# This file is part of Wolnelektury, licensed under GNU Affero GPLv3 or later.
# Copyright © Fundacja Nowoczesna Polska. See NOTICE for more information.
#
+import re
import sys
from cPickle import load, dump
from optparse import make_option
from catalogue.models import Book, Tag
+# extract text from text file
+re_text = re_text = re.compile(r'\n{3,}(.*?)\n*-----\n', re.S).search
+
class Command(BaseCommand):
option_list = BaseCommand.option_list + (
print self.style.NOTICE('%s has no TXT file' % book.slug)
skipped += 1
continue
- processed += 1
- last_word = ''
- for number, line in enumerate(book.txt_file):
- if number < 17:
+ with open(book.txt_file.path) as f:
+ m = re_text(f.read())
+ if not m:
+ print self.style.ERROR("Unknown text format: %s" % book.slug)
+ skipped += 1
continue
- line = unicode(line, 'utf-8').lower()
- for letter in line:
+
+ processed += 1
+ last_word = ''
+ text = unicode(m.group(1), 'utf-8').lower()
+ for letter in text:
mydict = lesmianator.setdefault(last_word, {})
myval = mydict.setdefault(letter, 0)
mydict[letter] += 1
try:
dump(lesmianator, open(path, 'w'))
except:
- print self.style.ERROR("Counldn't write to $s" % path)
+ print self.style.ERROR("Couldn't write to $s" % path)
return
dump(lesmianator, open(path, 'w'))
if verbose >= 1:
print "%d processed, %d skipped" % (processed, skipped)
- print "Results dumped do %s" % path
+ print "Results dumped to %s" % path