+ os.rename(cache_tmp, cache_item)
+ debug('Fetcher stored')
+ return data
+
+ def yoweb(self, kind, tail, max_age):
+ self.default_ocean()
+ url = 'http://%s.puzzlepirates.com/yoweb/%s%s' % (
+ self.ocean, kind, tail)
+ return self.fetch(url, max_age)
+
+#---------- logging assistance for troubled screenscrapers ----------
+
+class SoupLog:
+ def __init__(self):
+ self.msgs = [ ]
+ def msg(self, m):
+ self.msgs.append(m)
+ def soupm(self, obj, m):
+ self.msg(m + '; in ' + `obj`)
+ def needs_msgs(self, child_souplog):
+ self.msgs += child_souplog.msgs
+ child_souplog.msgs = [ ]
+
+def soup_text(obj):
+ str = ''.join(obj.findAll(text=True))
+ return str.strip()
+
+class SomethingSoupInfo(SoupLog):
+ def __init__(self, kind, tail, max_age):
+ SoupLog.__init__(self)
+ html = fetcher.yoweb(kind, tail, max_age)
+ self._soup = BeautifulSoup(html,
+ convertEntities=BeautifulSoup.HTML_ENTITIES
+ )
+
+#---------- scraper for pirate pages ----------
+
+class PirateInfo(SomethingSoupInfo):
+ # Public data members:
+ # pi.standings = { 'Treasure Haul': 'Able' ... }
+ # pi.name = name
+ # pi.crew = (id, name)
+ # pi.flag = (id, name)
+ # pi.msgs = [ 'message describing problem with scrape' ]
+
+ def __init__(self, pirate, max_age=300):
+ SomethingSoupInfo.__init__(self,
+ 'pirate.wm?target=', pirate, max_age)
+ self.name = pirate
+ self._find_standings()
+ self.crew = self._find_crewflag('crew',
+ '^/yoweb/crew/info\\.wm')
+ self.flag = self._find_crewflag('flag',
+ '^/yoweb/flag/info\\.wm')
+
+ def _find_standings(self):
+ imgs = self._soup.findAll('img',
+ src=regexp.compile('/yoweb/images/stat.*'))
+ re = regexp.compile(
+u'\\s*\\S*/([-A-Za-z]+)\\s*$|\\s*\\S*/\\S*\\s*\\(ocean\\-wide(?:\\s|\\xa0)+([-A-Za-z]+)\\)\\s*$'
+ )
+ standings = { }
+
+ for skill in puzzles:
+ standings[skill] = [ ]
+
+ skl = SoupLog()
+
+ for img in imgs:
+ try: puzzle = img['alt']
+ except KeyError: continue
+
+ if not puzzle in puzzles:
+ skl.soupm(img, 'unknown puzzle: "%s"' % puzzle)
+ continue
+ key = img.findParent('td')
+ if key is None:
+ skl.soupm(img, 'puzzle at root! "%s"' % puzzle)
+ continue
+ valelem = key.findNextSibling('td')
+ if valelem is None:
+ skl.soupm(key, 'puzzle missing sibling "%s"'
+ % puzzle)
+ continue
+ valstr = soup_text(valelem)
+ match = re.match(valstr)
+ if match is None:
+ skl.soupm(key, ('puzzle "%s" unparseable'+
+ ' standing "%s"') % (puzzle, valstr))
+ continue
+ standing = match.group(match.lastindex)
+ standings[puzzle].append(standing)
+
+ self.standings = { }
+
+ for puzzle in puzzles:
+ sl = standings[puzzle]
+ if len(sl) > 1:
+ skl.msg('puzzle "%s" multiple standings %s' %
+ (puzzle, `sl`))
+ continue
+ if not sl:
+ skl.msg('puzzle "%s" no standing found' % puzzle)
+ continue
+ standing = sl[0]
+ for i in range(0, len(standingvals)-1):
+ if standing == standingvals[i]:
+ self.standings[puzzle] = i
+ if not puzzle in self.standings:
+ skl.msg('puzzle "%s" unknown standing "%s"' %
+ (puzzle, standing))
+
+ all_standings_ok = True
+ for puzzle in puzzles:
+ if not puzzle in self.standings:
+ self.needs_msgs(skl)
+
+ def _find_crewflag(self, cf, yoweb_re):
+ things = self._soup.findAll('a', href=regexp.compile(yoweb_re))
+ if len(things) != 1:
+ self.msg('zero or several %s id references found' % cf)
+ return None
+ thing = things[0]
+ id_re = '\\b%sid\\=(\\w+)$' % cf
+ id_haystack = thing['href']
+ match = regexp.compile(id_re).search(id_haystack)
+ if match is None:
+ self.soupm(thing, ('incomprehensible %s id ref'+
+ ' (%s in %s)') % (cf, id_re, id_haystack))
+ return None
+ name = soup_text(thing)
+ return (match.group(1), name)
+
+ def __str__(self):
+ return `(self.crew, self.flag, self.standings, self.msgs)`
+
+#---------- scraper for crew pages ----------
+
+class CrewInfo(SomethingSoupInfo):
+ # Public data members:
+ # ci.crew = [ ('Captain', ['Pirate', ...]),
+ # ('Senior Officer', [...]),
+ # ... ]
+ # pi.msgs = [ 'message describing problem with scrape' ]
+
+ def __init__(self, crewid, max_age=300):
+ SomethingSoupInfo.__init__(self,
+ 'crew/info.wm?crewid=', crewid, max_age)
+ self._find_crew()
+
+ def _find_crew(self):
+ self.crew = []
+ capts = self._soup.findAll('img',
+ src='/yoweb/images/crew-captain.png')
+ if len(capts) != 1:
+ self.msg('crew members: no. of captain images != 1')
+ return
+ tbl = capts[0]
+ while not tbl.find('a', href=pirate_ref_re):
+ tbl = tbl.findParent('table')
+ if not tbl:
+ self.msg('crew members: cannot find table')
+ return
+ current_rank_crew = None
+ crew_rank_re = regexp.compile('/yoweb/images/crew')
+ for row in tbl.contents:
+ # findAll(recurse=False)
+ if isinstance(row,basestring):
+ continue
+
+ is_rank = row.find('img', attrs={'src': crew_rank_re})
+ if is_rank:
+ rank = soup_text(row)
+ current_rank_crew = []
+ self.crew.append((rank, current_rank_crew))
+ continue
+ for cell in row.findAll('a', href=pirate_ref_re):
+ if current_rank_crew is None:
+ self.soupm(cell, 'crew members: crew'
+ ' before rank')
+ continue
+ current_rank_crew.append(soup_text(cell))
+
+ def __str__(self):
+ return `(self.crew, self.msgs)`
+
+#---------- pretty-printer for tables of pirate puzzle standings ----------
+
+class StandingsTable:
+ def __init__(self, use_puzzles=None, col_width=6):
+ if use_puzzles is None:
+ if opts.ship_duty:
+ use_puzzles=[
+ 'Navigating','Battle Navigation',
+ 'Gunning',
+ ['Sailing','Rigging'],
+ 'Bilging',
+ 'Carpentry',
+ 'Treasure Haul'
+ ]
+ else:
+ use_puzzles=puzzles
+ self._puzzles = use_puzzles
+ self.s = ''
+ self._cw = col_width-1
+
+ def _pline(self, pirate, puzstrs, extra):
+ self.s += ' %-*s' % (max(max_pirate_namelen, 14), pirate)
+ for v in puzstrs:
+ self.s += ' %-*.*s' % (self._cw,self._cw, v)
+ if extra:
+ self.s += ' ' + extra
+ self.s += '\n'
+
+ def _puzstr(self, pi, puzzle):
+ if not isinstance(puzzle,list): puzzle = [puzzle]
+ try: standing = max([pi.standings[p] for p in puzzle])
+ except KeyError: return '?'
+ if not standing: return ''
+ s = ''
+ if self._cw > 4:
+ c1 = standingvals[standing][0]
+ if standing < 3: c1 = c1.lower() # 3 = Master
+ s += `standing`
+ if self._cw > 5:
+ s += ' '
+ s += '*' * (standing / 2)
+ s += '+' * (standing % 2)
+ return s
+
+ def headings(self):
+ def puzn_redact(name):
+ if isinstance(name,list):
+ return '/'.join(
+ ["%.*s" % (self._cw/2, puzn_redact(n))
+ for n in name])
+ spc = name.find(' ')
+ if spc < 0: return name
+ return name[0:min(4,spc)] + name[spc+1:]
+ self._pline('', map(puzn_redact, self._puzzles), None)
+ def literalline(self, line):
+ self.s += line + '\n'
+ def pirate_dummy(self, name, standingstring, extra=None):
+ self._pline(name, standingstring * len(self._puzzles), extra)
+ def pirate(self, pi, extra=None):
+ puzstrs = [self._puzstr(pi,puz) for puz in self._puzzles]
+ self._pline(pi.name, puzstrs, extra)
+
+ def results(self):
+ return self.s
+
+#---------- chat log parser ----------
+
+class PirateAboard:
+ # This is essentially a transparent, dumb, data class.
+ # pa.v
+ # pa.name
+ # pa.last_time
+ # pa.last_event
+ # pa.gunner
+ # pa.last_chat_time
+ # pa.last_chat_chan
+ # pa.pi
+
+ def __init__(pa, pn, v, time, event):
+ pa.name = 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(pa):
+ if not pa.pi and not fetcher.need_wait(time.time()):
+ pa.pi = PirateInfo(pa.name, 3600)
+ return pa.pi
+
+class ChatLogTracker:
+ # This is quite complex so we make it opaque. Use the
+ # official invokers, accessors etc.
+
+ def __init__(self, myself_pi, logfn):
+ self._pl = {} # self._pl['Pirate'] =
+ self._vl = {} # self._vl['Vessel']['Pirate'] = PirateAboard
+ # self._vl['Vessel']['#lastinfo']
+ self._v = None # self._v =
+ self._vessel = None # self._vl[self._vessel]
+ self._date = None
+ self._myself = myself_pi
+ self._need_redisplay = False
+ self._f = file(logfn)
+ self._lbuf = ''
+ self._progress = [0, os.fstat(self._f.fileno()).st_size]
+
+ def force_redisplay(self):
+ self._need_redisplay = True
+
+ def _onboard_event(self,v,timestamp,pirate,event):
+ pa = self._pl.get(pirate, None)
+ if pa is not None and pa.v is v:
+ pa.last_time = timestamp
+ pa.last_event = event
+ else:
+ if pa is not None: del pa.v[pirate]
+ pa = PirateAboard(pirate, v, timestamp, event)
+ self._pl[pirate] = pa
+ v[pirate] = pa
+ v['#lastinfo'] = timestamp
+ self.force_redisplay()
+ return pa
+
+ def _trash_vessel(self, v):
+ for pn in v:
+ if pn.startswith('#'): continue
+ del self._pl[pn]
+ self.force_redisplay()
+
+ def expire_garbage(self, timestamp):
+ for (vn,v) in list(self._vl.iteritems()):
+ la = v['#lastinfo']
+ if timestamp - la > opts.ship_reboard_clearout:
+ self._debug_line_disposition(timestamp,'',
+ 'stale reset '+vn)
+ self._trash_vessel(v)
+ del self._vl[vn]
+
+ def _create_vessel(self, vn, timestamp):
+ self._vl[vn] = v = { '#lastinfo': timestamp }
+ return v
+
+ def _update_vessel_lookup(self, vn, timestamp, dml):
+ v = self._vl.get(vn, None)
+ if v is None:
+ dml.append('new')
+ v = self._create_vessel(vn, timestamp)
+ elif timestamp - v['#lastinfo'] > opts.ship_reboard_clearout:
+ dml.append('stale')
+ self._trash_vessel(v)
+ v = self._create_vessel(vn, timestamp)
+ else:
+ dml.append('current')
+ return v
+
+ 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 = [int(x) for x in m.groups()]
+ self._previous_timestamp = None
+ return d('date '+`self._date`)
+
+ if self._date is None:
+ return d('date unset')
+
+ m = rm('\\[(\d\d):(\d\d):(\d\d)\\] ')
+ if not m:
+ return d('no timestamp')
+
+ while True:
+ time_tuple = (self._date +
+ [int(x) for x in m.groups()] +
+ [-1,-1,-1])
+ timestamp = time.mktime(time_tuple)
+ if timestamp >= self._previous_timestamp: break
+ self._date[2] += 1
+ self._debug_line_disposition(timestamp,'',
+ 'new date '+`self._date`)
+
+ self._previous_timestamp = timestamp
+
+ l = l[l.find(' ')+1:]
+
+ def ob_x(pirate,event):
+ return self._onboard_event(
+ self._v, timestamp, pirate, event)
+ def ob1(did): ob_x(m.group(1), did); return d(did)
+ def oba(did): return ob1('%s %s' % (did, m.group(2)))
+
+ def disembark(v, timestamp, pirate, event):
+ self._onboard_event(
+ v, timestamp, pirate, 'leaving '+event)
+ del v[pirate]
+ del self._pl[pirate]
+
+ m = rm('Going aboard the (\\S.*\\S)\\.\\.\\.$')
+ if m:
+ dm = ['boarding']
+ pn = self._myself.name
+ self._vessel = vn = m.group(1)
+ self._v = self._update_vessel_lookup(vn, timestamp, dm)
+
+ ob_x(pn, 'we boarded')
+ self.expire_garbage(timestamp)
+ return d(' '.join(dm))
+
+ if self._v is None:
+ return d('no vessel')