X-Git-Url: https://git.mdrn.pl/wolnelektury.git/blobdiff_plain/fab2510d92afec657693cdf6887ad9cee8376499..2910709b1c3b5e2b7d526323f8d8ed03b13895e5:/apps/lesmianator/models.py diff --git a/apps/lesmianator/models.py b/apps/lesmianator/models.py index 0a59ddee4..1b53efdcd 100644 --- a/apps/lesmianator/models.py +++ b/apps/lesmianator/models.py @@ -18,7 +18,6 @@ from django.contrib.contenttypes.models import ContentType from django.contrib.contenttypes import generic from django.conf import settings -from librarian import text from catalogue.fields import JSONField from catalogue.models import Book, Tag @@ -47,46 +46,53 @@ class Poem(models.Model): def __unicode__(self): return "%s (%s...)" % (self.slug, self.text[:20]) - @classmethod - def write(cls, continuations=None, length=3, maxlen=1000): - def choose_word(word, continuations): - try: - choices = sum((continuations[word][post] for post in continuations[word])) - r = randint(0, choices - 1) + @staticmethod + def choose_letter(word, continuations): + if word not in continuations: + return u'\n' - for post in continuations[word]: - r -= continuations[word][post] - if r < 0: - return post - except KeyError: - return '' + choices = sum((continuations[word][letter] + for letter in continuations[word])) + r = randint(0, choices - 1) + for letter in continuations[word]: + r -= continuations[word][letter] + if r < 0: + return letter + @classmethod + def write(cls, continuations=None, length=3, min_lines=2, maxlen=1000): if continuations is None: continuations = cls.global_dictionary + if not continuations: + return '' letters = [] word = u'' - empty = -10 - lines = 0 - if not continuations: - maxlen = 0 - # want at least two lines, but let Lesmianator end his stanzas - while (empty < 2 or lines < 2) and maxlen: - letter = choose_word(word, continuations) + + finished_stanza_verses = 0 + current_stanza_verses = 0 + verse_start = True + + char_count = 0 + + # do `min_lines' non-empty verses and then stop, + # but let Lesmianator finish his last stanza. + while finished_stanza_verses < min_lines and char_count < maxlen: + letter = cls.choose_letter(word, continuations) letters.append(letter) - word = word[-length+1:] + letter + word = word[-length + 1:] + letter + char_count += 1 + if letter == u'\n': - # count non-empty lines - if empty == 0: - lines += 1 - # - if lines >= 2: - empty += 1 - lines += 1 + if verse_start: + finished_stanza_verses += current_stanza_verses + current_stanza_verses = 0 + else: + current_stanza_verses += 1 + verse_start = True else: - empty = 0 - maxlen -= 1 + verse_start = False return ''.join(letters).strip() @@ -119,12 +125,13 @@ class Continuations(models.Model): def for_book(cls, book, length=3): # count from this book only output = StringIO() - f = open(book.xml_file.path) - text.transform(f, output, False, ('raw-text',)) - f.close() + wldoc = book.wldocument(parse_dublincore=False) + output = wldoc.as_text(('raw-text',)).get_string() + del wldoc + conts = {} last_word = '' - for letter in output.getvalue().decode('utf-8').strip().lower(): + for letter in output.decode('utf-8').strip().lower(): mydict = conts.setdefault(last_word, {}) mydict.setdefault(letter, 0) mydict[letter] += 1