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.slug, 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.slug,)
 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             old = None
 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         if old:
 282             remove_zip("%s_%s" % (old.book.slug, old.type))
 283         remove_zip("%s_%s" % (self.book.slug, self.type))
 284
 285         extra_info = self.get_extra_info_value()
 286         extra_info.update(self.read_meta())
 287         self.set_extra_info_value(extra_info)
 288         self.source_sha1 = self.read_source_sha1(self.file.path, self.type)
 289         return super(BookMedia, self).save(*args, **kwargs)
 290
 291     def read_meta(self):
 292         """
 293             Reads some metadata from the audiobook.
 294         """
 295         import mutagen
 296         from mutagen import id3
 297
 298         artist_name = director_name = project = funded_by = ''
 299         if self.type == 'mp3':
 300             try:
 301                 audio = id3.ID3(self.file.path)
 302                 artist_name = ', '.join(', '.join(tag.text) for tag in audio.getall('TPE1'))
 303                 director_name = ', '.join(', '.join(tag.text) for tag in audio.getall('TPE3'))
 304                 project = ", ".join([t.data for t in audio.getall('PRIV')
 305                         if t.owner=='wolnelektury.pl?project'])
 306                 funded_by = ", ".join([t.data for t in audio.getall('PRIV')
 307                         if t.owner=='wolnelektury.pl?funded_by'])
 308             except:
 309                 pass
 310         elif self.type == 'ogg':
 311             try:
 312                 audio = mutagen.File(self.file.path)
 313                 artist_name = ', '.join(audio.get('artist', []))
 314                 director_name = ', '.join(audio.get('conductor', []))
 315                 project = ", ".join(audio.get('project', []))
 316                 funded_by = ", ".join(audio.get('funded_by', []))
 317             except:
 318                 pass
 319         else:
 320             return {}
 321         return {'artist_name': artist_name, 'director_name': director_name,
 322                 'project': project, 'funded_by': funded_by}
 323
 324     @staticmethod
 325     def read_source_sha1(filepath, filetype):
 326         """
 327             Reads source file SHA1 from audiobok metadata.
 328         """
 329         import mutagen
 330         from mutagen import id3
 331
 332         if filetype == 'mp3':
 333             try:
 334                 audio = id3.ID3(filepath)
 335                 return [t.data for t in audio.getall('PRIV')
 336                         if t.owner=='wolnelektury.pl?flac_sha1'][0]
 337             except:
 338                 return None
 339         elif filetype == 'ogg':
 340             try:
 341                 audio = mutagen.File(filepath)
 342                 return audio.get('flac_sha1', [None])[0]
 343             except:
 344                 return None
 345         else:
 346             return None
 347
 348
 349 class Book(models.Model):
 350     title         = models.CharField(_('title'), max_length=120)
 351     sort_key = models.CharField(_('sort key'), max_length=120, db_index=True, editable=False)
 352     slug = models.SlugField(_('slug'), max_length=120, db_index=True,
 353             unique=True)
 354     common_slug = models.SlugField(_('slug'), max_length=120, db_index=True)
 355     language = models.CharField(_('language code'), max_length=3, db_index=True,
 356                     default=settings.CATALOGUE_DEFAULT_LANGUAGE)
 357     description   = models.TextField(_('description'), blank=True)
 358     created_at    = models.DateTimeField(_('creation date'), auto_now_add=True, db_index=True)
 359     changed_at    = models.DateTimeField(_('creation date'), auto_now=True, db_index=True)
 360     parent_number = models.IntegerField(_('parent number'), default=0)
 361     extra_info    = JSONField(_('extra information'), default='{}')
 362     gazeta_link   = models.CharField(blank=True, max_length=240)
 363     wiki_link     = models.CharField(blank=True, max_length=240)
 364     # files generated during publication
 365
 366     cover = models.FileField(_('cover'), upload_to=book_upload_path('png'),
 367                 null=True, blank=True)
 368     ebook_formats = ['pdf', 'epub', 'mobi', 'txt']
 369     formats = ebook_formats + ['html', 'xml']
 370
 371     parent        = models.ForeignKey('self', blank=True, null=True, related_name='children')
 372     objects  = models.Manager()
 373     tagged   = managers.ModelTaggedItemManager(Tag)
 374     tags     = managers.TagDescriptor(Tag)
 375
 376     html_built = django.dispatch.Signal()
 377     published = django.dispatch.Signal()
 378
 379     class AlreadyExists(Exception):
 380         pass
 381
 382     class Meta:
 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 save(self, force_insert=False, force_update=False, reset_short_html=True, **kwargs):
 391         from sortify import sortify
 392
 393         self.sort_key = sortify(self.title)
 394
 395         ret = super(Book, self).save(force_insert, force_update)
 396
 397         if reset_short_html:
 398             self.reset_short_html()
 399
 400         return ret
 401
 402     @permalink
 403     def get_absolute_url(self):
 404         return ('catalogue.views.book_detail', [self.slug])
 405
 406     @property
 407     def name(self):
 408         return self.title
 409
 410     def book_tag_slug(self):
 411         return ('l-' + self.slug)[:120]
 412
 413     def book_tag(self):
 414         slug = self.book_tag_slug()
 415         book_tag, created = Tag.objects.get_or_create(slug=slug, category='book')
 416         if created:
 417             book_tag.name = self.title[:50]
 418             book_tag.sort_key = self.title.lower()
 419             book_tag.save()
 420         return book_tag
 421
 422     def has_media(self, type):
 423         if type in Book.formats:
 424             return bool(getattr(self, "%s_file" % type))
 425         else:
 426             return self.media.filter(type=type).exists()
 427
 428     def get_media(self, type):
 429         if self.has_media(type):
 430             if type in Book.formats:
 431                 return getattr(self, "%s_file" % type)
 432             else:
 433                 return self.media.filter(type=type)
 434         else:
 435             return None
 436
 437     def get_mp3(self):
 438         return self.get_media("mp3")
 439     def get_odt(self):
 440         return self.get_media("odt")
 441     def get_ogg(self):
 442         return self.get_media("ogg")
 443     def get_daisy(self):
 444         return self.get_media("daisy")
 445
 446     def reset_short_html(self):
 447         if self.id is None:
 448             return
 449
 450         cache_key = "Book.short_html/%d/%s"
 451         for lang, langname in settings.LANGUAGES:
 452             cache.delete(cache_key % (self.id, lang))
 453         cache.delete("Book.mini_box/%d" % (self.id, ))
 454         # Fragment.short_html relies on book's tags, so reset it here too
 455         for fragm in self.fragments.all():
 456             fragm.reset_short_html()
 457
 458     def short_html(self):
 459         if self.id:
 460             cache_key = "Book.short_html/%d/%s" % (self.id, get_language())
 461             short_html = cache.get(cache_key)
 462         else:
 463             short_html = None
 464
 465         if short_html is not None:
 466             return mark_safe(short_html)
 467         else:
 468             tags = self.tags.filter(category__in=('author', 'kind', 'genre', 'epoch'))
 469             tags = split_tags(tags)
 470
 471             formats = []
 472             # files generated during publication
 473             for ebook_format in self.ebook_formats:
 474                 if self.has_media(ebook_format):
 475                     formats.append(u'<a href="%s">%s</a>' % (
 476                         self.get_media(ebook_format).url,
 477                         ebook_format.upper()
 478                     ))
 479
 480             formats = [mark_safe(format) for format in formats]
 481
 482             short_html = unicode(render_to_string('catalogue/book_short.html',
 483                 {'book': self, 'tags': tags, 'formats': formats}))
 484
 485             if self.id:
 486                 cache.set(cache_key, short_html, CACHE_FOREVER)
 487             return mark_safe(short_html)
 488
 489     def mini_box(self):
 490         if self.id:
 491             cache_key = "Book.mini_box/%d" % (self.id, )
 492             short_html = cache.get(cache_key)
 493         else:
 494             short_html = None
 495
 496         if short_html is None:
 497             authors = self.tags.filter(category='author')
 498
 499             short_html = unicode(render_to_string('catalogue/book_mini_box.html',
 500                 {'book': self, 'authors': authors, 'STATIC_URL': settings.STATIC_URL}))
 501
 502             if self.id:
 503                 cache.set(cache_key, short_html, CACHE_FOREVER)
 504         return mark_safe(short_html)
 505
 506     def has_description(self):
 507         return len(self.description) > 0
 508     has_description.short_description = _('description')
 509     has_description.boolean = True
 510
 511     # ugly ugly ugly
 512     def has_mp3_file(self):
 513         return bool(self.has_media("mp3"))
 514     has_mp3_file.short_description = 'MP3'
 515     has_mp3_file.boolean = True
 516
 517     def has_ogg_file(self):
 518         return bool(self.has_media("ogg"))
 519     has_ogg_file.short_description = 'OGG'
 520     has_ogg_file.boolean = True
 521
 522     def has_daisy_file(self):
 523         return bool(self.has_media("daisy"))
 524     has_daisy_file.short_description = 'DAISY'
 525     has_daisy_file.boolean = True
 526
 527     def wldocument(self, parse_dublincore=True):
 528         from catalogue.import_utils import ORMDocProvider
 529         from librarian.parser import WLDocument
 530
 531         return WLDocument.from_file(self.xml_file.path,
 532                 provider=ORMDocProvider(self),
 533                 parse_dublincore=parse_dublincore)
 534
 535     def build_cover(self, book_info=None):
 536         """(Re)builds the cover image."""
 537         from StringIO import StringIO
 538         from django.core.files.base import ContentFile
 539         from librarian.cover import WLCover
 540
 541         if book_info is None:
 542             book_info = self.wldocument().book_info
 543
 544         cover = WLCover(book_info).image()
 545         imgstr = StringIO()
 546         cover.save(imgstr, 'png')
 547         self.cover.save(None, ContentFile(imgstr.getvalue()))
 548
 549     def build_pdf(self, customizations=None, file_name=None):
 550         """ (Re)builds the pdf file.
 551         customizations - customizations which are passed to LaTeX class file.
 552         file_name - save the pdf file under a different name and DO NOT save it in db.
 553         """
 554         from os import unlink
 555         from django.core.files import File
 556         from catalogue.utils import remove_zip
 557
 558         pdf = self.wldocument().as_pdf(customizations=customizations)
 559
 560         if file_name is None:
 561             # we'd like to be sure not to overwrite changes happening while
 562             # (timely) pdf generation is taking place (async celery scenario)
 563             current_self = Book.objects.get(id=self.id)
 564             current_self.pdf_file.save('%s.pdf' % self.slug,
 565                     File(open(pdf.get_filename())))
 566             self.pdf_file = current_self.pdf_file
 567
 568             # remove cached downloadables
 569             remove_zip(settings.ALL_PDF_ZIP)
 570
 571             for customized_pdf in get_existing_customized_pdf(self):
 572                 unlink(customized_pdf)
 573         else:
 574             print "saving %s" % file_name
 575             print "to: %s" % DefaultStorage().path(file_name)
 576             DefaultStorage().save(file_name, File(open(pdf.get_filename())))
 577
 578     def build_mobi(self):
 579         """ (Re)builds the MOBI file.
 580
 581         """
 582         from django.core.files import File
 583         from catalogue.utils import remove_zip
 584
 585         mobi = self.wldocument().as_mobi()
 586
 587         self.mobi_file.save('%s.mobi' % self.slug, File(open(mobi.get_filename())))
 588
 589         # remove zip with all mobi files
 590         remove_zip(settings.ALL_MOBI_ZIP)
 591
 592     def build_epub(self):
 593         """(Re)builds the epub file."""
 594         from django.core.files import File
 595         from catalogue.utils import remove_zip
 596
 597         epub = self.wldocument().as_epub()
 598
 599         self.epub_file.save('%s.epub' % self.slug,
 600                 File(open(epub.get_filename())))
 601
 602         # remove zip package with all epub files
 603         remove_zip(settings.ALL_EPUB_ZIP)
 604
 605     def build_txt(self):
 606         from django.core.files.base import ContentFile
 607
 608         text = self.wldocument().as_text()
 609         self.txt_file.save('%s.txt' % self.slug, ContentFile(text.get_string()))
 610
 611
 612     def build_html(self):
 613         from markupstring import MarkupString
 614         from django.core.files.base import ContentFile
 615         from slughifi import slughifi
 616         from librarian import html
 617
 618         meta_tags = list(self.tags.filter(
 619             category__in=('author', 'epoch', 'genre', 'kind')))
 620         book_tag = self.book_tag()
 621
 622         html_output = self.wldocument(parse_dublincore=False).as_html()
 623         if html_output:
 624             self.html_file.save('%s.html' % self.slug,
 625                     ContentFile(html_output.get_string()))
 626
 627             # get ancestor l-tags for adding to new fragments
 628             ancestor_tags = []
 629             p = self.parent
 630             while p:
 631                 ancestor_tags.append(p.book_tag())
 632                 p = p.parent
 633
 634             # Delete old fragments and create them from scratch
 635             self.fragments.all().delete()
 636             # Extract fragments
 637             closed_fragments, open_fragments = html.extract_fragments(self.html_file.path)
 638             for fragment in closed_fragments.values():
 639                 try:
 640                     theme_names = [s.strip() for s in fragment.themes.split(',')]
 641                 except AttributeError:
 642                     continue
 643                 themes = []
 644                 for theme_name in theme_names:
 645                     if not theme_name:
 646                         continue
 647                     tag, created = Tag.objects.get_or_create(slug=slughifi(theme_name), category='theme')
 648                     if created:
 649                         tag.name = theme_name
 650                         tag.sort_key = theme_name.lower()
 651                         tag.save()
 652                     themes.append(tag)
 653                 if not themes:
 654                     continue
 655
 656                 text = fragment.to_string()
 657                 short_text = ''
 658                 if (len(MarkupString(text)) > 240):
 659                     short_text = unicode(MarkupString(text)[:160])
 660                 new_fragment = Fragment.objects.create(anchor=fragment.id, book=self,
 661                     text=text, short_text=short_text)
 662
 663                 new_fragment.save()
 664                 new_fragment.tags = set(meta_tags + themes + [book_tag] + ancestor_tags)
 665             self.save()
 666             self.html_built.send(sender=self)
 667             return True
 668         return False
 669
 670     @staticmethod
 671     def zip_format(format_):
 672         def pretty_file_name(book):
 673             return "%s/%s.%s" % (
 674                 b.get_extra_info_value()['author'],
 675                 b.slug,
 676                 format_)
 677
 678         field_name = "%s_file" % format_
 679         books = Book.objects.filter(parent=None).exclude(**{field_name: ""})
 680         paths = [(pretty_file_name(b), getattr(b, field_name).path)
 681                     for b in books]
 682         result = create_zip.delay(paths,
 683                     getattr(settings, "ALL_%s_ZIP" % format_.upper()))
 684         return result.wait()
 685
 686     def zip_audiobooks(self, format_):
 687         bm = BookMedia.objects.filter(book=self, type=format_)
 688         paths = map(lambda bm: (None, bm.file.path), bm)
 689         result = create_zip.delay(paths, "%s_%s" % (self.slug, format_))
 690         return result.wait()
 691
 692     @classmethod
 693     def from_xml_file(cls, xml_file, **kwargs):
 694         from django.core.files import File
 695         from librarian import dcparser
 696
 697         # use librarian to parse meta-data
 698         book_info = dcparser.parse(xml_file)
 699
 700         if not isinstance(xml_file, File):
 701             xml_file = File(open(xml_file))
 702
 703         try:
 704             return cls.from_text_and_meta(xml_file, book_info, **kwargs)
 705         finally:
 706             xml_file.close()
 707
 708     @classmethod
 709     def from_text_and_meta(cls, raw_file, book_info, overwrite=False,
 710             build_epub=True, build_txt=True, build_pdf=True, build_mobi=True):
 711         import re
 712         from sortify import sortify
 713
 714         # check for parts before we do anything
 715         children = []
 716         if hasattr(book_info, 'parts'):
 717             for part_url in book_info.parts:
 718                 try:
 719                     children.append(Book.objects.get(slug=part_url.slug))
 720                 except Book.DoesNotExist, e:
 721                     raise Book.DoesNotExist(_('Book "%s" does not exist.') %
 722                             part_url.slug)
 723
 724
 725         # Read book metadata
 726         book_slug = book_info.url.slug
 727         if re.search(r'[^a-z0-9-]', book_slug):
 728             raise ValueError('Invalid characters in slug')
 729         book, created = Book.objects.get_or_create(slug=book_slug)
 730
 731         if created:
 732             book_shelves = []
 733         else:
 734             if not overwrite:
 735                 raise Book.AlreadyExists(_('Book %s already exists') % (
 736                         book_slug))
 737             # Save shelves for this book
 738             book_shelves = list(book.tags.filter(category='set'))
 739
 740         book.language = book_info.language
 741         book.title = book_info.title
 742         if book_info.variant_of:
 743             book.common_slug = book_info.variant_of.slug
 744         else:
 745             book.common_slug = book.slug
 746         book.set_extra_info_value(book_info.to_dict())
 747         book.save()
 748
 749         meta_tags = Tag.tags_from_info(book_info)
 750
 751         book.tags = set(meta_tags + book_shelves)
 752
 753         book_tag = book.book_tag()
 754
 755         for n, child_book in enumerate(children):
 756             child_book.parent = book
 757             child_book.parent_number = n
 758             child_book.save()
 759
 760         # Save XML and HTML files
 761         book.xml_file.save('%s.xml' % book.slug, raw_file, save=False)
 762
 763         # delete old fragments when overwriting
 764         book.fragments.all().delete()
 765
 766         if book.build_html():
 767             if not settings.NO_BUILD_TXT and build_txt:
 768                 book.build_txt()
 769
 770         book.build_cover(book_info)
 771
 772         if not settings.NO_BUILD_EPUB and build_epub:
 773             book.build_epub()
 774
 775         if not settings.NO_BUILD_PDF and build_pdf:
 776             book.build_pdf()
 777
 778         if not settings.NO_BUILD_MOBI and build_mobi:
 779             book.build_mobi()
 780
 781         book_descendants = list(book.children.all())
 782         descendants_tags = set()
 783         # add l-tag to descendants and their fragments
 784         while len(book_descendants) > 0:
 785             child_book = book_descendants.pop(0)
 786             descendants_tags.update(child_book.tags)
 787             child_book.tags = list(child_book.tags) + [book_tag]
 788             child_book.save()
 789             for fragment in child_book.fragments.all():
 790                 fragment.tags = set(list(fragment.tags) + [book_tag])
 791             book_descendants += list(child_book.children.all())
 792
 793         for tag in descendants_tags:
 794             touch_tag.delay(tag)
 795
 796         book.save()
 797
 798         # refresh cache
 799         book.reset_tag_counter()
 800         book.reset_theme_counter()
 801
 802         cls.published.send(sender=book)
 803         return book
 804
 805     def reset_tag_counter(self):
 806         if self.id is None:
 807             return
 808
 809         cache_key = "Book.tag_counter/%d" % self.id
 810         cache.delete(cache_key)
 811         if self.parent:
 812             self.parent.reset_tag_counter()
 813
 814     @property
 815     def tag_counter(self):
 816         if self.id:
 817             cache_key = "Book.tag_counter/%d" % self.id
 818             tags = cache.get(cache_key)
 819         else:
 820             tags = None
 821
 822         if tags is None:
 823             tags = {}
 824             for child in self.children.all().order_by():
 825                 for tag_pk, value in child.tag_counter.iteritems():
 826                     tags[tag_pk] = tags.get(tag_pk, 0) + value
 827             for tag in self.tags.exclude(category__in=('book', 'theme', 'set')).order_by():
 828                 tags[tag.pk] = 1
 829
 830             if self.id:
 831                 cache.set(cache_key, tags, CACHE_FOREVER)
 832         return tags
 833
 834     def reset_theme_counter(self):
 835         if self.id is None:
 836             return
 837
 838         cache_key = "Book.theme_counter/%d" % self.id
 839         cache.delete(cache_key)
 840         if self.parent:
 841             self.parent.reset_theme_counter()
 842
 843     @property
 844     def theme_counter(self):
 845         if self.id:
 846             cache_key = "Book.theme_counter/%d" % self.id
 847             tags = cache.get(cache_key)
 848         else:
 849             tags = None
 850
 851         if tags is None:
 852             tags = {}
 853             for fragment in Fragment.tagged.with_any([self.book_tag()]).order_by():
 854                 for tag in fragment.tags.filter(category='theme').order_by():
 855                     tags[tag.pk] = tags.get(tag.pk, 0) + 1
 856
 857             if self.id:
 858                 cache.set(cache_key, tags, CACHE_FOREVER)
 859         return tags
 860
 861     def pretty_title(self, html_links=False):
 862         book = self
 863         names = list(book.tags.filter(category='author'))
 864
 865         books = []
 866         while book:
 867             books.append(book)
 868             book = book.parent
 869         names.extend(reversed(books))
 870
 871         if html_links:
 872             names = ['<a href="%s">%s</a>' % (tag.get_absolute_url(), tag.name) for tag in names]
 873         else:
 874             names = [tag.name for tag in names]
 875
 876         return ', '.join(names)
 877
 878     @classmethod
 879     def tagged_top_level(cls, tags):
 880         """ Returns top-level books tagged with `tags'.
 881
 882         It only returns those books which don't have ancestors which are
 883         also tagged with those tags.
 884
 885         """
 886         # get relevant books and their tags
 887         objects = cls.tagged.with_all(tags)
 888         # eliminate descendants
 889         l_tags = Tag.objects.filter(category='book', slug__in=[book.book_tag_slug() for book in objects])
 890         descendants_keys = [book.pk for book in cls.tagged.with_any(l_tags)]
 891         if descendants_keys:
 892             objects = objects.exclude(pk__in=descendants_keys)
 893
 894         return objects
 895
 896     @classmethod
 897     def book_list(cls, filter=None):
 898         """Generates a hierarchical listing of all books.
 899
 900         Books are optionally filtered with a test function.
 901
 902         """
 903
 904         books_by_parent = {}
 905         books = cls.objects.all().order_by('parent_number', 'sort_key').only(
 906                 'title', 'parent', 'slug')
 907         if filter:
 908             books = books.filter(filter).distinct()
 909             book_ids = set((book.pk for book in books))
 910             for book in books:
 911                 parent = book.parent_id
 912                 if parent not in book_ids:
 913                     parent = None
 914                 books_by_parent.setdefault(parent, []).append(book)
 915         else:
 916             for book in books:
 917                 books_by_parent.setdefault(book.parent_id, []).append(book)
 918
 919         orphans = []
 920         books_by_author = SortedDict()
 921         for tag in Tag.objects.filter(category='author'):
 922             books_by_author[tag] = []
 923
 924         for book in books_by_parent.get(None,()):
 925             authors = list(book.tags.filter(category='author'))
 926             if authors:
 927                 for author in authors:
 928                     books_by_author[author].append(book)
 929             else:
 930                 orphans.append(book)
 931
 932         return books_by_author, orphans, books_by_parent
 933
 934     _audiences_pl = {
 935         "SP1": (1, u"szkoła podstawowa"),
 936         "SP2": (1, u"szkoła podstawowa"),
 937         "P": (1, u"szkoła podstawowa"),
 938         "G": (2, u"gimnazjum"),
 939         "L": (3, u"liceum"),
 940         "LP": (3, u"liceum"),
 941     }
 942     def audiences_pl(self):
 943         audiences = self.get_extra_info_value().get('audiences', [])
 944         audiences = sorted(set([self._audiences_pl[a] for a in audiences]))
 945         return [a[1] for a in audiences]
 946
 947
 948 def _has_factory(ftype):
 949     has = lambda self: bool(getattr(self, "%s_file" % ftype))
 950     has.short_description = t.upper()
 951     has.boolean = True
 952     has.__name__ = "has_%s_file" % ftype
 953     return has
 954
 955
 956 # add the file fields
 957 for t in Book.formats:
 958     field_name = "%s_file" % t
 959     models.FileField(_("%s file" % t.upper()),
 960             upload_to=book_upload_path(t),
 961             blank=True).contribute_to_class(Book, field_name)
 962
 963     setattr(Book, "has_%s_file" % t, _has_factory(t))
 964
 965
 966 class Fragment(models.Model):
 967     text = models.TextField()
 968     short_text = models.TextField(editable=False)
 969     anchor = models.CharField(max_length=120)
 970     book = models.ForeignKey(Book, related_name='fragments')
 971
 972     objects = models.Manager()
 973     tagged = managers.ModelTaggedItemManager(Tag)
 974     tags = managers.TagDescriptor(Tag)
 975
 976     class Meta:
 977         ordering = ('book', 'anchor',)
 978         verbose_name = _('fragment')
 979         verbose_name_plural = _('fragments')
 980
 981     def get_absolute_url(self):
 982         return '%s#m%s' % (self.book.get_html_url(), self.anchor)
 983
 984     def reset_short_html(self):
 985         if self.id is None:
 986             return
 987
 988         cache_key = "Fragment.short_html/%d/%s"
 989         for lang, langname in settings.LANGUAGES:
 990             cache.delete(cache_key % (self.id, lang))
 991
 992     def short_html(self):
 993         if self.id:
 994             cache_key = "Fragment.short_html/%d/%s" % (self.id, get_language())
 995             short_html = cache.get(cache_key)
 996         else:
 997             short_html = None
 998
 999         if short_html is not None:
1000             return mark_safe(short_html)
1001         else:
1002             short_html = unicode(render_to_string('catalogue/fragment_short.html',
1003                 {'fragment': self}))
1004             if self.id:
1005                 cache.set(cache_key, short_html, CACHE_FOREVER)
1006             return mark_safe(short_html)
1007
1008
1009 ###########
1010 #
1011 # SIGNALS
1012 #
1013 ###########
1014
1015
1016 def _tags_updated_handler(sender, affected_tags, **kwargs):
1017     # reset tag global counter
1018     # we want Tag.changed_at updated for API to know the tag was touched
1019     for tag in affected_tags:
1020         touch_tag.delay(tag)
1021
1022     # if book tags changed, reset book tag counter
1023     if isinstance(sender, Book) and \
1024                 Tag.objects.filter(pk__in=(tag.pk for tag in affected_tags)).\
1025                     exclude(category__in=('book', 'theme', 'set')).count():
1026         sender.reset_tag_counter()
1027     # if fragment theme changed, reset book theme counter
1028     elif isinstance(sender, Fragment) and \
1029                 Tag.objects.filter(pk__in=(tag.pk for tag in affected_tags)).\
1030                     filter(category='theme').count():
1031         sender.book.reset_theme_counter()
1032 tags_updated.connect(_tags_updated_handler)
1033
1034
1035 def _pre_delete_handler(sender, instance, **kwargs):
1036     """ refresh Book on BookMedia delete """
1037     if sender == BookMedia:
1038         instance.book.save()
1039 pre_delete.connect(_pre_delete_handler)
1040
1041 def _post_save_handler(sender, instance, **kwargs):
1042     """ refresh all the short_html stuff on BookMedia update """
1043     if sender == BookMedia:
1044         instance.book.save()
1045 post_save.connect(_post_save_handler)