Limnoria/plugins/RSS.py

431 lines
18 KiB
Python
Raw Normal View History

2003-04-06 13:42:41 +02:00
###
# Copyright (c) 2002-2004, Jeremiah Fincher
2003-04-06 13:42:41 +02:00
# All rights reserved.
#
# Redistribution and use in source and binary forms, with or without
# modification, are permitted provided that the following conditions are met:
#
# * Redistributions of source code must retain the above copyright notice,
# this list of conditions, and the following disclaimer.
# * Redistributions in binary form must reproduce the above copyright notice,
# this list of conditions, and the following disclaimer in the
# documentation and/or other materials provided with the distribution.
# * Neither the name of the author of this software nor the name of
# contributors to this software may be used to endorse or promote products
# derived from this software without specific prior written consent.
#
# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
# AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
# IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
# ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
# LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
# CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
# SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
# INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
# CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
# ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
# POSSIBILITY OF SUCH DAMAGE.
###
"""
Provides basic functionality for handling RSS/RDF feeds.
2003-04-06 13:42:41 +02:00
"""
2004-09-02 06:06:20 +02:00
import supybot
2003-11-25 09:23:47 +01:00
__revision__ = "$Id$"
2004-09-02 06:06:20 +02:00
__author__ = supybot.authors.jemfinch
2003-11-25 09:23:47 +01:00
2004-07-24 07:18:26 +02:00
import supybot.plugins as plugins
2003-04-06 13:42:41 +02:00
2003-12-03 22:21:43 +01:00
import sets
2003-04-06 13:49:53 +02:00
import time
2004-07-21 21:12:57 +02:00
import getopt
import sgmllib
import threading
from itertools import imap
2003-04-06 13:49:53 +02:00
2003-04-06 13:42:41 +02:00
import rssparser
2004-07-24 07:18:26 +02:00
import supybot.conf as conf
import supybot.utils as utils
import supybot.world as world
2004-10-29 03:35:37 +02:00
from supybot.commands import *
2004-07-24 07:18:26 +02:00
import supybot.ircmsgs as ircmsgs
import supybot.ircutils as ircutils
import supybot.privmsgs as privmsgs
import supybot.registry as registry
import supybot.callbacks as callbacks
2003-04-06 13:42:41 +02:00
class AnnouncedFeeds(registry.SpaceSeparatedListOfStrings):
List = callbacks.CanonicalNameSet
conf.registerPlugin('RSS')
conf.registerChannelValue(conf.supybot.plugins.RSS, 'bold', registry.Boolean(
True, """Determines whether the bot will bold the title of the feed when it
announces new news."""))
conf.registerChannelValue(conf.supybot.plugins.RSS, 'headlineSeparator',
registry.StringSurroundedBySpaces(' || ', """Determines what string is used
to separate headlines in new feeds."""))
conf.registerChannelValue(conf.supybot.plugins.RSS, 'announcementPrefix',
registry.StringWithSpaceOnRight('New news from ', """Determines what prefix
is prepended (if any) to the new news item announcements made in the
channel."""))
conf.registerChannelValue(conf.supybot.plugins.RSS, 'announce',
AnnouncedFeeds([], """Determines which RSS feeds should be announced in the
channel; valid input is a list of strings (either registered RSS feeds or
RSS feed URLs) separated by spaces."""))
conf.registerGlobalValue(conf.supybot.plugins.RSS, 'waitPeriod',
registry.PositiveInteger(1800, """Indicates how many seconds the bot will
wait between retrieving RSS feeds; requests made within this period will
return cached results."""))
conf.registerGlobalValue(conf.supybot.plugins.RSS, 'feeds',
AnnouncedFeeds([], """Determines what feeds should be accessible as
commands."""))
conf.registerChannelValue(conf.supybot.plugins.RSS, 'showLinks',
registry.Boolean(False, """Determines whether the bot will list the link
along with the title of the feed when the rss command is called.
supybot.plugins.RSS.announce.showLinks affects whether links will be
listed when a feed is automatically announced."""))
conf.registerGroup(conf.supybot.plugins.RSS, 'announce')
conf.registerChannelValue(conf.supybot.plugins.RSS.announce, 'showLinks',
registry.Boolean(False, """Determines whether the bot will list the link
along with the title of the feed when a feed is automatically
announced."""))
2004-07-21 21:12:57 +02:00
class RSS(callbacks.Privmsg):
"""This plugin is useful both for announcing updates to RSS feeds in a
channel, and for retrieving the headlines of RSS feeds via command. Use
the "add" command to add feeds to this plugin, and use the "announce"
command to determine what feeds should be announced in a given channel."""
2003-04-06 13:42:41 +02:00
threaded = True
def __init__(self):
2004-10-29 03:35:37 +02:00
self.__parent = super(RSS, self)
self.__parent.__init__()
2004-09-02 06:06:20 +02:00
self.feedNames = callbacks.CanonicalNameSet()
self.locks = {}
self.lastRequest = {}
self.cachedFeeds = {}
2004-04-13 20:25:49 +02:00
self.gettingLockLock = threading.Lock()
for name in self.registryValue('feeds'):
self._registerFeed(name)
try:
url = self.registryValue('feeds.%s' % name)
except registry.NonExistentRegistryEntry:
self.log.warning('%s is not a registered feed, removing.',name)
continue
self.makeFeedCommand(name, url)
self.getFeed(url) # So announced feeds don't announce on startup.
def _registerFeed(self, name, url=''):
self.registryValue('feeds').add(name)
group = self.registryValue('feeds', value=False)
group.register(name, registry.String(url, ''))
def __call__(self, irc, msg):
2004-10-29 03:35:37 +02:00
self.__parent.__call__(irc, msg)
irc = callbacks.SimpleProxy(irc, msg)
newFeeds = {}
for channel in irc.state.channels:
feeds = self.registryValue('announce', channel)
for name in feeds:
2004-04-15 16:17:07 +02:00
commandName = callbacks.canonicalName(name)
if self.isCommand(commandName):
name = commandName
url = self.getCommand(name).url
else:
url = name
if self.willGetNewFeed(url):
newFeeds.setdefault((url, name), []).append(channel)
for ((url, name), channels) in newFeeds.iteritems():
# We check if we can acquire the lock right here because if we
# don't, we'll possibly end up spawning a lot of threads to get
# the feed, because this thread may run for a number of bytecodes
# before it switches to a thread that'll get the lock in
# _newHeadlines.
2004-04-13 20:25:49 +02:00
if self.acquireLock(url, blocking=False):
try:
t = threading.Thread(target=self._newHeadlines,
name='Fetching <%s>' % url,
args=(irc, channels, name, url))
2004-12-06 04:29:45 +01:00
self.log.info('Checking for announcements at <%s>', url)
world.threadsSpawned += 1
t.setDaemon(True)
t.start()
finally:
2004-04-13 20:25:49 +02:00
self.releaseLock(url)
2004-04-09 17:13:26 +02:00
time.sleep(0.1) # So other threads can run.
def buildHeadlines(self, headlines, channel, config='announce.showLinks'):
newheadlines = []
if self.registryValue('%s' % config, channel):
for headline in headlines:
if headline[1]:
newheadlines.append('%s <%s>' % headline)
else:
newheadlines.append('%s' % headline[0])
else:
for headline in headlines:
newheadlines = ['%s' % h[0] for h in headlines]
return newheadlines
def _newHeadlines(self, irc, channels, name, url):
try:
# We acquire the lock here so there's only one announcement thread
# in this code at any given time. Otherwise, several announcement
# threads will getFeed (all blocking, in turn); then they'll all
# want to sent their news messages to the appropriate channels.
2004-04-09 17:13:26 +02:00
# Note that we're allowed to acquire this lock twice within the
# same thread because it's an RLock and not just a normal Lock.
2004-04-13 20:25:49 +02:00
self.acquireLock(url)
try:
oldresults = self.cachedFeeds[url]
oldheadlines = self.getHeadlines(oldresults)
except KeyError:
oldheadlines = []
newresults = self.getFeed(url)
newheadlines = self.getHeadlines(newresults)
2004-06-01 18:22:48 +02:00
def canonicalize(headline):
return (tuple(headline[0].lower().split()), headline[1])
2004-06-01 18:22:48 +02:00
oldheadlines = sets.Set(map(canonicalize, oldheadlines))
for (i, headline) in enumerate(newheadlines):
if canonicalize(headline) in oldheadlines:
2004-04-20 11:51:55 +02:00
newheadlines[i] = None
newheadlines = filter(None, newheadlines) # Removes Nones.
if newheadlines:
for channel in channels:
bold = self.registryValue('bold', channel)
sep = self.registryValue('headlineSeparator', channel)
prefix = self.registryValue('announcementPrefix', channel)
pre = '%s%s: ' % (prefix, name)
if bold:
pre = ircutils.bold(pre)
sep = ircutils.bold(sep)
newheadlines = self.buildHeadlines(newheadlines, channel)
irc.replies(newheadlines, prefixer=pre, joiner=sep,
to=channel, prefixName=False, private=True)
finally:
2004-04-13 20:25:49 +02:00
self.releaseLock(url)
2004-07-21 21:12:57 +02:00
def willGetNewFeed(self, url):
now = time.time()
wait = self.registryValue('waitPeriod')
if url not in self.lastRequest or now - self.lastRequest[url] > wait:
return True
else:
return False
2004-02-10 04:14:46 +01:00
2004-04-13 20:25:49 +02:00
def acquireLock(self, url, blocking=True):
try:
self.gettingLockLock.acquire()
try:
lock = self.locks[url]
except KeyError:
lock = threading.RLock()
self.locks[url] = lock
return lock.acquire(blocking=blocking)
finally:
self.gettingLockLock.release()
def releaseLock(self, url):
self.locks[url].release()
2004-07-21 21:12:57 +02:00
def getFeed(self, url):
def error(s):
return {'items': [{'title': s}]}
try:
# This is the most obvious place to acquire the lock, because a
# malicious user could conceivably flood the bot with rss commands
# and DoS the website in question.
2004-04-13 20:25:49 +02:00
self.acquireLock(url)
if self.willGetNewFeed(url):
2004-02-10 04:14:46 +01:00
try:
2004-08-30 05:33:46 +02:00
self.log.debug('Downloading new feed from <%s>', url)
2004-02-10 04:14:46 +01:00
results = rssparser.parse(url)
except sgmllib.SGMLParseError:
self.log.exception('Uncaught exception from rssparser:')
2004-08-19 01:15:27 +02:00
raise callbacks.Error, 'Invalid (unparsable) RSS feed.'
except socket.timeout:
return error('Timeout downloading feed.')
2004-02-10 04:14:46 +01:00
self.cachedFeeds[url] = results
self.lastRequest[url] = time.time()
try:
return self.cachedFeeds[url]
except KeyError:
self.lastRequest[url] = 0
return error('Unable to download feed.')
finally:
2004-04-13 20:25:49 +02:00
self.releaseLock(url)
def getHeadlines(self, feed):
headlines = []
for d in feed['items']:
if 'title' in d:
title = utils.htmlToText(d['title']).strip()
2004-09-02 06:06:20 +02:00
link = d.get('link')
if link:
headlines.append((title, link))
else:
headlines.append((title, None))
return headlines
2004-04-29 13:49:24 +02:00
def _validFeedName(self, name):
if not registry.isValidRegistryName(name):
raise ValueError, name
return callbacks.canonicalName(name)
def makeFeedCommand(self, name, url):
2004-10-29 03:35:37 +02:00
docstring = """[<number of headlines>]
2003-12-03 22:21:43 +01:00
Reports the titles for %s at the RSS feed <%s>. If
<number of headlines> is given, returns only that many headlines.
RSS feeds are only looked up every supybot.plugins.RSS.waitPeriod
seconds, which defaults to 1800 (30 minutes) since that's what most
websites prefer.
2003-12-03 22:21:43 +01:00
""" % (name, url)
2004-04-29 13:49:24 +02:00
assert name == self._validFeedName(name)
2004-04-13 20:25:49 +02:00
if url not in self.locks:
self.locks[url] = threading.RLock()
2004-09-30 10:41:07 +02:00
if hasattr(self.__class__, name) and \
not hasattr(getattr(self, name), 'url'):
2004-09-13 03:07:04 +02:00
s = 'I already have a command in this plugin named %s.' % name
raise callbacks.Error, s
2003-12-03 22:21:43 +01:00
def f(self, irc, msg, args):
args.insert(0, url)
self.rss(irc, msg, args)
f = utils.changeFunctionName(f, name, docstring)
f.url = url # Used by __call__.
2003-12-03 22:21:43 +01:00
self.feedNames.add(name)
setattr(self.__class__, name, f)
self._registerFeed(name, url)
2004-07-21 21:12:57 +02:00
2004-10-29 03:35:37 +02:00
def add(self, irc, msg, args, name, url):
"""<name> <url>
Adds a command to this plugin that will look up the RSS feed at the
given URL.
"""
2004-04-29 13:49:24 +02:00
try:
name = self._validFeedName(name)
except ValueError:
irc.errorInvalid('feed name', name, 'Feed names must not '
'include dots, colons, or spaces.')
self.makeFeedCommand(name, url)
irc.replySuccess()
2004-10-29 03:35:37 +02:00
add = wrap(add, ['something', 'url'])
2003-12-03 22:21:43 +01:00
2004-10-29 03:35:37 +02:00
def remove(self, irc, msg, args, name):
2003-12-03 22:21:43 +01:00
"""<name>
Removes the command for looking up RSS feeds at <name> from
this plugin.
"""
if name not in self.feedNames:
irc.error('That\'s not a valid RSS feed command name.')
2003-12-03 22:21:43 +01:00
return
self.feedNames.remove(name)
2003-12-03 22:21:43 +01:00
delattr(self.__class__, name)
conf.supybot.plugins.RSS.feeds.unregister(name)
irc.replySuccess()
2004-10-29 03:35:37 +02:00
remove = wrap(remove, ['commandName'])
2004-02-07 23:38:49 +01:00
2004-10-29 03:35:37 +02:00
def announce(self, irc, msg, args, channel, optlist, rest):
2004-07-21 21:12:57 +02:00
"""[<channel>] [--remove] [<name|url> ...]
2004-02-07 23:38:49 +01:00
Sets the current list of announced feeds in the channel to the feeds
given. Valid feeds include the names of registered feeds as well as
URLs for a RSS feeds. <channel> is only necessary if the message isn't
2004-07-21 21:12:57 +02:00
sent in the channel itself. If no arguments are specified, replies
with the current list of feeds to announce. If --remove is given,
the specified feeds will be removed from the list of feeds to announce.
2004-02-07 23:38:49 +01:00
"""
2004-07-21 21:12:57 +02:00
remove = False
announce = conf.supybot.plugins.RSS.announce
for (option, _) in optlist:
2004-10-29 03:35:37 +02:00
if option == 'remove':
2004-07-21 21:12:57 +02:00
remove = True
if remove:
if rest:
feeds = announce.get(channel)()
for feed in rest:
if feed in feeds:
feeds.remove(feed)
announce.get(channel).setValue(feeds)
irc.replySuccess()
return
else:
raise callbacks.ArgumentError
elif not rest:
feeds = utils.commaAndify(announce.get(channel)())
irc.reply(feeds or 'I am currently not announcing any feeds.')
return
else:
2004-07-21 21:12:57 +02:00
announce.get(channel).setValue(rest)
irc.replySuccess()
2004-07-21 21:12:57 +02:00
return
2004-10-29 03:35:37 +02:00
announce = wrap(announce, [('checkChannelCapability', 'op'),
getopts({'remove':''}), any('something')])
2004-07-21 21:12:57 +02:00
2004-10-29 03:35:37 +02:00
def rss(self, irc, msg, args, url, n):
"""<url> [<number of headlines>]
Gets the title components of the given RSS feed.
If <number of headlines> is given, return only that many headlines.
"""
self.log.debug('Fetching %s', url)
feed = self.getFeed(url)
if irc.isChannel(msg.args[0]):
channel = msg.args[0]
else:
channel = None
headlines = self.getHeadlines(feed)
if not headlines:
2004-09-28 09:43:55 +02:00
irc.error('Couldn\'t get RSS feed.')
return
headlines = self.buildHeadlines(headlines, channel, 'showLinks')
if n:
headlines = headlines[:n]
sep = self.registryValue('headlineSeparator', channel)
if self.registryValue('bold', channel):
sep = ircutils.bold(sep)
irc.reply(sep.join(headlines))
2004-10-29 03:35:37 +02:00
rss = wrap(rss, ['url', additional('int')])
2003-08-20 18:26:23 +02:00
2004-10-29 03:35:37 +02:00
def info(self, irc, msg, args, url):
"""<url|feed>
Returns information from the given RSS feed, namely the title,
URL, description, and last update date, if available.
"""
try:
url = self.registryValue('feeds.%s' % url)
except registry.NonExistentRegistryEntry:
pass
feed = self.getFeed(url)
info = feed['channel']
if not info:
irc.error('I couldn\'t retrieve that RSS feed.')
return
# check the 'modified' key, if it's there, convert it here first
2003-04-19 23:40:39 +02:00
if 'modified' in feed:
seconds = time.mktime(feed['modified'])
2003-08-28 19:55:01 +02:00
now = time.mktime(time.gmtime())
when = utils.timeElapsed(now - seconds) + ' ago'
else:
2003-04-19 23:40:39 +02:00
when = "time unavailable"
# The rest of the entries are all available in the channel key
2003-04-19 23:40:39 +02:00
response = 'Title: %s; URL: <%s>; ' \
'Description: %s; Last updated %s.' % (
2003-08-20 18:26:23 +02:00
info.get('title', 'unavailable').strip(),
2003-04-19 23:40:39 +02:00
info.get('link', 'unavailable').strip(),
info.get('description', 'unavailable').strip(),
when)
irc.reply(' '.join(response.split()))
2004-10-29 03:35:37 +02:00
info = wrap(info, ['something'])
2003-04-06 13:42:41 +02:00
Class = RSS
# vim:set shiftwidth=4 tabstop=4 expandtab textwidth=78: