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 get_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, post_delete
  20 import jsonfield
  21
  22 from django.conf import settings
  23
  24 from newtagging.models import TagBase, tags_updated
  25 from newtagging import managers
  26 from catalogue.fields import JSONField, OverwritingFileField
  27 from catalogue.utils import create_zip, split_tags, truncate_html_words
  28 from catalogue.tasks import touch_tag, index_book
  29 from shutil import copy
  30 from glob import glob
  31 import re
  32 from os import path
  33
  34 import search
  35
  36 # Those are hard-coded here so that makemessages sees them.
  37 TAG_CATEGORIES = (
  38     ('author', _('author')),
  39     ('epoch', _('epoch')),
  40     ('kind', _('kind')),
  41     ('genre', _('genre')),
  42     ('theme', _('theme')),
  43     ('set', _('set')),
  44     ('book', _('book')),
  45 )
  46
  47
  48 permanent_cache = get_cache('permanent')
  49
  50
  51 class TagSubcategoryManager(models.Manager):
  52     def __init__(self, subcategory):
  53         super(TagSubcategoryManager, self).__init__()
  54         self.subcategory = subcategory
  55
  56     def get_query_set(self):
  57         return super(TagSubcategoryManager, self).get_query_set().filter(category=self.subcategory)
  58
  59
  60 class Tag(TagBase):
  61     name = models.CharField(_('name'), max_length=50, db_index=True)
  62     slug = models.SlugField(_('slug'), max_length=120, db_index=True)
  63     sort_key = models.CharField(_('sort key'), max_length=120, db_index=True)
  64     category = models.CharField(_('category'), max_length=50, blank=False, null=False,
  65         db_index=True, choices=TAG_CATEGORIES)
  66     description = models.TextField(_('description'), blank=True)
  67
  68     user = models.ForeignKey(User, blank=True, null=True)
  69     book_count = models.IntegerField(_('book count'), blank=True, null=True)
  70     gazeta_link = models.CharField(blank=True, max_length=240)
  71     wiki_link = models.CharField(blank=True, max_length=240)
  72
  73     created_at    = models.DateTimeField(_('creation date'), auto_now_add=True, db_index=True)
  74     changed_at    = models.DateTimeField(_('creation date'), auto_now=True, db_index=True)
  75
  76     class UrlDeprecationWarning(DeprecationWarning):
  77         pass
  78
  79     categories_rev = {
  80         'autor': 'author',
  81         'epoka': 'epoch',
  82         'rodzaj': 'kind',
  83         'gatunek': 'genre',
  84         'motyw': 'theme',
  85         'polka': 'set',
  86     }
  87     categories_dict = dict((item[::-1] for item in categories_rev.iteritems()))
  88
  89     class Meta:
  90         ordering = ('sort_key',)
  91         verbose_name = _('tag')
  92         verbose_name_plural = _('tags')
  93         unique_together = (("slug", "category"),)
  94
  95     def __unicode__(self):
  96         return self.name
  97
  98     def __repr__(self):
  99         return "Tag(slug=%r)" % self.slug
 100
 101     @permalink
 102     def get_absolute_url(self):
 103         return ('catalogue.views.tagged_object_list', [self.url_chunk])
 104
 105     def has_description(self):
 106         return len(self.description) > 0
 107     has_description.short_description = _('description')
 108     has_description.boolean = True
 109
 110     def get_count(self):
 111         """Returns global book count for book tags, fragment count for themes."""
 112
 113         if self.category == 'book':
 114             # never used
 115             objects = Book.objects.none()
 116         elif self.category == 'theme':
 117             objects = Fragment.tagged.with_all((self,))
 118         else:
 119             objects = Book.tagged.with_all((self,)).order_by()
 120             if self.category != 'set':
 121                 # eliminate descendants
 122                 l_tags = Tag.objects.filter(slug__in=[book.book_tag_slug() for book in objects])
 123                 descendants_keys = [book.pk for book in Book.tagged.with_any(l_tags)]
 124                 if descendants_keys:
 125                     objects = objects.exclude(pk__in=descendants_keys)
 126         return objects.count()
 127
 128     @staticmethod
 129     def get_tag_list(tags):
 130         if isinstance(tags, basestring):
 131             real_tags = []
 132             ambiguous_slugs = []
 133             category = None
 134             deprecated = False
 135             tags_splitted = tags.split('/')
 136             for name in tags_splitted:
 137                 if category:
 138                     real_tags.append(Tag.objects.get(slug=name, category=category))
 139                     category = None
 140                 elif name in Tag.categories_rev:
 141                     category = Tag.categories_rev[name]
 142                 else:
 143                     try:
 144                         real_tags.append(Tag.objects.exclude(category='book').get(slug=name))
 145                         deprecated = True
 146                     except Tag.MultipleObjectsReturned, e:
 147                         ambiguous_slugs.append(name)
 148
 149             if category:
 150                 # something strange left off
 151                 raise Tag.DoesNotExist()
 152             if ambiguous_slugs:
 153                 # some tags should be qualified
 154                 e = Tag.MultipleObjectsReturned()
 155                 e.tags = real_tags
 156                 e.ambiguous_slugs = ambiguous_slugs
 157                 raise e
 158             if deprecated:
 159                 e = Tag.UrlDeprecationWarning()
 160                 e.tags = real_tags
 161                 raise e
 162             return real_tags
 163         else:
 164             return TagBase.get_tag_list(tags)
 165
 166     @property
 167     def url_chunk(self):
 168         return '/'.join((Tag.categories_dict[self.category], self.slug))
 169
 170     @staticmethod
 171     def tags_from_info(info):
 172         from slughifi import slughifi
 173         from sortify import sortify
 174         meta_tags = []
 175         categories = (('kinds', 'kind'), ('genres', 'genre'), ('authors', 'author'), ('epochs', 'epoch'))
 176         for field_name, category in categories:
 177             try:
 178                 tag_names = getattr(info, field_name)
 179             except:
 180                 try:
 181                     tag_names = [getattr(info, category)]
 182                 except:
 183                     # For instance, Pictures do not have 'genre' field.
 184                     continue
 185             for tag_name in tag_names:
 186                 tag_sort_key = tag_name
 187                 if category == 'author':
 188                     tag_sort_key = tag_name.last_name
 189                     tag_name = tag_name.readable()
 190                 tag, created = Tag.objects.get_or_create(slug=slughifi(tag_name), category=category)
 191                 if created:
 192                     tag.name = tag_name
 193                     tag.sort_key = sortify(tag_sort_key.lower())
 194                     tag.save()
 195                 meta_tags.append(tag)
 196         return meta_tags
 197
 198
 199
 200 def get_dynamic_path(media, filename, ext=None, maxlen=100):
 201     from slughifi import slughifi
 202
 203     # how to put related book's slug here?
 204     if not ext:
 205         # BookMedia case
 206         ext = media.formats[media.type].ext
 207     if media is None or not media.name:
 208         name = slughifi(filename.split(".")[0])
 209     else:
 210         name = slughifi(media.name)
 211     return 'book/%s/%s.%s' % (ext, name[:maxlen-len('book/%s/.%s' % (ext, ext))-4], ext)
 212
 213
 214 # TODO: why is this hard-coded ?
 215 def book_upload_path(ext=None, maxlen=100):
 216     return lambda *args: get_dynamic_path(*args, ext=ext, maxlen=maxlen)
 217
 218
 219 def get_customized_pdf_path(book, customizations):
 220     """
 221     Returns a MEDIA_ROOT relative path for a customized pdf. The name will contain a hash of customization options.
 222     """
 223     customizations.sort()
 224     h = hash(tuple(customizations))
 225
 226     pdf_name = '%s-custom-%s' % (book.slug, h)
 227     pdf_file = get_dynamic_path(None, pdf_name, ext='pdf')
 228
 229     return pdf_file
 230
 231
 232 def get_existing_customized_pdf(book):
 233     """
 234     Returns a list of paths to generated customized pdf of a book
 235     """
 236     pdf_glob = '%s-custom-' % (book.slug,)
 237     pdf_glob = get_dynamic_path(None, pdf_glob, ext='pdf')
 238     pdf_glob = re.sub(r"[.]([a-z0-9]+)$", "*.\\1", pdf_glob)
 239     return glob(path.join(settings.MEDIA_ROOT, pdf_glob))
 240
 241
 242 class BookMedia(models.Model):
 243     FileFormat = namedtuple("FileFormat", "name ext")
 244     formats = SortedDict([
 245         ('mp3', FileFormat(name='MP3', ext='mp3')),
 246         ('ogg', FileFormat(name='Ogg Vorbis', ext='ogg')),
 247         ('daisy', FileFormat(name='DAISY', ext='daisy.zip')),
 248     ])
 249     format_choices = [(k, _('%s file') % t.name)
 250             for k, t in formats.items()]
 251
 252     type        = models.CharField(_('type'), choices=format_choices, max_length="100")
 253     name        = models.CharField(_('name'), max_length="100")
 254     file        = OverwritingFileField(_('file'), upload_to=book_upload_path())
 255     uploaded_at = models.DateTimeField(_('creation date'), auto_now_add=True, editable=False)
 256     extra_info  = JSONField(_('extra information'), default='{}', editable=False)
 257     book = models.ForeignKey('Book', related_name='media')
 258     source_sha1 = models.CharField(null=True, blank=True, max_length=40, editable=False)
 259
 260     def __unicode__(self):
 261         return "%s (%s)" % (self.name, self.file.name.split("/")[-1])
 262
 263     class Meta:
 264         ordering            = ('type', 'name')
 265         verbose_name        = _('book media')
 266         verbose_name_plural = _('book media')
 267
 268     def save(self, *args, **kwargs):
 269         from slughifi import slughifi
 270         from catalogue.utils import ExistingFile, remove_zip
 271
 272         try:
 273             old = BookMedia.objects.get(pk=self.pk)
 274         except BookMedia.DoesNotExist, e:
 275             old = None
 276         else:
 277             # if name changed, change the file name, too
 278             if slughifi(self.name) != slughifi(old.name):
 279                 self.file.save(None, ExistingFile(self.file.path), save=False, leave=True)
 280
 281         super(BookMedia, self).save(*args, **kwargs)
 282
 283         # remove the zip package for book with modified media
 284         if old:
 285             remove_zip("%s_%s" % (old.book.slug, old.type))
 286         remove_zip("%s_%s" % (self.book.slug, self.type))
 287
 288         extra_info = self.get_extra_info_value()
 289         extra_info.update(self.read_meta())
 290         self.set_extra_info_value(extra_info)
 291         self.source_sha1 = self.read_source_sha1(self.file.path, self.type)
 292         return super(BookMedia, self).save(*args, **kwargs)
 293
 294     def read_meta(self):
 295         """
 296             Reads some metadata from the audiobook.
 297         """
 298         import mutagen
 299         from mutagen import id3
 300
 301         artist_name = director_name = project = funded_by = ''
 302         if self.type == 'mp3':
 303             try:
 304                 audio = id3.ID3(self.file.path)
 305                 artist_name = ', '.join(', '.join(tag.text) for tag in audio.getall('TPE1'))
 306                 director_name = ', '.join(', '.join(tag.text) for tag in audio.getall('TPE3'))
 307                 project = ", ".join([t.data for t in audio.getall('PRIV')
 308                         if t.owner=='wolnelektury.pl?project'])
 309                 funded_by = ", ".join([t.data for t in audio.getall('PRIV')
 310                         if t.owner=='wolnelektury.pl?funded_by'])
 311             except:
 312                 pass
 313         elif self.type == 'ogg':
 314             try:
 315                 audio = mutagen.File(self.file.path)
 316                 artist_name = ', '.join(audio.get('artist', []))
 317                 director_name = ', '.join(audio.get('conductor', []))
 318                 project = ", ".join(audio.get('project', []))
 319                 funded_by = ", ".join(audio.get('funded_by', []))
 320             except:
 321                 pass
 322         else:
 323             return {}
 324         return {'artist_name': artist_name, 'director_name': director_name,
 325                 'project': project, 'funded_by': funded_by}
 326
 327     @staticmethod
 328     def read_source_sha1(filepath, filetype):
 329         """
 330             Reads source file SHA1 from audiobok metadata.
 331         """
 332         import mutagen
 333         from mutagen import id3
 334
 335         if filetype == 'mp3':
 336             try:
 337                 audio = id3.ID3(filepath)
 338                 return [t.data for t in audio.getall('PRIV')
 339                         if t.owner=='wolnelektury.pl?flac_sha1'][0]
 340             except:
 341                 return None
 342         elif filetype == 'ogg':
 343             try:
 344                 audio = mutagen.File(filepath)
 345                 return audio.get('flac_sha1', [None])[0]
 346             except:
 347                 return None
 348         else:
 349             return None
 350
 351
 352 class Book(models.Model):
 353     title         = models.CharField(_('title'), max_length=120)
 354     sort_key = models.CharField(_('sort key'), max_length=120, db_index=True, editable=False)
 355     slug = models.SlugField(_('slug'), max_length=120, db_index=True,
 356             unique=True)
 357     common_slug = models.SlugField(_('slug'), max_length=120, db_index=True)
 358     language = models.CharField(_('language code'), max_length=3, db_index=True,
 359                     default=settings.CATALOGUE_DEFAULT_LANGUAGE)
 360     description   = models.TextField(_('description'), blank=True)
 361     created_at    = models.DateTimeField(_('creation date'), auto_now_add=True, db_index=True)
 362     changed_at    = models.DateTimeField(_('creation date'), auto_now=True, db_index=True)
 363     parent_number = models.IntegerField(_('parent number'), default=0)
 364     extra_info    = JSONField(_('extra information'), default='{}')
 365     gazeta_link   = models.CharField(blank=True, max_length=240)
 366     wiki_link     = models.CharField(blank=True, max_length=240)
 367     # files generated during publication
 368
 369     cover = models.FileField(_('cover'), upload_to=book_upload_path('png'),
 370                 null=True, blank=True)
 371     ebook_formats = ['pdf', 'epub', 'mobi', 'txt']
 372     formats = ebook_formats + ['html', 'xml']
 373
 374     parent        = models.ForeignKey('self', blank=True, null=True, related_name='children')
 375
 376     _related_info = jsonfield.JSONField(blank=True, null=True, editable=False)
 377
 378     objects  = models.Manager()
 379     tagged   = managers.ModelTaggedItemManager(Tag)
 380     tags     = managers.TagDescriptor(Tag)
 381
 382     html_built = django.dispatch.Signal()
 383     published = django.dispatch.Signal()
 384
 385     class AlreadyExists(Exception):
 386         pass
 387
 388     class Meta:
 389         ordering = ('sort_key',)
 390         verbose_name = _('book')
 391         verbose_name_plural = _('books')
 392
 393     def __unicode__(self):
 394         return self.title
 395
 396     def save(self, force_insert=False, force_update=False, reset_short_html=True, **kwargs):
 397         from sortify import sortify
 398
 399         self.sort_key = sortify(self.title)
 400
 401         ret = super(Book, self).save(force_insert, force_update)
 402
 403         if reset_short_html:
 404             self.reset_short_html()
 405
 406         return ret
 407
 408     @permalink
 409     def get_absolute_url(self):
 410         return ('catalogue.views.book_detail', [self.slug])
 411
 412     @property
 413     def name(self):
 414         return self.title
 415
 416     def book_tag_slug(self):
 417         return ('l-' + self.slug)[:120]
 418
 419     def book_tag(self):
 420         slug = self.book_tag_slug()
 421         book_tag, created = Tag.objects.get_or_create(slug=slug, category='book')
 422         if created:
 423             book_tag.name = self.title[:50]
 424             book_tag.sort_key = self.title.lower()
 425             book_tag.save()
 426         return book_tag
 427
 428     def has_media(self, type):
 429         if type in Book.formats:
 430             return bool(getattr(self, "%s_file" % type))
 431         else:
 432             return self.media.filter(type=type).exists()
 433
 434     def get_media(self, type):
 435         if self.has_media(type):
 436             if type in Book.formats:
 437                 return getattr(self, "%s_file" % type)
 438             else:
 439                 return self.media.filter(type=type)
 440         else:
 441             return None
 442
 443     def get_mp3(self):
 444         return self.get_media("mp3")
 445     def get_odt(self):
 446         return self.get_media("odt")
 447     def get_ogg(self):
 448         return self.get_media("ogg")
 449     def get_daisy(self):
 450         return self.get_media("daisy")
 451
 452     def reset_short_html(self):
 453         if self.id is None:
 454             return
 455
 456         type(self).objects.filter(pk=self.pk).update(_related_info=None)
 457         # Fragment.short_html relies on book's tags, so reset it here too
 458         for fragm in self.fragments.all():
 459             fragm.reset_short_html()
 460
 461     def has_description(self):
 462         return len(self.description) > 0
 463     has_description.short_description = _('description')
 464     has_description.boolean = True
 465
 466     # ugly ugly ugly
 467     def has_mp3_file(self):
 468         return bool(self.has_media("mp3"))
 469     has_mp3_file.short_description = 'MP3'
 470     has_mp3_file.boolean = True
 471
 472     def has_ogg_file(self):
 473         return bool(self.has_media("ogg"))
 474     has_ogg_file.short_description = 'OGG'
 475     has_ogg_file.boolean = True
 476
 477     def has_daisy_file(self):
 478         return bool(self.has_media("daisy"))
 479     has_daisy_file.short_description = 'DAISY'
 480     has_daisy_file.boolean = True
 481
 482     def wldocument(self, parse_dublincore=True):
 483         from catalogue.import_utils import ORMDocProvider
 484         from librarian.parser import WLDocument
 485
 486         return WLDocument.from_file(self.xml_file.path,
 487                 provider=ORMDocProvider(self),
 488                 parse_dublincore=parse_dublincore)
 489
 490     def build_cover(self, book_info=None):
 491         """(Re)builds the cover image."""
 492         from StringIO import StringIO
 493         from django.core.files.base import ContentFile
 494         from librarian.cover import WLCover
 495
 496         if book_info is None:
 497             book_info = self.wldocument().book_info
 498
 499         cover = WLCover(book_info).image()
 500         imgstr = StringIO()
 501         cover.save(imgstr, 'png')
 502         self.cover.save(None, ContentFile(imgstr.getvalue()))
 503
 504     def build_pdf(self, customizations=None, file_name=None):
 505         """ (Re)builds the pdf file.
 506         customizations - customizations which are passed to LaTeX class file.
 507         file_name - save the pdf file under a different name and DO NOT save it in db.
 508         """
 509         from os import unlink
 510         from django.core.files import File
 511         from catalogue.utils import remove_zip
 512
 513         pdf = self.wldocument().as_pdf(customizations=customizations)
 514
 515         if file_name is None:
 516             # we'd like to be sure not to overwrite changes happening while
 517             # (timely) pdf generation is taking place (async celery scenario)
 518             current_self = Book.objects.get(id=self.id)
 519             current_self.pdf_file.save('%s.pdf' % self.slug,
 520                     File(open(pdf.get_filename())))
 521             self.pdf_file = current_self.pdf_file
 522
 523             # remove cached downloadables
 524             remove_zip(settings.ALL_PDF_ZIP)
 525
 526             for customized_pdf in get_existing_customized_pdf(self):
 527                 unlink(customized_pdf)
 528         else:
 529             print "saving %s" % file_name
 530             print "to: %s" % DefaultStorage().path(file_name)
 531             DefaultStorage().save(file_name, File(open(pdf.get_filename())))
 532
 533     def build_mobi(self):
 534         """ (Re)builds the MOBI file.
 535
 536         """
 537         from django.core.files import File
 538         from catalogue.utils import remove_zip
 539
 540         mobi = self.wldocument().as_mobi()
 541
 542         self.mobi_file.save('%s.mobi' % self.slug, File(open(mobi.get_filename())))
 543
 544         # remove zip with all mobi files
 545         remove_zip(settings.ALL_MOBI_ZIP)
 546
 547     def build_epub(self):
 548         """(Re)builds the epub file."""
 549         from django.core.files import File
 550         from catalogue.utils import remove_zip
 551
 552         epub = self.wldocument().as_epub()
 553
 554         self.epub_file.save('%s.epub' % self.slug,
 555                 File(open(epub.get_filename())))
 556
 557         # remove zip package with all epub files
 558         remove_zip(settings.ALL_EPUB_ZIP)
 559
 560     def build_txt(self):
 561         from django.core.files.base import ContentFile
 562
 563         text = self.wldocument().as_text()
 564         self.txt_file.save('%s.txt' % self.slug, ContentFile(text.get_string()))
 565
 566
 567     def build_html(self):
 568         from django.core.files.base import ContentFile
 569         from slughifi import slughifi
 570         from librarian import html
 571
 572         meta_tags = list(self.tags.filter(
 573             category__in=('author', 'epoch', 'genre', 'kind')))
 574         book_tag = self.book_tag()
 575
 576         html_output = self.wldocument(parse_dublincore=False).as_html()
 577         if html_output:
 578             self.html_file.save('%s.html' % self.slug,
 579                     ContentFile(html_output.get_string()))
 580
 581             # get ancestor l-tags for adding to new fragments
 582             ancestor_tags = []
 583             p = self.parent
 584             while p:
 585                 ancestor_tags.append(p.book_tag())
 586                 p = p.parent
 587
 588             # Delete old fragments and create them from scratch
 589             self.fragments.all().delete()
 590             # Extract fragments
 591             closed_fragments, open_fragments = html.extract_fragments(self.html_file.path)
 592             for fragment in closed_fragments.values():
 593                 try:
 594                     theme_names = [s.strip() for s in fragment.themes.split(',')]
 595                 except AttributeError:
 596                     continue
 597                 themes = []
 598                 for theme_name in theme_names:
 599                     if not theme_name:
 600                         continue
 601                     tag, created = Tag.objects.get_or_create(slug=slughifi(theme_name), category='theme')
 602                     if created:
 603                         tag.name = theme_name
 604                         tag.sort_key = theme_name.lower()
 605                         tag.save()
 606                     themes.append(tag)
 607                 if not themes:
 608                     continue
 609
 610                 text = fragment.to_string()
 611                 short_text = truncate_html_words(text, 15)
 612                 if text == short_text:
 613                     short_text = ''
 614                 new_fragment = Fragment.objects.create(anchor=fragment.id, book=self,
 615                     text=text, short_text=short_text)
 616
 617                 new_fragment.save()
 618                 new_fragment.tags = set(meta_tags + themes + [book_tag] + ancestor_tags)
 619             self.save()
 620             self.html_built.send(sender=self)
 621             return True
 622         return False
 623
 624     @staticmethod
 625     def zip_format(format_):
 626         def pretty_file_name(book):
 627             return "%s/%s.%s" % (
 628                 b.get_extra_info_value()['author'],
 629                 b.slug,
 630                 format_)
 631
 632         field_name = "%s_file" % format_
 633         books = Book.objects.filter(parent=None).exclude(**{field_name: ""})
 634         paths = [(pretty_file_name(b), getattr(b, field_name).path)
 635                     for b in books]
 636         result = create_zip.delay(paths,
 637                     getattr(settings, "ALL_%s_ZIP" % format_.upper()))
 638         return result.wait()
 639
 640     def zip_audiobooks(self, format_):
 641         bm = BookMedia.objects.filter(book=self, type=format_)
 642         paths = map(lambda bm: (None, bm.file.path), bm)
 643         result = create_zip.delay(paths, "%s_%s" % (self.slug, format_))
 644         return result.wait()
 645
 646     def search_index(self, book_info=None, reuse_index=False, index_tags=True):
 647         if reuse_index:
 648             idx = search.ReusableIndex()
 649         else:
 650             idx = search.Index()
 651
 652         idx.open()
 653         try:
 654             idx.index_book(self, book_info)
 655             if index_tags:
 656                 idx.index_tags()
 657         finally:
 658             idx.close()
 659
 660     @classmethod
 661     def from_xml_file(cls, xml_file, **kwargs):
 662         from django.core.files import File
 663         from librarian import dcparser
 664
 665         # use librarian to parse meta-data
 666         book_info = dcparser.parse(xml_file)
 667
 668         if not isinstance(xml_file, File):
 669             xml_file = File(open(xml_file))
 670
 671         try:
 672             return cls.from_text_and_meta(xml_file, book_info, **kwargs)
 673         finally:
 674             xml_file.close()
 675
 676     @classmethod
 677     def from_text_and_meta(cls, raw_file, book_info, overwrite=False,
 678             build_epub=True, build_txt=True, build_pdf=True, build_mobi=True,
 679             search_index=True, search_index_tags=True, search_index_reuse=False):
 680         import re
 681         from sortify import sortify
 682
 683         # check for parts before we do anything
 684         children = []
 685         if hasattr(book_info, 'parts'):
 686             for part_url in book_info.parts:
 687                 try:
 688                     children.append(Book.objects.get(slug=part_url.slug))
 689                 except Book.DoesNotExist, e:
 690                     raise Book.DoesNotExist(_('Book "%s" does not exist.') %
 691                             part_url.slug)
 692
 693
 694         # Read book metadata
 695         book_slug = book_info.url.slug
 696         if re.search(r'[^a-z0-9-]', book_slug):
 697             raise ValueError('Invalid characters in slug')
 698         book, created = Book.objects.get_or_create(slug=book_slug)
 699
 700         if created:
 701             book_shelves = []
 702         else:
 703             if not overwrite:
 704                 raise Book.AlreadyExists(_('Book %s already exists') % (
 705                         book_slug))
 706             # Save shelves for this book
 707             book_shelves = list(book.tags.filter(category='set'))
 708
 709         book.language = book_info.language
 710         book.title = book_info.title
 711         if book_info.variant_of:
 712             book.common_slug = book_info.variant_of.slug
 713         else:
 714             book.common_slug = book.slug
 715         book.set_extra_info_value(book_info.to_dict())
 716         book.save()
 717
 718         meta_tags = Tag.tags_from_info(book_info)
 719
 720         book.tags = set(meta_tags + book_shelves)
 721
 722         book_tag = book.book_tag()
 723
 724         for n, child_book in enumerate(children):
 725             child_book.parent = book
 726             child_book.parent_number = n
 727             child_book.save()
 728
 729         # Save XML and HTML files
 730         book.xml_file.save('%s.xml' % book.slug, raw_file, save=False)
 731
 732         # delete old fragments when overwriting
 733         book.fragments.all().delete()
 734
 735         if book.build_html():
 736             if not settings.NO_BUILD_TXT and build_txt:
 737                 book.build_txt()
 738
 739         book.build_cover(book_info)
 740
 741         if not settings.NO_BUILD_EPUB and build_epub:
 742             book.build_epub()
 743
 744         if not settings.NO_BUILD_PDF and build_pdf:
 745             book.build_pdf()
 746
 747         if not settings.NO_BUILD_MOBI and build_mobi:
 748             book.build_mobi()
 749
 750         if not settings.NO_SEARCH_INDEX and search_index:
 751             book.search_index(index_tags=search_index_tags, reuse_index=search_index_reuse)
 752             #index_book.delay(book.id, book_info)
 753
 754         book_descendants = list(book.children.all())
 755         descendants_tags = set()
 756         # add l-tag to descendants and their fragments
 757         while len(book_descendants) > 0:
 758             child_book = book_descendants.pop(0)
 759             descendants_tags.update(child_book.tags)
 760             child_book.tags = list(child_book.tags) + [book_tag]
 761             child_book.save()
 762             for fragment in child_book.fragments.all():
 763                 fragment.tags = set(list(fragment.tags) + [book_tag])
 764             book_descendants += list(child_book.children.all())
 765
 766         for tag in descendants_tags:
 767             touch_tag(tag)
 768
 769         book.save()
 770
 771         # refresh cache
 772         book.reset_tag_counter()
 773         book.reset_theme_counter()
 774
 775         cls.published.send(sender=book)
 776         return book
 777
 778     def related_info(self):
 779         """Keeps info about related objects (tags, media) in cache field."""
 780         if self._related_info is not None:
 781             return self._related_info
 782         else:
 783             rel = {'tags': {}, 'media': {}}
 784
 785             tags = self.tags.filter(category__in=(
 786                     'author', 'kind', 'genre', 'epoch'))
 787             tags = split_tags(tags)
 788             for category in tags:
 789                 rel['tags'][category] = [
 790                         (t.name, t.slug) for t in tags[category]]
 791
 792             for media_format in BookMedia.formats:
 793                 rel['media'][media_format] = self.has_media(media_format)
 794
 795             book = self
 796             parents = []
 797             while book.parent:
 798                 parents.append((book.parent.title, book.parent.slug))
 799                 book = book.parent
 800             parents = parents[::-1]
 801             if parents:
 802                 rel['parents'] = parents
 803
 804             if self.pk:
 805                 type(self).objects.filter(pk=self.pk).update(_related_info=rel)
 806             return rel
 807
 808     def reset_tag_counter(self):
 809         if self.id is None:
 810             return
 811
 812         cache_key = "Book.tag_counter/%d" % self.id
 813         permanent_cache.delete(cache_key)
 814         if self.parent:
 815             self.parent.reset_tag_counter()
 816
 817     @property
 818     def tag_counter(self):
 819         if self.id:
 820             cache_key = "Book.tag_counter/%d" % self.id
 821             tags = permanent_cache.get(cache_key)
 822         else:
 823             tags = None
 824
 825         if tags is None:
 826             tags = {}
 827             for child in self.children.all().order_by():
 828                 for tag_pk, value in child.tag_counter.iteritems():
 829                     tags[tag_pk] = tags.get(tag_pk, 0) + value
 830             for tag in self.tags.exclude(category__in=('book', 'theme', 'set')).order_by():
 831                 tags[tag.pk] = 1
 832
 833             if self.id:
 834                 permanent_cache.set(cache_key, tags)
 835         return tags
 836
 837     def reset_theme_counter(self):
 838         if self.id is None:
 839             return
 840
 841         cache_key = "Book.theme_counter/%d" % self.id
 842         permanent_cache.delete(cache_key)
 843         if self.parent:
 844             self.parent.reset_theme_counter()
 845
 846     @property
 847     def theme_counter(self):
 848         if self.id:
 849             cache_key = "Book.theme_counter/%d" % self.id
 850             tags = permanent_cache.get(cache_key)
 851         else:
 852             tags = None
 853
 854         if tags is None:
 855             tags = {}
 856             for fragment in Fragment.tagged.with_any([self.book_tag()]).order_by():
 857                 for tag in fragment.tags.filter(category='theme').order_by():
 858                     tags[tag.pk] = tags.get(tag.pk, 0) + 1
 859
 860             if self.id:
 861                 permanent_cache.set(cache_key, tags)
 862         return tags
 863
 864     def pretty_title(self, html_links=False):
 865         book = self
 866         names = list(book.tags.filter(category='author'))
 867
 868         books = []
 869         while book:
 870             books.append(book)
 871             book = book.parent
 872         names.extend(reversed(books))
 873
 874         if html_links:
 875             names = ['<a href="%s">%s</a>' % (tag.get_absolute_url(), tag.name) for tag in names]
 876         else:
 877             names = [tag.name for tag in names]
 878
 879         return ', '.join(names)
 880
 881     @classmethod
 882     def tagged_top_level(cls, tags):
 883         """ Returns top-level books tagged with `tags'.
 884
 885         It only returns those books which don't have ancestors which are
 886         also tagged with those tags.
 887
 888         """
 889         # get relevant books and their tags
 890         objects = cls.tagged.with_all(tags)
 891         # eliminate descendants
 892         l_tags = Tag.objects.filter(category='book', slug__in=[book.book_tag_slug() for book in objects])
 893         descendants_keys = [book.pk for book in cls.tagged.with_any(l_tags)]
 894         if descendants_keys:
 895             objects = objects.exclude(pk__in=descendants_keys)
 896
 897         return objects
 898
 899     @classmethod
 900     def book_list(cls, filter=None):
 901         """Generates a hierarchical listing of all books.
 902
 903         Books are optionally filtered with a test function.
 904
 905         """
 906
 907         books_by_parent = {}
 908         books = cls.objects.all().order_by('parent_number', 'sort_key').only(
 909                 'title', 'parent', 'slug')
 910         if filter:
 911             books = books.filter(filter).distinct()
 912             book_ids = set((book.pk for book in books))
 913             for book in books:
 914                 parent = book.parent_id
 915                 if parent not in book_ids:
 916                     parent = None
 917                 books_by_parent.setdefault(parent, []).append(book)
 918         else:
 919             for book in books:
 920                 books_by_parent.setdefault(book.parent_id, []).append(book)
 921
 922         orphans = []
 923         books_by_author = SortedDict()
 924         for tag in Tag.objects.filter(category='author'):
 925             books_by_author[tag] = []
 926
 927         for book in books_by_parent.get(None,()):
 928             authors = list(book.tags.filter(category='author'))
 929             if authors:
 930                 for author in authors:
 931                     books_by_author[author].append(book)
 932             else:
 933                 orphans.append(book)
 934
 935         return books_by_author, orphans, books_by_parent
 936
 937     _audiences_pl = {
 938         "SP1": (1, u"szkoła podstawowa"),
 939         "SP2": (1, u"szkoła podstawowa"),
 940         "P": (1, u"szkoła podstawowa"),
 941         "G": (2, u"gimnazjum"),
 942         "L": (3, u"liceum"),
 943         "LP": (3, u"liceum"),
 944     }
 945     def audiences_pl(self):
 946         audiences = self.get_extra_info_value().get('audiences', [])
 947         audiences = sorted(set([self._audiences_pl[a] for a in audiences]))
 948         return [a[1] for a in audiences]
 949
 950     def choose_fragment(self):
 951         tag = self.book_tag()
 952         fragments = Fragment.tagged.with_any([tag])
 953         if fragments.exists():
 954             return fragments.order_by('?')[0]
 955         elif self.parent:
 956             return self.parent.choose_fragment()
 957         else:
 958             return None
 959
 960
 961 def _has_factory(ftype):
 962     has = lambda self: bool(getattr(self, "%s_file" % ftype))
 963     has.short_description = t.upper()
 964     has.boolean = True
 965     has.__name__ = "has_%s_file" % ftype
 966     return has
 967
 968
 969 # add the file fields
 970 for t in Book.formats:
 971     field_name = "%s_file" % t
 972     models.FileField(_("%s file" % t.upper()),
 973             upload_to=book_upload_path(t),
 974             blank=True).contribute_to_class(Book, field_name)
 975
 976     setattr(Book, "has_%s_file" % t, _has_factory(t))
 977
 978
 979 class Fragment(models.Model):
 980     text = models.TextField()
 981     short_text = models.TextField(editable=False)
 982     anchor = models.CharField(max_length=120)
 983     book = models.ForeignKey(Book, related_name='fragments')
 984
 985     objects = models.Manager()
 986     tagged = managers.ModelTaggedItemManager(Tag)
 987     tags = managers.TagDescriptor(Tag)
 988
 989     class Meta:
 990         ordering = ('book', 'anchor',)
 991         verbose_name = _('fragment')
 992         verbose_name_plural = _('fragments')
 993
 994     def get_absolute_url(self):
 995         return '%s#m%s' % (reverse('book_text', args=[self.book.slug]), self.anchor)
 996
 997     def reset_short_html(self):
 998         if self.id is None:
 999             return
1000
1001         cache_key = "Fragment.short_html/%d/%s"
1002         for lang, langname in settings.LANGUAGES:
1003             permanent_cache.delete(cache_key % (self.id, lang))
1004
1005     def get_short_text(self):
1006         """Returns short version of the fragment."""
1007         return self.short_text if self.short_text else self.text
1008
1009     def short_html(self):
1010         if self.id:
1011             cache_key = "Fragment.short_html/%d/%s" % (self.id, get_language())
1012             short_html = permanent_cache.get(cache_key)
1013         else:
1014             short_html = None
1015
1016         if short_html is not None:
1017             return mark_safe(short_html)
1018         else:
1019             short_html = unicode(render_to_string('catalogue/fragment_short.html',
1020                 {'fragment': self}))
1021             if self.id:
1022                 permanent_cache.set(cache_key, short_html)
1023             return mark_safe(short_html)
1024
1025
1026 class Collection(models.Model):
1027     """A collection of books, which might be defined before publishing them."""
1028     title = models.CharField(_('title'), max_length=120, db_index=True)
1029     slug = models.SlugField(_('slug'), max_length=120, primary_key=True)
1030     description = models.TextField(_('description'), null=True, blank=True)
1031
1032     models.SlugField(_('slug'), max_length=120, unique=True, db_index=True)
1033     book_slugs = models.TextField(_('book slugs'))
1034
1035     class Meta:
1036         ordering = ('title',)
1037         verbose_name = _('collection')
1038         verbose_name_plural = _('collections')
1039
1040     def __unicode__(self):
1041         return self.title
1042
1043
1044 ###########
1045 #
1046 # SIGNALS
1047 #
1048 ###########
1049
1050
1051 def _tags_updated_handler(sender, affected_tags, **kwargs):
1052     # reset tag global counter
1053     # we want Tag.changed_at updated for API to know the tag was touched
1054     for tag in affected_tags:
1055         touch_tag(tag)
1056
1057     # if book tags changed, reset book tag counter
1058     if isinstance(sender, Book) and \
1059                 Tag.objects.filter(pk__in=(tag.pk for tag in affected_tags)).\
1060                     exclude(category__in=('book', 'theme', 'set')).count():
1061         sender.reset_tag_counter()
1062     # if fragment theme changed, reset book theme counter
1063     elif isinstance(sender, Fragment) and \
1064                 Tag.objects.filter(pk__in=(tag.pk for tag in affected_tags)).\
1065                     filter(category='theme').count():
1066         sender.book.reset_theme_counter()
1067 tags_updated.connect(_tags_updated_handler)
1068
1069
1070 def _pre_delete_handler(sender, instance, **kwargs):
1071     """ refresh Book on BookMedia delete """
1072     if sender == BookMedia:
1073         instance.book.save()
1074 pre_delete.connect(_pre_delete_handler)
1075
1076
1077 def _post_save_handler(sender, instance, **kwargs):
1078     """ refresh all the short_html stuff on BookMedia update """
1079     if sender == BookMedia:
1080         instance.book.save()
1081 post_save.connect(_post_save_handler)
1082
1083
1084 @django.dispatch.receiver(post_delete, sender=Book)
1085 def _remove_book_from_index_handler(sender, instance, **kwargs):
1086     """ remove the book from search index, when it is deleted."""
1087     idx = search.Index()
1088     idx.open(timeout=10000)  # 10 seconds timeout.
1089     try:
1090         idx.remove_book(instance)
1091     finally:
1092         idx.close()