="remove", commands={ "link": Command.multiaction_command( level=500, delay_all=0, delay_user=0, default=None, commands={ "blacklist": Command.raw_command( self.remove_link_blacklist, level=500, delay_all=0, delay_user=0, description="Remove a link from the blacklist.", examples=[ CommandExample( None, "Remove a link from the blacklist.", chat="user:!remove link blacklist 20\n" "bot>user:Successfully removed blacklisted link with id 20", description= "Remove a link from the blacklist with an ID", ).parse() ], ), "whitelist": Command.raw_command( self.remove_link_whitelist, level=500, delay_all=0, delay_user=0, description="Remove a link from the whitelist.", examples=[ CommandExample( None, "Remove a link from the whitelist.", chat="user:!remove link whitelist 12\n" "bot>user:Successfully removed blacklisted link with id 12", description= "Remove a link from the whitelist with an ID", ).parse() ], ), }, ) }, ) def add_link_blacklist(self, **options): bot = options["bot"] message = options["message"] source = options["source"] options, new_links = self.parse_link_blacklist_arguments(message) if new_links: parts = new_links.split(" ") try: for link in parts: if len(link) > 1: self.blacklist_url(link, **options) AdminLogManager.post("Blacklist link added", source, link) bot.whisper(source.username, "Successfully added your links") return True except: log.exception("Unhandled exception in add_link_blacklist") bot.whisper(source.username, "Some error occurred while adding your links") return False else: bot.whisper(source.username, "Usage: !add link blacklist LINK") return False def add_link_whitelist(self, **options): bot = options["bot"] message = options["message"] source = options["source"] parts = message.split(" ") try: for link in parts: self.whitelist_url(link) AdminLogManager.post("Whitelist link added", source, link) except: log.exception("Unhandled exception in add_link") bot.whisper(source.username, "Some error occurred white adding your links") return False bot.whisper(source.username, "Successfully added your links") def remove_link_blacklist(self, **options): message = options["message"] bot = options["bot"] source = options["source"] if message: id = None try: id = int(message) except ValueError: pass link = self.db_session.query(BlacklistedLink).filter_by( id=id).one_or_none() if link: self.blacklisted_links.remove(link) self.db_session.delete(link) self.db_session.commit() else: bot.whisper(source.username, "No link with the given id found") return False AdminLogManager.post("Blacklist link removed", source, link.domain) bot.whisper( source.username, "Successfully removed blacklisted link with id {0}".format( link.id)) else: bot.whisper(source.username, "Usage: !remove link blacklist ID") return False def remove_link_whitelist(self, **options): message = options["message"] bot = options["bot"] source = options["source"] if message: id = None try: id = int(message) except ValueError: pass link = self.db_session.query(WhitelistedLink).filter_by( id=id).one_or_none() if link: self.whitelisted_links.remove(link) self.db_session.delete(link) self.db_session.commit() else: bot.whisper(source.username, "No link with the given id found") return False AdminLogManager.post("Whitelist link removed", source, link.domain) bot.whisper( source.username, "Successfully removed whitelisted link with id {0}".format( link.id)) else: bot.whisper(source.username, "Usage: !remove link whitelist ID") return False @staticmethod def parse_link_blacklist_arguments(message): parser = argparse.ArgumentParser() parser.add_argument("--deep", dest="level", action="store_true") parser.add_argument("--shallow", dest="level", action="store_false") parser.set_defaults(level=False) try: args, unknown = parser.parse_known_args(message.split()) except SystemExit: return False, False except: log.exception("Unhandled exception in add_link_blacklist") return False, False # Strip options of any values that are set as None options = {k: v for k, v in vars(args).items() if v is not None} response = " ".join(unknown) return options, response
class FollowAgeModule(BaseModule): ID = __name__.split('.')[-1] NAME = 'Follow age' DESCRIPTION = 'Makes two commands available: !followage and !followsince' CATEGORY = 'Feature' def __init__(self): super().__init__() self.action_queue = ActionQueue() self.action_queue.start() def load_commands(self, **options): # TODO: Have delay modifiable in settings self.commands['followage'] = pajbot.models.command.Command.raw_command( self.follow_age, delay_all=4, delay_user=8, description='Check your or someone elses follow age for a channel', examples=[ pajbot.models.command.CommandExample( None, 'Check your own follow age', chat='user:!followage\n' 'bot:pajlada, you have been following Karl_Kons for 4 months and 24 days', description= 'Check how long you have been following the current streamer (Karl_Kons in this case)' ).parse(), pajbot.models.command.CommandExample( None, 'Check someone elses follow age', chat='user:!followage NightNacht\n' 'bot:pajlada, NightNacht has been following Karl_Kons for 5 months and 4 days', description= 'Check how long any user has been following the current streamer (Karl_Kons in this case)' ).parse(), pajbot.models.command.CommandExample( None, 'Check someones follow age for a certain streamer', chat='user:!followage NightNacht forsenlol\n' 'bot:pajlada, NightNacht has been following forsenlol for 1 year and 4 months', description= 'Check how long NightNacht has been following forsenlol'). parse(), pajbot.models.command.CommandExample( None, 'Check your own follow age for a certain streamer', chat='user:!followage pajlada forsenlol\n' 'bot:pajlada, you have been following forsenlol for 1 year and 3 months', description= 'Check how long you have been following forsenlol').parse( ), ], ) self.commands['followsince'] = pajbot.models.command.Command.raw_command( self.follow_since, delay_all=4, delay_user=8, description= 'Check from when you or someone else first followed a channel', examples=[ pajbot.models.command.CommandExample( None, 'Check your own follow since', chat='user:!followsince\n' 'bot:pajlada, you have been following Karl_Kons since 04 March 2015, 07:02:01 UTC', description= 'Check when you first followed the current streamer (Karl_Kons in this case)' ).parse(), pajbot.models.command.CommandExample( None, 'Check someone elses follow since', chat='user:!followsince NightNacht\n' 'bot:pajlada, NightNacht has been following Karl_Kons since 03 July 2014, 04:12:42 UTC', description= 'Check when NightNacht first followed the current streamer (Karl_Kons in this case)' ).parse(), pajbot.models.command.CommandExample( None, 'Check someone elses follow since for another streamer', chat='user:!followsince NightNacht forsenlol\n' 'bot:pajlada, NightNacht has been following forsenlol since 13 June 2013, 13:10:51 UTC', description= 'Check when NightNacht first followed the given streamer (forsenlol)' ).parse(), pajbot.models.command.CommandExample( None, 'Check your follow since for another streamer', chat='user:!followsince pajlada forsenlol\n' 'bot:pajlada, you have been following forsenlol since 16 December 1990, 03:06:51 UTC', description= 'Check when you first followed the given streamer (forsenlol)' ).parse(), ], ) def check_follow_age(self, bot, source, username, streamer): streamer = bot.streamer if streamer is None else streamer.lower() age = bot.twitchapi.get_follow_relationship(username, streamer) if source.username == username: if age is False: bot.say('{}, you are not following {}'.format( source.username_raw, streamer)) else: bot.say('{}, you have been following {} for {}'.format( source.username_raw, streamer, time_since( datetime.datetime.now().timestamp() - age.timestamp(), 0))) else: if age is False: bot.say('{}, {} is not following {}'.format( source.username_raw, username, streamer)) else: bot.say('{}, {} has been following {} for {}'.format( source.username_raw, username, streamer, time_since( datetime.datetime.now().timestamp() - age.timestamp(), 0))) def check_follow_since(self, bot, source, username, streamer): streamer = bot.streamer if streamer is None else streamer.lower() follow_since = bot.twitchapi.get_follow_relationship( username, streamer) if source.username == username: if follow_since is False: bot.say('{}, you are not following {}'.format( source.username_raw, streamer)) else: bot.say('{}, you have been following {} since {} UTC'.format( source.username_raw, streamer, follow_since.strftime('%d %B %Y, %X'))) else: if follow_since is False: bot.say('{}, {} is not following {}'.format( source.username_raw, username, streamer)) else: bot.say('{}, {} has been following {} since {} UTC'.format( source.username_raw, username, streamer, follow_since.strftime('%d %B %Y, %X'))) def follow_age(self, **options): source = options['source'] message = options['message'] bot = options['bot'] username, streamer = self.parse_message(bot, source, message) self.action_queue.add(self.check_follow_age, args=[bot, source, username, streamer]) def follow_since(self, **options): bot = options['bot'] source = options['source'] message = options['message'] username, streamer = self.parse_message(bot, source, message) self.action_queue.add(self.check_follow_since, args=[bot, source, username, streamer]) def parse_message(self, bot, source, message): username = source.username streamer = None if message is not None and len(message) > 0: message_split = message.split(' ') if len(message_split[0]) and message_split[0].replace( '_', '').isalnum(): username = message_split[0].lower() if len(message_split) > 1 and message_split[1].replace( '_', '').isalnum(): streamer = message_split[1] return username, streamer
class Bot: """ Main class for the twitch bot """ version = "1.35" date_fmt = "%H:%M" admin = None url_regex_str = r"\(?(?:(http|https):\/\/)?(?:((?:[^\W\s]|\.|-|[:]{1})+)@{1})?((?:www.)?(?:[^\W\s]|\.|-)+[\.][^\W\s]{2,4}|localhost(?=\/)|\d{1,3}\.\d{1,3}\.\d{1,3}\.\d{1,3})(?::(\d*))?([\/]?[^\s\?]*[\/]{1})*(?:\/?([^\s\n\?\[\]\{\}\#]*(?:(?=\.)){1}|[^\s\n\?\[\]\{\}\.\#]*)?([\.]{1}[^\s\?\#]*)?)?(?:\?{1}([^\s\n\#\[\]]*))?([\#][^\s\n]*)?\)?" last_ping = pajbot.utils.now() last_pong = pajbot.utils.now() @staticmethod def parse_args(): parser = argparse.ArgumentParser() parser.add_argument( "--config", "-c", default="config.ini", help="Specify which config file to use (default: config.ini)") parser.add_argument( "--silent", action="count", help="Decides whether the bot should be silent or not") # TODO: Add a log level argument. return parser.parse_args() bot_token = None @property def password(self): if "password" in self.config["main"]: log.warning( "DEPRECATED - Using bot password/oauth token from file. " "You should authenticate in web gui using route /bot_login " "and remove password from config file") return self.config["main"]["password"] if self.bot_token is None: self.bot_token = BotToken(self.config) t = self.bot_token.access_token() return "oauth:{}".format(t) def load_config(self, config): self.config = config DBManager.init(self.config["main"]["db"]) redis_options = {} if "redis" in config: redis_options = dict(config.items("redis")) RedisManager.init(**redis_options) pajbot.models.user.Config.se_sync_token = config["main"].get( "se_sync_token", None) pajbot.models.user.Config.se_channel = config["main"].get( "se_channel", None) self.domain = config["web"].get("domain", "localhost") self.nickname = config["main"].get("nickname", "pajbot") self.timezone = config["main"].get("timezone", "UTC") if config["main"].getboolean("verified", False): TMI.promote_to_verified() self.trusted_mods = config.getboolean("main", "trusted_mods") self.phrases = { "welcome": ["{nickname} {version} running!"], "quit": ["{nickname} {version} shutting down..."] } if "phrases" in config: phrases = config["phrases"] if "welcome" in phrases: self.phrases["welcome"] = phrases["welcome"].splitlines() if "quit" in phrases: self.phrases["quit"] = phrases["quit"].splitlines() TimeManager.init_timezone(self.timezone) if "streamer" in config["main"]: self.streamer = config["main"]["streamer"] self.channel = "#" + self.streamer elif "target" in config["main"]: self.channel = config["main"]["target"] self.streamer = self.channel[1:] StreamHelper.init_streamer(self.streamer) self.silent = False self.dev = False if "flags" in config: self.silent = True if "silent" in config["flags"] and config[ "flags"]["silent"] == "1" else self.silent self.dev = True if "dev" in config["flags"] and config["flags"][ "dev"] == "1" else self.dev def __init__(self, config, args=None): # Load various configuration variables from the given config object # The config object that should be passed through should # come from pajbot.utils.load_config self.load_config(config) log.debug("Loaded config") # streamer is additionally initialized here so streamer can be accessed by the DB migrations # before StreamHelper.init_bot() is called later (which depends on an upgraded DB because # StreamManager accesses the DB) StreamHelper.init_streamer(self.streamer) # Update the database (and partially redis) scheme if necessary using alembic # In case of errors, i.e. if the database is out of sync or the alembic # binary can't be called, we will shut down the bot. pajbot.utils.alembic_upgrade() log.debug("ran db upgrade") # Actions in this queue are run in a separate thread. # This means actions should NOT access any database-related stuff. self.action_queue = ActionQueue() self.action_queue.start() self.reactor = irc.client.Reactor(self.on_connect) self.start_time = pajbot.utils.now() ActionParser.bot = self HandlerManager.init_handlers() self.socket_manager = SocketManager(self.streamer) self.stream_manager = StreamManager(self) StreamHelper.init_bot(self, self.stream_manager) ScheduleManager.init() self.users = UserManager() self.decks = DeckManager() self.banphrase_manager = BanphraseManager(self).load() self.timer_manager = TimerManager(self).load() self.kvi = KVIManager() twitch_client_id = None twitch_oauth = None if "twitchapi" in self.config: twitch_client_id = self.config["twitchapi"].get("client_id", None) twitch_oauth = self.config["twitchapi"].get("oauth", None) # A client ID is required for the bot to work properly now, give an error for now if twitch_client_id is None: log.error( 'MISSING CLIENT ID, SET "client_id" VALUE UNDER [twitchapi] SECTION IN CONFIG FILE' ) self.twitchapi = TwitchAPI(twitch_client_id, twitch_oauth) self.emote_manager = EmoteManager(twitch_client_id) self.epm_manager = EpmManager() self.ecount_manager = EcountManager() self.twitter_manager = TwitterManager(self) self.module_manager = ModuleManager(self.socket_manager, bot=self).load() self.commands = CommandManager(socket_manager=self.socket_manager, module_manager=self.module_manager, bot=self).load() HandlerManager.trigger("on_managers_loaded") # Reloadable managers self.reloadable = {} # Commitable managers self.commitable = { "commands": self.commands, "banphrases": self.banphrase_manager } self.execute_every(10 * 60, self.commit_all) self.execute_every(1, self.do_tick) try: self.admin = self.config["main"]["admin"] except KeyError: log.warning( "No admin user specified. See the [main] section in config.example.ini for its usage." ) if self.admin: with self.users.get_user_context(self.admin) as user: user.level = 2000 self.parse_version() relay_host = self.config["main"].get("relay_host", None) relay_password = self.config["main"].get("relay_password", None) if relay_host is None or relay_password is None: self.irc = MultiIRCManager(self) else: self.irc = SingleIRCManager(self) self.reactor.add_global_handler("all_events", self.irc._dispatcher, -10) self.data = {} self.data_cb = {} self.url_regex = re.compile(self.url_regex_str, re.IGNORECASE) self.data["broadcaster"] = self.streamer self.data["version"] = self.version self.data["version_brief"] = self.version_brief self.data["bot_name"] = self.nickname self.data_cb["status_length"] = self.c_status_length self.data_cb["stream_status"] = self.c_stream_status self.data_cb["bot_uptime"] = self.c_uptime self.data_cb["current_time"] = self.c_current_time self.silent = True if args.silent else self.silent if self.silent: log.info("Silent mode enabled") """ For actions that need to access the main thread, we can use the mainthread_queue. """ self.mainthread_queue = ActionQueue() self.execute_every(1, self.mainthread_queue.parse_action) self.websocket_manager = WebSocketManager(self) try: if self.config["twitchapi"]["update_subscribers"] == "1": self.execute_every(30 * 60, self.action_queue.add, (self.update_subscribers_stage1, )) except: pass def on_connect(self, sock): return self.irc.on_connect(sock) def update_subscribers_stage1(self): limit = 25 offset = 0 subscribers = [] log.info("Starting stage1 subscribers update") try: retry_same = 0 while True: log.debug("Beginning sub request limit=%s offset=%s", limit, offset) subs, retry_same, error = self.twitchapi.get_subscribers( self.streamer, 0, offset, 0 if retry_same is False else retry_same) log.debug("got em!") if error is True: log.error("Too many attempts, aborting") return False if retry_same is False: offset += limit if not subs: # We don't need to retry, and the last query finished propery # Break out of the loop and start fiddling with the subs! log.debug("Done!") break log.debug("Fetched %d subs", len(subs)) subscribers.extend(subs) if retry_same is not False: # In case the next attempt is a retry, wait for 3 seconds log.debug("waiting for 3 seconds...") time.sleep(3) log.debug("waited enough!") log.debug("Finished with the while True loop!") except: log.exception( "Caught an exception while trying to get subscribers") return None log.info("Ended stage1 subscribers update") if subscribers: log.info( "Got some subscribers, so we are pushing them to stage 2!") self.mainthread_queue.add(self.update_subscribers_stage2, args=[subscribers]) log.info("Pushed them now.") def update_subscribers_stage2(self, subscribers): self.kvi["active_subs"].set(len(subscribers) - 1) self.users.reset_subs() self.users.update_subs(subscribers) def start(self): """Start the IRC client.""" self.reactor.process_forever() def get_kvi_value(self, key, extra={}): return self.kvi[key].get() def get_last_tweet(self, key, extra={}): return self.twitter_manager.get_last_tweet(key) def get_emote_epm(self, key, extra={}): val = self.epm_manager.get_emote_epm(key) if val is None: return None # formats the number with grouping (e.g. 112,556) and zero decimal places return "{0:,.0f}".format(val) def get_emote_epm_record(self, key, extra={}): val = self.epm_manager.get_emote_epm_record(key) if val is None: return None # formats the number with grouping (e.g. 112,556) and zero decimal places return "{0:,.0f}".format(val) def get_emote_count(self, key, extra={}): val = self.ecount_manager.get_emote_count(key) if val is None: return None # formats the number with grouping (e.g. 112,556) and zero decimal places return "{0:,.0f}".format(val) @staticmethod def get_source_value(key, extra={}): try: return getattr(extra["source"], key) except: log.exception("Caught exception in get_source_value") return None def get_user_value(self, key, extra={}): try: user = self.users.find(extra["argument"]) if user: return getattr(user, key) except: log.exception("Caught exception in get_source_value") return None @staticmethod def get_command_value(key, extra={}): try: return getattr(extra["command"].data, key) except: log.exception("Caught exception in get_source_value") return None def get_usersource_value(self, key, extra={}): try: user = self.users.find(extra["argument"]) if user: return getattr(user, key) return getattr(extra["source"], key) except: log.exception("Caught exception in get_source_value") return None def get_time_value(self, key, extra={}): try: tz = timezone(key) return datetime.datetime.now(tz).strftime(self.date_fmt) except: log.exception("Unhandled exception in get_time_value") return None def get_current_song_value(self, key, extra={}): if self.stream_manager.online: current_song = PleblistManager.get_current_song( self.stream_manager.current_stream.id) inner_keys = key.split(".") val = current_song for inner_key in inner_keys: val = getattr(val, inner_key, None) if val is None: return None if val is not None: return val return None def get_strictargs_value(self, key, extra={}): ret = self.get_args_value(key, extra) if not ret: return None return ret @staticmethod def get_args_value(key, extra={}): r = None try: msg_parts = extra["message"].split(" ") except (KeyError, AttributeError): msg_parts = [""] try: if "-" in key: range_str = key.split("-") if len(range_str) == 2: r = (int(range_str[0]), int(range_str[1])) if r is None: r = (int(key), len(msg_parts)) except (TypeError, ValueError): r = (0, len(msg_parts)) try: return " ".join(msg_parts[r[0]:r[1]]) except AttributeError: return "" except: log.exception("UNHANDLED ERROR IN get_args_value") return "" def get_notify_value(self, key, extra={}): payload = { "message": extra["message"] or "", "trigger": extra["trigger"], "user": extra["source"].username_raw } self.websocket_manager.emit("notify", payload) return "" def get_value(self, key, extra={}): if key in extra: return extra[key] if key in self.data: return self.data[key] if key in self.data_cb: return self.data_cb[key]() log.warning("Unknown key passed to get_value: %s", key) return None def privmsg_arr(self, arr, target=None): for msg in arr: self.privmsg(msg, target) def privmsg_from_file(self, url, per_chunk=35, chunk_delay=30, target=None): try: r = requests.get(url) r.raise_for_status() content_type = r.headers["Content-Type"] if content_type is not None and cgi.parse_header( content_type)[0] != "text/plain": log.error( "privmsg_from_file should be fed with a text/plain URL. Refusing to send." ) return lines = r.text.splitlines() i = 0 while lines: if i == 0: self.privmsg_arr(lines[:per_chunk], target) else: self.execute_delayed(chunk_delay * i, self.privmsg_arr, (lines[:per_chunk], target)) del lines[:per_chunk] i = i + 1 except: log.exception("error in privmsg_from_file") # event is an event to clone and change the text from. # Usage: !eval bot.eval_from_file(event, 'https://pastebin.com/raw/LhCt8FLh') def eval_from_file(self, event, url): try: r = requests.get(url) r.raise_for_status() content_type = r.headers["Content-Type"] if content_type is not None and cgi.parse_header( content_type)[0] != "text/plain": log.error( "privmsg_from_file should be fed with a text/plain URL. Refusing to send." ) return lines = r.text.splitlines() import copy for msg in lines: cloned_event = copy.deepcopy(event) cloned_event.arguments = [msg] # omit the source connection as None (since its not used) self.on_pubmsg(None, cloned_event) self.whisper(event.source.user.lower(), "Successfully evaluated {0} lines".format(len(lines))) except: log.exception("BabyRage") self.whisper(event.source.user.lower(), "Exception BabyRage") def privmsg(self, message, channel=None, increase_message=True): if channel is None: channel = self.channel return self.irc.privmsg(message, channel, increase_message=increase_message) def c_uptime(self): return time_ago(self.start_time) @staticmethod def c_current_time(): return pajbot.utils.now() @property def is_online(self): return self.stream_manager.online def c_stream_status(self): return "online" if self.stream_manager.online else "offline" def c_status_length(self): if self.stream_manager.online: return time_ago(self.stream_manager.current_stream.stream_start) if self.stream_manager.last_stream is not None: return time_ago(self.stream_manager.last_stream.stream_end) return "No recorded stream FeelsBadMan " def execute_at(self, at, function, arguments=()): self.reactor.scheduler.execute_at(at, lambda: function(*arguments)) def execute_delayed(self, delay, function, arguments=()): self.reactor.scheduler.execute_after(delay, lambda: function(*arguments)) def execute_every(self, period, function, arguments=()): self.reactor.scheduler.execute_every(period, lambda: function(*arguments)) def _ban(self, username, reason=""): self.privmsg(".ban {0} {1}".format(username, reason), increase_message=False) def ban(self, username, reason=""): log.debug("Banning %s", username) self._timeout(username, 30, reason) self.execute_delayed(1, self._ban, (username, reason)) def ban_user(self, user, reason=""): self._timeout(user.username, 30, reason) self.execute_delayed(1, self._ban, (user.username, reason)) def unban(self, username): self.privmsg(".unban {0}".format(username), increase_message=False) def _timeout(self, username, duration, reason=""): self.privmsg(".timeout {0} {1} {2}".format(username, duration, reason), increase_message=False) def timeout(self, username, duration, reason=""): # log.debug("Timing out %s for %d seconds", username, duration) self._timeout(username, duration, reason) def timeout_warn(self, user, duration, reason=""): duration, punishment = user.timeout( duration, warning_module=self.module_manager["warning"]) self.timeout(user.username, duration, reason) return (duration, punishment) def timeout_user(self, user, duration, reason=""): self._timeout(user.username, duration, reason) def _timeout_user(self, user, duration, reason=""): self._timeout(user.username, duration, reason) def whisper(self, username, *messages, separator=". ", **rest): """ Takes a sequence of strings and concatenates them with separator. Then sends that string as a whisper to username """ if len(messages) < 0: return False message = separator.join(messages) return self.irc.whisper(username, message) def send_message_to_user(self, user, message, event, separator=". ", method="say"): if method == "say": self.say(user.username + ", " + lowercase_first_letter(message), separator=separator) elif method == "whisper": self.whisper(user.username, message, separator=separator) elif method == "me": self.me(message) elif method == "reply": if event.type in ["action", "pubmsg"]: self.say(message, separator=separator) elif event.type == "whisper": self.whisper(user.username, message, separator=separator) else: log.warning("Unknown send_message method: %s", method) def safe_privmsg(self, message, channel=None, increase_message=True): # Check for banphrases res = self.banphrase_manager.check_message(message, None) if res is not False: self.privmsg("filtered message ({})".format(res.id), channel, increase_message) return self.privmsg(message, channel, increase_message) def say(self, *messages, channel=None, separator=". "): """ Takes a sequence of strings and concatenates them with separator. Then sends that string to the given channel. """ if len(messages) < 0: return False if not self.silent: message = separator.join(messages).strip() message = clean_up_message(message) if not message: return False # log.info("Sending message: %s", message) self.privmsg(message[:510], channel) def is_bad_message(self, message): return self.banphrase_manager.check_message(message, None) is not False def safe_me(self, message, channel=None): if not self.is_bad_message(message): self.me(message, channel) def me(self, message, channel=None): self.say(".me " + message[:500], channel=channel) def parse_version(self): self.version = self.version self.version_brief = self.version if self.dev: try: current_branch = (subprocess.check_output( ["git", "rev-parse", "--abbrev-ref", "HEAD"]).decode("utf8").strip()) latest_commit = subprocess.check_output( ["git", "rev-parse", "HEAD"]).decode("utf8").strip()[:8] commit_number = subprocess.check_output( ["git", "rev-list", "HEAD", "--count"]).decode("utf8").strip() self.version = "{0} DEV ({1}, {2}, commit {3})".format( self.version, current_branch, latest_commit, commit_number) except: log.exception("hmm") def on_welcome(self, chatconn, event): return self.irc.on_welcome(chatconn, event) def connect(self): return self.irc.start() def on_disconnect(self, chatconn, event): self.irc.on_disconnect(chatconn, event) def parse_message(self, message, source, event, tags={}, whisper=False): msg_lower = message.lower() emote_tag = None for tag in tags: if tag["key"] == "subscriber" and event.target == self.channel: source.subscriber = tag["value"] == "1" elif tag["key"] == "emotes": emote_tag = tag["value"] elif tag["key"] == "display-name" and tag["value"]: source.username_raw = tag["value"] elif tag["key"] == "user-type": source.moderator = tag[ "value"] == "mod" or source.username == self.streamer # source.num_lines += 1 if source is None: log.error("No valid user passed to parse_message") return False if source.banned: self.ban(source.username) return False # If a user types when timed out, we assume he's been unbanned for a good reason and remove his flag. if source.timed_out is True: source.timed_out = False # Parse emotes in the message emote_instances, emote_counts = self.emote_manager.parse_all_emotes( message, emote_tag) self.epm_manager.handle_emotes(emote_counts) self.ecount_manager.handle_emotes(emote_counts) urls = self.find_unique_urls(message) # log.debug("{2}{0}: {1}".format(source.username, message, "<w>" if whisper else "")) res = HandlerManager.trigger( "on_message", source=source, message=message, emote_instances=emote_instances, emote_counts=emote_counts, whisper=whisper, urls=urls, event=event, ) if res is False: return False source.last_seen = pajbot.utils.now() source.last_active = pajbot.utils.now() if source.ignored: return False if whisper: self.whisper('datguy1', '{} said: {}'.format(source.username_raw, message)) if msg_lower[:1] == "!": msg_lower_parts = msg_lower.split(" ") trigger = msg_lower_parts[0][1:] msg_raw_parts = message.split(" ") remaining_message = " ".join( msg_raw_parts[1:]) if len(msg_raw_parts) > 1 else None if trigger in self.commands: command = self.commands[trigger] extra_args = { "emote_instances": emote_instances, "emote_counts": emote_counts, "trigger": trigger } command.run(self, source, remaining_message, event=event, args=extra_args, whisper=whisper) def on_whisper(self, chatconn, event): # We use .lower() in case twitch ever starts sending non-lowercased usernames username = event.source.user.lower() with self.users.get_user_context(username) as source: self.parse_message(event.arguments[0], source, event, whisper=True, tags=event.tags) def on_ping(self, chatconn, event): # self.say('Received a ping. Last ping received {} ago'.format(time_since(pajbot.utils.now().timestamp(), self.last_ping.timestamp()))) log.info("Received a ping. Last ping received %s ago", time_ago(self.last_ping)) self.last_ping = pajbot.utils.now() def on_pong(self, chatconn, event): # self.say('Received a pong. Last pong received {} ago'.format(time_since(pajbot.utils.now().timestamp(), self.last_pong.timestamp()))) log.info("Received a pong. Last pong received %s ago", time_ago(self.last_pong)) self.last_pong = pajbot.utils.now() def on_pubnotice(self, chatconn, event): return type = "whisper" if chatconn in self.whisper_manager else "normal" log.debug("NOTICE {}@{}: {}".format(type, event.target, event.arguments)) def on_usernotice(self, chatconn, event): # We use .lower() in case twitch ever starts sending non-lowercased usernames tags = {} for d in event.tags: tags[d["key"]] = d["value"] if "login" not in tags: return username = tags["login"] with self.users.get_user_context(username) as source: msg = "" if event.arguments: msg = event.arguments[0] HandlerManager.trigger("on_usernotice", source=source, message=msg, tags=tags) def on_action(self, chatconn, event): self.on_pubmsg(chatconn, event) def on_pubmsg(self, chatconn, event): if event.source.user == self.nickname: return False username = event.source.user.lower() # We use .lower() in case twitch ever starts sending non-lowercased usernames with self.users.get_user_context(username) as source: res = HandlerManager.trigger("on_pubmsg", source=source, message=event.arguments[0]) if res is False: return False self.parse_message(event.arguments[0], source, event, tags=event.tags) @time_method def reload_all(self): log.info("Reloading all...") for key, manager in self.reloadable.items(): log.debug("Reloading %s", key) manager.reload() log.debug("Done with %s", key) log.info("ok!") @time_method def commit_all(self): # log.info("Commiting all...") for key, manager in self.commitable.items(): # log.info("Commiting %s", key) manager.commit() # log.info("Done with %s", key) # log.info("ok!") HandlerManager.trigger("on_commit", stop_on_false=False) @staticmethod def do_tick(): HandlerManager.trigger("on_tick") def quit(self, message, event, **options): quit_chub = self.config["main"].get("control_hub", None) quit_delay = 0 if quit_chub is not None and event.target == ("#{}".format(quit_chub)): quit_delay_random = 300 try: if message is not None and int(message.split()[0]) >= 1: quit_delay_random = int(message.split()[0]) except (IndexError, ValueError, TypeError): pass quit_delay = random.randint(0, quit_delay_random) log.info("%s is restarting in %d seconds.", self.nickname, quit_delay) self.execute_delayed(quit_delay, self.quit_bot) def quit_bot(self, **options): HandlerManager.trigger("on_quit") self.commit_all() phrase_data = {"nickname": self.nickname, "version": self.version} try: ScheduleManager.base_scheduler.print_jobs() ScheduleManager.base_scheduler.shutdown(wait=False) except: log.exception("Error while shutting down the apscheduler") try: for p in self.phrases["quit"]: if not self.silent: self.privmsg(p.format(**phrase_data)) except Exception: log.exception("Exception caught while trying to say quit phrase") self.twitter_manager.quit() self.socket_manager.quit() self.irc.quit() sys.exit(0) @staticmethod def apply_filter(resp, f): available_filters = { "strftime": _filter_strftime, "lower": lambda var, args: var.lower(), "upper": lambda var, args: var.upper(), "time_since_minutes": lambda var, args: "no time" if var == 0 else time_since(var * 60, 0, time_format="long"), "time_since": lambda var, args: "no time" if var == 0 else time_since(var, 0, time_format="long"), "time_since_dt": _filter_time_since_dt, "urlencode": _filter_urlencode, "join": _filter_join, "number_format": _filter_number_format, "add": _filter_add, } if f.name in available_filters: return available_filters[f.name](resp, f.arguments) return resp def find_unique_urls(self, message): from pajbot.modules.linkchecker import find_unique_urls return find_unique_urls(self.url_regex, message)
class Bot: """ Main class for the twitch bot """ version = '2.6.3' date_fmt = '%H:%M' update_chatters_interval = 5 admin = None url_regex_str = r'\(?(?:(http|https):\/\/)?(?:((?:[^\W\s]|\.|-|[:]{1})+)@{1})?((?:www.)?(?:[^\W\s]|\.|-)+[\.][^\W\s]{2,4}|localhost(?=\/)|\d{1,3}\.\d{1,3}\.\d{1,3}\.\d{1,3})(?::(\d*))?([\/]?[^\s\?]*[\/]{1})*(?:\/?([^\s\n\?\[\]\{\}\#]*(?:(?=\.)){1}|[^\s\n\?\[\]\{\}\.\#]*)?([\.]{1}[^\s\?\#]*)?)?(?:\?{1}([^\s\n\#\[\]]*))?([\#][^\s\n]*)?\)?' def parse_args(): parser = argparse.ArgumentParser() parser.add_argument('--config', '-c', default='config.ini', help='Specify which config file to use ' '(default: config.ini)') parser.add_argument('--silent', action='count', help='Decides whether the bot should be ' 'silent or not') # TODO: Add a log level argument. return parser.parse_args() def load_default_phrases(self): default_phrases = { 'welcome': False, 'quit': False, 'nl': '{username} has typed {num_lines} messages in this channel!', 'nl_0': '{username} has not typed any messages in this channel BibleThump', 'nl_pos': '{username} is rank {nl_pos} line-farmer in this channel!', 'new_sub': 'Sub hype! {username} just subscribed PogChamp', 'resub': 'Resub hype! {username} just subscribed, {num_months} months in a row PogChamp <3 PogChamp', 'point_pos': '{username_w_verb} rank {point_pos} point-hoarder in this channel with {points} points.', } if 'phrases' in self.config: self.phrases = {} for phrase_key, phrase_value in self.config['phrases'].items(): if len(phrase_value.strip()) <= 0: self.phrases[phrase_key] = False else: self.phrases[phrase_key] = phrase_value for phrase_key, phrase_value in default_phrases.items(): if phrase_key not in self.phrases: self.phrases[phrase_key] = phrase_value else: self.phrases = default_phrases def load_config(self, config): self.config = config self.nickname = config['main'].get('nickname', 'pajbot') self.password = config['main'].get('password', 'abcdef') self.timezone = config['main'].get('timezone', 'UTC') self.trusted_mods = config.getboolean('main', 'trusted_mods') TimeManager.init_timezone(self.timezone) if 'streamer' in config['main']: self.streamer = config['main']['streamer'] self.channel = '#' + self.streamer elif 'target' in config['main']: self.channel = config['main']['target'] self.streamer = self.channel[1:] self.wolfram = None try: if 'wolfram' in config['main']: import wolframalpha self.wolfram = wolframalpha.Client(config['main']['wolfram']) except: pass self.silent = False self.dev = False if 'flags' in config: self.silent = True if 'silent' in config['flags'] and config['flags']['silent'] == '1' else self.silent self.dev = True if 'dev' in config['flags'] and config['flags']['dev'] == '1' else self.dev DBManager.init(self.config['main']['db']) redis_options = {} if 'redis' in config: log.info(config._sections['redis']) redis_options = config._sections['redis'] RedisManager.init(**redis_options) def __init__(self, config, args=None): self.load_config(config) self.last_ping = datetime.datetime.now() self.last_pong = datetime.datetime.now() self.load_default_phrases() self.db_session = DBManager.create_session() try: subprocess.check_call(['alembic', 'upgrade', 'head'] + ['--tag="{0}"'.format(' '.join(sys.argv[1:]))]) except subprocess.CalledProcessError: log.exception('aaaa') log.error('Unable to call `alembic upgrade head`, this means the database could be out of date. Quitting.') sys.exit(1) except PermissionError: log.error('No permission to run `alembic upgrade head`. This means your user probably doesn\'t have execution rights on the `alembic` binary.') log.error('The error can also occur if it can\'t find `alembic` in your PATH, and instead tries to execute the alembic folder.') sys.exit(1) except FileNotFoundError: log.error('Could not found an installation of alembic. Please install alembic to continue.') sys.exit(1) except: log.exception('Unhandled exception when calling db update') sys.exit(1) # Actions in this queue are run in a separate thread. # This means actions should NOT access any database-related stuff. self.action_queue = ActionQueue() self.action_queue.start() self.reactor = irc.client.Reactor() self.start_time = datetime.datetime.now() ActionParser.bot = self HandlerManager.init_handlers() self.socket_manager = SocketManager(self) self.stream_manager = StreamManager(self) StreamHelper.init_bot(self, self.stream_manager) self.users = UserManager() self.decks = DeckManager().reload() self.module_manager = ModuleManager(self.socket_manager, bot=self).load() self.commands = CommandManager( socket_manager=self.socket_manager, module_manager=self.module_manager, bot=self).load() self.filters = FilterManager().reload() self.banphrase_manager = BanphraseManager(self).load() self.timer_manager = TimerManager(self).load() self.kvi = KVIManager().reload() self.emotes = EmoteManager(self).reload() self.twitter_manager = TwitterManager(self).reload() self.duel_manager = DuelManager(self) HandlerManager.trigger('on_managers_loaded') # Reloadable managers self.reloadable = { 'filters': self.filters, 'kvi': self.kvi, 'emotes': self.emotes, 'twitter': self.twitter_manager, 'decks': self.decks, } # Commitable managers self.commitable = { 'commands': self.commands, 'filters': self.filters, 'kvi': self.kvi, 'emotes': self.emotes, 'twitter': self.twitter_manager, 'decks': self.decks, 'users': self.users, 'banphrases': self.banphrase_manager, } self.execute_every(10 * 60, self.commit_all) self.execute_every(30, lambda: self.connection_manager.get_main_conn().ping('tmi.twitch.tv')) try: self.admin = self.config['main']['admin'] except KeyError: log.warning('No admin user specified. See the [main] section in config.example.ini for its usage.') if self.admin: self.users[self.admin].level = 2000 self.parse_version() self.connection_manager = ConnectionManager(self.reactor, self, TMI.message_limit, streamer=self.streamer) chub = self.config['main'].get('control_hub', None) if chub is not None: self.control_hub = ConnectionManager(self.reactor, self, TMI.message_limit, streamer=chub, backup_conns=1) log.info('start pls') else: self.control_hub = None twitch_client_id = None twitch_oauth = None if 'twitchapi' in self.config: twitch_client_id = self.config['twitchapi'].get('client_id', None) twitch_oauth = self.config['twitchapi'].get('oauth', None) self.twitchapi = TwitchAPI(twitch_client_id, twitch_oauth) self.reactor.add_global_handler('all_events', self._dispatcher, -10) self.whisper_manager = WhisperConnectionManager(self.reactor, self, self.streamer, TMI.whispers_message_limit, TMI.whispers_limit_interval) self.whisper_manager.start(accounts=[{'username': self.nickname, 'oauth': self.password, 'can_send_whispers': self.config.getboolean('main', 'add_self_as_whisper_account')}]) self.ascii_timeout_duration = 120 self.msg_length_timeout_duration = 120 self.data = {} self.data_cb = {} self.url_regex = re.compile(self.url_regex_str, re.IGNORECASE) self.data['broadcaster'] = self.streamer self.data['version'] = self.version self.data_cb['status_length'] = self.c_status_length self.data_cb['stream_status'] = self.c_stream_status self.data_cb['bot_uptime'] = self.c_uptime self.data_cb['current_time'] = self.c_current_time self.silent = True if args.silent else self.silent if self.silent: log.info('Silent mode enabled') self.reconnection_interval = 5 """ For actions that need to access the main thread, we can use the mainthread_queue. """ self.mainthread_queue = ActionQueue() self.execute_every(1, self.mainthread_queue.parse_action) self.websocket_manager = WebSocketManager(self) """ Update chatters every `update_chatters_interval' minutes. By default, this is set to run every 5 minutes. """ self.execute_every(self.update_chatters_interval * 60, self.action_queue.add, (self.update_chatters_stage1, )) try: if self.config['twitchapi']['update_subscribers'] == '1': self.execute_every(30 * 60, self.action_queue.add, (self.update_subscribers_stage1, )) except: pass # XXX: TEMPORARY UGLY CODE HandlerManager.add_handler('on_user_gain_tokens', self.on_user_gain_tokens) def on_user_gain_tokens(self, user, tokens_gained): self.whisper(user.username, 'You finished todays quest! You have been awarded with {} tokens.'.format(tokens_gained)) def update_subscribers_stage1(self): limit = 100 offset = 0 subscribers = [] log.info('Starting stage1 subscribers update') try: retry_same = 0 while True: log.debug('Beginning sub request {0} {1}'.format(limit, offset)) subs, retry_same, error = self.twitchapi.get_subscribers(self.streamer, limit, offset, 0 if retry_same is False else retry_same) log.debug('got em!') if error is True: log.error('Too many attempts, aborting') return False if retry_same is False: offset += limit if len(subs) == 0: # We don't need to retry, and the last query finished propery # Break out of the loop and start fiddling with the subs! log.debug('Done!') break else: log.debug('Fetched {0} subs'.format(len(subs))) subscribers.extend(subs) if retry_same is not False: # In case the next attempt is a retry, wait for 3 seconds log.debug('waiting for 3 seconds...') time.sleep(3) log.debug('waited enough!') log.debug('Finished with the while True loop!') except: log.exception('Caught an exception while trying to get subscribers') return log.info('Ended stage1 subscribers update') if len(subscribers) > 0: log.info('Got some subscribers, so we are pushing them to stage 2!') self.mainthread_queue.add(self.update_subscribers_stage2, args=[subscribers]) log.info('Pushed them now.') def update_subscribers_stage2(self, subscribers): log.debug('begiunning stage 2 of update subs') self.kvi['active_subs'].set(len(subscribers) - 1) log.debug('Bulk loading subs...') loaded_subscribers = self.users.bulk_load(subscribers) log.debug('ok!') log.debug('settings all loaded users as non-subs') self.users.reset_subs() """ for username, user in self.users.items(): if user.subscriber: user.subscriber = False """ log.debug('ok!, setting loaded subs as subs') for user in loaded_subscribers: user.subscriber = True log.debug('end of stage 2 of update subs') def update_chatters_stage1(self): chatters = self.twitchapi.get_chatters(self.streamer) if len(chatters) > 0: self.mainthread_queue.add(self.update_chatters_stage2, args=[chatters]) def update_chatters_stage2(self, chatters): points = 1 if self.is_online else 0 log.debug('Updating {0} chatters'.format(len(chatters))) u_chatters = self.users.bulk_load(chatters) for user in u_chatters: if self.is_online: user.minutes_in_chat_online += self.update_chatters_interval else: user.minutes_in_chat_offline += self.update_chatters_interval num_points = points if user.subscriber: num_points *= 5 if self.streamer == 'forsenlol' and 'trump_sub' in user.tags: num_points *= 0.5 user.touch(num_points) def _dispatcher(self, connection, event): if connection == self.connection_manager.get_main_conn() or connection in self.whisper_manager or (self.control_hub is not None and connection == self.control_hub.get_main_conn()): method = getattr(self, 'on_' + event.type, do_nothing) method(connection, event) def start(self): """Start the IRC client.""" self.reactor.process_forever() def get_kvi_value(self, key, extra={}): if key in self.kvi.data: # We check if the value exists first. # We don't want to create a bunch of unneccesary KVIData's return self.kvi[key].get() return 0 def get_last_tweet(self, key, extra={}): return self.twitter_manager.get_last_tweet(key) def get_emote_tm(self, key, extra={}): emote = self.emotes.find(key) if emote: return emote.tm return None def get_emote_count(self, key, extra={}): emote = self.emotes.find(key) if emote: return '{0:,d}'.format(emote.count) return None def get_emote_tm_record(self, key, extra={}): emote = self.emotes.find(key) if emote: return '{0:,d}'.format(emote.tm_record) return None def get_source_value(self, key, extra={}): try: return getattr(extra['source'], key) except: log.exception('Caught exception in get_source_value') return None def get_user_value(self, key, extra={}): try: user = self.users.find(extra['argument']) if user: return getattr(user, key) except: log.exception('Caught exception in get_source_value') return None def get_usersource_value(self, key, extra={}): try: user = self.users.find(extra['argument']) if user: return getattr(user, key) else: return getattr(extra['source'], key) except: log.exception('Caught exception in get_source_value') return None def get_time_value(self, key, extra={}): try: tz = timezone(key) return datetime.datetime.now(tz).strftime(self.date_fmt) except: log.exception('Unhandled exception in get_time_value') return None def get_current_song_value(self, key, extra={}): if self.stream_manager.online: current_song = PleblistManager.get_current_song(self.stream_manager.current_stream.id) inner_keys = key.split('.') val = current_song for inner_key in inner_keys: val = getattr(val, inner_key, None) if val is None: return None if val is not None: return val return None def get_strictargs_value(self, key, extra={}): ret = self.get_args_value(key, extra) if len(ret) == 0: return None return ret def get_args_value(self, key, extra={}): range = None try: msg_parts = extra['message'].split(' ') except (KeyError, AttributeError): msg_parts = [''] try: if '-' in key: range_str = key.split('-') if len(range_str) == 2: range = (int(range_str[0]), int(range_str[1])) if range is None: range = (int(key), len(msg_parts)) except (TypeError, ValueError): range = (0, len(msg_parts)) try: return ' '.join(msg_parts[range[0]:range[1]]) except AttributeError: return '' except: log.exception('UNHANDLED ERROR IN get_args_value') return '' def get_notify_value(self, key, extra={}): payload = { 'message': extra['message'] or '', 'trigger': extra['trigger'], 'user': extra['source'].username_raw, } self.websocket_manager.emit('notify', payload) return '' def get_value(self, key, extra={}): if key in extra: return extra[key] elif key in self.data: return self.data[key] elif key in self.data_cb: return self.data_cb[key]() log.warning('Unknown key passed to get_value: {0}'.format(key)) return None def privmsg(self, message, channel=None, increase_message=True): try: if channel is None: channel = self.channel if self.control_hub is not None and self.control_hub.channel == channel: self.control_hub.privmsg(channel, message) else: self.connection_manager.privmsg(channel, message, increase_message=increase_message) except Exception: log.exception('Exception caught while sending privmsg') def c_uptime(self): return time_since(datetime.datetime.now().timestamp(), self.start_time.timestamp()) def c_current_time(self): return datetime.datetime.now() @property def is_online(self): return self.stream_manager.online def c_stream_status(self): return 'online' if self.stream_manager.online else 'offline' def c_status_length(self): if self.stream_manager.online: return time_since(time.time(), self.stream_manager.current_stream.stream_start.timestamp()) else: if self.stream_manager.last_stream is not None: return time_since(time.time(), self.stream_manager.last_stream.stream_end.timestamp()) else: return 'No recorded stream FeelsBadMan ' def _ban(self, username): self.privmsg('.ban {0}'.format(username), increase_message=False) def execute_at(self, at, function, arguments=()): self.reactor.execute_at(at, function, arguments) def execute_delayed(self, delay, function, arguments=()): self.reactor.execute_delayed(delay, function, arguments) def execute_every(self, period, function, arguments=()): self.reactor.execute_every(period, function, arguments) def ban(self, username): log.debug('Banning {}'.format(username)) self._timeout(username, 30) self.execute_delayed(1, self._ban, (username, )) def ban_user(self, user): if not user.ban_immune: self._timeout(user.username, 30) self.execute_delayed(1, self._ban, (user.username, )) def unban(self, username): self.privmsg('.unban {0}'.format(username), increase_message=False) def _timeout(self, username, duration): self.privmsg('.timeout {0} {1}'.format(username, duration), increase_message=False) def timeout(self, username, duration): log.debug('Timing out {} for {} seconds'.format(username, duration)) self._timeout(username, duration) self.execute_delayed(1, self._timeout, (username, duration)) def timeout_warn(self, user, duration): duration, punishment = user.timeout(duration, warning_module=self.module_manager['warning']) if not user.ban_immune: self.timeout(user.username, duration) return (duration, punishment) return (0, punishment) def timeout_user(self, user, duration): if not user.ban_immune: self._timeout(user.username, duration) self.execute_delayed(1, self._timeout, (user.username, duration)) def whisper(self, username, *messages, separator='. '): """ Takes a sequence of strings and concatenates them with separator. Then sends that string as a whisper to username """ if len(messages) < 0: return False if self.whisper_manager: self.whisper_manager.whisper(username, separator.join(messages)) else: log.debug('No whisper conn set up.') def say(self, *messages, channel=None, separator='. '): """ Takes a sequence of strings and concatenates them with separator. Then sends that string to the given channel. """ if len(messages) < 0: return False if not self.silent: message = separator.join(messages).strip() if len(message) >= 1: if (message[0] == '.' or message[0] == '/') and not message[1:3] == 'me': log.warning('Message we attempted to send started with . or /, skipping.') return log.info('Sending message: {0}'.format(message)) self.privmsg(message[:510], channel) def me(self, message, channel=None): if not self.silent: message = message.strip() if len(message) >= 1: if message[0] == '.' or message[0] == '/': log.warning('Message we attempted to send started with . or /, skipping.') return log.info('Sending message: {0}'.format(message)) self.privmsg('.me ' + message[:500], channel) def parse_version(self): self.version = self.version if self.dev: try: current_branch = subprocess.check_output(['git', 'rev-parse', '--abbrev-ref', 'HEAD']).decode('utf8').strip() latest_commit = subprocess.check_output(['git', 'rev-parse', 'HEAD']).decode('utf8').strip()[:8] commit_number = subprocess.check_output(['git', 'rev-list', 'HEAD', '--count']).decode('utf8').strip() self.version = '{0} DEV ({1}, {2}, commit {3})'.format(self.version, current_branch, latest_commit, commit_number) except: pass def on_welcome(self, chatconn, event): if chatconn in self.whisper_manager: log.debug('Connected to Whisper server.') else: log.debug('Connected to IRC server.') def connect(self): return self.connection_manager.start() def on_disconnect(self, chatconn, event): if chatconn in self.whisper_manager: log.debug('Whispers: Disconnecting from Whisper server') self.whisper_manager.on_disconnect(chatconn) else: log.debug('Disconnected from IRC server') self.connection_manager.on_disconnect(chatconn) def parse_message(self, msg_raw, source, event, tags={}, whisper=False): msg_lower = msg_raw.lower() if source is None: log.error('No valid user passed to parse_message') return False if source.banned: self.ban(source.username) return False # If a user types when timed out, we assume he's been unbanned for a good reason and remove his flag. if source.timed_out is True: source.timed_out = False message_emotes = [] for tag in tags: if tag['key'] == 'subscriber' and event.target == self.channel: if source.subscriber and tag['value'] == '0': source.subscriber = False elif not source.subscriber and tag['value'] == '1': source.subscriber = True elif tag['key'] == 'emotes' and tag['value']: emote_data = tag['value'].split('/') for emote in emote_data: try: emote_id, emote_occurrence = emote.split(':') emote_indices = emote_occurrence.split(',') emote_count = len(emote_indices) emote = self.emotes[int(emote_id)] first_index, last_index = emote_indices[0].split('-') first_index = int(first_index) last_index = int(last_index) emote_code = msg_raw[first_index:last_index + 1] if emote_code[0] == ':': emote_code = emote_code.upper() message_emotes.append({ 'code': emote_code, 'twitch_id': emote_id, 'start': first_index, 'end': last_index, }) tag_as = None if emote_code.startswith('trump'): tag_as = 'trump_sub' elif emote_code.startswith('eloise'): tag_as = 'eloise_sub' elif emote_code.startswith('forsen'): tag_as = 'forsen_sub' elif emote_code.startswith('nostam'): tag_as = 'nostam_sub' elif emote_code.startswith('reynad'): tag_as = 'reynad_sub' elif emote_code.startswith('athene'): tag_as = 'athene_sub' elif emote_id in [12760, 35600, 68498, 54065, 59411, 59412, 59413, 62683, 70183, 70181, 68499, 70429, 70432, 71432, 71433]: tag_as = 'massan_sub' if tag_as is not None: if source.tag_as(tag_as) is True: self.execute_delayed(60 * 60 * 24, source.remove_tag, (tag_as, )) if emote.id is None and emote.code is None: # The emote we just detected is new, set its code. emote.code = emote_code if emote.code not in self.emotes: self.emotes[emote.code] = emote emote.add(emote_count, self.reactor) except: log.exception('Exception caught while splitting emote data') log.error('Emote data: {}'.format(emote_data)) log.error('msg_raw: {}'.format(msg_raw)) elif tag['key'] == 'display-name' and tag['value']: try: source.update_username(tag['value']) except: log.exception('Exception caught while updating a users username') elif tag['key'] == 'user-type': source.moderator = tag['value'] == 'mod' or source.username == self.streamer for emote in self.emotes.custom_data: num = 0 for match in emote.regex.finditer(msg_raw): num += 1 message_emotes.append({ 'code': emote.code, 'bttv_hash': emote.emote_hash, 'start': match.span()[0], 'end': match.span()[1] - 1, # don't ask me }) if num > 0: emote.add(num, self.reactor) urls = self.find_unique_urls(msg_raw) log.debug('{2}{0}: {1}'.format(source.username, msg_raw, '<w>' if whisper else '')) res = HandlerManager.trigger('on_message', source, msg_raw, message_emotes, whisper, urls, event, stop_on_false=True) if res is False: return False source.last_seen = datetime.datetime.now() source.last_active = datetime.datetime.now() if source.ignored: return False if msg_lower[:1] == '!': msg_lower_parts = msg_lower.split(' ') trigger = msg_lower_parts[0][1:] msg_raw_parts = msg_raw.split(' ') remaining_message = ' '.join(msg_raw_parts[1:]) if len(msg_raw_parts) > 1 else None if trigger in self.commands: command = self.commands[trigger] extra_args = { 'emotes': message_emotes, 'trigger': trigger, } command.run(self, source, remaining_message, event=event, args=extra_args, whisper=whisper) def on_whisper(self, chatconn, event): # We use .lower() in case twitch ever starts sending non-lowercased usernames source = self.users[event.source.user.lower()] self.parse_message(event.arguments[0], source, event, whisper=True, tags=event.tags) def on_ping(self, chatconn, event): # self.say('Received a ping. Last ping received {} ago'.format(time_since(datetime.datetime.now().timestamp(), self.last_ping.timestamp()))) log.info('Received a ping. Last ping received {} ago'.format(time_since(datetime.datetime.now().timestamp(), self.last_ping.timestamp()))) self.last_ping = datetime.datetime.now() def on_pong(self, chatconn, event): # self.say('Received a pong. Last pong received {} ago'.format(time_since(datetime.datetime.now().timestamp(), self.last_pong.timestamp()))) log.info('Received a pong. Last pong received {} ago'.format(time_since(datetime.datetime.now().timestamp(), self.last_pong.timestamp()))) self.last_pong = datetime.datetime.now() def on_pubnotice(self, chatconn, event): type = 'whisper' if chatconn in self.whisper_manager else 'normal' log.debug('NOTICE {}@{}: {}'.format(type, event.target, event.arguments)) def on_action(self, chatconn, event): self.on_pubmsg(chatconn, event) def on_pubmsg(self, chatconn, event): if event.source.user == self.nickname: return False # We use .lower() in case twitch ever starts sending non-lowercased usernames source = self.users[event.source.user.lower()] res = HandlerManager.trigger('on_pubmsg', source, event.arguments[0], stop_on_false=True) if res is False: return False self.parse_message(event.arguments[0], source, event, tags=event.tags) @time_method def reload_all(self): log.info('Reloading all...') for key, manager in self.reloadable.items(): log.debug('Reloading {0}'.format(key)) manager.reload() log.debug('Done with {0}'.format(key)) log.info('ok!') @time_method def commit_all(self): log.info('Commiting all...') for key, manager in self.commitable.items(): log.info('Commiting {0}'.format(key)) manager.commit() log.info('Done with {0}'.format(key)) log.info('ok!') HandlerManager.trigger('on_commit', stop_on_false=False) def quit(self, message, event, **options): quit_chub = self.config['main'].get('control_hub', None) quit_delay = 1 if quit_chub is not None and event.target == ('#{}'.format(quit_chub)): quit_delay_random = 300 try: if message is not None and int(message.split()[0]) >= 1: quit_delay_random = int(message.split()[0]) except (IndexError, ValueError, TypeError): pass quit_delay = random.randint(0, quit_delay_random) log.info('{} is restarting in {} seconds.'.format(self.nickname, quit_delay)) self.execute_delayed(quit_delay, self.quit_bot) def quit_bot(self, **options): self.commit_all() if self.phrases['quit']: phrase_data = { 'nickname': self.nickname, 'version': self.version, } try: self.say(self.phrases['quit'].format(**phrase_data)) except Exception: log.exception('Exception caught while trying to say quit phrase') self.twitter_manager.quit() self.socket_manager.quit() if self.whisper_manager: self.whisper_manager.quit() sys.exit(0) def apply_filter(self, resp, filter): available_filters = { 'strftime': _filter_strftime, 'lower': lambda var, args: var.lower(), 'upper': lambda var, args: var.upper(), 'time_since_minutes': lambda var, args: 'no time' if var == 0 else time_since(var * 60, 0, format='long'), 'time_since': lambda var, args: 'no time' if var == 0 else time_since(var, 0, format='long'), 'time_since_dt': _filter_time_since_dt, 'urlencode': lambda var, args: urllib.parse.urlencode(var), 'join': _filter_join, 'number_format': _filter_number_format, } if filter.name in available_filters: return available_filters[filter.name](resp, filter.arguments) return resp def find_unique_urls(self, message): from pajbot.modules.linkchecker import find_unique_urls return find_unique_urls(self.url_regex, message)
class FollowAgeModule(BaseModule): ID = __name__.split('.')[-1] NAME = 'Follow age' DESCRIPTION = 'Makes two commands available: !followage and !followsince' CATEGORY = 'Feature' SETTINGS = [ ModuleSetting(key='action_followage', label='MessageAction for !followage', type='options', required=True, default='say', options=['say', 'whisper', 'reply']), ModuleSetting(key='action_followsince', label='MessageAction for !followsince', type='options', required=True, default='say', options=['say', 'whisper', 'reply']), ] def __init__(self): super().__init__() self.action_queue = ActionQueue() self.action_queue.start() def load_commands(self, **options): # TODO: Have delay modifiable in settings self.commands['followage'] = pajbot.models.command.Command.raw_command( self.follow_age, delay_all=4, delay_user=8, description='Check your or someone elses follow age for a channel', can_execute_with_whisper=True, examples=[ pajbot.models.command.CommandExample( None, 'Check your own follow age', chat='user:!followage\n' 'bot:pajlada, you have been following Karl_Kons for 4 months and 24 days', description= 'Check how long you have been following the current streamer (Karl_Kons in this case)' ).parse(), pajbot.models.command.CommandExample( None, 'Check someone elses follow age', chat='user:!followage NightNacht\n' 'bot:pajlada, NightNacht has been following Karl_Kons for 5 months and 4 days', description= 'Check how long any user has been following the current streamer (Karl_Kons in this case)' ).parse(), pajbot.models.command.CommandExample( None, 'Check someones follow age for a certain streamer', chat='user:!followage NightNacht forsenlol\n' 'bot:pajlada, NightNacht has been following forsenlol for 1 year and 4 months', description= 'Check how long NightNacht has been following forsenlol'). parse(), pajbot.models.command.CommandExample( None, 'Check your own follow age for a certain streamer', chat='user:!followage pajlada forsenlol\n' 'bot:pajlada, you have been following forsenlol for 1 year and 3 months', description= 'Check how long you have been following forsenlol').parse( ), ], ) self.commands['followsince'] = pajbot.models.command.Command.raw_command( self.follow_since, delay_all=4, delay_user=8, description= 'Check from when you or someone else first followed a channel', can_execute_with_whisper=True, examples=[ pajbot.models.command.CommandExample( None, 'Check your own follow since', chat='user:!followsince\n' 'bot:pajlada, you have been following Karl_Kons since 04 March 2015, 07:02:01 UTC', description= 'Check when you first followed the current streamer (Karl_Kons in this case)' ).parse(), pajbot.models.command.CommandExample( None, 'Check someone elses follow since', chat='user:!followsince NightNacht\n' 'bot:pajlada, NightNacht has been following Karl_Kons since 03 July 2014, 04:12:42 UTC', description= 'Check when NightNacht first followed the current streamer (Karl_Kons in this case)' ).parse(), pajbot.models.command.CommandExample( None, 'Check someone elses follow since for another streamer', chat='user:!followsince NightNacht forsenlol\n' 'bot:pajlada, NightNacht has been following forsenlol since 13 June 2013, 13:10:51 UTC', description= 'Check when NightNacht first followed the given streamer (forsenlol)' ).parse(), pajbot.models.command.CommandExample( None, 'Check your follow since for another streamer', chat='user:!followsince pajlada forsenlol\n' 'bot:pajlada, you have been following forsenlol since 16 December 1990, 03:06:51 UTC', description= 'Check when you first followed the given streamer (forsenlol)' ).parse(), ], ) def check_follow_age(self, bot, source, username, streamer, event): streamer = bot.streamer if streamer is None else streamer.lower() age = bot.twitchapi.get_follow_relationship(username, streamer) is_self = source.username == username message = '' if age: # Following human_age = time_since( datetime.datetime.now().timestamp() - age.timestamp(), 0) suffix = 'been following {} for {}'.format(streamer, human_age) if is_self: message = 'You have ' + suffix else: message = username + ' has ' + suffix else: # Not following suffix = 'not following {}'.format(streamer) if is_self: message = 'You are ' + suffix else: message = username + ' is ' + suffix bot.send_message_to_user(source, message, event, method=self.settings['action_followage']) def check_follow_since(self, bot, source, username, streamer, event): streamer = bot.streamer if streamer is None else streamer.lower() follow_since = bot.twitchapi.get_follow_relationship( username, streamer) is_self = source.username == username message = '' if follow_since: # Following human_age = follow_since.strftime('%d %B %Y, %X') suffix = 'been following {} since {} UTC'.format( streamer, human_age) if is_self: message = 'You have ' + suffix else: message = username + ' has ' + suffix else: # Not following suffix = 'not following {}'.format(streamer) if is_self: message = 'You are ' + suffix else: message = username + ' is ' + suffix bot.send_message_to_user(source, message, event, method=self.settings['action_followsince']) def follow_age(self, **options): source = options['source'] message = options['message'] bot = options['bot'] event = options['event'] username, streamer = self.parse_message(bot, source, message) self.action_queue.add(self.check_follow_age, args=[bot, source, username, streamer, event]) def follow_since(self, **options): bot = options['bot'] source = options['source'] message = options['message'] event = options['event'] username, streamer = self.parse_message(bot, source, message) self.action_queue.add(self.check_follow_since, args=[bot, source, username, streamer, event]) def parse_message(self, bot, source, message): username = source.username streamer = None if message is not None and len(message) > 0: message_split = message.split(' ') if len(message_split[0]) and message_split[0].replace( '_', '').isalnum(): username = message_split[0].lower() if len(message_split) > 1 and message_split[1].replace( '_', '').isalnum(): streamer = message_split[1] return username, streamer
class Bot: """ Main class for the twitch bot """ version = '2.7.3' date_fmt = '%H:%M' admin = None url_regex_str = r'\(?(?:(http|https):\/\/)?(?:((?:[^\W\s]|\.|-|[:]{1})+)@{1})?((?:www.)?(?:[^\W\s]|\.|-)+[\.][^\W\s]{2,4}|localhost(?=\/)|\d{1,3}\.\d{1,3}\.\d{1,3}\.\d{1,3})(?::(\d*))?([\/]?[^\s\?]*[\/]{1})*(?:\/?([^\s\n\?\[\]\{\}\#]*(?:(?=\.)){1}|[^\s\n\?\[\]\{\}\.\#]*)?([\.]{1}[^\s\?\#]*)?)?(?:\?{1}([^\s\n\#\[\]]*))?([\#][^\s\n]*)?\)?' def parse_args(): parser = argparse.ArgumentParser() parser.add_argument('--config', '-c', default='config.ini', help='Specify which config file to use ' '(default: config.ini)') parser.add_argument('--silent', action='count', help='Decides whether the bot should be ' 'silent or not') # TODO: Add a log level argument. return parser.parse_args() def load_default_phrases(self): default_phrases = { 'welcome': False, 'quit': False, 'nl': '{username} has typed {num_lines} messages in this channel!', 'nl_0': '{username} has not typed any messages in this channel BibleThump', 'nl_pos': '{username} is rank {nl_pos} line-farmer in this channel!', 'new_sub': 'Sub hype! {username} just subscribed PogChamp', 'resub': 'Resub hype! {username} just subscribed, {num_months} months in a row PogChamp <3 PogChamp', 'point_pos': '{username_w_verb} rank {point_pos} point-hoarder in this channel with {points} points.', } if 'phrases' in self.config: self.phrases = {} for phrase_key, phrase_value in self.config['phrases'].items(): if len(phrase_value.strip()) <= 0: self.phrases[phrase_key] = False else: self.phrases[phrase_key] = phrase_value for phrase_key, phrase_value in default_phrases.items(): if phrase_key not in self.phrases: self.phrases[phrase_key] = phrase_value else: self.phrases = default_phrases def load_config(self, config): self.config = config self.nickname = config['main'].get('nickname', 'pajbot') self.password = config['main'].get('password', 'abcdef') self.timezone = config['main'].get('timezone', 'UTC') self.trusted_mods = config.getboolean('main', 'trusted_mods') TimeManager.init_timezone(self.timezone) if 'streamer' in config['main']: self.streamer = config['main']['streamer'] self.channel = '#' + self.streamer elif 'target' in config['main']: self.channel = config['main']['target'] self.streamer = self.channel[1:] self.wolfram = None try: if 'wolfram' in config['main']: import wolframalpha self.wolfram = wolframalpha.Client(config['main']['wolfram']) except: pass self.silent = False self.dev = False if 'flags' in config: self.silent = True if 'silent' in config['flags'] and config[ 'flags']['silent'] == '1' else self.silent self.dev = True if 'dev' in config['flags'] and config['flags'][ 'dev'] == '1' else self.dev DBManager.init(self.config['main']['db']) redis_options = {} if 'redis' in config: log.info(config._sections['redis']) redis_options = config._sections['redis'] RedisManager.init(**redis_options) def __init__(self, config, args=None): self.load_config(config) self.last_ping = datetime.datetime.now() self.last_pong = datetime.datetime.now() self.load_default_phrases() self.db_session = DBManager.create_session() try: subprocess.check_call( ['alembic', 'upgrade', 'head'] + ['--tag="{0}"'.format(' '.join(sys.argv[1:]))]) except subprocess.CalledProcessError: log.exception('aaaa') log.error( 'Unable to call `alembic upgrade head`, this means the database could be out of date. Quitting.' ) sys.exit(1) except PermissionError: log.error( 'No permission to run `alembic upgrade head`. This means your user probably doesn\'t have execution rights on the `alembic` binary.' ) log.error( 'The error can also occur if it can\'t find `alembic` in your PATH, and instead tries to execute the alembic folder.' ) sys.exit(1) except FileNotFoundError: log.error( 'Could not found an installation of alembic. Please install alembic to continue.' ) sys.exit(1) except: log.exception('Unhandled exception when calling db update') sys.exit(1) # Actions in this queue are run in a separate thread. # This means actions should NOT access any database-related stuff. self.action_queue = ActionQueue() self.action_queue.start() self.reactor = irc.client.Reactor(self.on_connect) self.start_time = datetime.datetime.now() ActionParser.bot = self HandlerManager.init_handlers() self.socket_manager = SocketManager(self) self.stream_manager = StreamManager(self) StreamHelper.init_bot(self, self.stream_manager) ScheduleManager.init() self.users = UserManager() self.decks = DeckManager() self.module_manager = ModuleManager(self.socket_manager, bot=self).load() self.commands = CommandManager(socket_manager=self.socket_manager, module_manager=self.module_manager, bot=self).load() self.filters = FilterManager().reload() self.banphrase_manager = BanphraseManager(self).load() self.timer_manager = TimerManager(self).load() self.kvi = KVIManager() self.emotes = EmoteManager(self).reload() self.twitter_manager = TwitterManager(self) self.duel_manager = DuelManager(self) HandlerManager.trigger('on_managers_loaded') # Reloadable managers self.reloadable = { 'filters': self.filters, 'emotes': self.emotes, } # Commitable managers self.commitable = { 'commands': self.commands, 'filters': self.filters, 'emotes': self.emotes, 'users': self.users, 'banphrases': self.banphrase_manager, } self.execute_every(10 * 60, self.commit_all) try: self.admin = self.config['main']['admin'] except KeyError: log.warning( 'No admin user specified. See the [main] section in config.example.ini for its usage.' ) if self.admin: self.users[self.admin].level = 2000 self.parse_version() relay_host = self.config['main'].get('relay_host', None) relay_password = self.config['main'].get('relay_password', None) if relay_host is None or relay_password is None: self.irc = MultiIRCManager(self) else: self.irc = SingleIRCManager(self) self.reactor.add_global_handler('all_events', self.irc._dispatcher, -10) twitch_client_id = None twitch_oauth = None if 'twitchapi' in self.config: twitch_client_id = self.config['twitchapi'].get('client_id', None) twitch_oauth = self.config['twitchapi'].get('oauth', None) self.twitchapi = TwitchAPI(twitch_client_id, twitch_oauth) self.ascii_timeout_duration = 120 self.msg_length_timeout_duration = 120 self.data = {} self.data_cb = {} self.url_regex = re.compile(self.url_regex_str, re.IGNORECASE) self.data['broadcaster'] = self.streamer self.data['version'] = self.version self.data_cb['status_length'] = self.c_status_length self.data_cb['stream_status'] = self.c_stream_status self.data_cb['bot_uptime'] = self.c_uptime self.data_cb['current_time'] = self.c_current_time self.silent = True if args.silent else self.silent if self.silent: log.info('Silent mode enabled') self.reconnection_interval = 5 """ For actions that need to access the main thread, we can use the mainthread_queue. """ self.mainthread_queue = ActionQueue() self.execute_every(1, self.mainthread_queue.parse_action) self.websocket_manager = WebSocketManager(self) try: if self.config['twitchapi']['update_subscribers'] == '1': self.execute_every(30 * 60, self.action_queue.add, (self.update_subscribers_stage1, )) except: pass # XXX: TEMPORARY UGLY CODE HandlerManager.add_handler('on_user_gain_tokens', self.on_user_gain_tokens) def on_connect(self, sock): return self.irc.on_connect(sock) def on_user_gain_tokens(self, user, tokens_gained): self.whisper( user.username, 'You finished todays quest! You have been awarded with {} tokens.'. format(tokens_gained)) def update_subscribers_stage1(self): limit = 100 offset = 0 subscribers = [] log.info('Starting stage1 subscribers update') try: retry_same = 0 while True: log.debug('Beginning sub request {0} {1}'.format( limit, offset)) subs, retry_same, error = self.twitchapi.get_subscribers( self.streamer, limit, offset, 0 if retry_same is False else retry_same) log.debug('got em!') if error is True: log.error('Too many attempts, aborting') return False if retry_same is False: offset += limit if len(subs) == 0: # We don't need to retry, and the last query finished propery # Break out of the loop and start fiddling with the subs! log.debug('Done!') break else: log.debug('Fetched {0} subs'.format(len(subs))) subscribers.extend(subs) if retry_same is not False: # In case the next attempt is a retry, wait for 3 seconds log.debug('waiting for 3 seconds...') time.sleep(3) log.debug('waited enough!') log.debug('Finished with the while True loop!') except: log.exception( 'Caught an exception while trying to get subscribers') return log.info('Ended stage1 subscribers update') if len(subscribers) > 0: log.info( 'Got some subscribers, so we are pushing them to stage 2!') self.mainthread_queue.add(self.update_subscribers_stage2, args=[subscribers]) log.info('Pushed them now.') def update_subscribers_stage2(self, subscribers): log.debug('begiunning stage 2 of update subs') self.kvi['active_subs'].set(len(subscribers) - 1) log.debug('Bulk loading subs...') loaded_subscribers = self.users.bulk_load(subscribers) log.debug('ok!') log.debug('settings all loaded users as non-subs') self.users.reset_subs() """ for username, user in self.users.items(): if user.subscriber: user.subscriber = False """ log.debug('ok!, setting loaded subs as subs') for user in loaded_subscribers: user.subscriber = True log.debug('end of stage 2 of update subs') def start(self): """Start the IRC client.""" self.reactor.process_forever() def get_kvi_value(self, key, extra={}): return self.kvi[key].get() def get_last_tweet(self, key, extra={}): return self.twitter_manager.get_last_tweet(key) def get_emote_tm(self, key, extra={}): emote = self.emotes.find(key) if emote: return emote.tm return None def get_emote_count(self, key, extra={}): emote = self.emotes.find(key) if emote: return '{0:,d}'.format(emote.count) return None def get_emote_tm_record(self, key, extra={}): emote = self.emotes.find(key) if emote: return '{0:,d}'.format(emote.tm_record) return None def get_source_value(self, key, extra={}): try: return getattr(extra['source'], key) except: log.exception('Caught exception in get_source_value') return None def get_user_value(self, key, extra={}): try: user = self.users.find(extra['argument']) if user: return getattr(user, key) except: log.exception('Caught exception in get_source_value') return None def get_usersource_value(self, key, extra={}): try: user = self.users.find(extra['argument']) if user: return getattr(user, key) else: return getattr(extra['source'], key) except: log.exception('Caught exception in get_source_value') return None def get_time_value(self, key, extra={}): try: tz = timezone(key) return datetime.datetime.now(tz).strftime(self.date_fmt) except: log.exception('Unhandled exception in get_time_value') return None def get_current_song_value(self, key, extra={}): if self.stream_manager.online: current_song = PleblistManager.get_current_song( self.stream_manager.current_stream.id) inner_keys = key.split('.') val = current_song for inner_key in inner_keys: val = getattr(val, inner_key, None) if val is None: return None if val is not None: return val return None def get_strictargs_value(self, key, extra={}): ret = self.get_args_value(key, extra) if len(ret) == 0: return None return ret def get_args_value(self, key, extra={}): range = None try: msg_parts = extra['message'].split(' ') except (KeyError, AttributeError): msg_parts = [''] try: if '-' in key: range_str = key.split('-') if len(range_str) == 2: range = (int(range_str[0]), int(range_str[1])) if range is None: range = (int(key), len(msg_parts)) except (TypeError, ValueError): range = (0, len(msg_parts)) try: return ' '.join(msg_parts[range[0]:range[1]]) except AttributeError: return '' except: log.exception('UNHANDLED ERROR IN get_args_value') return '' def get_notify_value(self, key, extra={}): payload = { 'message': extra['message'] or '', 'trigger': extra['trigger'], 'user': extra['source'].username_raw, } self.websocket_manager.emit('notify', payload) return '' def get_value(self, key, extra={}): if key in extra: return extra[key] elif key in self.data: return self.data[key] elif key in self.data_cb: return self.data_cb[key]() log.warning('Unknown key passed to get_value: {0}'.format(key)) return None def privmsg(self, message, channel=None, increase_message=True): if channel is None: channel = self.channel return self.irc.privmsg(message, channel, increase_message=increase_message) def c_uptime(self): return time_since(datetime.datetime.now().timestamp(), self.start_time.timestamp()) def c_current_time(self): return datetime.datetime.now() @property def is_online(self): return self.stream_manager.online def c_stream_status(self): return 'online' if self.stream_manager.online else 'offline' def c_status_length(self): if self.stream_manager.online: return time_since( time.time(), self.stream_manager.current_stream.stream_start.timestamp()) else: if self.stream_manager.last_stream is not None: return time_since( time.time(), self.stream_manager.last_stream.stream_end.timestamp()) else: return 'No recorded stream FeelsBadMan ' def _ban(self, username): self.privmsg('.ban {0}'.format(username), increase_message=False) def execute_at(self, at, function, arguments=()): self.reactor.execute_at(at, function, arguments) def execute_delayed(self, delay, function, arguments=()): self.reactor.execute_delayed(delay, function, arguments) def execute_every(self, period, function, arguments=()): self.reactor.execute_every(period, function, arguments) def ban(self, username): log.debug('Banning {}'.format(username)) self._timeout(username, 30) self.execute_delayed(1, self._ban, (username, )) def ban_user(self, user): if not user.ban_immune: self._timeout(user.username, 30) self.execute_delayed(1, self._ban, (user.username, )) def unban(self, username): self.privmsg('.unban {0}'.format(username), increase_message=False) def _timeout(self, username, duration): self.privmsg('.timeout {0} {1}'.format(username, duration), increase_message=False) def timeout(self, username, duration): log.debug('Timing out {} for {} seconds'.format(username, duration)) self._timeout(username, duration) self.execute_delayed(1, self._timeout, (username, duration)) def timeout_warn(self, user, duration): duration, punishment = user.timeout( duration, warning_module=self.module_manager['warning']) if not user.ban_immune: self.timeout(user.username, duration) return (duration, punishment) return (0, punishment) def timeout_user(self, user, duration): if not user.ban_immune: self._timeout(user.username, duration) self.execute_delayed(1, self._timeout, (user.username, duration)) def whisper(self, username, *messages, separator='. '): """ Takes a sequence of strings and concatenates them with separator. Then sends that string as a whisper to username """ if len(messages) < 0: return False message = separator.join(messages) return self.irc.whisper(username, message) def say(self, *messages, channel=None, separator='. '): """ Takes a sequence of strings and concatenates them with separator. Then sends that string to the given channel. """ if len(messages) < 0: return False if not self.silent: message = separator.join(messages).strip() if len(message) >= 1: if (message[0] == '.' or message[0] == '/') and not message[1:3] == 'me': log.warning( 'Message we attempted to send started with . or /, skipping.' ) return log.info('Sending message: {0}'.format(message)) self.privmsg(message[:510], channel) def me(self, message, channel=None): if not self.silent: message = message.strip() if len(message) >= 1: if message[0] == '.' or message[0] == '/': log.warning( 'Message we attempted to send started with . or /, skipping.' ) return log.info('Sending message: {0}'.format(message)) self.privmsg('.me ' + message[:500], channel) def parse_version(self): self.version = self.version if self.dev: try: current_branch = subprocess.check_output( ['git', 'rev-parse', '--abbrev-ref', 'HEAD']).decode('utf8').strip() latest_commit = subprocess.check_output( ['git', 'rev-parse', 'HEAD']).decode('utf8').strip()[:8] commit_number = subprocess.check_output( ['git', 'rev-list', 'HEAD', '--count']).decode('utf8').strip() self.version = '{0} DEV ({1}, {2}, commit {3})'.format( self.version, current_branch, latest_commit, commit_number) except: pass def on_welcome(self, chatconn, event): return self.irc.on_welcome(chatconn, event) def connect(self): return self.irc.start() def on_disconnect(self, chatconn, event): self.irc.on_disconnect(chatconn, event) def parse_message(self, msg_raw, source, event, tags={}, whisper=False): msg_lower = msg_raw.lower() if source is None: log.error('No valid user passed to parse_message') return False if source.banned: self.ban(source.username) return False # If a user types when timed out, we assume he's been unbanned for a good reason and remove his flag. if source.timed_out is True: source.timed_out = False message_emotes = [] for tag in tags: if tag['key'] == 'subscriber' and event.target == self.channel: if source.subscriber and tag['value'] == '0': source.subscriber = False elif not source.subscriber and tag['value'] == '1': source.subscriber = True elif tag['key'] == 'emotes' and tag['value']: emote_data = tag['value'].split('/') for emote in emote_data: try: emote_id, emote_occurrence = emote.split(':') emote_indices = emote_occurrence.split(',') emote_count = len(emote_indices) emote = self.emotes[int(emote_id)] first_index, last_index = emote_indices[0].split('-') first_index = int(first_index) last_index = int(last_index) emote_code = msg_raw[first_index:last_index + 1] if emote_code[0] == ':': emote_code = emote_code.upper() message_emotes.append({ 'code': emote_code, 'twitch_id': emote_id, 'start': first_index, 'end': last_index, }) tag_as = None if emote_code.startswith('trump'): tag_as = 'trump_sub' elif emote_code.startswith('eloise'): tag_as = 'eloise_sub' elif emote_code.startswith('forsen'): tag_as = 'forsen_sub' elif emote_code.startswith('nostam'): tag_as = 'nostam_sub' elif emote_code.startswith('reynad'): tag_as = 'reynad_sub' elif emote_code.startswith('athene'): tag_as = 'athene_sub' elif emote_id in [ 12760, 35600, 68498, 54065, 59411, 59412, 59413, 62683, 70183, 70181, 68499, 70429, 70432, 71432, 71433 ]: tag_as = 'massan_sub' if tag_as is not None: if source.tag_as(tag_as) is True: self.execute_delayed(60 * 60 * 24, source.remove_tag, (tag_as, )) if emote.id is None and emote.code is None: # The emote we just detected is new, set its code. emote.code = emote_code if emote.code not in self.emotes: self.emotes[emote.code] = emote emote.add(emote_count, self.reactor) except: log.exception( 'Exception caught while splitting emote data') log.error('Emote data: {}'.format(emote_data)) log.error('msg_raw: {}'.format(msg_raw)) elif tag['key'] == 'display-name' and tag['value']: try: source.update_username(tag['value']) except: log.exception( 'Exception caught while updating a users username') elif tag['key'] == 'user-type': source.moderator = tag[ 'value'] == 'mod' or source.username == self.streamer for emote in self.emotes.custom_data: num = 0 for match in emote.regex.finditer(msg_raw): num += 1 message_emotes.append({ 'code': emote.code, 'bttv_hash': emote.emote_hash, 'start': match.span()[0], 'end': match.span()[1] - 1, # don't ask me }) if num > 0: emote.add(num, self.reactor) urls = self.find_unique_urls(msg_raw) log.debug('{2}{0}: {1}'.format(source.username, msg_raw, '<w>' if whisper else '')) res = HandlerManager.trigger('on_message', source, msg_raw, message_emotes, whisper, urls, event, stop_on_false=True) if res is False: return False source.last_seen = datetime.datetime.now() source.last_active = datetime.datetime.now() if source.ignored: return False if msg_lower[:1] == '!': msg_lower_parts = msg_lower.split(' ') trigger = msg_lower_parts[0][1:] msg_raw_parts = msg_raw.split(' ') remaining_message = ' '.join( msg_raw_parts[1:]) if len(msg_raw_parts) > 1 else None if trigger in self.commands: command = self.commands[trigger] extra_args = { 'emotes': message_emotes, 'trigger': trigger, } command.run(self, source, remaining_message, event=event, args=extra_args, whisper=whisper) def on_whisper(self, chatconn, event): # We use .lower() in case twitch ever starts sending non-lowercased usernames source = self.users[event.source.user.lower()] self.parse_message(event.arguments[0], source, event, whisper=True, tags=event.tags) def on_ping(self, chatconn, event): # self.say('Received a ping. Last ping received {} ago'.format(time_since(datetime.datetime.now().timestamp(), self.last_ping.timestamp()))) log.info('Received a ping. Last ping received {} ago'.format( time_since(datetime.datetime.now().timestamp(), self.last_ping.timestamp()))) self.last_ping = datetime.datetime.now() def on_pong(self, chatconn, event): # self.say('Received a pong. Last pong received {} ago'.format(time_since(datetime.datetime.now().timestamp(), self.last_pong.timestamp()))) log.info('Received a pong. Last pong received {} ago'.format( time_since(datetime.datetime.now().timestamp(), self.last_pong.timestamp()))) self.last_pong = datetime.datetime.now() def on_pubnotice(self, chatconn, event): return type = 'whisper' if chatconn in self.whisper_manager else 'normal' log.debug('NOTICE {}@{}: {}'.format(type, event.target, event.arguments)) def on_action(self, chatconn, event): self.on_pubmsg(chatconn, event) def on_pubmsg(self, chatconn, event): if event.source.user == self.nickname: return False # We use .lower() in case twitch ever starts sending non-lowercased usernames source = self.users[event.source.user.lower()] res = HandlerManager.trigger('on_pubmsg', source, event.arguments[0], stop_on_false=True) if res is False: return False self.parse_message(event.arguments[0], source, event, tags=event.tags) @time_method def reload_all(self): log.info('Reloading all...') for key, manager in self.reloadable.items(): log.debug('Reloading {0}'.format(key)) manager.reload() log.debug('Done with {0}'.format(key)) log.info('ok!') @time_method def commit_all(self): log.info('Commiting all...') for key, manager in self.commitable.items(): log.info('Commiting {0}'.format(key)) manager.commit() log.info('Done with {0}'.format(key)) log.info('ok!') HandlerManager.trigger('on_commit', stop_on_false=False) def quit(self, message, event, **options): quit_chub = self.config['main'].get('control_hub', None) quit_delay = 1 if quit_chub is not None and event.target == ('#{}'.format(quit_chub)): quit_delay_random = 300 try: if message is not None and int(message.split()[0]) >= 1: quit_delay_random = int(message.split()[0]) except (IndexError, ValueError, TypeError): pass quit_delay = random.randint(0, quit_delay_random) log.info('{} is restarting in {} seconds.'.format( self.nickname, quit_delay)) self.execute_delayed(quit_delay, self.quit_bot) def quit_bot(self, **options): self.commit_all() if self.phrases['quit']: phrase_data = { 'nickname': self.nickname, 'version': self.version, } try: self.say(self.phrases['quit'].format(**phrase_data)) except Exception: log.exception( 'Exception caught while trying to say quit phrase') self.twitter_manager.quit() self.socket_manager.quit() self.irc.quit() sys.exit(0) def apply_filter(self, resp, filter): available_filters = { 'strftime': _filter_strftime, 'lower': lambda var, args: var.lower(), 'upper': lambda var, args: var.upper(), 'time_since_minutes': lambda var, args: 'no time' if var == 0 else time_since(var * 60, 0, format='long'), 'time_since': lambda var, args: 'no time' if var == 0 else time_since(var, 0, format='long'), 'time_since_dt': _filter_time_since_dt, 'urlencode': lambda var, args: urllib.parse.urlencode(var), 'join': _filter_join, 'number_format': _filter_number_format, } if filter.name in available_filters: return available_filters[filter.name](resp, filter.arguments) return resp def find_unique_urls(self, message): from pajbot.modules.linkchecker import find_unique_urls return find_unique_urls(self.url_regex, message)