chiark / gitweb /
Reorg chat log to make ship aid nice
[ypp-sc-tools.web-live.git] / yoweb-scrape
index ca174e27647efdba100a326e8559e9f6fd62d5c4..2d968921ab1eed4023fea7344ce90951f2c7c21d 100755 (executable)
@@ -33,7 +33,7 @@ max_pirate_namelen = 12
 
 def debug(m):
        if opts.debug:
-               print >>sys.stderr, m
+               print m
 
 class Fetcher:
        def __init__(self, ocean, cachedir):
@@ -70,7 +70,7 @@ class Fetcher:
                        ages.append(age)
                return ages
 
-       def _rate_limit_cache_clean(self, now):
+       def need_wait(self, now):
                ages = self._cache_scan(now)
                ages.sort()
                debug('Fetcher   ages ' + `ages`)
@@ -83,6 +83,10 @@ class Fetcher:
                                need_wait = max(need_wait, min_age - age)
                        min_age += 3
                        min_age *= 1.25
+               return need_wait
+
+       def _rate_limit_cache_clean(self, now):
+               need_wait = self.need_wait(now)
                if need_wait > 0:
                        debug('Fetcher   wait %d' % need_wait)
                        time.sleep(need_wait)
@@ -280,7 +284,7 @@ class CrewInfo(SomethingSoupInfo):
                crew_rank_re = regexp.compile('/yoweb/images/crew')
                for row in tbl.contents:
                        # findAll(recurse=False)
-                       if isinstance(row, unicode):
+                       if isinstance(row,basestring):
                                continue
 
                        is_rank = row.find('img', attrs={'src': crew_rank_re})
@@ -327,6 +331,7 @@ class StandingsTable:
                if not isinstance(puzzle,list): puzzle = [puzzle]
                try: standing = max([pi.standings[p] for p in puzzle])
                except KeyError: return '?'
+               if isinstance(standing,basestring): return standing
                if not standing: return ''
                s = ''
                if self._cw > 4:
@@ -391,14 +396,27 @@ class PirateAboard:
        # pa.last_time
        # pa.last_event
        # pa.gunner
-       def __init__(pa, v, time, event):
+       # pa.last_chat_time
+       # pa.last_chat_chan
+       # pa.pi
+
+       def __init__(pa, pn, v, time, event):
+               pa.pn = pn
                pa.v = v
                pa.last_time = time
                pa.last_event = event
+               pa.last_chat_time = None
+               pa.last_chat_chan = None
                pa.gunner = False
+               pa.pi = None
+
+       def pirate_info(self):
+               if not pa.pi and not fetcher.need_wait(time.time):
+                       pa.pi = PirateInfo(pa.name, 3600)
+               return pa.pi
 
-class ShipCrewTracker:
-       def __init__(self, myself_pi):
+class ChatLogTracker:
+       def __init__(self, myself_pi, logfn):
                self._pl = {}   # self._pl['Pirate'] =
                self._vl = {}   #   self._vl['Vessel']['Pirate'] = PirateAboard
                                # self._vl['Vessel']['#lastaboard']
@@ -406,10 +424,13 @@ class ShipCrewTracker:
                self._vessel = None     #       self._vl[self._vessel]
                self._date = None
                self._myself = myself_pi
-               self._need_redisplay = False
+               self.need_redisplay = False
+               self._f = file(logfn)
+               self._lbuf = ''
+               self._progress = [0, os.fstat(self._f.fileno()).st_size]
 
        def _refresh(self):
-               self._need_redisplay = True
+               self.need_redisplay = True
 
        def _onboard_event(self,timestamp,pirate,event):
                try: pa = self._pl[pirate]
@@ -419,109 +440,234 @@ class ShipCrewTracker:
                        pa.last_event = event
                else:
                        if pa is not None: del pa.v[pirate]
-                       pa = PirateAboard(self._v, timestamp, event)
+                       pa = PirateAboard(pirate, self._v, timestamp, event)
                        self._pl[pirate] = pa
                        self._v[pirate] = pa
                self._v['#lastaboard'] = timestamp
                self._refresh()
                return pa
 
+       def _trash_vessel(self, v):
+               for pn in v:
+                       if pn.startswith('#'): continue
+                       del self._pl[pn]
+               self._refresh()
+
+       def expire_garbage(self, timestamp):
+               for (vn,v) in list(self._vl.iteritems()):
+                       la = v['#lastaboard']
+                       if timestamp - la > opts.ship_reboard_clearout:
+                               self._debug_line_disposition(timestamp,'',
+                                       'stale reset '+vn)
+                               self._trash_vessel(v)
+                               del self._vl[vn]
+
        def clear_vessel(self, timestamp):
                if self._v is not None:
-                       for p in self._v:
-                               if p.startswith('#'): continue
-                               del self._pl[p]
+                       self._trash_vessel(self._v)
                self._v = {'#lastaboard': timestamp}
+               self._vl[self._vessel] = self._v
 
-       def _warn(self, m):
-               pass
+       def _debug_line_disposition(self,timestamp,l,m):
+               debug('CLT %13s %-30s %s' % (timestamp,m,l))
 
        def chatline(self,l):
                rm = lambda re: regexp.match(re,l)
+               d = lambda m: self._debug_line_disposition(timestamp,l,m)
+               timestamp = None
+
                m = rm('=+ (\\d+)/(\\d+)/(\\d+) =+$')
                if m:
                        self._date = m.groups()
-                       return
+                       return d('date '+`self._date`)
+
                if self._date is None:
-                       return
+                       return d('date unset')
+
                m = rm('\\[(\d\d):(\d\d):(\d\d)\\] ')
                if not m:
-                       self._warn('undated? '+l)
-                       return
+                       return d('no timestamp')
 
                time_tuple = [int(x) for x in self._date + m.groups()]
                time_tuple += (-1,-1,-1)
-               print `time_tuple`
                timestamp = time.mktime(time_tuple)
                l = l[l.find(' ')+1:]
 
-               ob = lambda who, event: self._onboard_event(
-                       timestamp, who, event)
-               oba = lambda m, did: ob(
-                       m.group(1), '%s %s' % (did, m.group(2)))
+               def ob_x(who,event):
+                       return self._onboard_event(timestamp, who, event)
+               def ob1(did): ob_x(m.group(1), did); return d(did)
+               def oba(did): return ob1('%s %s' % (did, m.group(2)))
 
                m = rm('Going aboard the (\\S.*\\S)\\.\\.\\.$')
                if m:
+                       pn = self._myself.name
                        self._vessel = m.group(1)
+                       dm = 'boarding'
+
                        try:             self._v = self._vl[self._vessel]
-                       except KeyError: self._v = None
+                       except KeyError: self._v = None; dm += ' new'
+                       
                        if self._v is not None:  la = self._v['#lastaboard']
-                       else:                    la = 0
-                       if timestamp - la > 3600:
+                       else:                    la = 0; dm += ' ?la'
+
+                       if timestamp - la > opts.ship_reboard_clearout:
                                self.clear_vessel(timestamp)
-                       ob(self._myself.name, 'we boarded')
+                               dm += ' stale'
+
+                       ob_x(pn, 'we boarded')
+                       self.expire_garbage(timestamp)
+                       return d(dm)
+
+               if self._v is None:
+                       return d('no vessel')
 
-               if self._v is None: return
+               m = rm('(\\w+) has come aboard\\.$')
+               if m: return ob1('boarded');
 
                m = rm('You have ordered (\\w+) to do some (\\S.*\\S)\\.$')
                if m:
-                       pa = oba(m, 'ordered')
-                       if m.group(2) == 'Gunning':
+                       (who,what) = m.groups()
+                       pa = ob_x(who,'ordered '+what)
+                       if what == 'Gunning':
                                pa.gunner = True
-                       return
+                       return d('duty order')
 
                m = rm('(\\w+) abandoned a (\\S.*\\S) station\\.$')
-               if m: oba(m,'abandoned'); return
+               if m: oba('stopped'); return d('stopped')
+
+               def chat(what):
+                       who = m.group(1)
+                       try: pa = self._pl[who]
+                       except KeyError: return d('chat mystery')
+                       if pa.v is self._v:
+                               pa.last_chat_time = timestamp
+                               pa.last_chat_chan = what
+                               self._refresh()
+                               return d(what+' chat')
+
+               m = rm('(\\w+) (?:issued an order|ordered everyone) "')
+               if m: return ob1('general order');
 
                m = rm('(\\w+) says, "')
-               if m: ob(m.group(1), 'talked'); return
+               if m: return chat('public')
+
+               m = rm('(\\w+) tells ye, "')
+               if m: return chat('private')
+
+               m = rm('(\\w+) flag officer chats, "')
+               if m: return chat('flag officer')
+
+               m = rm('(\\w+) officer chats, "')
+               if m: return chat('officer')
+
+               m = rm('Game over\\.  Winners: ([A-Za-z, ]+)\\.$')
+               if m:
+                       pl = m.group(1).split(', ')
+                       if not self._myself.name in pl:
+                               return d('lost boarding battle')
+                       for pn in pl:
+                               if ' ' in pn: continue
+                               ob_x(pn,'won boarding battle')
+                       return d('won boarding battle')
+
+               m = rm('(\\w+) is eliminated\\!')
+               if m: return ob1('eliminated in fray');
 
                m = rm('(\\w+) has left the vessel\.')
                if m:
                        who = m.group(1)
-                       ob(who, 'disembarked')
+                       ob_x(who, 'disembarked')
                        del self._v[who]
                        del self._pl[who]
-                       return
+                       return d('disembarked')
+
+               return d('not matched')
+
+       def _str_vessel(self, vn, v):
+               s = ' vessel %s\n' % vn
+               s += ' '*20 + "%-*s   %13s\n" % (
+                               max_pirate_namelen, '#lastaboard',
+                               v['#lastaboard'])
+               for pn in sorted(v.keys()):
+                       if pn.startswith('#'): continue
+                       pa = v[pn]
+                       assert pa.v == v
+                       assert self._pl[pn] == pa
+                       s += ' '*20 + "%s %-*s %13s %-30s %13s %s\n" % (
+                               (' ','G')[pa.gunner],
+                               max_pirate_namelen, pn,
+                               pa.last_time, pa.last_event,
+                               pa.last_chat_time, pa.last_chat_chan)
+               return s
 
-               return
+       def __str__(self):
+               s = '''<ChatLogTracker
+ myself %s
+ vessel %s
+'''                    % (self._myself.name, self._vessel)
+               assert ((self._v is None and self._vessel is None) or
+                       (self._v is self._vl[self._vessel]))
+               if self._vessel is not None:
+                       s += self._str_vessel(self._vessel, self._v)
+               for vn in sorted(self._vl.keys()):
+                       if vn == self._vessel: continue
+                       s += self._str_vessel(vn, self._vl[vn])
+               for p in self._pl:
+                       pa = self._pl[p]
+                       assert pa.v[p] is pa
+                       assert pa.v in self._vl.values()
+               s += '>\n'
+               return s
 
-def do_ship_aid(args, bu):
-       if len(args) != 1: bu('ship-aid takes only chat log filename')
+       def catchup(self, progress=None):
+               while True:
+                       more = self._f.readline()
+                       if not more: break
+
+                       self._progress[0] += len(more)
+                       if progress: progress.progress(*self._progress)
+
+                       self._lbuf += more
+                       if self._lbuf.endswith('\n'):
+                               self.chatline(self._lbuf.rstrip())
+                               self._lbuf = ''
+               if progress: progress.caughtup()
+
+class ProgressPrintPercentage:
+       def __init__(self, f=sys.stdout): self._f = f
+       def progress(self,done,total):
+               self._f.write("scan chat logs %3d%%\r" % ((done*100) / total))
+               self._f.flush()
+       def caughtup(self):
+               self._f.write('                   \r')
+               self._f.flush()
+
+def simple_printer(x): print x
+
+def run_chat_log(args, bu, progress=ProgressPrintPercentage(),
+               display=simple_printer, max_myself_age=3600):
+       if len(args) != 1: bu('this action takes only chat log filename')
        logfn = args[0]
        logfn_re = '(?:.*/)?([A-Z][a-z]+)_([a-z]+)_chat-log-\\w+$'
        match = regexp.match(logfn_re, logfn)
-       if not match: bu('ship-aid chat log filename is not in default format')
+       if not match: bu('chat log filename is not in default format')
        (pirate, fetcher.ocean) = match.groups()
-       myself_pi = PirateInfo(pirate,3600)
-       track = ShipCrewTracker(myself_pi)
-       f = file(logfn)
-       l = ''
+       
+       myself_pi = PirateInfo(pirate,max_myself_age)
+       track = ChatLogTracker(myself_pi, logfn)
+
+       track.catchup(progress)
        while True:
-               print "1>"+l+"<"
-               l += f.readline()
-               if l.endswith('\n'):
-                       l.rstrip()
-                       print "2>"+l+"<"
-                       track.chatline(l)
-#                      print `track.__dict__`
-                       l = ''
-                       continue
-               if l:
-                       continue
-               print "3>EOF<"
-               print `track.__dict__`
-               os.sleep(1)
+               track.catchup()
+               if track.need_redisplay:
+                       display(track)
+                       track.need_redisplay = False
+               time.sleep(1)
+
+def do_track_chat_log(args, bu):
+       run_chat_log(args, bu)
+
+#def do_ship_aid(args, bu):
 
 def main():
        global opts, fetcher
@@ -532,6 +678,7 @@ actions:
  yoweb-scrape [--ocean OCEAN ...] pirate PIRATE
  yoweb-scrape [--ocean OCEAN ...] crew-of PIRATE
  yoweb-scrape [--ocean OCEAN ...] standings-crew-of PIRATE
+ yoweb-scrape [--ocean OCEAN ...] track-chat-log CHAT-LOG
  yoweb-scrape [--ocean OCEAN ...] ship-aid CHAT-LOG
 ''')
        ao = pa.add_option
@@ -547,6 +694,8 @@ actions:
 
        ao('--ship-duty', action='store_true', dest='ship_duty',
                help='show ship duty station puzzles')
+       ao('--all-puzzles', action='store_false', dest='ship_duty',
+               help='show all puzzles, not just ship duty stations')
 
        (opts,args) = pa.parse_args()
        random.seed()
@@ -562,6 +711,7 @@ actions:
        # fixed parameters
        opts.min_max_age = 60
        opts.expire_age = 3600
+       opts.ship_reboard_clearout = 3600
 
        if opts.cache_dir.startswith('~/'):
                opts.cache_dir = os.getenv('HOME') + opts.cache_dir[1:]