yoweb-scrape

   1 #!/usr/bin/python
   2
   3 import signal
   4 signal.signal(signal.SIGINT, signal.SIG_DFL)
   5
   6 import os
   7 import time
   8 import urllib
   9 import urllib2
  10 import errno
  11 import sys
  12 import re as regexp
  13 from optparse import OptionParser
  14
  15 from BeautifulSoup import BeautifulSoup
  16
  17 opts = None
  18
  19
  20 puzzles = ('Swordfighting/Bilging/Sailing/Rigging/Navigating'+
  21         '/Battle Navigation/Gunning/Carpentry/Rumble/Treasure Haul'+
  22         '/Drinking/Spades/Hearts/Treasure Drop/Poker/Distilling'+
  23         '/Alchemistry/Shipwrightery/Blacksmithing/Foraging').split('/')
  24
  25 standingvals = ('Able/Distinguished/Respected/Master'+
  26                 '/Renowned/Grand-Master/Legendary/Ultimate').split('/')
  27
  28 pirate_ref_re = regexp.compile('^/yoweb/pirate\\.wm')
  29
  30 max_pirate_namelen = 20
  31
  32
  33 def debug(m):
  34         if opts.debug:
  35                 print >>sys.stderr, m
  36
  37 class Fetcher:
  38         def __init__(self, ocean, cachedir):
  39                 debug('Fetcher init %s' % cachedir)
  40                 self.ocean = ocean
  41                 self.cachedir = cachedir
  42                 try: os.mkdir(cachedir)
  43                 except (OSError,IOError), oe:
  44                         if oe.errno != errno.EEXIST: raise
  45                 self._cache_scan(time.time())
  46
  47         def _cache_scan(self, now):
  48                 # returns list of ages, unsorted
  49                 ages = []
  50                 debug('Fetcher   scan_cache')
  51                 for leaf in os.listdir(self.cachedir):
  52                         if not leaf.startswith('#'): continue
  53                         path = self.cachedir + '/' + leaf
  54                         try: s = os.stat(path)
  55                         except (OSError,IOError), oe:
  56                                 if oe.errno != errno.ENOENT: raise
  57                                 continue
  58                         age = now - s.st_mtime
  59                         if age > opts.expire_age:
  60                                 debug('Fetcher    expire %d %s' % (age, path))
  61                                 try: os.remove(path)
  62                                 except (OSError,IOError), oe:
  63                                         if oe.errno != errno.ENOENT: raise
  64                                 continue
  65                         ages.append(age)
  66                 return ages
  67
  68         def _rate_limit_cache_clean(self, now):
  69                 ages = self._cache_scan(now)
  70                 ages.sort()
  71                 debug('Fetcher   ages ' + `ages`)
  72                 min_age = 1
  73                 need_wait = 0
  74                 for age in ages:
  75                         if age < min_age:
  76                                 debug('Fetcher   morewait min=%d age=%d' %
  77                                         (min_age, age))
  78                                 need_wait = max(need_wait, min_age - age)
  79                         min_age += 1
  80                         min_age *= 1.5
  81                 if need_wait > 0:
  82                         debug('Fetcher   wait %d' % need_wait)
  83                         time.sleep(need_wait)
  84
  85         def fetch(self, url, max_age):
  86                 debug('Fetcher fetch %s' % url)
  87                 cache_corename = urllib.quote_plus(url)
  88                 cache_item = "%s/#%s#" % (self.cachedir, cache_corename)
  89                 try: f = file(cache_item, 'r')
  90                 except (OSError,IOError), oe:
  91                         if oe.errno != errno.ENOENT: raise
  92                         f = None
  93                 now = time.time()
  94                 max_age = max(opts.min_max_age, max(max_age, opts.expire_age))
  95                 if f is not None:
  96                         s = os.fstat(f.fileno())
  97                         age = now - s.st_mtime
  98                         if age > max_age:
  99                                 debug('Fetcher  stale')
 100                                 f = None
 101                 if f is not None:
 102                         data = f.read()
 103                         f.close()
 104                         debug('Fetcher  cached %d > %d' % (max_age, age))
 105                         return data
 106
 107                 debug('Fetcher  fetch')
 108                 self._rate_limit_cache_clean(now)
 109
 110                 stream = urllib2.urlopen(url)
 111                 data = stream.read()
 112                 cache_tmp = "%s/#%s~%d#" % (
 113                         self.cachedir, cache_corename, os.getpid())
 114                 f = file(cache_tmp, 'w')
 115                 f.write(data)
 116                 f.close()
 117                 os.rename(cache_tmp, cache_item)
 118                 debug('Fetcher  stored')
 119                 return data
 120
 121         def yoweb(self, kind, tail, max_age):
 122                 url = 'http://%s.puzzlepirates.com/yoweb/%s%s' % (
 123                         self.ocean, kind, tail)
 124                 return self.fetch(url, max_age)
 125
 126 class SoupLog:
 127         def __init__(self):
 128                 self.msgs = [ ]
 129         def msg(self, m):
 130                 self.msgs.append(m)
 131         def soupm(self, obj, m):
 132                 self.msg(m + '; in ' + `obj`)
 133         def needs_msgs(self, child_souplog):
 134                 self.msgs += child_souplog.msgs
 135                 child_souplog.msgs = [ ]
 136
 137 def soup_text(obj):
 138         str = ''.join(obj.findAll(text=True))
 139         return str.strip()
 140
 141 class SomethingSoupInfo(SoupLog):
 142         def __init__(self, kind, tail, max_age):
 143                 SoupLog.__init__(self)
 144                 html = fetcher.yoweb(kind, tail, max_age)
 145                 self._soup = BeautifulSoup(html,
 146                         convertEntities=BeautifulSoup.HTML_ENTITIES
 147                         )
 148
 149 class PirateInfo(SomethingSoupInfo):
 150         # Public data members:
 151         #  pi.standings = { 'Treasure Haul': 'Able' ... }
 152         #  pi.crew = (id, name)
 153         #  pi.flag = (id, name)
 154         #  pi.msgs = [ 'message describing problem with scrape' ]
 155
 156         def __init__(self, pirate, max_age=300):
 157                 SomethingSoupInfo.__init__(self,
 158                         'pirate.wm?target=', pirate, max_age)
 159                 self._find_standings()
 160                 self.crew = self._find_crewflag('crew',
 161                         '^/yoweb/crew/info\\.wm')
 162                 self.flag = self._find_crewflag('flag',
 163                         '^/yoweb/flag/info\\.wm')
 164
 165         def _find_standings(self):
 166                 imgs = self._soup.findAll('img',
 167                         src=regexp.compile('/yoweb/images/stat.*'))
 168                 re = regexp.compile(
 169 u'\\s*\\S*/([-A-Za-z]+)\\s*$|\\s*\\S*/\\S*\\s*\\(ocean\\-wide(?:\\s|\\xa0)+([-A-Za-z]+)\\)\\s*$'
 170                         )
 171                 standings = { }
 172
 173                 for skill in puzzles:
 174                         standings[skill] = [ ]
 175
 176                 skl = SoupLog()
 177
 178                 for img in imgs:
 179                         try: puzzle = img['alt']
 180                         except KeyError: continue
 181
 182                         if not puzzle in puzzles:
 183                                 skl.soupm(img, 'unknown puzzle: "%s"' % puzzle)
 184                                 continue
 185                         key = img.findParent('td')
 186                         if key is None:
 187                                 skl.soupm(img, 'puzzle at root! "%s"' % puzzle)
 188                                 continue
 189                         valelem = key.findNextSibling('td')
 190                         if valelem is None:
 191                                 skl.soupm(key, 'puzzle missing sibling "%s"'
 192                                         % puzzle)
 193                                 continue
 194                         valstr = soup_text(valelem)
 195                         match = re.match(valstr)
 196                         if match is None:
 197                                 skl.soupm(key, ('puzzle "%s" unparseable'+
 198                                         ' standing "%s"') % (puzzle, valstr))
 199                                 continue
 200                         standing = match.group(match.lastindex)
 201                         standings[puzzle].append(standing)
 202
 203                 self.standings = { }
 204
 205                 for puzzle in puzzles:
 206                         sl = standings[puzzle]
 207                         if len(sl) > 1:
 208                                 skl.msg('puzzle "%s" multiple standings %s' %
 209                                                 (puzzle, `sl`))
 210                                 continue
 211                         if not len(sl):
 212                                 skl.msg('puzzle "%s" no standing found' % puzzle)
 213                                 continue
 214                         standing = sl[0]
 215                         for i in range(0, len(standingvals)-1):
 216                                 if standing == standingvals[i]:
 217                                         self.standings[puzzle] = i
 218                         if not puzzle in self.standings:
 219                                 skl.msg('puzzle "%s" unknown standing "%s"' %
 220                                         (puzzle, standing))
 221
 222                 all_standings_ok = True
 223                 for puzzle in puzzles:
 224                         if not puzzle in self.standings:
 225                                 self.needs_msgs(skl)
 226
 227         def _find_crewflag(self, cf, yoweb_re):
 228                 things = self._soup.findAll('a', href=regexp.compile(yoweb_re))
 229                 if len(things) != 1:
 230                         self.msg('zero or several %s id references found' % cf)
 231                         return None
 232                 thing = things[0]
 233                 id_re = '\\b%sid\\=(\\w+)$' % cf
 234                 id_haystack = thing['href']
 235                 match = regexp.compile(id_re).search(id_haystack)
 236                 if match is None:
 237                         self.soupm(thing, ('incomprehensible %s id ref'+
 238                                 ' (%s in %s)') % (cf, id_re, id_haystack))
 239                         return None
 240                 name = soup_text(thing)
 241                 return (match.group(1), name)
 242
 243         def __str__(self):
 244                 return `(self.crew, self.flag, self.standings, self.msgs)`
 245
 246 class CrewInfo(SomethingSoupInfo):
 247         # Public data members:
 248         #  ci.crew = [ ('Captain',        ['Pirate', ...]),
 249         #              ('Senior Officer', [...]),
 250         #               ... ]
 251         #  pi.msgs = [ 'message describing problem with scrape' ]
 252
 253         def __init__(self, crewid, max_age=300):
 254                 SomethingSoupInfo.__init__(self,
 255                         'crew/info.wm?crewid=', crewid, max_age)
 256                 self._find_crew()
 257
 258         def _find_crew(self):
 259                 self.crew = []
 260                 capts = self._soup.findAll('img',
 261                         src='/yoweb/images/crew-captain.png')
 262                 if len(capts) != 1:
 263                         self.msg('crew members: no. of captain images != 1')
 264                         return
 265                 tbl = capts[0]
 266                 while not tbl.find('a', href=pirate_ref_re):
 267                         tbl = tbl.findParent('table')
 268                         if not tbl:
 269                                 self.msg('crew members: cannot find table')
 270                                 return
 271                 current_rank_crew = None
 272                 crew_rank_re = regexp.compile('/yoweb/images/crew')
 273                 for row in tbl.contents:
 274                         # findAll(recurse=False)
 275                         if isinstance(row, unicode):
 276                                 continue
 277
 278                         is_rank = row.find('img', attrs={'src': crew_rank_re})
 279                         if is_rank:
 280                                 rank = soup_text(row)
 281                                 current_rank_crew = []
 282                                 self.crew.append((rank, current_rank_crew))
 283                                 continue
 284                         for cell in row.findAll('a', href=pirate_ref_re):
 285                                 if current_rank_crew is None:
 286                                         self.soupm(cell, 'crew members: crew'
 287                                                 ' before rank')
 288                                         continue
 289                                 current_rank_crew.append(soup_text(cell))
 290
 291         def __str__(self):
 292                 return `(self.crew, self.msgs)`
 293
 294 class StandingsTable:
 295         def __init__(self, use_puzzles=puzzles):
 296                 self._puzzles = use_puzzles
 297                 self.s = ''
 298
 299         def _pline(self, pirate, puzstrs):
 300                 self.s += '%-*s' % (max_pirate_namelen, pirate)
 301                 for v in puzstrs:
 302                         self.s += ' %-*.*s' % (5,5, v)
 303                 self.s += '\n'
 304
 305         def _puzstr(self, pi, puzzle):
 306                 if not isinstance(puzzle,list): puzzle = [puzzle]
 307                 try: standing = max([pi.standings[p] for p in puzzle])
 308                 except KeyError: return '?'
 309                 c1 = standingvals[standing][0]
 310                 if standing < 3: c1 = c1.lower() # 3 = Master
 311                 hashes = '*' * (standing / 2)
 312                 equals = '+' * (standing % 2)
 313                 return c1 + hashes + equals
 314
 315         def headings(self):
 316                 def puzn_redact(name):
 317                         if isinstance(name,list):
 318                                 return '/'.join(
 319                                         ["%.2s" % puzn_redact(n)
 320                                          for n in name])
 321                         spc = name.find(' ')
 322                         if spc < 0: return name
 323                         return name[0:min(4,spc)] + name[spc+1:]
 324                 self._pline('', map(puzn_redact, self._puzzles))
 325         def literalline(self, line):
 326                 self.s += line + '\n'
 327         def pirate(self, pirate):
 328                 pi = PirateInfo(pirate, 600)
 329                 puzstrs = [self._puzstr(pi,puz) for puz in self._puzzles]
 330                 self._pline(pirate, puzstrs)
 331
 332         def results(self):
 333                 return self.s
 334
 335 def do_pirate(pirates, bu):
 336         print '{'
 337         for pirate in pirates:
 338                 info = PirateInfo(pirate)
 339                 print '%s: %s,' % (`pirate`, info)
 340         print '}'
 341
 342 def prep_crew_of(args, bu, max_age=300):
 343         if len(args) != 1: bu('crew-of takes one pirate name')
 344         pi = PirateInfo(args[0])
 345         return CrewInfo(pi.crew[0])
 346
 347 def do_crew_of(args, bu):
 348         ci = prep_crew_of(args, bu)
 349         print ci
 350
 351 def do_standings_crew_of(args, bu):
 352         ci = prep_crew_of(args, bu, 60)
 353         tab = StandingsTable()
 354         tab.headings()
 355         for (rank, members) in ci.crew:
 356                 if not members: continue
 357                 tab.literalline('%s:' % rank)
 358                 for p in members:
 359                         tab.pirate(p)
 360         print tab.results()
 361
 362 def main():
 363         global opts, fetcher
 364
 365         pa = OptionParser(
 366 '''usage: .../yoweb-scrape [OPTION...] ACTION [ARGS...]
 367 actions:
 368  yoweb-scrape [--ocean OCEAN ...] pirate PIRATE
 369  yoweb-scrape [--ocean OCEAN ...] crew-of PIRATE
 370  yoweb-scrape [--ocean OCEAN ...] standings-crew-of PIRATE
 371 ''')
 372         ao = pa.add_option
 373         ao('-O','--ocean',dest='ocean', metavar='OCEAN',
 374                 default='ice',
 375                 help='select ocean OCEAN')
 376         ao('--cache-dir', dest='cache_dir', metavar='DIR',
 377                 default='~/.yoweb-scrape-cache',
 378                 help='cache yoweb pages in DIR')
 379         ao('-D','--debug', action='store_true', dest='debug', default=False,
 380                 help='enable debugging output')
 381         ao('-q','--quiet', action='store_true', dest='quiet',
 382                 help='suppress warning output')
 383         (opts,args) = pa.parse_args()
 384
 385         if len(args) < 1:
 386                 pa.error('need a mode argument')
 387
 388         mode = args[0]
 389         mode_fn_name = 'do_' + mode.replace('_','#').replace('-','_')
 390         try: mode_fn = globals()[mode_fn_name]
 391         except KeyError: pa.error('unknown mode "%s"' % mode)
 392
 393         # fixed parameters
 394         opts.min_max_age = 60
 395         opts.expire_age = 3600
 396
 397         if opts.cache_dir.startswith('~/'):
 398                 opts.cache_dir = os.getenv('HOME') + opts.cache_dir[1:]
 399
 400         fetcher = Fetcher(opts.ocean, opts.cache_dir)
 401
 402         mode_fn(args[1:], pa.error)
 403
 404 main()