apps/catalogue/models.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 from collections import namedtuple
   6
   7 from django.db import models
   8 from django.db.models import permalink, Q
   9 import django.dispatch
  10 from django.core.cache import cache
  11 from django.core.files.storage import DefaultStorage
  12 from django.utils.translation import ugettext_lazy as _
  13 from django.contrib.auth.models import User
  14 from django.template.loader import render_to_string
  15 from django.utils.datastructures import SortedDict
  16 from django.utils.safestring import mark_safe
  17 from django.utils.translation import get_language
  18 from django.core.urlresolvers import reverse
  19 from django.db.models.signals import post_save, m2m_changed, pre_delete
  20
  21 from django.conf import settings
  22
  23 from newtagging.models import TagBase, tags_updated
  24 from newtagging import managers
  25 from catalogue.fields import JSONField, OverwritingFileField
  26 from catalogue.utils import create_zip, split_tags
  27 from catalogue.tasks import touch_tag
  28 from shutil import copy
  29 from glob import glob
  30 import re
  31 from os import path
  32
  33
  34 TAG_CATEGORIES = (
  35     ('author', _('author')),
  36     ('epoch', _('epoch')),
  37     ('kind', _('kind')),
  38     ('genre', _('genre')),
  39     ('theme', _('theme')),
  40     ('set', _('set')),
  41     ('book', _('book')),
  42 )
  43
  44 # not quite, but Django wants you to set a timeout
  45 CACHE_FOREVER = 2419200  # 28 days
  46
  47
  48 class TagSubcategoryManager(models.Manager):
  49     def __init__(self, subcategory):
  50         super(TagSubcategoryManager, self).__init__()
  51         self.subcategory = subcategory
  52
  53     def get_query_set(self):
  54         return super(TagSubcategoryManager, self).get_query_set().filter(category=self.subcategory)
  55
  56
  57 class Tag(TagBase):
  58     name = models.CharField(_('name'), max_length=50, db_index=True)
  59     slug = models.SlugField(_('slug'), max_length=120, db_index=True)
  60     sort_key = models.CharField(_('sort key'), max_length=120, db_index=True)
  61     category = models.CharField(_('category'), max_length=50, blank=False, null=False,
  62         db_index=True, choices=TAG_CATEGORIES)
  63     description = models.TextField(_('description'), blank=True)
  64
  65     user = models.ForeignKey(User, blank=True, null=True)
  66     book_count = models.IntegerField(_('book count'), blank=True, null=True)
  67     gazeta_link = models.CharField(blank=True, max_length=240)
  68     wiki_link = models.CharField(blank=True, max_length=240)
  69
  70     created_at    = models.DateTimeField(_('creation date'), auto_now_add=True, db_index=True)
  71     changed_at    = models.DateTimeField(_('creation date'), auto_now=True, db_index=True)
  72
  73     class UrlDeprecationWarning(DeprecationWarning):
  74         pass
  75
  76     categories_rev = {
  77         'autor': 'author',
  78         'epoka': 'epoch',
  79         'rodzaj': 'kind',
  80         'gatunek': 'genre',
  81         'motyw': 'theme',
  82         'polka': 'set',
  83     }
  84     categories_dict = dict((item[::-1] for item in categories_rev.iteritems()))
  85
  86     class Meta:
  87         ordering = ('sort_key',)
  88         verbose_name = _('tag')
  89         verbose_name_plural = _('tags')
  90         unique_together = (("slug", "category"),)
  91
  92     def __unicode__(self):
  93         return self.name
  94
  95     def __repr__(self):
  96         return "Tag(slug=%r)" % self.slug
  97
  98     @permalink
  99     def get_absolute_url(self):
 100         return ('catalogue.views.tagged_object_list', [self.url_chunk])
 101
 102     def has_description(self):
 103         return len(self.description) > 0
 104     has_description.short_description = _('description')
 105     has_description.boolean = True
 106
 107     def get_count(self):
 108         """Returns global book count for book tags, fragment count for themes."""
 109
 110         if self.category == 'book':
 111             # never used
 112             objects = Book.objects.none()
 113         elif self.category == 'theme':
 114             objects = Fragment.tagged.with_all((self,))
 115         else:
 116             objects = Book.tagged.with_all((self,)).order_by()
 117             if self.category != 'set':
 118                 # eliminate descendants
 119                 l_tags = Tag.objects.filter(slug__in=[book.book_tag_slug() for book in objects])
 120                 descendants_keys = [book.pk for book in Book.tagged.with_any(l_tags)]
 121                 if descendants_keys:
 122                     objects = objects.exclude(pk__in=descendants_keys)
 123         return objects.count()
 124
 125     @staticmethod
 126     def get_tag_list(tags):
 127         if isinstance(tags, basestring):
 128             real_tags = []
 129             ambiguous_slugs = []
 130             category = None
 131             deprecated = False
 132             tags_splitted = tags.split('/')
 133             for name in tags_splitted:
 134                 if category:
 135                     real_tags.append(Tag.objects.get(slug=name, category=category))
 136                     category = None
 137                 elif name in Tag.categories_rev:
 138                     category = Tag.categories_rev[name]
 139                 else:
 140                     try:
 141                         real_tags.append(Tag.objects.exclude(category='book').get(slug=name))
 142                         deprecated = True
 143                     except Tag.MultipleObjectsReturned, e:
 144                         ambiguous_slugs.append(name)
 145
 146             if category:
 147                 # something strange left off
 148                 raise Tag.DoesNotExist()
 149             if ambiguous_slugs:
 150                 # some tags should be qualified
 151                 e = Tag.MultipleObjectsReturned()
 152                 e.tags = real_tags
 153                 e.ambiguous_slugs = ambiguous_slugs
 154                 raise e
 155             if deprecated:
 156                 e = Tag.UrlDeprecationWarning()
 157                 e.tags = real_tags
 158                 raise e
 159             return real_tags
 160         else:
 161             return TagBase.get_tag_list(tags)
 162
 163     @property
 164     def url_chunk(self):
 165         return '/'.join((Tag.categories_dict[self.category], self.slug))
 166
 167     @staticmethod
 168     def tags_from_info(info):
 169         from slughifi import slughifi
 170         from sortify import sortify
 171         meta_tags = []
 172         categories = (('kinds', 'kind'), ('genres', 'genre'), ('authors', 'author'), ('epochs', 'epoch'))
 173         for field_name, category in categories:
 174             try:
 175                 tag_names = getattr(info, field_name)
 176             except:
 177                 try:
 178                     tag_names = [getattr(info, category)]
 179                 except:
 180                     # For instance, Pictures do not have 'genre' field.
 181                     continue
 182             for tag_name in tag_names:
 183                 tag_sort_key = tag_name
 184                 if category == 'author':
 185                     tag_sort_key = tag_name.last_name
 186                     tag_name = tag_name.readable()
 187                 tag, created = Tag.objects.get_or_create(slug=slughifi(tag_name), category=category)
 188                 if created:
 189                     tag.name = tag_name
 190                     tag.sort_key = sortify(tag_sort_key.lower())
 191                     tag.save()
 192                 meta_tags.append(tag)
 193         return meta_tags
 194
 195
 196
 197 def get_dynamic_path(media, filename, ext=None, maxlen=100):
 198     from slughifi import slughifi
 199
 200     # how to put related book's slug here?
 201     if not ext:
 202         # BookMedia case
 203         ext = media.formats[media.type].ext
 204     if media is None or not media.name:
 205         name = slughifi(filename.split(".")[0])
 206     else:
 207         name = slughifi(media.name)
 208     return 'book/%s/%s.%s' % (ext, name[:maxlen-len('book/%s/.%s' % (ext, ext))-4], ext)
 209
 210
 211 # TODO: why is this hard-coded ?
 212 def book_upload_path(ext=None, maxlen=100):
 213     return lambda *args: get_dynamic_path(*args, ext=ext, maxlen=maxlen)
 214
 215
 216 def get_customized_pdf_path(book, customizations):
 217     """
 218     Returns a MEDIA_ROOT relative path for a customized pdf. The name will contain a hash of customization options.
 219     """
 220     customizations.sort()
 221     h = hash(tuple(customizations))
 222
 223     pdf_name = '%s-custom-%s' % (book.fileid(), h)
 224     pdf_file = get_dynamic_path(None, pdf_name, ext='pdf')
 225
 226     return pdf_file
 227
 228
 229 def get_existing_customized_pdf(book):
 230     """
 231     Returns a list of paths to generated customized pdf of a book
 232     """
 233     pdf_glob = '%s-custom-' % (book.fileid(),)
 234     pdf_glob = get_dynamic_path(None, pdf_glob, ext='pdf')
 235     pdf_glob = re.sub(r"[.]([a-z0-9]+)$", "*.\\1", pdf_glob)
 236     return glob(path.join(settings.MEDIA_ROOT, pdf_glob))
 237
 238
 239 class BookMedia(models.Model):
 240     FileFormat = namedtuple("FileFormat", "name ext")
 241     formats = SortedDict([
 242         ('mp3', FileFormat(name='MP3', ext='mp3')),
 243         ('ogg', FileFormat(name='Ogg Vorbis', ext='ogg')),
 244         ('daisy', FileFormat(name='DAISY', ext='daisy.zip')),
 245     ])
 246     format_choices = [(k, _('%s file') % t.name)
 247             for k, t in formats.items()]
 248
 249     type        = models.CharField(_('type'), choices=format_choices, max_length="100")
 250     name        = models.CharField(_('name'), max_length="100")
 251     file        = OverwritingFileField(_('file'), upload_to=book_upload_path())
 252     uploaded_at = models.DateTimeField(_('creation date'), auto_now_add=True, editable=False)
 253     extra_info  = JSONField(_('extra information'), default='{}', editable=False)
 254     book = models.ForeignKey('Book', related_name='media')
 255     source_sha1 = models.CharField(null=True, blank=True, max_length=40, editable=False)
 256
 257     def __unicode__(self):
 258         return "%s (%s)" % (self.name, self.file.name.split("/")[-1])
 259
 260     class Meta:
 261         ordering            = ('type', 'name')
 262         verbose_name        = _('book media')
 263         verbose_name_plural = _('book media')
 264
 265     def save(self, *args, **kwargs):
 266         from slughifi import slughifi
 267         from catalogue.utils import ExistingFile, remove_zip
 268
 269         try:
 270             old = BookMedia.objects.get(pk=self.pk)
 271         except BookMedia.DoesNotExist, e:
 272             pass
 273         else:
 274             # if name changed, change the file name, too
 275             if slughifi(self.name) != slughifi(old.name):
 276                 self.file.save(None, ExistingFile(self.file.path), save=False, leave=True)
 277
 278         super(BookMedia, self).save(*args, **kwargs)
 279
 280         # remove the zip package for book with modified media
 281         remove_zip(self.book.fileid())
 282
 283         extra_info = self.get_extra_info_value()
 284         extra_info.update(self.read_meta())
 285         self.set_extra_info_value(extra_info)
 286         self.source_sha1 = self.read_source_sha1(self.file.path, self.type)
 287         return super(BookMedia, self).save(*args, **kwargs)
 288
 289     def read_meta(self):
 290         """
 291             Reads some metadata from the audiobook.
 292         """
 293         import mutagen
 294         from mutagen import id3
 295
 296         artist_name = director_name = project = funded_by = ''
 297         if self.type == 'mp3':
 298             try:
 299                 audio = id3.ID3(self.file.path)
 300                 artist_name = ', '.join(', '.join(tag.text) for tag in audio.getall('TPE1'))
 301                 director_name = ', '.join(', '.join(tag.text) for tag in audio.getall('TPE3'))
 302                 project = ", ".join([t.data for t in audio.getall('PRIV')
 303                         if t.owner=='wolnelektury.pl?project'])
 304                 funded_by = ", ".join([t.data for t in audio.getall('PRIV')
 305                         if t.owner=='wolnelektury.pl?funded_by'])
 306             except:
 307                 pass
 308         elif self.type == 'ogg':
 309             try:
 310                 audio = mutagen.File(self.file.path)
 311                 artist_name = ', '.join(audio.get('artist', []))
 312                 director_name = ', '.join(audio.get('conductor', []))
 313                 project = ", ".join(audio.get('project', []))
 314                 funded_by = ", ".join(audio.get('funded_by', []))
 315             except:
 316                 pass
 317         else:
 318             return {}
 319         return {'artist_name': artist_name, 'director_name': director_name,
 320                 'project': project, 'funded_by': funded_by}
 321
 322     @staticmethod
 323     def read_source_sha1(filepath, filetype):
 324         """
 325             Reads source file SHA1 from audiobok metadata.
 326         """
 327         import mutagen
 328         from mutagen import id3
 329
 330         if filetype == 'mp3':
 331             try:
 332                 audio = id3.ID3(filepath)
 333                 return [t.data for t in audio.getall('PRIV')
 334                         if t.owner=='wolnelektury.pl?flac_sha1'][0]
 335             except:
 336                 return None
 337         elif filetype == 'ogg':
 338             try:
 339                 audio = mutagen.File(filepath)
 340                 return audio.get('flac_sha1', [None])[0]
 341             except:
 342                 return None
 343         else:
 344             return None
 345
 346
 347 class Book(models.Model):
 348     title         = models.CharField(_('title'), max_length=120)
 349     sort_key = models.CharField(_('sort key'), max_length=120, db_index=True, editable=False)
 350     slug          = models.SlugField(_('slug'), max_length=120, db_index=True)
 351     language = models.CharField(_('language code'), max_length=3, db_index=True,
 352                     default=settings.CATALOGUE_DEFAULT_LANGUAGE)
 353     description   = models.TextField(_('description'), blank=True)
 354     created_at    = models.DateTimeField(_('creation date'), auto_now_add=True, db_index=True)
 355     changed_at    = models.DateTimeField(_('creation date'), auto_now=True, db_index=True)
 356     parent_number = models.IntegerField(_('parent number'), default=0)
 357     extra_info    = JSONField(_('extra information'), default='{}')
 358     gazeta_link   = models.CharField(blank=True, max_length=240)
 359     wiki_link     = models.CharField(blank=True, max_length=240)
 360     # files generated during publication
 361
 362     cover = models.FileField(_('cover'), upload_to=book_upload_path('png'),
 363                 null=True, blank=True)
 364     ebook_formats = ['pdf', 'epub', 'mobi', 'txt']
 365     formats = ebook_formats + ['html', 'xml']
 366
 367     parent        = models.ForeignKey('self', blank=True, null=True, related_name='children')
 368     objects  = models.Manager()
 369     tagged   = managers.ModelTaggedItemManager(Tag)
 370     tags     = managers.TagDescriptor(Tag)
 371
 372     html_built = django.dispatch.Signal()
 373     published = django.dispatch.Signal()
 374
 375     URLID_RE = r'[a-z0-9-]+(?:/[a-z]{3})?'
 376     FILEID_RE = r'[a-z0-9-]+(?:_[a-z]{3})?'
 377
 378     class AlreadyExists(Exception):
 379         pass
 380
 381     class Meta:
 382         unique_together = [['slug', 'language']]
 383         ordering = ('sort_key',)
 384         verbose_name = _('book')
 385         verbose_name_plural = _('books')
 386
 387     def __unicode__(self):
 388         return self.title
 389
 390     def urlid(self, sep='/'):
 391         stem = self.slug
 392         if self.language != settings.CATALOGUE_DEFAULT_LANGUAGE:
 393             stem += sep + self.language
 394         return stem
 395
 396     def fileid(self):
 397         return self.urlid('_')
 398
 399     @staticmethod
 400     def split_urlid(urlid, sep='/', default_lang=settings.CATALOGUE_DEFAULT_LANGUAGE):
 401         """Splits a URL book id into slug and language code.
 402
 403         Returns a dictionary usable i.e. for object lookup, or None.
 404
 405         >>> Book.split_urlid("a-slug/pol", default_lang="eng")
 406         {'slug': 'a-slug', 'language': 'pol'}
 407         >>> Book.split_urlid("a-slug", default_lang="eng")
 408         {'slug': 'a-slug', 'language': 'eng'}
 409         >>> Book.split_urlid("a-slug_pol", "_", default_lang="eng")
 410         {'slug': 'a-slug', 'language': 'pol'}
 411         >>> Book.split_urlid("a-slug/eng", default_lang="eng")
 412
 413         """
 414         parts = urlid.rsplit(sep, 1)
 415         if len(parts) == 2:
 416             if parts[1] == default_lang:
 417                 return None
 418             return {'slug': parts[0], 'language': parts[1]}
 419         else:
 420             return {'slug': urlid, 'language': default_lang}
 421
 422     @classmethod
 423     def split_fileid(cls, fileid):
 424         return cls.split_urlid(fileid, '_')
 425
 426     def save(self, force_insert=False, force_update=False, reset_short_html=True, **kwargs):
 427         from sortify import sortify
 428
 429         self.sort_key = sortify(self.title)
 430
 431         ret = super(Book, self).save(force_insert, force_update)
 432
 433         if reset_short_html:
 434             self.reset_short_html()
 435
 436         return ret
 437
 438     @permalink
 439     def get_absolute_url(self):
 440         return ('catalogue.views.book_detail', [self.urlid()])
 441
 442     @property
 443     def name(self):
 444         return self.title
 445
 446     def book_tag_slug(self):
 447         stem = 'l-' + self.slug
 448         if self.language != settings.CATALOGUE_DEFAULT_LANGUAGE:
 449             return stem[:116] + ' ' + self.language
 450         else:
 451             return stem[:120]
 452
 453     def book_tag(self):
 454         slug = self.book_tag_slug()
 455         book_tag, created = Tag.objects.get_or_create(slug=slug, category='book')
 456         if created:
 457             book_tag.name = self.title[:50]
 458             book_tag.sort_key = self.title.lower()
 459             book_tag.save()
 460         return book_tag
 461
 462     def has_media(self, type):
 463         if type in Book.formats:
 464             return bool(getattr(self, "%s_file" % type))
 465         else:
 466             return self.media.filter(type=type).exists()
 467
 468     def get_media(self, type):
 469         if self.has_media(type):
 470             if type in Book.formats:
 471                 return getattr(self, "%s_file" % type)
 472             else:
 473                 return self.media.filter(type=type)
 474         else:
 475             return None
 476
 477     def get_mp3(self):
 478         return self.get_media("mp3")
 479     def get_odt(self):
 480         return self.get_media("odt")
 481     def get_ogg(self):
 482         return self.get_media("ogg")
 483     def get_daisy(self):
 484         return self.get_media("daisy")
 485
 486     def reset_short_html(self):
 487         if self.id is None:
 488             return
 489
 490         cache_key = "Book.short_html/%d/%s"
 491         for lang, langname in settings.LANGUAGES:
 492             cache.delete(cache_key % (self.id, lang))
 493         cache.delete("Book.mini_box/%d" % (self.id, ))
 494         # Fragment.short_html relies on book's tags, so reset it here too
 495         for fragm in self.fragments.all():
 496             fragm.reset_short_html()
 497
 498     def short_html(self):
 499         if self.id:
 500             cache_key = "Book.short_html/%d/%s" % (self.id, get_language())
 501             short_html = cache.get(cache_key)
 502         else:
 503             short_html = None
 504
 505         if short_html is not None:
 506             return mark_safe(short_html)
 507         else:
 508             tags = self.tags.filter(category__in=('author', 'kind', 'genre', 'epoch'))
 509             tags = split_tags(tags)
 510
 511             formats = []
 512             # files generated during publication
 513             for ebook_format in self.ebook_formats:
 514                 if self.has_media(ebook_format):
 515                     formats.append(u'<a href="%s">%s</a>' % (
 516                         self.get_media(ebook_format).url,
 517                         ebook_format.upper()
 518                     ))
 519
 520             formats = [mark_safe(format) for format in formats]
 521
 522             short_html = unicode(render_to_string('catalogue/book_short.html',
 523                 {'book': self, 'tags': tags, 'formats': formats}))
 524
 525             if self.id:
 526                 cache.set(cache_key, short_html, CACHE_FOREVER)
 527             return mark_safe(short_html)
 528
 529     def mini_box(self):
 530         if self.id:
 531             cache_key = "Book.mini_box/%d" % (self.id, )
 532             short_html = cache.get(cache_key)
 533         else:
 534             short_html = None
 535
 536         if short_html is None:
 537             authors = self.tags.filter(category='author')
 538
 539             short_html = unicode(render_to_string('catalogue/book_mini_box.html',
 540                 {'book': self, 'authors': authors, 'STATIC_URL': settings.STATIC_URL}))
 541
 542             if self.id:
 543                 cache.set(cache_key, short_html, CACHE_FOREVER)
 544         return mark_safe(short_html)
 545
 546     def has_description(self):
 547         return len(self.description) > 0
 548     has_description.short_description = _('description')
 549     has_description.boolean = True
 550
 551     # ugly ugly ugly
 552     def has_mp3_file(self):
 553         return bool(self.has_media("mp3"))
 554     has_mp3_file.short_description = 'MP3'
 555     has_mp3_file.boolean = True
 556
 557     def has_ogg_file(self):
 558         return bool(self.has_media("ogg"))
 559     has_ogg_file.short_description = 'OGG'
 560     has_ogg_file.boolean = True
 561
 562     def has_daisy_file(self):
 563         return bool(self.has_media("daisy"))
 564     has_daisy_file.short_description = 'DAISY'
 565     has_daisy_file.boolean = True
 566
 567     def wldocument(self, parse_dublincore=True):
 568         from catalogue.utils import ORMDocProvider
 569         from librarian.parser import WLDocument
 570
 571         return WLDocument.from_file(self.xml_file.path,
 572                 provider=ORMDocProvider(self),
 573                 parse_dublincore=parse_dublincore)
 574
 575     def build_cover(self, book_info=None):
 576         """(Re)builds the cover image."""
 577         from StringIO import StringIO
 578         from django.core.files.base import ContentFile
 579         from librarian.cover import WLCover
 580
 581         if book_info is None:
 582             book_info = self.wldocument().book_info
 583
 584         cover = WLCover(book_info).image()
 585         imgstr = StringIO()
 586         cover.save(imgstr, 'png')
 587         self.cover.save(None, ContentFile(imgstr.getvalue()))
 588
 589     def build_pdf(self, customizations=None, file_name=None):
 590         """ (Re)builds the pdf file.
 591         customizations - customizations which are passed to LaTeX class file.
 592         file_name - save the pdf file under a different name and DO NOT save it in db.
 593         """
 594         from os import unlink
 595         from django.core.files import File
 596         from catalogue.utils import remove_zip
 597
 598         pdf = self.wldocument().as_pdf(customizations=customizations)
 599
 600         if file_name is None:
 601             # we'd like to be sure not to overwrite changes happening while
 602             # (timely) pdf generation is taking place (async celery scenario)
 603             current_self = Book.objects.get(id=self.id)
 604             current_self.pdf_file.save('%s.pdf' % self.fileid(),
 605                     File(open(pdf.get_filename())))
 606             self.pdf_file = current_self.pdf_file
 607
 608             # remove cached downloadables
 609             remove_zip(settings.ALL_PDF_ZIP)
 610
 611             for customized_pdf in get_existing_customized_pdf(self):
 612                 unlink(customized_pdf)
 613         else:
 614             print "saving %s" % file_name
 615             print "to: %s" % DefaultStorage().path(file_name)
 616             DefaultStorage().save(file_name, File(open(pdf.get_filename())))
 617
 618     def build_mobi(self):
 619         """ (Re)builds the MOBI file.
 620
 621         """
 622         from django.core.files import File
 623         from catalogue.utils import remove_zip
 624
 625         mobi = self.wldocument().as_mobi()
 626
 627         self.mobi_file.save('%s.mobi' % self.fileid(), File(open(mobi.get_filename())))
 628
 629         # remove zip with all mobi files
 630         remove_zip(settings.ALL_MOBI_ZIP)
 631
 632     def build_epub(self):
 633         """(Re)builds the epub file."""
 634         from django.core.files import File
 635         from catalogue.utils import remove_zip
 636
 637         epub = self.wldocument().as_epub()
 638
 639         self.epub_file.save('%s.epub' % self.fileid(),
 640                 File(open(epub.get_filename())))
 641
 642         # remove zip package with all epub files
 643         remove_zip(settings.ALL_EPUB_ZIP)
 644
 645     def build_txt(self):
 646         from django.core.files.base import ContentFile
 647
 648         text = self.wldocument().as_text()
 649         self.txt_file.save('%s.txt' % self.fileid(), ContentFile(text.get_string()))
 650
 651
 652     def build_html(self):
 653         from markupstring import MarkupString
 654         from django.core.files.base import ContentFile
 655         from slughifi import slughifi
 656         from librarian import html
 657
 658         meta_tags = list(self.tags.filter(
 659             category__in=('author', 'epoch', 'genre', 'kind')))
 660         book_tag = self.book_tag()
 661
 662         html_output = self.wldocument(parse_dublincore=False).as_html()
 663         if html_output:
 664             self.html_file.save('%s.html' % self.fileid(),
 665                     ContentFile(html_output.get_string()))
 666
 667             # get ancestor l-tags for adding to new fragments
 668             ancestor_tags = []
 669             p = self.parent
 670             while p:
 671                 ancestor_tags.append(p.book_tag())
 672                 p = p.parent
 673
 674             # Delete old fragments and create them from scratch
 675             self.fragments.all().delete()
 676             # Extract fragments
 677             closed_fragments, open_fragments = html.extract_fragments(self.html_file.path)
 678             for fragment in closed_fragments.values():
 679                 try:
 680                     theme_names = [s.strip() for s in fragment.themes.split(',')]
 681                 except AttributeError:
 682                     continue
 683                 themes = []
 684                 for theme_name in theme_names:
 685                     if not theme_name:
 686                         continue
 687                     tag, created = Tag.objects.get_or_create(slug=slughifi(theme_name), category='theme')
 688                     if created:
 689                         tag.name = theme_name
 690                         tag.sort_key = theme_name.lower()
 691                         tag.save()
 692                     themes.append(tag)
 693                 if not themes:
 694                     continue
 695
 696                 text = fragment.to_string()
 697                 short_text = ''
 698                 if (len(MarkupString(text)) > 240):
 699                     short_text = unicode(MarkupString(text)[:160])
 700                 new_fragment = Fragment.objects.create(anchor=fragment.id, book=self,
 701                     text=text, short_text=short_text)
 702
 703                 new_fragment.save()
 704                 new_fragment.tags = set(meta_tags + themes + [book_tag] + ancestor_tags)
 705             self.save()
 706             self.html_built.send(sender=self)
 707             return True
 708         return False
 709
 710     @staticmethod
 711     def zip_format(format_):
 712         def pretty_file_name(book):
 713             return "%s/%s.%s" % (
 714                 b.get_extra_info_value()['author'],
 715                 b.fileid(),
 716                 format_)
 717
 718         field_name = "%s_file" % format_
 719         books = Book.objects.filter(parent=None).exclude(**{field_name: ""})
 720         paths = [(pretty_file_name(b), getattr(b, field_name).path)
 721                     for b in books]
 722         result = create_zip.delay(paths,
 723                     getattr(settings, "ALL_%s_ZIP" % format_.upper()))
 724         return result.wait()
 725
 726     def zip_audiobooks(self, format_):
 727         bm = BookMedia.objects.filter(book=self, type=format_)
 728         paths = map(lambda bm: (None, bm.file.path), bm)
 729         result = create_zip.delay(paths, "%s_%s" % (self.fileid(), format_))
 730         return result.wait()
 731
 732     @classmethod
 733     def from_xml_file(cls, xml_file, **kwargs):
 734         from django.core.files import File
 735         from librarian import dcparser
 736
 737         # use librarian to parse meta-data
 738         book_info = dcparser.parse(xml_file)
 739
 740         if not isinstance(xml_file, File):
 741             xml_file = File(open(xml_file))
 742
 743         try:
 744             return cls.from_text_and_meta(xml_file, book_info, **kwargs)
 745         finally:
 746             xml_file.close()
 747
 748     @classmethod
 749     def from_text_and_meta(cls, raw_file, book_info, overwrite=False,
 750             build_epub=True, build_txt=True, build_pdf=True, build_mobi=True):
 751         import re
 752         from sortify import sortify
 753
 754         # check for parts before we do anything
 755         children = []
 756         if hasattr(book_info, 'parts'):
 757             for part_url in book_info.parts:
 758                 try:
 759                     children.append(Book.objects.get(
 760                         slug=part_url.slug, language=part_url.language))
 761                 except Book.DoesNotExist, e:
 762                     raise Book.DoesNotExist(_('Book "%s/%s" does not exist.') %
 763                             (part_url.slug, part_url.language))
 764
 765
 766         # Read book metadata
 767         book_slug = book_info.url.slug
 768         language = book_info.language
 769         if re.search(r'[^a-zA-Z0-9-]', book_slug):
 770             raise ValueError('Invalid characters in slug')
 771         book, created = Book.objects.get_or_create(slug=book_slug, language=language)
 772
 773         if created:
 774             book_shelves = []
 775         else:
 776             if not overwrite:
 777                 raise Book.AlreadyExists(_('Book %s/%s already exists') % (
 778                         book_slug, language))
 779             # Save shelves for this book
 780             book_shelves = list(book.tags.filter(category='set'))
 781
 782         book.title = book_info.title
 783         book.set_extra_info_value(book_info.to_dict())
 784         book.save()
 785
 786         meta_tags = Tag.tags_from_info(book_info)
 787
 788         book.tags = set(meta_tags + book_shelves)
 789
 790         book_tag = book.book_tag()
 791
 792         for n, child_book in enumerate(children):
 793             child_book.parent = book
 794             child_book.parent_number = n
 795             child_book.save()
 796
 797         # Save XML and HTML files
 798         book.xml_file.save('%s.xml' % book.slug, raw_file, save=False)
 799
 800         # delete old fragments when overwriting
 801         book.fragments.all().delete()
 802
 803         if book.build_html():
 804             if not settings.NO_BUILD_TXT and build_txt:
 805                 book.build_txt()
 806
 807         book.build_cover(book_info)
 808
 809         if not settings.NO_BUILD_EPUB and build_epub:
 810             book.build_epub()
 811
 812         if not settings.NO_BUILD_PDF and build_pdf:
 813             book.build_pdf()
 814
 815         if not settings.NO_BUILD_MOBI and build_mobi:
 816             book.build_mobi()
 817
 818         book_descendants = list(book.children.all())
 819         descendants_tags = set()
 820         # add l-tag to descendants and their fragments
 821         while len(book_descendants) > 0:
 822             child_book = book_descendants.pop(0)
 823             descendants_tags.update(child_book.tags)
 824             child_book.tags = list(child_book.tags) + [book_tag]
 825             child_book.save()
 826             for fragment in child_book.fragments.all():
 827                 fragment.tags = set(list(fragment.tags) + [book_tag])
 828             book_descendants += list(child_book.children.all())
 829
 830         for tag in descendants_tags:
 831             touch_tag.delay(tag)
 832
 833         book.save()
 834
 835         # refresh cache
 836         book.reset_tag_counter()
 837         book.reset_theme_counter()
 838
 839         cls.published.send(sender=book)
 840         return book
 841
 842     def reset_tag_counter(self):
 843         if self.id is None:
 844             return
 845
 846         cache_key = "Book.tag_counter/%d" % self.id
 847         cache.delete(cache_key)
 848         if self.parent:
 849             self.parent.reset_tag_counter()
 850
 851     @property
 852     def tag_counter(self):
 853         if self.id:
 854             cache_key = "Book.tag_counter/%d" % self.id
 855             tags = cache.get(cache_key)
 856         else:
 857             tags = None
 858
 859         if tags is None:
 860             tags = {}
 861             for child in self.children.all().order_by():
 862                 for tag_pk, value in child.tag_counter.iteritems():
 863                     tags[tag_pk] = tags.get(tag_pk, 0) + value
 864             for tag in self.tags.exclude(category__in=('book', 'theme', 'set')).order_by():
 865                 tags[tag.pk] = 1
 866
 867             if self.id:
 868                 cache.set(cache_key, tags, CACHE_FOREVER)
 869         return tags
 870
 871     def reset_theme_counter(self):
 872         if self.id is None:
 873             return
 874
 875         cache_key = "Book.theme_counter/%d" % self.id
 876         cache.delete(cache_key)
 877         if self.parent:
 878             self.parent.reset_theme_counter()
 879
 880     @property
 881     def theme_counter(self):
 882         if self.id:
 883             cache_key = "Book.theme_counter/%d" % self.id
 884             tags = cache.get(cache_key)
 885         else:
 886             tags = None
 887
 888         if tags is None:
 889             tags = {}
 890             for fragment in Fragment.tagged.with_any([self.book_tag()]).order_by():
 891                 for tag in fragment.tags.filter(category='theme').order_by():
 892                     tags[tag.pk] = tags.get(tag.pk, 0) + 1
 893
 894             if self.id:
 895                 cache.set(cache_key, tags, CACHE_FOREVER)
 896         return tags
 897
 898     def pretty_title(self, html_links=False):
 899         book = self
 900         names = list(book.tags.filter(category='author'))
 901
 902         books = []
 903         while book:
 904             books.append(book)
 905             book = book.parent
 906         names.extend(reversed(books))
 907
 908         if html_links:
 909             names = ['<a href="%s">%s</a>' % (tag.get_absolute_url(), tag.name) for tag in names]
 910         else:
 911             names = [tag.name for tag in names]
 912
 913         return ', '.join(names)
 914
 915     @classmethod
 916     def tagged_top_level(cls, tags):
 917         """ Returns top-level books tagged with `tags'.
 918
 919         It only returns those books which don't have ancestors which are
 920         also tagged with those tags.
 921
 922         """
 923         # get relevant books and their tags
 924         objects = cls.tagged.with_all(tags)
 925         # eliminate descendants
 926         l_tags = Tag.objects.filter(category='book', slug__in=[book.book_tag_slug() for book in objects])
 927         descendants_keys = [book.pk for book in cls.tagged.with_any(l_tags)]
 928         if descendants_keys:
 929             objects = objects.exclude(pk__in=descendants_keys)
 930
 931         return objects
 932
 933     @classmethod
 934     def book_list(cls, filter=None):
 935         """Generates a hierarchical listing of all books.
 936
 937         Books are optionally filtered with a test function.
 938
 939         """
 940
 941         books_by_parent = {}
 942         books = cls.objects.all().order_by('parent_number', 'sort_key').only(
 943                 'title', 'parent', 'slug', 'language')
 944         if filter:
 945             books = books.filter(filter).distinct()
 946             book_ids = set((book.pk for book in books))
 947             for book in books:
 948                 parent = book.parent_id
 949                 if parent not in book_ids:
 950                     parent = None
 951                 books_by_parent.setdefault(parent, []).append(book)
 952         else:
 953             for book in books:
 954                 books_by_parent.setdefault(book.parent_id, []).append(book)
 955
 956         orphans = []
 957         books_by_author = SortedDict()
 958         for tag in Tag.objects.filter(category='author'):
 959             books_by_author[tag] = []
 960
 961         for book in books_by_parent.get(None,()):
 962             authors = list(book.tags.filter(category='author'))
 963             if authors:
 964                 for author in authors:
 965                     books_by_author[author].append(book)
 966             else:
 967                 orphans.append(book)
 968
 969         return books_by_author, orphans, books_by_parent
 970
 971     _audiences_pl = {
 972         "SP1": (1, u"szkoła podstawowa"),
 973         "SP2": (1, u"szkoła podstawowa"),
 974         "P": (1, u"szkoła podstawowa"),
 975         "G": (2, u"gimnazjum"),
 976         "L": (3, u"liceum"),
 977         "LP": (3, u"liceum"),
 978     }
 979     def audiences_pl(self):
 980         audiences = self.get_extra_info_value().get('audiences', [])
 981         audiences = sorted(set([self._audiences_pl[a] for a in audiences]))
 982         return [a[1] for a in audiences]
 983
 984
 985 def _has_factory(ftype):
 986     has = lambda self: bool(getattr(self, "%s_file" % ftype))
 987     has.short_description = t.upper()
 988     has.boolean = True
 989     has.__name__ = "has_%s_file" % ftype
 990     return has
 991
 992
 993 # add the file fields
 994 for t in Book.formats:
 995     field_name = "%s_file" % t
 996     models.FileField(_("%s file" % t.upper()),
 997             upload_to=book_upload_path(t),
 998             blank=True).contribute_to_class(Book, field_name)
 999
1000     setattr(Book, "has_%s_file" % t, _has_factory(t))
1001
1002
1003 class Fragment(models.Model):
1004     text = models.TextField()
1005     short_text = models.TextField(editable=False)
1006     anchor = models.CharField(max_length=120)
1007     book = models.ForeignKey(Book, related_name='fragments')
1008
1009     objects = models.Manager()
1010     tagged = managers.ModelTaggedItemManager(Tag)
1011     tags = managers.TagDescriptor(Tag)
1012
1013     class Meta:
1014         ordering = ('book', 'anchor',)
1015         verbose_name = _('fragment')
1016         verbose_name_plural = _('fragments')
1017
1018     def get_absolute_url(self):
1019         return '%s#m%s' % (self.book.get_html_url(), self.anchor)
1020
1021     def reset_short_html(self):
1022         if self.id is None:
1023             return
1024
1025         cache_key = "Fragment.short_html/%d/%s"
1026         for lang, langname in settings.LANGUAGES:
1027             cache.delete(cache_key % (self.id, lang))
1028
1029     def short_html(self):
1030         if self.id:
1031             cache_key = "Fragment.short_html/%d/%s" % (self.id, get_language())
1032             short_html = cache.get(cache_key)
1033         else:
1034             short_html = None
1035
1036         if short_html is not None:
1037             return mark_safe(short_html)
1038         else:
1039             short_html = unicode(render_to_string('catalogue/fragment_short.html',
1040                 {'fragment': self}))
1041             if self.id:
1042                 cache.set(cache_key, short_html, CACHE_FOREVER)
1043             return mark_safe(short_html)
1044
1045
1046 ###########
1047 #
1048 # SIGNALS
1049 #
1050 ###########
1051
1052
1053 def _tags_updated_handler(sender, affected_tags, **kwargs):
1054     # reset tag global counter
1055     # we want Tag.changed_at updated for API to know the tag was touched
1056     for tag in affected_tags:
1057         touch_tag.delay(tag)
1058
1059     # if book tags changed, reset book tag counter
1060     if isinstance(sender, Book) and \
1061                 Tag.objects.filter(pk__in=(tag.pk for tag in affected_tags)).\
1062                     exclude(category__in=('book', 'theme', 'set')).count():
1063         sender.reset_tag_counter()
1064     # if fragment theme changed, reset book theme counter
1065     elif isinstance(sender, Fragment) and \
1066                 Tag.objects.filter(pk__in=(tag.pk for tag in affected_tags)).\
1067                     filter(category='theme').count():
1068         sender.book.reset_theme_counter()
1069 tags_updated.connect(_tags_updated_handler)
1070
1071
1072 def _pre_delete_handler(sender, instance, **kwargs):
1073     """ refresh Book on BookMedia delete """
1074     if sender == BookMedia:
1075         instance.book.save()
1076 pre_delete.connect(_pre_delete_handler)
1077
1078 def _post_save_handler(sender, instance, **kwargs):
1079     """ refresh all the short_html stuff on BookMedia update """
1080     if sender == BookMedia:
1081         instance.book.save()
1082 post_save.connect(_post_save_handler)