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