pyfpdb/SplitHandHistory.py

   1 #!/usr/bin/env python
   2 # -*- coding: utf-8 -*-
   3
   4 #Copyright 2010-2011 Chaz Littlejohn
   5 #This program is free software: you can redistribute it and/or modify
   6 #it under the terms of the GNU Affero General Public License as published by
   7 #the Free Software Foundation, version 3 of the License.
   8 #
   9 #This program is distributed in the hope that it will be useful,
  10 #but WITHOUT ANY WARRANTY; without even the implied warranty of
  11 #MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  12 #GNU General Public License for more details.
  13 #
  14 #You should have received a copy of the GNU Affero General Public License
  15 #along with this program. If not, see <http://www.gnu.org/licenses/>.
  16 #In the "official" distribution you can find the license in agpl-3.0.txt.
  17
  18 import L10n
  19 _ = L10n.get_translation()
  20
  21 # This code is based heavily on stars-support-hh-split.py by Mika Boström
  22
  23 import os
  24 import sys
  25 import re
  26 import codecs
  27 import Options
  28 import Configuration
  29 from Exceptions import *
  30 from cStringIO import StringIO
  31
  32 (options, argv) = Options.fpdb_options()
  33
  34 __ARCHIVE_PRE_HEADER_REGEX='^Hand #(\d+)\s*$|\*{20}\s#\s\d+\s\*+\s+'
  35 re_SplitArchive = re.compile(__ARCHIVE_PRE_HEADER_REGEX)
  36 codepage = ["utf-16", "utf-8", "cp1252"]
  37
  38
  39 class SplitHandHistory:
  40     def __init__(self, config, in_path = '-', out_path = None, hands = 100, filter = "PokerStarsToFpdb", archive = False):
  41         self.config = config
  42         self.in_path = in_path
  43         self.out_path = out_path
  44         if not self.out_path:
  45             self.out_path = os.path.dirname(self.in_path)
  46         self.hands = hands
  47         self.archive = archive
  48         self.re_SplitHands = None
  49         self.line_delimiter = None
  50         self.line_addendum = None
  51         self.filedone = False
  52
  53         #Acquire re_SplitHands for this hh
  54         filter_name = filter.replace("ToFpdb", "")
  55         mod = __import__(filter)
  56         obj = getattr(mod, filter_name, None)
  57         self.re_SplitHands = obj.re_SplitHands
  58
  59         #Determine line delimiter type if any
  60         if self.re_SplitHands.match('\n\n'):
  61             self.line_delimiter = '\n\n'
  62         if self.re_SplitHands.match('\n\n\n'):
  63             self.line_delimiter = '\n\n\n'
  64
  65         #Add new line addendum for sites which match SplitHand to next line as well
  66         if filter_name == 'OnGame':
  67             self.line_addendum = '*'
  68         if filter_name == 'Carbon':
  69             self.line_addendum = '<game'
  70
  71         #Open the gargantuan file
  72         for kodec in self.__listof(codepage):
  73             try:
  74                 infile = codecs.open(self.in_path, 'r', kodec)
  75             except IOError:
  76                 print (_('File not found'))
  77                 sys.exit(2)
  78
  79         #Split with do_hands_per_file if archive and paragraphs if a regular hh
  80         if self.archive:
  81             nn = 0
  82             while True:
  83                 nn += 1
  84                 check = self.do_hands_per_file(infile, nn)
  85                 if check is None:
  86                     print (_('%s processed') % self.in_path)
  87                     break
  88         else:
  89             filenum = 0
  90             while not self.filedone:
  91                 filenum += 1
  92                 outfile = self.new_file(filenum)
  93                 handnum = 0
  94                 for hand in self.paragraphs(infile, None, self.line_addendum):
  95                     outfile.write(hand)
  96                     if self.line_delimiter:
  97                         outfile.write(self.line_delimiter)
  98                     handnum += 1
  99                     if handnum >= self.hands:
 100                         break
 101                 outfile.close()
 102
 103     def new_file(self, fileno=-1):
 104         if fileno < 1:
 105             print (_('Nope, will not work (fileno=%d)') % fileno)
 106             sys.exit(2)
 107         basename = os.path.splitext(os.path.basename(self.in_path))[0]
 108         name = os.path.join(self.out_path, basename+'-%06d.txt' % fileno)
 109         print ('-> %s' % name)
 110         newfile = file(name, 'w')
 111         return newfile
 112
 113     #Archive Hand Splitter
 114     def do_hands_per_file(self, infile, num=-1):
 115         done = False
 116         n = 0
 117         outfile = self.new_file(num)
 118         while n < self.hands:
 119             try:
 120                 infile = self.next_hand(infile)
 121                 infile = self.process_hand(infile, outfile)
 122             except FpdbEndOfFile:
 123                 done = True
 124                 break
 125             except:
 126                 print _("Unexpected error processing file")
 127                 sys.exit(2)
 128             n += 1
 129         outfile.close()
 130         if not done:
 131             return infile
 132         else:
 133             return None
 134
 135     #Non-Archive Hand Splitter
 136     def paragraphs(self, file, separator=None, addendum=None):
 137         if not callable(separator) and self.line_delimiter:
 138             def separator(line): return line == '\n'
 139         else:
 140             def separator(line): return self.re_SplitHands.search(line)
 141         file_str = StringIO()
 142         print file_str.getvalue()
 143         for line in file:
 144             if separator(line+addendum):
 145                 if file_str.getvalue():
 146                     if not self.line_delimiter:
 147                         file_str.write(line)
 148                     yield file_str.getvalue()
 149                     file_str = None
 150                     file_str = StringIO()
 151             else:
 152                 file_str.write(line)
 153         if file_str.getvalue(): yield file_str.getvalue()
 154         self.filedone = True
 155
 156
 157     # Finds pre-hand header (Hand #<num>)
 158     def next_hand(self, infile):
 159         m = None
 160         while not m:
 161             l = infile.readline()
 162             #print l, len(l)
 163             # Catch EOF
 164             if len(l) == 0:
 165                 raise FpdbEndOfFile(_("End of file reached"))
 166             m = re_SplitArchive.search(l)
 167         # There is an empty line after pre-hand header and actual HH entry
 168         l = infile.readline()
 169
 170         return infile
 171
 172     # Each individual hand is written separately
 173     def process_hand(self, infile=None, outfile=None):
 174         l = infile.readline()
 175         l = l.replace('\r\n', '\n')
 176         outfile.write(l)
 177         l = infile.readline()
 178
 179         while len(l) < 3:
 180             l = infile.readline()
 181
 182         while len(l) > 2:
 183             l = l.replace('\r\n', '\n')
 184             outfile.write(l)
 185             l = infile.readline()
 186         outfile.write(self.line_delimiter)
 187         return infile
 188
 189     def __listof(self, x):
 190         if isinstance(x, list) or isinstance(x, tuple):
 191             return x
 192         else:
 193             return [x]
 194
 195 def main(argv=None):
 196     if argv is None:
 197         argv = sys.argv[1:]
 198
 199     if not options.config:
 200         options.config = Configuration.Config(file = "HUD_config.test.xml")
 201
 202     if options.filename:
 203         SplitHH = SplitHandHistory(options.config, options.filename, options.outpath, options.hands,
 204                                    options.hhc, options.archive)
 205
 206 if __name__ == '__main__':
 207     sys.exit(main())