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