X-Git-Url: http://www.chiark.greenend.org.uk/ucgi/~matthewv/git?a=blobdiff_plain;f=commands.py;h=a8499587f222c016b408000b72059b5e7314f431;hb=129d1b9256f7020515e6ccfd2dcdab26d0908eec;hp=cea3a2a6d2d3ef627735184bcfdfab3cce978a9f;hpb=632390ccdb0fba803e44b341c22c994cfbe5d0ae;p=irc.git diff --git a/commands.py b/commands.py index cea3a2a..a849958 100755 --- a/commands.py +++ b/commands.py @@ -1,8 +1,13 @@ # Part of Acrobat. -import string, cPickle, random, urllib, sys, time, re, os, twitter, subprocess, datetime, urlparse +import string, cPickle, random, urllib, sys, time, re, os, twitter, subprocess, datetime, urlparse, hashlib from collections import defaultdict from irclib import irc_lower, nm_to_n +try: + from blame_filter import bfd +except ImportError: + bfd = None + # query karma def karmaq(bot, cmd, nick, conn, public, karma): try: @@ -89,6 +94,7 @@ def troutq(bot, cmd, nick, conn, public, cfg): me = bot.connection.get_nickname() trout_msg = random.choice(fishlist) fishpond.last=trout_msg + fishpond.last_cfg=cfg # The bot won't trout or flirt with itself; if irc_lower(me) == irc_lower(target) or irc_lower(target) in synonyms: target = nick @@ -130,6 +136,7 @@ def slashq(bot, cmd, nick, conn, public, cfg): me = bot.connection.get_nickname() slash_msg = random.choice(fishlist) fishpond.last=slash_msg + fishpond.last_cfg=cfg # The bot won't slash people with themselves if irc_lower(who[0]) == irc_lower(who[1]): conn.notice(nick, "oooooh no missus!") @@ -279,8 +286,12 @@ def __getcommitinfo(commit): return(err) ts,mes=out.split('|') + mes=mes.strip() + md5mes=hashlib.md5(mes).hexdigest() + if bfd and md5mes in bfd: + mes=bfd[md5mes] when=datetime.date.fromtimestamp(float(ts)) - return mes.strip(), when + return mes, when ###Return an array of commit messages and timestamps for lines in db that match what def __getcommits(db,keys,what): @@ -303,7 +314,10 @@ def __getall(tdb,tdbk,fdb,fdbk,sdb,sdbk,what): sans=__getcommits(sdb,sdbk,what) return tans+fans+sans -def blameq(bot,cmd,nick,conn,public,fish,tdb,tdbk,fdb,fdbk,sdb,sdbk): +def blameq(bot,cmd,nick,conn,public,fish,cfgs): + tdb,tdbk = cfgs[0][7] # urgh, magic, to support magic knowledge below + fdb,fdbk = cfgs[1][7] + sdb,sdbk = cfgs[2][7] clist=cmd.split() if len(clist) < 2: bot.automsg(public,nick,"Who or what do you want to blame?") @@ -393,23 +407,33 @@ class UrlLog: self.nick=nick self.url=url self.first=time.time() + self.localfirst=time.localtime(self.first) self.count=1 self.lastseen=time.time() self.lastasked=time.time() def recenttime(self): return max(self.lastseen,self.lastasked) def firstmen(self): - return nicetime(time.time()-self.first) + n=time.localtime(time.time()) + s="%02d:%02d" % (self.localfirst.tm_hour,self.localfirst.tm_min) + if n.tm_yday != self.localfirst.tm_yday: + s+=time.strftime(" on %d %B", self.localfirst) + return s def urltype(self): - z=min(len(urlcomplaints)-1, self.count-1) - return urlcomplaints[z] + z=min(len(urlinfos)-1, self.count-1) + return urlinfos[z] #(?:) is a regexp that doesn't group urlre = re.compile(r"((?:(?:http)|(?:nsfw))s?://[^ ]+)( |$)") hturlre= re.compile(r"(http)(s?://[^ ]+)( |$)") #matches \bre\:?\s+ before a regexp; (?i)==case insensitive match shibboleth = re.compile(r"(?i)\bre\:?\s+((?:(?:http)|(?:nsfw))s?://[^ ]+)( |$)") -urlcomplaints = ["a contemporary","an interesting","a fascinating","an overused","a vastly overused"] +#How long (in s) to wait since the most recent mention before commenting +url_repeat_time = 300 +urlinfos = ["a new", + "a fascinating", + "an interesting", + "a popular"] ### Deal with /msg bot url or ~url in channel def urlq(bot, cmd, nick, conn, public,urldb): @@ -421,12 +445,12 @@ def urlq(bot, cmd, nick, conn, public,urldb): url=canonical_url(urlstring) if (url in urldb): T = urldb[url] - complaint="That's %s URL that was first mentioned %s by %s" % \ - (T.urltype(),T.firstmen(),T.nick) + comment="I saw that URL in scrool, first mentioned by %s at %s" % \ + (T.nick,T.firstmen()) if (public): - complaint=complaint+". Furthermore it defeats the point of this command to use it other than via /msg." + comment=comment+". Furthermore it defeats the point of this command to use it other than via /msg." T.count+=1 - bot.automsg(False,nick,complaint) + bot.automsg(False,nick,comment) T.lastasked=time.time() #URL suppressed, so mention in #urls if urlstring != cmd.split()[1]: #first argument to URL was not the url @@ -450,9 +474,10 @@ def dourl(bot,conn,nick,command,urldb): if urlstring in urldb: T=urldb[urlstring] - message="observes %s URL, first mentioned %s by %s" % \ - (T.urltype(),T.firstmen(),T.nick) - if shibboleth.search(command)==None: + message="saw that URL in scrool, first mentioned by %s at %s" % \ + (T.nick,T.firstmen()) + if shibboleth.search(command)==None and \ + time.time() - T.lastseen > url_repeat_time: conn.action(bot.channel, message) T.lastseen=time.time() T.count+=1 @@ -498,15 +523,17 @@ def twitterq(bot,cmd,nick,conn,public,twitapi): urlstring = urlre.search(cmd).group(1) if (urlstring.find("twitter.com") !=-1): - stringout = getTweet(urlstring,twitapi) - bot.automsg(public, nick, stringout) + stringsout = getTweet(urlstring,twitapi) + for stringout in stringsout: + bot.automsg(public, nick, stringout) -def getTweet(urlstring,twitapi,inclusion=False): +def getTweet(urlstring,twitapi,inclusion=False,recurlvl=0): unobfuscate_urls=True expand_included_tweets=True + stringsout=[] - parts = string.split(urlstring,'/') - tweetID = parts[-1] + path = urlparse.urlparse(urlstring).path + tweetID = path.split('/')[-1] try: status = twitapi.GetStatus(tweetID) if status == {}: @@ -521,7 +548,7 @@ def getTweet(urlstring,twitapi,inclusion=False): else: tweeter_screen = "[not returned]" ; tweeter_name = "[not returned]" tweeter_name = tweeter_name + " RTing " + status.user.name #.encode('UTF-8', 'replace') - tweetText = status.text + tweetText = status.full_text if status.media: replacements = defaultdict( list ) for medium in status.media: @@ -529,7 +556,7 @@ def getTweet(urlstring,twitapi,inclusion=False): for k,v in replacements.items(): - v = [re.sub(r"/tweet_video_thumb/(\w+).jpg", r"/tweet_video/\1.mp4", link) for link in v] + v = [re.sub(r"/tweet_video_thumb/([\w\-]+).jpg", r"/tweet_video/\1.mp4", link) for link in v] if len(v) > 1: replacementstring = "[" + " ; ".join(v) +"]" else: @@ -574,10 +601,18 @@ def getTweet(urlstring,twitapi,inclusion=False): else: toReplace = '%s://%s%s' % (rv.scheme, rv.hostname, rv.path) # leave off the final '?' - if expand_included_tweets and not inclusion: + if expand_included_tweets: if rv.hostname == 'twitter.com' and re.search(r'status/\d+',rv.path): - quotedtweet = getTweet(toReplace, twitapi, inclusion=True) # inclusion parameter limits recursion. - tweetText += " Q{" + quotedtweet + "}" + if recurlvl > 2: + stringsout = [ "{{ Recursion level too high }}" ] + stringsout + else: + quotedtweet = getTweet(toReplace, twitapi, inclusion=True, recurlvl=recurlvl+1) # inclusion parameter limits recursion. + if not quotedtweet: + quotedtweet = [""] + quotedtweet[0] = "Q{ " + quotedtweet[0] + quotedtweet[-1] += " }" + stringsout = quotedtweet + stringsout + tweetText = tweetText.replace(url.url, toReplace) tweetText = tweetText.replace(">",">") @@ -591,6 +626,8 @@ def getTweet(urlstring,twitapi,inclusion=False): except Exception: terror = sys.exc_info() stringout = "Error: %s" % terror[1].__str__() + stringsout = [stringout] + stringsout if inclusion: - return stringout # don't want to double-encode it, so just pass it on for now and encode later - return stringout.encode('UTF-8', 'replace') + return stringsout # don't want to double-encode it, so just pass it on for now and encode later + + return map(lambda x: x.encode('UTF-8', 'replace'), stringsout)