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
  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     main_page = models.BooleanField(_('main page'), default=False, db_index=True, help_text=_('Show tag on main page'))
  65
  66     user = models.ForeignKey(User, blank=True, null=True)
  67     book_count = models.IntegerField(_('book count'), blank=True, null=True)
  68     gazeta_link = models.CharField(blank=True, max_length=240)
  69     wiki_link = models.CharField(blank=True, max_length=240)
  70
  71     created_at    = models.DateTimeField(_('creation date'), auto_now_add=True, db_index=True)
  72     changed_at    = models.DateTimeField(_('creation date'), auto_now=True, db_index=True)
  73
  74     class UrlDeprecationWarning(DeprecationWarning):
  75         pass
  76
  77     categories_rev = {
  78         'autor': 'author',
  79         'epoka': 'epoch',
  80         'rodzaj': 'kind',
  81         'gatunek': 'genre',
  82         'motyw': 'theme',
  83         'polka': 'set',
  84     }
  85     categories_dict = dict((item[::-1] for item in categories_rev.iteritems()))
  86
  87     class Meta:
  88         ordering = ('sort_key',)
  89         verbose_name = _('tag')
  90         verbose_name_plural = _('tags')
  91         unique_together = (("slug", "category"),)
  92
  93     def __unicode__(self):
  94         return self.name
  95
  96     def __repr__(self):
  97         return "Tag(slug=%r)" % self.slug
  98
  99     @permalink
 100     def get_absolute_url(self):
 101         return ('catalogue.views.tagged_object_list', [self.url_chunk])
 102
 103     def has_description(self):
 104         return len(self.description) > 0
 105     has_description.short_description = _('description')
 106     has_description.boolean = True
 107
 108     def get_count(self):
 109         """Returns global book count for book tags, fragment count for themes."""
 110
 111         if self.category == 'book':
 112             # never used
 113             objects = Book.objects.none()
 114         elif self.category == 'theme':
 115             objects = Fragment.tagged.with_all((self,))
 116         else:
 117             objects = Book.tagged.with_all((self,)).order_by()
 118             if self.category != 'set':
 119                 # eliminate descendants
 120                 l_tags = Tag.objects.filter(slug__in=[book.book_tag_slug() for book in objects])
 121                 descendants_keys = [book.pk for book in Book.tagged.with_any(l_tags)]
 122                 if descendants_keys:
 123                     objects = objects.exclude(pk__in=descendants_keys)
 124         return objects.count()
 125
 126     @staticmethod
 127     def get_tag_list(tags):
 128         if isinstance(tags, basestring):
 129             real_tags = []
 130             ambiguous_slugs = []
 131             category = None
 132             deprecated = False
 133             tags_splitted = tags.split('/')
 134             for name in tags_splitted:
 135                 if category:
 136                     real_tags.append(Tag.objects.get(slug=name, category=category))
 137                     category = None
 138                 elif name in Tag.categories_rev:
 139                     category = Tag.categories_rev[name]
 140                 else:
 141                     try:
 142                         real_tags.append(Tag.objects.exclude(category='book').get(slug=name))
 143                         deprecated = True
 144                     except Tag.MultipleObjectsReturned, e:
 145                         ambiguous_slugs.append(name)
 146
 147             if category:
 148                 # something strange left off
 149                 raise Tag.DoesNotExist()
 150             if ambiguous_slugs:
 151                 # some tags should be qualified
 152                 e = Tag.MultipleObjectsReturned()
 153                 e.tags = real_tags
 154                 e.ambiguous_slugs = ambiguous_slugs
 155                 raise e
 156             if deprecated:
 157                 e = Tag.UrlDeprecationWarning()
 158                 e.tags = real_tags
 159                 raise e
 160             return real_tags
 161         else:
 162             return TagBase.get_tag_list(tags)
 163
 164     @property
 165     def url_chunk(self):
 166         return '/'.join((Tag.categories_dict[self.category], self.slug))
 167
 168     @staticmethod
 169     def tags_from_info(info):
 170         from slughifi import slughifi
 171         from sortify import sortify
 172         meta_tags = []
 173         categories = (('kinds', 'kind'), ('genres', 'genre'), ('authors', 'author'), ('epochs', 'epoch'))
 174         for field_name, category in categories:
 175             try:
 176                 tag_names = getattr(info, field_name)
 177             except:
 178                 tag_names = [getattr(info, category)]
 179             for tag_name in tag_names:
 180                 tag_sort_key = tag_name
 181                 if category == 'author':
 182                     tag_sort_key = tag_name.last_name
 183                     tag_name = tag_name.readable()
 184                 tag, created = Tag.objects.get_or_create(slug=slughifi(tag_name), category=category)
 185                 if created:
 186                     tag.name = tag_name
 187                     tag.sort_key = sortify(tag_sort_key.lower())
 188                     tag.save()
 189                 meta_tags.append(tag)
 190         return meta_tags
 191
 192
 193
 194 def get_dynamic_path(media, filename, ext=None, maxlen=100):
 195     from slughifi import slughifi
 196
 197     # how to put related book's slug here?
 198     if not ext:
 199         # BookMedia case
 200         ext = media.formats[media.type].ext
 201     if media is None or not media.name:
 202         name = slughifi(filename.split(".")[0])
 203     else:
 204         name = slughifi(media.name)
 205     return 'book/%s/%s.%s' % (ext, name[:maxlen-len('book/%s/.%s' % (ext, ext))-4], ext)
 206
 207
 208 # TODO: why is this hard-coded ?
 209 def book_upload_path(ext=None, maxlen=100):
 210     return lambda *args: get_dynamic_path(*args, ext=ext, maxlen=maxlen)
 211
 212
 213 def get_customized_pdf_path(book, customizations):
 214     """
 215     Returns a MEDIA_ROOT relative path for a customized pdf. The name will contain a hash of customization options.
 216     """
 217     customizations.sort()
 218     h = hash(tuple(customizations))
 219
 220     pdf_name = '%s-custom-%s' % (book.fileid(), h)
 221     pdf_file = get_dynamic_path(None, pdf_name, ext='pdf')
 222
 223     return pdf_file
 224
 225
 226 def get_existing_customized_pdf(book):
 227     """
 228     Returns a list of paths to generated customized pdf of a book
 229     """
 230     pdf_glob = '%s-custom-' % (book.fileid(),)
 231     pdf_glob = get_dynamic_path(None, pdf_glob, ext='pdf')
 232     pdf_glob = re.sub(r"[.]([a-z0-9]+)$", "*.\\1", pdf_glob)
 233     return glob(path.join(settings.MEDIA_ROOT, pdf_glob))
 234
 235
 236 class BookMedia(models.Model):
 237     FileFormat = namedtuple("FileFormat", "name ext")
 238     formats = SortedDict([
 239         ('mp3', FileFormat(name='MP3', ext='mp3')),
 240         ('ogg', FileFormat(name='Ogg Vorbis', ext='ogg')),
 241         ('daisy', FileFormat(name='DAISY', ext='daisy.zip')),
 242     ])
 243     format_choices = [(k, _('%s file') % t.name)
 244             for k, t in formats.items()]
 245
 246     type        = models.CharField(_('type'), choices=format_choices, max_length="100")
 247     name        = models.CharField(_('name'), max_length="100")
 248     file        = OverwritingFileField(_('file'), upload_to=book_upload_path())
 249     uploaded_at = models.DateTimeField(_('creation date'), auto_now_add=True, editable=False)
 250     extra_info  = JSONField(_('extra information'), default='{}', editable=False)
 251     book = models.ForeignKey('Book', related_name='media')
 252     source_sha1 = models.CharField(null=True, blank=True, max_length=40, editable=False)
 253
 254     def __unicode__(self):
 255         return "%s (%s)" % (self.name, self.file.name.split("/")[-1])
 256
 257     class Meta:
 258         ordering            = ('type', 'name')
 259         verbose_name        = _('book media')
 260         verbose_name_plural = _('book media')
 261
 262     def save(self, *args, **kwargs):
 263         from slughifi import slughifi
 264         from catalogue.utils import ExistingFile, remove_zip
 265
 266         try:
 267             old = BookMedia.objects.get(pk=self.pk)
 268         except BookMedia.DoesNotExist, e:
 269             pass
 270         else:
 271             # if name changed, change the file name, too
 272             if slughifi(self.name) != slughifi(old.name):
 273                 self.file.save(None, ExistingFile(self.file.path), save=False, leave=True)
 274
 275         super(BookMedia, self).save(*args, **kwargs)
 276
 277         # remove the zip package for book with modified media
 278         remove_zip(self.book.fileid())
 279
 280         extra_info = self.get_extra_info_value()
 281         extra_info.update(self.read_meta())
 282         self.set_extra_info_value(extra_info)
 283         self.source_sha1 = self.read_source_sha1(self.file.path, self.type)
 284         return super(BookMedia, self).save(*args, **kwargs)
 285
 286     def read_meta(self):
 287         """
 288             Reads some metadata from the audiobook.
 289         """
 290         import mutagen
 291         from mutagen import id3
 292
 293         artist_name = director_name = project = funded_by = ''
 294         if self.type == 'mp3':
 295             try:
 296                 audio = id3.ID3(self.file.path)
 297                 artist_name = ', '.join(', '.join(tag.text) for tag in audio.getall('TPE1'))
 298                 director_name = ', '.join(', '.join(tag.text) for tag in audio.getall('TPE3'))
 299                 project = ", ".join([t.data for t in audio.getall('PRIV')
 300                         if t.owner=='wolnelektury.pl?project'])
 301                 funded_by = ", ".join([t.data for t in audio.getall('PRIV')
 302                         if t.owner=='wolnelektury.pl?funded_by'])
 303             except:
 304                 pass
 305         elif self.type == 'ogg':
 306             try:
 307                 audio = mutagen.File(self.file.path)
 308                 artist_name = ', '.join(audio.get('artist', []))
 309                 director_name = ', '.join(audio.get('conductor', []))
 310                 project = ", ".join(audio.get('project', []))
 311                 funded_by = ", ".join(audio.get('funded_by', []))
 312             except:
 313                 pass
 314         else:
 315             return {}
 316         return {'artist_name': artist_name, 'director_name': director_name,
 317                 'project': project, 'funded_by': funded_by}
 318
 319     @staticmethod
 320     def read_source_sha1(filepath, filetype):
 321         """
 322             Reads source file SHA1 from audiobok metadata.
 323         """
 324         import mutagen
 325         from mutagen import id3
 326
 327         if filetype == 'mp3':
 328             try:
 329                 audio = id3.ID3(filepath)
 330                 return [t.data for t in audio.getall('PRIV')
 331                         if t.owner=='wolnelektury.pl?flac_sha1'][0]
 332             except:
 333                 return None
 334         elif filetype == 'ogg':
 335             try:
 336                 audio = mutagen.File(filepath)
 337                 return audio.get('flac_sha1', [None])[0]
 338             except:
 339                 return None
 340         else:
 341             return None
 342
 343
 344 class Book(models.Model):
 345     title         = models.CharField(_('title'), max_length=120)
 346     sort_key = models.CharField(_('sort key'), max_length=120, db_index=True, editable=False)
 347     slug          = models.SlugField(_('slug'), max_length=120, db_index=True)
 348     language = models.CharField(_('language code'), max_length=3, db_index=True,
 349                     default=settings.CATALOGUE_DEFAULT_LANGUAGE)
 350     description   = models.TextField(_('description'), blank=True)
 351     created_at    = models.DateTimeField(_('creation date'), auto_now_add=True, db_index=True)
 352     changed_at    = models.DateTimeField(_('creation date'), auto_now=True, db_index=True)
 353     parent_number = models.IntegerField(_('parent number'), default=0)
 354     extra_info    = JSONField(_('extra information'), default='{}')
 355     gazeta_link   = models.CharField(blank=True, max_length=240)
 356     wiki_link     = models.CharField(blank=True, max_length=240)
 357     # files generated during publication
 358
 359     cover = models.FileField(_('cover'), upload_to=book_upload_path('png'),
 360                 null=True, blank=True)
 361     ebook_formats = ['pdf', 'epub', 'mobi', 'txt']
 362     formats = ebook_formats + ['html', 'xml']
 363
 364     parent        = models.ForeignKey('self', blank=True, null=True, related_name='children')
 365     objects  = models.Manager()
 366     tagged   = managers.ModelTaggedItemManager(Tag)
 367     tags     = managers.TagDescriptor(Tag)
 368
 369     html_built = django.dispatch.Signal()
 370     published = django.dispatch.Signal()
 371
 372     URLID_RE = r'[a-z0-9-]+(?:/[a-z]{3})?'
 373     FILEID_RE = r'[a-z0-9-]+(?:_[a-z]{3})?'
 374
 375     class AlreadyExists(Exception):
 376         pass
 377
 378     class Meta:
 379         unique_together = [['slug', 'language']]
 380         ordering = ('sort_key',)
 381         verbose_name = _('book')
 382         verbose_name_plural = _('books')
 383
 384     def __unicode__(self):
 385         return self.title
 386
 387     def urlid(self, sep='/'):
 388         stem = self.slug
 389         if self.language != settings.CATALOGUE_DEFAULT_LANGUAGE:
 390             stem += sep + self.language
 391         return stem
 392
 393     def fileid(self):
 394         return self.urlid('_')
 395
 396     @staticmethod
 397     def split_urlid(urlid, sep='/', default_lang=settings.CATALOGUE_DEFAULT_LANGUAGE):
 398         """Splits a URL book id into slug and language code.
 399
 400         Returns a dictionary usable i.e. for object lookup, or None.
 401
 402         >>> Book.split_urlid("a-slug/pol", default_lang="eng")
 403         {'slug': 'a-slug', 'language': 'pol'}
 404         >>> Book.split_urlid("a-slug", default_lang="eng")
 405         {'slug': 'a-slug', 'language': 'eng'}
 406         >>> Book.split_urlid("a-slug_pol", "_", default_lang="eng")
 407         {'slug': 'a-slug', 'language': 'pol'}
 408         >>> Book.split_urlid("a-slug/eng", default_lang="eng")
 409
 410         """
 411         parts = urlid.rsplit(sep, 1)
 412         if len(parts) == 2:
 413             if parts[1] == default_lang:
 414                 return None
 415             return {'slug': parts[0], 'language': parts[1]}
 416         else:
 417             return {'slug': urlid, 'language': default_lang}
 418
 419     @classmethod
 420     def split_fileid(cls, fileid):
 421         return cls.split_urlid(fileid, '_')
 422
 423     def save(self, force_insert=False, force_update=False, reset_short_html=True, **kwargs):
 424         from sortify import sortify
 425
 426         self.sort_key = sortify(self.title)
 427
 428         ret = super(Book, self).save(force_insert, force_update)
 429
 430         if reset_short_html:
 431             self.reset_short_html()
 432
 433         return ret
 434
 435     @permalink
 436     def get_absolute_url(self):
 437         return ('catalogue.views.book_detail', [self.urlid()])
 438
 439     @property
 440     def name(self):
 441         return self.title
 442
 443     def book_tag_slug(self):
 444         stem = 'l-' + self.slug
 445         if self.language != settings.CATALOGUE_DEFAULT_LANGUAGE:
 446             return stem[:116] + ' ' + self.language
 447         else:
 448             return stem[:120]
 449
 450     def book_tag(self):
 451         slug = self.book_tag_slug()
 452         book_tag, created = Tag.objects.get_or_create(slug=slug, category='book')
 453         if created:
 454             book_tag.name = self.title[:50]
 455             book_tag.sort_key = self.title.lower()
 456             book_tag.save()
 457         return book_tag
 458
 459     def has_media(self, type):
 460         if type in Book.formats:
 461             return bool(getattr(self, "%s_file" % type))
 462         else:
 463             return self.media.filter(type=type).exists()
 464
 465     def get_media(self, type):
 466         if self.has_media(type):
 467             if type in Book.formats:
 468                 return getattr(self, "%s_file" % type)
 469             else:
 470                 return self.media.filter(type=type)
 471         else:
 472             return None
 473
 474     def get_mp3(self):
 475         return self.get_media("mp3")
 476     def get_odt(self):
 477         return self.get_media("odt")
 478     def get_ogg(self):
 479         return self.get_media("ogg")
 480     def get_daisy(self):
 481         return self.get_media("daisy")
 482
 483     def reset_short_html(self):
 484         if self.id is None:
 485             return
 486
 487         cache_key = "Book.short_html/%d/%s"
 488         for lang, langname in settings.LANGUAGES:
 489             cache.delete(cache_key % (self.id, lang))
 490         # Fragment.short_html relies on book's tags, so reset it here too
 491         for fragm in self.fragments.all():
 492             fragm.reset_short_html()
 493
 494     def short_html(self):
 495         if self.id:
 496             cache_key = "Book.short_html/%d/%s" % (self.id, get_language())
 497             short_html = cache.get(cache_key)
 498         else:
 499             short_html = None
 500
 501         if short_html is not None:
 502             return mark_safe(short_html)
 503         else:
 504             tags = self.tags.filter(~Q(category__in=('set', 'theme', 'book')))
 505             tags = [mark_safe(u'<a href="%s">%s</a>' % (tag.get_absolute_url(), tag.name)) for tag in tags]
 506
 507             formats = []
 508             # files generated during publication
 509             if self.has_media("html"):
 510                 formats.append(u'<a href="%s">%s</a>' % (reverse('book_text', args=[self.fileid()]), _('Read online')))
 511             for ebook_format in self.ebook_formats:
 512                 if self.has_media(ebook_format):
 513                     formats.append(u'<a href="%s">%s</a>' % (
 514                         self.get_media(ebook_format).url,
 515                         ebook_format.upper()
 516                     ))
 517             # other files
 518             for m in self.media.order_by('type'):
 519                 formats.append(u'<a href="%s">%s</a>' % (m.file.url, m.type.upper()))
 520
 521             formats = [mark_safe(format) for format in formats]
 522
 523             short_html = unicode(render_to_string('catalogue/book_short.html',
 524                 {'book': self, 'tags': tags, 'formats': formats}))
 525
 526             if self.id:
 527                 cache.set(cache_key, short_html, CACHE_FOREVER)
 528             return mark_safe(short_html)
 529
 530     def mini_box(self):
 531         if self.id:
 532             cache_key = "Book.mini_boxs/%d" % (self.id, )
 533             short_html = cache.get(cache_key)
 534         else:
 535             short_html = None
 536
 537         if short_html is None:
 538             authors = self.tags.filter(category='author')
 539
 540             short_html = unicode(render_to_string('catalogue/book_mini_box.html',
 541                 {'book': self, 'authors': authors, 'STATIC_URL': settings.STATIC_URL}))
 542
 543             if self.id:
 544                 cache.set(cache_key, short_html, CACHE_FOREVER)
 545         return mark_safe(short_html)
 546
 547     def has_description(self):
 548         return len(self.description) > 0
 549     has_description.short_description = _('description')
 550     has_description.boolean = True
 551
 552     # ugly ugly ugly
 553     def has_mp3_file(self):
 554         return bool(self.has_media("mp3"))
 555     has_mp3_file.short_description = 'MP3'
 556     has_mp3_file.boolean = True
 557
 558     def has_ogg_file(self):
 559         return bool(self.has_media("ogg"))
 560     has_ogg_file.short_description = 'OGG'
 561     has_ogg_file.boolean = True
 562
 563     def has_daisy_file(self):
 564         return bool(self.has_media("daisy"))
 565     has_daisy_file.short_description = 'DAISY'
 566     has_daisy_file.boolean = True
 567
 568     def wldocument(self, parse_dublincore=True):
 569         from catalogue.utils import ORMDocProvider
 570         from librarian.parser import WLDocument
 571
 572         return WLDocument.from_file(self.xml_file.path,
 573                 provider=ORMDocProvider(self),
 574                 parse_dublincore=parse_dublincore)
 575
 576     def build_cover(self, book_info=None):
 577         """(Re)builds the cover image."""
 578         from StringIO import StringIO
 579         from django.core.files.base import ContentFile
 580         from librarian.cover import WLCover
 581
 582         if book_info is None:
 583             book_info = self.wldocument().book_info
 584
 585         cover = WLCover(book_info).image()
 586         imgstr = StringIO()
 587         cover.save(imgstr, 'png')
 588         self.cover.save(None, ContentFile(imgstr.getvalue()))
 589
 590     def build_pdf(self, customizations=None, file_name=None):
 591         """ (Re)builds the pdf file.
 592         customizations - customizations which are passed to LaTeX class file.
 593         file_name - save the pdf file under a different name and DO NOT save it in db.
 594         """
 595         from os import unlink
 596         from django.core.files import File
 597         from catalogue.utils import remove_zip
 598
 599         pdf = self.wldocument().as_pdf(customizations=customizations)
 600
 601         if file_name is None:
 602             # we'd like to be sure not to overwrite changes happening while
 603             # (timely) pdf generation is taking place (async celery scenario)
 604             current_self = Book.objects.get(id=self.id)
 605             current_self.pdf_file.save('%s.pdf' % self.fileid(),
 606                     File(open(pdf.get_filename())))
 607             self.pdf_file = current_self.pdf_file
 608
 609             # remove cached downloadables
 610             remove_zip(settings.ALL_PDF_ZIP)
 611
 612             for customized_pdf in get_existing_customized_pdf(self):
 613                 unlink(customized_pdf)
 614         else:
 615             print "saving %s" % file_name
 616             print "to: %s" % DefaultStorage().path(file_name)
 617             DefaultStorage().save(file_name, File(open(pdf.get_filename())))
 618
 619     def build_mobi(self):
 620         """ (Re)builds the MOBI file.
 621
 622         """
 623         from django.core.files import File
 624         from catalogue.utils import remove_zip
 625
 626         mobi = self.wldocument().as_mobi()
 627
 628         self.mobi_file.save('%s.mobi' % self.fileid(), File(open(mobi.get_filename())))
 629
 630         # remove zip with all mobi files
 631         remove_zip(settings.ALL_MOBI_ZIP)
 632
 633     def build_epub(self):
 634         """(Re)builds the epub file."""
 635         from django.core.files import File
 636         from catalogue.utils import remove_zip
 637
 638         epub = self.wldocument().as_epub()
 639
 640         self.epub_file.save('%s.epub' % self.fileid(),
 641                 File(open(epub.get_filename())))
 642
 643         # remove zip package with all epub files
 644         remove_zip(settings.ALL_EPUB_ZIP)
 645
 646     def build_txt(self):
 647         from django.core.files.base import ContentFile
 648
 649         text = self.wldocument().as_text()
 650         self.txt_file.save('%s.txt' % self.fileid(), ContentFile(text.get_string()))
 651
 652
 653     def build_html(self):
 654         from markupstring import MarkupString
 655         from django.core.files.base import ContentFile
 656         from slughifi import slughifi
 657         from librarian import html
 658
 659         meta_tags = list(self.tags.filter(
 660             category__in=('author', 'epoch', 'genre', 'kind')))
 661         book_tag = self.book_tag()
 662
 663         html_output = self.wldocument(parse_dublincore=False).as_html()
 664         if html_output:
 665             self.html_file.save('%s.html' % self.fileid(),
 666                     ContentFile(html_output.get_string()))
 667
 668             # get ancestor l-tags for adding to new fragments
 669             ancestor_tags = []
 670             p = self.parent
 671             while p:
 672                 ancestor_tags.append(p.book_tag())
 673                 p = p.parent
 674
 675             # Delete old fragments and create them from scratch
 676             self.fragments.all().delete()
 677             # Extract fragments
 678             closed_fragments, open_fragments = html.extract_fragments(self.html_file.path)
 679             for fragment in closed_fragments.values():
 680                 try:
 681                     theme_names = [s.strip() for s in fragment.themes.split(',')]
 682                 except AttributeError:
 683                     continue
 684                 themes = []
 685                 for theme_name in theme_names:
 686                     if not theme_name:
 687                         continue
 688                     tag, created = Tag.objects.get_or_create(slug=slughifi(theme_name), category='theme')
 689                     if created:
 690                         tag.name = theme_name
 691                         tag.sort_key = theme_name.lower()
 692                         tag.save()
 693                     themes.append(tag)
 694                 if not themes:
 695                     continue
 696
 697                 text = fragment.to_string()
 698                 short_text = ''
 699                 if (len(MarkupString(text)) > 240):
 700                     short_text = unicode(MarkupString(text)[:160])
 701                 new_fragment = Fragment.objects.create(anchor=fragment.id, book=self,
 702                     text=text, short_text=short_text)
 703
 704                 new_fragment.save()
 705                 new_fragment.tags = set(meta_tags + themes + [book_tag] + ancestor_tags)
 706             self.save()
 707             self.html_built.send(sender=self)
 708             return True
 709         return False
 710
 711     @staticmethod
 712     def zip_format(format_):
 713         def pretty_file_name(book):
 714             return "%s/%s.%s" % (
 715                 b.get_extra_info_value()['author'],
 716                 b.fileid(),
 717                 format_)
 718
 719         field_name = "%s_file" % format_
 720         books = Book.objects.filter(parent=None).exclude(**{field_name: ""})
 721         paths = [(pretty_file_name(b), getattr(b, field_name).path)
 722                     for b in books]
 723         result = create_zip.delay(paths,
 724                     getattr(settings, "ALL_%s_ZIP" % format_.upper()))
 725         return result.wait()
 726
 727     def zip_audiobooks(self):
 728         bm = BookMedia.objects.filter(book=self, type='mp3')
 729         paths = map(lambda bm: (None, bm.file.path), bm)
 730         result = create_zip.delay(paths, self.fileid())
 731         return result.wait()
 732
 733     @classmethod
 734     def from_xml_file(cls, xml_file, **kwargs):
 735         from django.core.files import File
 736         from librarian import dcparser
 737
 738         # use librarian to parse meta-data
 739         book_info = dcparser.parse(xml_file)
 740
 741         if not isinstance(xml_file, File):
 742             xml_file = File(open(xml_file))
 743
 744         try:
 745             return cls.from_text_and_meta(xml_file, book_info, **kwargs)
 746         finally:
 747             xml_file.close()
 748
 749     @classmethod
 750     def from_text_and_meta(cls, raw_file, book_info, overwrite=False,
 751             build_epub=True, build_txt=True, build_pdf=True, build_mobi=True):
 752         import re
 753         from sortify import sortify
 754
 755         # check for parts before we do anything
 756         children = []
 757         if hasattr(book_info, 'parts'):
 758             for part_url in book_info.parts:
 759                 try:
 760                     children.append(Book.objects.get(
 761                         slug=part_url.slug, language=part_url.language))
 762                 except Book.DoesNotExist, e:
 763                     raise Book.DoesNotExist(_('Book "%s/%s" does not exist.') %
 764                             (part_url.slug, part_url.language))
 765
 766
 767         # Read book metadata
 768         book_slug = book_info.url.slug
 769         language = book_info.language
 770         if re.search(r'[^a-zA-Z0-9-]', book_slug):
 771             raise ValueError('Invalid characters in slug')
 772         book, created = Book.objects.get_or_create(slug=book_slug, language=language)
 773
 774         if created:
 775             book_shelves = []
 776         else:
 777             if not overwrite:
 778                 raise Book.AlreadyExists(_('Book %s/%s already exists') % (
 779                         book_slug, language))
 780             # Save shelves for this book
 781             book_shelves = list(book.tags.filter(category='set'))
 782
 783         book.title = book_info.title
 784         book.set_extra_info_value(book_info.to_dict())
 785         book.save()
 786
 787         meta_tags = Tag.tags_from_info(book_info)
 788
 789         book.tags = set(meta_tags + book_shelves)
 790
 791         book_tag = book.book_tag()
 792
 793         for n, child_book in enumerate(children):
 794             child_book.parent = book
 795             child_book.parent_number = n
 796             child_book.save()
 797
 798         # Save XML and HTML files
 799         book.xml_file.save('%s.xml' % book.slug, raw_file, save=False)
 800
 801         # delete old fragments when overwriting
 802         book.fragments.all().delete()
 803
 804         if book.build_html():
 805             if not settings.NO_BUILD_TXT and build_txt:
 806                 book.build_txt()
 807
 808         book.build_cover(book_info)
 809
 810         if not settings.NO_BUILD_EPUB and build_epub:
 811             book.build_epub()
 812
 813         if not settings.NO_BUILD_PDF and build_pdf:
 814             book.build_pdf()
 815
 816         if not settings.NO_BUILD_MOBI and build_mobi:
 817             book.build_mobi()
 818
 819         book_descendants = list(book.children.all())
 820         descendants_tags = set()
 821         # add l-tag to descendants and their fragments
 822         while len(book_descendants) > 0:
 823             child_book = book_descendants.pop(0)
 824             descendants_tags.update(child_book.tags)
 825             child_book.tags = list(child_book.tags) + [book_tag]
 826             child_book.save()
 827             for fragment in child_book.fragments.all():
 828                 fragment.tags = set(list(fragment.tags) + [book_tag])
 829             book_descendants += list(child_book.children.all())
 830
 831         for tag in descendants_tags:
 832             touch_tag.delay(tag)
 833
 834         book.save()
 835
 836         # refresh cache
 837         book.reset_tag_counter()
 838         book.reset_theme_counter()
 839
 840         cls.published.send(sender=book)
 841         return book
 842
 843     def reset_tag_counter(self):
 844         if self.id is None:
 845             return
 846
 847         cache_key = "Book.tag_counter/%d" % self.id
 848         cache.delete(cache_key)
 849         if self.parent:
 850             self.parent.reset_tag_counter()
 851
 852     @property
 853     def tag_counter(self):
 854         if self.id:
 855             cache_key = "Book.tag_counter/%d" % self.id
 856             tags = cache.get(cache_key)
 857         else:
 858             tags = None
 859
 860         if tags is None:
 861             tags = {}
 862             for child in self.children.all().order_by():
 863                 for tag_pk, value in child.tag_counter.iteritems():
 864                     tags[tag_pk] = tags.get(tag_pk, 0) + value
 865             for tag in self.tags.exclude(category__in=('book', 'theme', 'set')).order_by():
 866                 tags[tag.pk] = 1
 867
 868             if self.id:
 869                 cache.set(cache_key, tags, CACHE_FOREVER)
 870         return tags
 871
 872     def reset_theme_counter(self):
 873         if self.id is None:
 874             return
 875
 876         cache_key = "Book.theme_counter/%d" % self.id
 877         cache.delete(cache_key)
 878         if self.parent:
 879             self.parent.reset_theme_counter()
 880
 881     @property
 882     def theme_counter(self):
 883         if self.id:
 884             cache_key = "Book.theme_counter/%d" % self.id
 885             tags = cache.get(cache_key)
 886         else:
 887             tags = None
 888
 889         if tags is None:
 890             tags = {}
 891             for fragment in Fragment.tagged.with_any([self.book_tag()]).order_by():
 892                 for tag in fragment.tags.filter(category='theme').order_by():
 893                     tags[tag.pk] = tags.get(tag.pk, 0) + 1
 894
 895             if self.id:
 896                 cache.set(cache_key, tags, CACHE_FOREVER)
 897         return tags
 898
 899     def pretty_title(self, html_links=False):
 900         book = self
 901         names = list(book.tags.filter(category='author'))
 902
 903         books = []
 904         while book:
 905             books.append(book)
 906             book = book.parent
 907         names.extend(reversed(books))
 908
 909         if html_links:
 910             names = ['<a href="%s">%s</a>' % (tag.get_absolute_url(), tag.name) for tag in names]
 911         else:
 912             names = [tag.name for tag in names]
 913
 914         return ', '.join(names)
 915
 916     @classmethod
 917     def tagged_top_level(cls, tags):
 918         """ Returns top-level books tagged with `tags'.
 919
 920         It only returns those books which don't have ancestors which are
 921         also tagged with those tags.
 922
 923         """
 924         # get relevant books and their tags
 925         objects = cls.tagged.with_all(tags)
 926         # eliminate descendants
 927         l_tags = Tag.objects.filter(category='book', slug__in=[book.book_tag_slug() for book in objects])
 928         descendants_keys = [book.pk for book in cls.tagged.with_any(l_tags)]
 929         if descendants_keys:
 930             objects = objects.exclude(pk__in=descendants_keys)
 931
 932         return objects
 933
 934     @classmethod
 935     def book_list(cls, filter=None):
 936         """Generates a hierarchical listing of all books.
 937
 938         Books are optionally filtered with a test function.
 939
 940         """
 941
 942         books_by_parent = {}
 943         books = cls.objects.all().order_by('parent_number', 'sort_key').only(
 944                 'title', 'parent', 'slug', 'language')
 945         if filter:
 946             books = books.filter(filter).distinct()
 947             book_ids = set((book.pk for book in books))
 948             for book in books:
 949                 parent = book.parent_id
 950                 if parent not in book_ids:
 951                     parent = None
 952                 books_by_parent.setdefault(parent, []).append(book)
 953         else:
 954             for book in books:
 955                 books_by_parent.setdefault(book.parent_id, []).append(book)
 956
 957         orphans = []
 958         books_by_author = SortedDict()
 959         for tag in Tag.objects.filter(category='author'):
 960             books_by_author[tag] = []
 961
 962         for book in books_by_parent.get(None,()):
 963             authors = list(book.tags.filter(category='author'))
 964             if authors:
 965                 for author in authors:
 966                     books_by_author[author].append(book)
 967             else:
 968                 orphans.append(book)
 969
 970         return books_by_author, orphans, books_by_parent
 971
 972     _audiences_pl = {
 973         "SP1": (1, u"szkoła podstawowa"),
 974         "SP2": (1, u"szkoła podstawowa"),
 975         "P": (1, u"szkoła podstawowa"),
 976         "G": (2, u"gimnazjum"),
 977         "L": (3, u"liceum"),
 978         "LP": (3, u"liceum"),
 979     }
 980     def audiences_pl(self):
 981         audiences = self.get_extra_info_value().get('audiences', [])
 982         audiences = sorted(set([self._audiences_pl[a] for a in audiences]))
 983         return [a[1] for a in audiences]
 984
 985
 986 def _has_factory(ftype):
 987     has = lambda self: bool(getattr(self, "%s_file" % ftype))
 988     has.short_description = t.upper()
 989     has.boolean = True
 990     has.__name__ = "has_%s_file" % ftype
 991     return has
 992
 993
 994 # add the file fields
 995 for t in Book.formats:
 996     field_name = "%s_file" % t
 997     models.FileField(_("%s file" % t.upper()),
 998             upload_to=book_upload_path(t),
 999             blank=True).contribute_to_class(Book, field_name)
1000
1001     setattr(Book, "has_%s_file" % t, _has_factory(t))
1002
1003
1004 class Fragment(models.Model):
1005     text = models.TextField()
1006     short_text = models.TextField(editable=False)
1007     anchor = models.CharField(max_length=120)
1008     book = models.ForeignKey(Book, related_name='fragments')
1009
1010     objects = models.Manager()
1011     tagged = managers.ModelTaggedItemManager(Tag)
1012     tags = managers.TagDescriptor(Tag)
1013
1014     class Meta:
1015         ordering = ('book', 'anchor',)
1016         verbose_name = _('fragment')
1017         verbose_name_plural = _('fragments')
1018
1019     def get_absolute_url(self):
1020         return '%s#m%s' % (self.book.get_html_url(), self.anchor)
1021
1022     def reset_short_html(self):
1023         if self.id is None:
1024             return
1025
1026         cache_key = "Fragment.short_html/%d/%s"
1027         for lang, langname in settings.LANGUAGES:
1028             cache.delete(cache_key % (self.id, lang))
1029
1030     def short_html(self):
1031         if self.id:
1032             cache_key = "Fragment.short_html/%d/%s" % (self.id, get_language())
1033             short_html = cache.get(cache_key)
1034         else:
1035             short_html = None
1036
1037         if short_html is not None:
1038             return mark_safe(short_html)
1039         else:
1040             short_html = unicode(render_to_string('catalogue/fragment_short.html',
1041                 {'fragment': self}))
1042             if self.id:
1043                 cache.set(cache_key, short_html, CACHE_FOREVER)
1044             return mark_safe(short_html)
1045
1046
1047 ###########
1048 #
1049 # SIGNALS
1050 #
1051 ###########
1052
1053
1054 def _tags_updated_handler(sender, affected_tags, **kwargs):
1055     # reset tag global counter
1056     # we want Tag.changed_at updated for API to know the tag was touched
1057     for tag in affected_tags:
1058         touch_tag.delay(tag)
1059
1060     # if book tags changed, reset book tag counter
1061     if isinstance(sender, Book) and \
1062                 Tag.objects.filter(pk__in=(tag.pk for tag in affected_tags)).\
1063                     exclude(category__in=('book', 'theme', 'set')).count():
1064         sender.reset_tag_counter()
1065     # if fragment theme changed, reset book theme counter
1066     elif isinstance(sender, Fragment) and \
1067                 Tag.objects.filter(pk__in=(tag.pk for tag in affected_tags)).\
1068                     filter(category='theme').count():
1069         sender.book.reset_theme_counter()
1070 tags_updated.connect(_tags_updated_handler)
1071
1072
1073 def _pre_delete_handler(sender, instance, **kwargs):
1074     """ refresh Book on BookMedia delete """
1075     if sender == BookMedia:
1076         instance.book.save()
1077 pre_delete.connect(_pre_delete_handler)
1078
1079 def _post_save_handler(sender, instance, **kwargs):
1080     """ refresh all the short_html stuff on BookMedia update """
1081     if sender == BookMedia:
1082         instance.book.save()
1083 post_save.connect(_post_save_handler)