Search in stagin phase - waiting for final guidelines on new wl
[wolnelektury.git] / apps / lesmianator / management / commands / lesmianator.py
1 # -*- coding: utf-8 -*-
2 # This file is part of Wolnelektury, licensed under GNU Affero GPLv3 or later.
3 # Copyright © Fundacja Nowoczesna Polska. See NOTICE for more information.
4 #
5 import re
6 import sys
7 from cPickle import load, dump
8 from optparse import make_option
9
10 from django.core.management.base import BaseCommand
11 from django.core.management.color import color_style
12 from django.conf import settings
13
14 from catalogue.models import Book, Tag
15
16 # extract text from text file
17 re_text = re.compile(r'\n{3,}(.*?)\n*-----\n', re.S).search
18
19
20 class Command(BaseCommand):
21     option_list = BaseCommand.option_list + (
22         make_option('-t', '--tags', dest='tags', metavar='SLUG,...',
23             help='Use only books tagged with this tags'),
24         make_option('-i', '--include', dest='include', metavar='SLUG,...',
25             help='Include specific books by slug'),
26         make_option('-e', '--exclude', dest='exclude', metavar='SLUG,...',
27             help='Exclude specific books by slug')
28     )
29     help = 'Prepare data for Lesmianator.'
30
31     def handle(self, *args, **options):
32         self.style = color_style()
33         verbose = int(options.get('verbosity'))
34         tags = options.get('tags')
35         include = options.get('include')
36         exclude = options.get('exclude')
37
38         try:
39             path = settings.LESMIANATOR_PICKLE
40         except:
41             print self.style.ERROR('LESMIANATOR_PICKLE not set in the settings.')
42             return
43
44         books = []
45
46         if include:
47             books += list(Book.objects.filter(slug__in=include.split(',')).only('slug', 'txt_file'))
48
49         if tags:
50             books += list(Book.tagged.with_all(Tag.objects.filter(slug__in=tags.split(','))).only('slug', 'txt_file'))
51         elif not include:
52             books = list(Book.objects.all().only('slug', 'txt_file'))
53
54         if exclude:
55             books = [book for book in books if book.slug not in exclude.split(',')]
56
57         books = set(books)
58
59         lesmianator = {}
60         processed = skipped = 0
61         for book in books:
62             if verbose >= 2:
63                 print 'Parsing', book.slug
64             if not book.txt_file:
65                 if verbose >= 1:
66                     print self.style.NOTICE('%s has no TXT file' % book.slug)
67                 skipped += 1
68                 continue
69             f = open(book.txt_file.path)
70             m = re_text(f.read())
71             if not m:
72                 print self.style.ERROR("Unknown text format: %s" % book.slug)
73                 skipped += 1
74                 continue
75
76             processed += 1
77             last_word = ''
78             text = unicode(m.group(1), 'utf-8').lower()
79             for letter in text:
80                 mydict = lesmianator.setdefault(last_word, {})
81                 myval = mydict.setdefault(letter, 0)
82                 mydict[letter] += 1
83                 last_word = last_word[-2:] + letter
84             f.close()
85
86         if not processed:
87             if skipped:
88                 print self.style.ERROR("No books with TXT files found")
89             else:
90                 print self.style.ERROR("No books found")
91             return
92
93         try:
94             dump(lesmianator, open(path, 'w'))
95         except:
96             print self.style.ERROR("Couldn't write to $s" % path)
97             return
98
99         dump(lesmianator, open(path, 'w'))
100         if verbose >= 1:
101             print "%d processed, %d skipped" % (processed, skipped)
102             print "Results dumped to %s" % path