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 datetime import datetime
   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 shutil import copy
  28 from glob import glob
  29 import re
  30 from os import path
  31
  32
  33 TAG_CATEGORIES = (
  34     ('author', _('author')),
  35     ('epoch', _('epoch')),
  36     ('kind', _('kind')),
  37     ('genre', _('genre')),
  38     ('theme', _('theme')),
  39     ('set', _('set')),
  40     ('book', _('book')),
  41 )
  42
  43 MEDIA_FORMATS = (
  44     ('odt', _('ODT file')),
  45     ('mp3', _('MP3 file')),
  46     ('ogg', _('OGG file')),
  47     ('daisy', _('DAISY file')),
  48 )
  49
  50 # not quite, but Django wants you to set a timeout
  51 CACHE_FOREVER = 2419200  # 28 days
  52
  53
  54 class TagSubcategoryManager(models.Manager):
  55     def __init__(self, subcategory):
  56         super(TagSubcategoryManager, self).__init__()
  57         self.subcategory = subcategory
  58
  59     def get_query_set(self):
  60         return super(TagSubcategoryManager, self).get_query_set().filter(category=self.subcategory)
  61
  62
  63 class Tag(TagBase):
  64     name = models.CharField(_('name'), max_length=50, db_index=True)
  65     slug = models.SlugField(_('slug'), max_length=120, db_index=True)
  66     sort_key = models.CharField(_('sort key'), max_length=120, db_index=True)
  67     category = models.CharField(_('category'), max_length=50, blank=False, null=False,
  68         db_index=True, choices=TAG_CATEGORIES)
  69     description = models.TextField(_('description'), blank=True)
  70     main_page = models.BooleanField(_('main page'), default=False, db_index=True, help_text=_('Show tag on main page'))
  71
  72     user = models.ForeignKey(User, blank=True, null=True)
  73     book_count = models.IntegerField(_('book count'), blank=True, null=True)
  74     gazeta_link = models.CharField(blank=True, max_length=240)
  75     wiki_link = models.CharField(blank=True, max_length=240)
  76
  77     created_at    = models.DateTimeField(_('creation date'), auto_now_add=True, db_index=True)
  78     changed_at    = models.DateTimeField(_('creation date'), auto_now=True, db_index=True)
  79
  80     class UrlDeprecationWarning(DeprecationWarning):
  81         pass
  82
  83     categories_rev = {
  84         'autor': 'author',
  85         'epoka': 'epoch',
  86         'rodzaj': 'kind',
  87         'gatunek': 'genre',
  88         'motyw': 'theme',
  89         'polka': 'set',
  90     }
  91     categories_dict = dict((item[::-1] for item in categories_rev.iteritems()))
  92
  93     class Meta:
  94         ordering = ('sort_key',)
  95         verbose_name = _('tag')
  96         verbose_name_plural = _('tags')
  97         unique_together = (("slug", "category"),)
  98
  99     def __unicode__(self):
 100         return self.name
 101
 102     def __repr__(self):
 103         return "Tag(slug=%r)" % self.slug
 104
 105     @permalink
 106     def get_absolute_url(self):
 107         return ('catalogue.views.tagged_object_list', [self.url_chunk])
 108
 109     def has_description(self):
 110         return len(self.description) > 0
 111     has_description.short_description = _('description')
 112     has_description.boolean = True
 113
 114     def get_count(self):
 115         """ returns global book count for book tags, fragment count for themes """
 116
 117         if self.book_count is None:
 118             if self.category == 'book':
 119                 # never used
 120                 objects = Book.objects.none()
 121             elif self.category == 'theme':
 122                 objects = Fragment.tagged.with_all((self,))
 123             else:
 124                 objects = Book.tagged.with_all((self,)).order_by()
 125                 if self.category != 'set':
 126                     # eliminate descendants
 127                     l_tags = Tag.objects.filter(slug__in=[book.book_tag_slug() for book in objects])
 128                     descendants_keys = [book.pk for book in Book.tagged.with_any(l_tags)]
 129                     if descendants_keys:
 130                         objects = objects.exclude(pk__in=descendants_keys)
 131             self.book_count = objects.count()
 132             self.save()
 133         return self.book_count
 134
 135     @staticmethod
 136     def get_tag_list(tags):
 137         if isinstance(tags, basestring):
 138             real_tags = []
 139             ambiguous_slugs = []
 140             category = None
 141             deprecated = False
 142             tags_splitted = tags.split('/')
 143             for name in tags_splitted:
 144                 if category:
 145                     real_tags.append(Tag.objects.get(slug=name, category=category))
 146                     category = None
 147                 elif name in Tag.categories_rev:
 148                     category = Tag.categories_rev[name]
 149                 else:
 150                     try:
 151                         real_tags.append(Tag.objects.exclude(category='book').get(slug=name))
 152                         deprecated = True
 153                     except Tag.MultipleObjectsReturned, e:
 154                         ambiguous_slugs.append(name)
 155
 156             if category:
 157                 # something strange left off
 158                 raise Tag.DoesNotExist()
 159             if ambiguous_slugs:
 160                 # some tags should be qualified
 161                 e = Tag.MultipleObjectsReturned()
 162                 e.tags = real_tags
 163                 e.ambiguous_slugs = ambiguous_slugs
 164                 raise e
 165             if deprecated:
 166                 e = Tag.UrlDeprecationWarning()
 167                 e.tags = real_tags
 168                 raise e
 169             return real_tags
 170         else:
 171             return TagBase.get_tag_list(tags)
 172
 173     @property
 174     def url_chunk(self):
 175         return '/'.join((Tag.categories_dict[self.category], self.slug))
 176
 177
 178 def get_dynamic_path(media, filename, ext=None, maxlen=100):
 179     from slughifi import slughifi
 180
 181     # how to put related book's slug here?
 182     if not ext:
 183         if media.type == 'daisy':
 184             ext = 'daisy.zip'
 185         else:
 186             ext = media.type
 187     if media is None or not media.name:
 188         name = slughifi(filename.split(".")[0])
 189     else:
 190         name = slughifi(media.name)
 191     return 'book/%s/%s.%s' % (ext, name[:maxlen-len('book/%s/.%s' % (ext, ext))-4], ext)
 192
 193
 194 # TODO: why is this hard-coded ?
 195 def book_upload_path(ext=None, maxlen=100):
 196     return lambda *args: get_dynamic_path(*args, ext=ext, maxlen=maxlen)
 197
 198
 199 def get_customized_pdf_path(book, customizations):
 200     """
 201     Returns a MEDIA_ROOT relative path for a customized pdf. The name will contain a hash of customization options.
 202     """
 203     customizations.sort()
 204     h = hash(tuple(customizations))
 205     pdf_name = '%s-custom-%s' % (book.fileid(), h)
 206     pdf_file = models.get_dynamic_path(None, pdf_name, ext='pdf')
 207     return pdf_file
 208
 209
 210 def get_existing_customized_pdf(book):
 211     """
 212     Returns a list of paths to generated customized pdf of a book
 213     """
 214     pdf_glob = '%s-custom-' % (book.fileid(),)
 215     pdf_glob = get_dynamic_path(None, pdf_glob, ext='pdf')
 216     pdf_glob = re.sub(r"[.]([a-z0-9]+)$", "*.\\1", pdf_glob)
 217     return glob(path.join(settings.MEDIA_ROOT, pdf_glob))
 218
 219
 220 class BookMedia(models.Model):
 221     type        = models.CharField(_('type'), choices=MEDIA_FORMATS, max_length="100")
 222     name        = models.CharField(_('name'), max_length="100")
 223     file        = OverwritingFileField(_('file'), upload_to=book_upload_path())
 224     uploaded_at = models.DateTimeField(_('creation date'), auto_now_add=True, editable=False)
 225     extra_info  = JSONField(_('extra information'), default='{}', editable=False)
 226     book = models.ForeignKey('Book', related_name='media')
 227     source_sha1 = models.CharField(null=True, blank=True, max_length=40, editable=False)
 228
 229     def __unicode__(self):
 230         return "%s (%s)" % (self.name, self.file.name.split("/")[-1])
 231
 232     class Meta:
 233         ordering            = ('type', 'name')
 234         verbose_name        = _('book media')
 235         verbose_name_plural = _('book media')
 236
 237     def save(self, *args, **kwargs):
 238         from slughifi import slughifi
 239         from catalogue.utils import ExistingFile, remove_zip
 240
 241         try:
 242             old = BookMedia.objects.get(pk=self.pk)
 243         except BookMedia.DoesNotExist, e:
 244             pass
 245         else:
 246             # if name changed, change the file name, too
 247             if slughifi(self.name) != slughifi(old.name):
 248                 self.file.save(None, ExistingFile(self.file.path), save=False, leave=True)
 249
 250         super(BookMedia, self).save(*args, **kwargs)
 251
 252         # remove the zip package for book with modified media
 253         remove_zip(self.book.fileid())
 254
 255         extra_info = self.get_extra_info_value()
 256         extra_info.update(self.read_meta())
 257         self.set_extra_info_value(extra_info)
 258         self.source_sha1 = self.read_source_sha1(self.file.path, self.type)
 259         return super(BookMedia, self).save(*args, **kwargs)
 260
 261     def read_meta(self):
 262         """
 263             Reads some metadata from the audiobook.
 264         """
 265         import mutagen
 266         from mutagen import id3
 267
 268         artist_name = director_name = project = funded_by = ''
 269         if self.type == 'mp3':
 270             try:
 271                 audio = id3.ID3(self.file.path)
 272                 artist_name = ', '.join(', '.join(tag.text) for tag in audio.getall('TPE1'))
 273                 director_name = ', '.join(', '.join(tag.text) for tag in audio.getall('TPE3'))
 274                 project = ", ".join([t.data for t in audio.getall('PRIV')
 275                         if t.owner=='wolnelektury.pl?project'])
 276                 funded_by = ", ".join([t.data for t in audio.getall('PRIV')
 277                         if t.owner=='wolnelektury.pl?funded_by'])
 278             except:
 279                 pass
 280         elif self.type == 'ogg':
 281             try:
 282                 audio = mutagen.File(self.file.path)
 283                 artist_name = ', '.join(audio.get('artist', []))
 284                 director_name = ', '.join(audio.get('conductor', []))
 285                 project = ", ".join(audio.get('project', []))
 286                 funded_by = ", ".join(audio.get('funded_by', []))
 287             except:
 288                 pass
 289         else:
 290             return {}
 291         return {'artist_name': artist_name, 'director_name': director_name,
 292                 'project': project, 'funded_by': funded_by}
 293
 294     @staticmethod
 295     def read_source_sha1(filepath, filetype):
 296         """
 297             Reads source file SHA1 from audiobok metadata.
 298         """
 299         import mutagen
 300         from mutagen import id3
 301
 302         if filetype == 'mp3':
 303             try:
 304                 audio = id3.ID3(filepath)
 305                 return [t.data for t in audio.getall('PRIV')
 306                         if t.owner=='wolnelektury.pl?flac_sha1'][0]
 307             except:
 308                 return None
 309         elif filetype == 'ogg':
 310             try:
 311                 audio = mutagen.File(filepath)
 312                 return audio.get('flac_sha1', [None])[0]
 313             except:
 314                 return None
 315         else:
 316             return None
 317
 318
 319 class Book(models.Model):
 320     title         = models.CharField(_('title'), max_length=120)
 321     sort_key = models.CharField(_('sort key'), max_length=120, db_index=True, editable=False)
 322     slug          = models.SlugField(_('slug'), max_length=120, db_index=True)
 323     language = models.CharField(_('language code'), max_length=3, db_index=True,
 324                     default=settings.CATALOGUE_DEFAULT_LANGUAGE)
 325     description   = models.TextField(_('description'), blank=True)
 326     created_at    = models.DateTimeField(_('creation date'), auto_now_add=True, db_index=True)
 327     changed_at    = models.DateTimeField(_('creation date'), auto_now=True, db_index=True)
 328     parent_number = models.IntegerField(_('parent number'), default=0)
 329     extra_info    = JSONField(_('extra information'), default='{}')
 330     gazeta_link   = models.CharField(blank=True, max_length=240)
 331     wiki_link     = models.CharField(blank=True, max_length=240)
 332     # files generated during publication
 333
 334     file_types = ['epub', 'html', 'mobi', 'pdf', 'txt', 'xml']
 335
 336     parent        = models.ForeignKey('self', blank=True, null=True, related_name='children')
 337     objects  = models.Manager()
 338     tagged   = managers.ModelTaggedItemManager(Tag)
 339     tags     = managers.TagDescriptor(Tag)
 340
 341     html_built = django.dispatch.Signal()
 342     published = django.dispatch.Signal()
 343
 344     URLID_RE = r'[a-z0-9-]+(?:/[a-z]{3})?'
 345     FILEID_RE = r'[a-z0-9-]+(?:_[a-z]{3})?'
 346
 347     class AlreadyExists(Exception):
 348         pass
 349
 350     class Meta:
 351         unique_together = [['slug', 'language']]
 352         ordering = ('sort_key',)
 353         verbose_name = _('book')
 354         verbose_name_plural = _('books')
 355
 356     def __unicode__(self):
 357         return self.title
 358
 359     def urlid(self, sep='/'):
 360         stem = self.slug
 361         if self.language != settings.CATALOGUE_DEFAULT_LANGUAGE:
 362             stem += sep + self.language
 363         return stem
 364
 365     def fileid(self):
 366         return self.urlid('_')
 367
 368     @staticmethod
 369     def split_urlid(urlid, sep='/', default_lang=settings.CATALOGUE_DEFAULT_LANGUAGE):
 370         """Splits a URL book id into slug and language code.
 371
 372         Returns a dictionary usable i.e. for object lookup, or None.
 373
 374         >>> Book.split_urlid("a-slug/pol", default_lang="eng")
 375         {'slug': 'a-slug', 'language': 'pol'}
 376         >>> Book.split_urlid("a-slug", default_lang="eng")
 377         {'slug': 'a-slug', 'language': 'eng'}
 378         >>> Book.split_urlid("a-slug_pol", "_", default_lang="eng")
 379         {'slug': 'a-slug', 'language': 'pol'}
 380         >>> Book.split_urlid("a-slug/eng", default_lang="eng")
 381
 382         """
 383         parts = urlid.rsplit(sep, 1)
 384         if len(parts) == 2:
 385             if parts[1] == default_lang:
 386                 return None
 387             return {'slug': parts[0], 'language': parts[1]}
 388         else:
 389             return {'slug': urlid, 'language': default_lang}
 390
 391     @classmethod
 392     def split_fileid(cls, fileid):
 393         return cls.split_urlid(fileid, '_')
 394
 395     def save(self, force_insert=False, force_update=False, reset_short_html=True, **kwargs):
 396         from sortify import sortify
 397
 398         self.sort_key = sortify(self.title)
 399
 400         ret = super(Book, self).save(force_insert, force_update)
 401
 402         if reset_short_html:
 403             self.reset_short_html()
 404
 405         return ret
 406
 407     @permalink
 408     def get_absolute_url(self):
 409         return ('catalogue.views.book_detail', [self.urlid()])
 410
 411     @property
 412     def name(self):
 413         return self.title
 414
 415     def book_tag_slug(self):
 416         stem = 'l-' + self.slug
 417         if self.language != settings.CATALOGUE_DEFAULT_LANGUAGE:
 418             return stem[:116] + ' ' + self.language
 419         else:
 420             return stem[:120]
 421
 422     def book_tag(self):
 423         slug = self.book_tag_slug()
 424         book_tag, created = Tag.objects.get_or_create(slug=slug, category='book')
 425         if created:
 426             book_tag.name = self.title[:50]
 427             book_tag.sort_key = self.title.lower()
 428             book_tag.save()
 429         return book_tag
 430
 431     def has_media(self, type):
 432         if type in Book.file_types:
 433             return bool(getattr(self, "%s_file" % type))
 434         else:
 435             return self.media.filter(type=type).exists()
 436
 437     def get_media(self, type):
 438         if self.has_media(type):
 439             if type in Book.file_types:
 440                 return getattr(self, "%s_file" % type)
 441             else:
 442                 return self.media.filter(type=type)
 443         else:
 444             return None
 445
 446     def get_mp3(self):
 447         return self.get_media("mp3")
 448     def get_odt(self):
 449         return self.get_media("odt")
 450     def get_ogg(self):
 451         return self.get_media("ogg")
 452     def get_daisy(self):
 453         return self.get_media("daisy")
 454
 455     def reset_short_html(self):
 456         if self.id is None:
 457             return
 458
 459         cache_key = "Book.short_html/%d/%s"
 460         for lang, langname in settings.LANGUAGES:
 461             cache.delete(cache_key % (self.id, lang))
 462         # Fragment.short_html relies on book's tags, so reset it here too
 463         for fragm in self.fragments.all():
 464             fragm.reset_short_html()
 465
 466     def short_html(self):
 467         if self.id:
 468             cache_key = "Book.short_html/%d/%s" % (self.id, get_language())
 469             short_html = cache.get(cache_key)
 470         else:
 471             short_html = None
 472
 473         if short_html is not None:
 474             return mark_safe(short_html)
 475         else:
 476             tags = self.tags.filter(~Q(category__in=('set', 'theme', 'book')))
 477             tags = [mark_safe(u'<a href="%s">%s</a>' % (tag.get_absolute_url(), tag.name)) for tag in tags]
 478
 479             formats = []
 480             # files generated during publication
 481             if self.has_media("html"):
 482                 formats.append(u'<a href="%s">%s</a>' % (reverse('book_text', [self.fileid()]), _('Read online')))
 483             if self.has_media("pdf"):
 484                 formats.append(u'<a href="%s">PDF</a>' % self.get_media('pdf').url)
 485             if self.has_media("mobi"):
 486                 formats.append(u'<a href="%s">MOBI</a>' % self.get_media('mobi').url)
 487             if self.root_ancestor.has_media("epub"):
 488                 formats.append(u'<a href="%s">EPUB</a>' % self.root_ancestor.get_media('epub').url)
 489             if self.has_media("txt"):
 490                 formats.append(u'<a href="%s">TXT</a>' % self.get_media('txt').url)
 491             # other files
 492             for m in self.media.order_by('type'):
 493                 formats.append(u'<a href="%s">%s</a>' % (m.file.url, m.type.upper()))
 494
 495             formats = [mark_safe(format) for format in formats]
 496
 497             short_html = unicode(render_to_string('catalogue/book_short.html',
 498                 {'book': self, 'tags': tags, 'formats': formats}))
 499
 500             if self.id:
 501                 cache.set(cache_key, short_html, CACHE_FOREVER)
 502             return mark_safe(short_html)
 503
 504     @property
 505     def root_ancestor(self):
 506         """ returns the oldest ancestor """
 507
 508         if not hasattr(self, '_root_ancestor'):
 509             book = self
 510             while book.parent:
 511                 book = book.parent
 512             self._root_ancestor = book
 513         return self._root_ancestor
 514
 515
 516     def has_description(self):
 517         return len(self.description) > 0
 518     has_description.short_description = _('description')
 519     has_description.boolean = True
 520
 521     # ugly ugly ugly
 522     def has_odt_file(self):
 523         return bool(self.has_media("odt"))
 524     has_odt_file.short_description = 'ODT'
 525     has_odt_file.boolean = True
 526
 527     def has_mp3_file(self):
 528         return bool(self.has_media("mp3"))
 529     has_mp3_file.short_description = 'MP3'
 530     has_mp3_file.boolean = True
 531
 532     def has_ogg_file(self):
 533         return bool(self.has_media("ogg"))
 534     has_ogg_file.short_description = 'OGG'
 535     has_ogg_file.boolean = True
 536
 537     def has_daisy_file(self):
 538         return bool(self.has_media("daisy"))
 539     has_daisy_file.short_description = 'DAISY'
 540     has_daisy_file.boolean = True
 541
 542     def wldocument(self, parse_dublincore=True):
 543         from catalogue.utils import ORMDocProvider
 544         from librarian.parser import WLDocument
 545
 546         return WLDocument.from_file(self.xml_file.path,
 547                 provider=ORMDocProvider(self),
 548                 parse_dublincore=parse_dublincore)
 549
 550     def build_pdf(self, customizations=None, file_name=None):
 551         """ (Re)builds the pdf file.
 552         customizations - customizations which are passed to LaTeX class file.
 553         file_name - save the pdf file under a different name and DO NOT save it in db.
 554         """
 555         from os import unlink
 556         from django.core.files import File
 557         from catalogue.utils import remove_zip
 558
 559         pdf = self.wldocument().as_pdf(customizations=customizations)
 560
 561         if file_name is None:
 562             # we'd like to be sure not to overwrite changes happening while
 563             # (timely) pdf generation is taking place (async celery scenario)
 564             current_self = Book.objects.get(id=self.id)
 565             current_self.pdf_file.save('%s.pdf' % self.fileid(),
 566                     File(open(pdf.get_filename())))
 567             self.pdf_file = current_self.pdf_file
 568         else:
 569             print "saving %s" % file_name
 570             print "to: %s" % DefaultStorage().path(file_name)
 571             DefaultStorage().save(file_name, File(open(pdf.get_filename())))
 572
 573         # remove cached downloadables
 574         remove_zip(settings.ALL_PDF_ZIP)
 575         for customized_pdf in get_existing_customized_pdf(self):
 576             unlink(customized_pdf)
 577
 578     def build_mobi(self):
 579         """ (Re)builds the MOBI file.
 580
 581         """
 582         from django.core.files import File
 583         from catalogue.utils import remove_zip
 584
 585         mobi = self.wldocument().as_mobi()
 586
 587         self.mobi_file.save('%s.mobi' % self.fileid(), File(open(mobi.get_filename())))
 588
 589         # remove zip with all mobi files
 590         remove_zip(settings.ALL_MOBI_ZIP)
 591
 592     def build_epub(self, remove_descendants=True):
 593         """ (Re)builds the epub file.
 594             If book has a parent, does nothing.
 595             Unless remove_descendants is False, descendants' epubs are removed.
 596         """
 597         from django.core.files import File
 598         from catalogue.utils import remove_zip
 599
 600         if self.parent:
 601             # don't need an epub
 602             return
 603
 604         epub = self.wldocument().as_epub()
 605
 606         try:
 607             epub.transform(ORMDocProvider(self), self.fileid(), output_file=epub_file)
 608             self.epub_file.save('%s.epub' % self.fileid(), File(open(epub.get_filename())))
 609         except NoDublinCore:
 610             pass
 611
 612         book_descendants = list(self.children.all())
 613         while len(book_descendants) > 0:
 614             child_book = book_descendants.pop(0)
 615             if remove_descendants and child_book.has_epub_file():
 616                 child_book.epub_file.delete()
 617             # save anyway, to refresh short_html
 618             child_book.save()
 619             book_descendants += list(child_book.children.all())
 620
 621         # remove zip package with all epub files
 622         remove_zip(settings.ALL_EPUB_ZIP)
 623
 624     def build_txt(self):
 625         from django.core.files.base import ContentFile
 626
 627         text = self.wldocument().as_text()
 628         self.txt_file.save('%s.txt' % self.fileid(), ContentFile(text.get_string()))
 629
 630
 631     def build_html(self):
 632         from markupstring import MarkupString
 633         from django.core.files.base import ContentFile
 634         from slughifi import slughifi
 635         from librarian import html
 636
 637         meta_tags = list(self.tags.filter(
 638             category__in=('author', 'epoch', 'genre', 'kind')))
 639         book_tag = self.book_tag()
 640
 641         html_output = self.wldocument(parse_dublincore=False).as_html()
 642         if html_output:
 643             self.html_file.save('%s.html' % self.fileid(),
 644                     ContentFile(html_output.get_string()))
 645
 646             # get ancestor l-tags for adding to new fragments
 647             ancestor_tags = []
 648             p = self.parent
 649             while p:
 650                 ancestor_tags.append(p.book_tag())
 651                 p = p.parent
 652
 653             # Delete old fragments and create them from scratch
 654             self.fragments.all().delete()
 655             # Extract fragments
 656             closed_fragments, open_fragments = html.extract_fragments(self.html_file.path)
 657             for fragment in closed_fragments.values():
 658                 try:
 659                     theme_names = [s.strip() for s in fragment.themes.split(',')]
 660                 except AttributeError:
 661                     continue
 662                 themes = []
 663                 for theme_name in theme_names:
 664                     if not theme_name:
 665                         continue
 666                     tag, created = Tag.objects.get_or_create(slug=slughifi(theme_name), category='theme')
 667                     if created:
 668                         tag.name = theme_name
 669                         tag.sort_key = theme_name.lower()
 670                         tag.save()
 671                     themes.append(tag)
 672                 if not themes:
 673                     continue
 674
 675                 text = fragment.to_string()
 676                 short_text = ''
 677                 if (len(MarkupString(text)) > 240):
 678                     short_text = unicode(MarkupString(text)[:160])
 679                 new_fragment = Fragment.objects.create(anchor=fragment.id, book=self,
 680                     text=text, short_text=short_text)
 681
 682                 new_fragment.save()
 683                 new_fragment.tags = set(meta_tags + themes + [book_tag] + ancestor_tags)
 684             self.save()
 685             self.html_built.send(sender=self)
 686             return True
 687         return False
 688
 689     @staticmethod
 690     def zip_format(format_):
 691         def pretty_file_name(book):
 692             return "%s/%s.%s" % (
 693                 b.get_extra_info_value()['author'],
 694                 b.fileid(),
 695                 format_)
 696
 697         field_name = "%s_file" % format_
 698         books = Book.objects.filter(parent=None).exclude(**{field_name: ""})
 699         paths = [(pretty_file_name(b), getattr(b, field_name).path)
 700                     for b in books]
 701         result = create_zip.delay(paths,
 702                     getattr(settings, "ALL_%s_ZIP" % format_.upper()))
 703         return result.wait()
 704
 705     def zip_audiobooks(self):
 706         bm = BookMedia.objects.filter(book=self, type='mp3')
 707         paths = map(lambda bm: (None, bm.file.path), bm)
 708         result = create_zip.delay(paths, self.fileid())
 709         return result.wait()
 710
 711     @classmethod
 712     def from_xml_file(cls, xml_file, **kwargs):
 713         from django.core.files import File
 714         from librarian import dcparser
 715
 716         # use librarian to parse meta-data
 717         book_info = dcparser.parse(xml_file)
 718
 719         if not isinstance(xml_file, File):
 720             xml_file = File(open(xml_file))
 721
 722         try:
 723             return cls.from_text_and_meta(xml_file, book_info, **kwargs)
 724         finally:
 725             xml_file.close()
 726
 727     @classmethod
 728     def from_text_and_meta(cls, raw_file, book_info, overwrite=False,
 729             build_epub=True, build_txt=True, build_pdf=True, build_mobi=True):
 730         import re
 731         from slughifi import slughifi
 732         from sortify import sortify
 733
 734         # check for parts before we do anything
 735         children = []
 736         if hasattr(book_info, 'parts'):
 737             for part_url in book_info.parts:
 738                 try:
 739                     children.append(Book.objects.get(
 740                         slug=part_url.slug, language=part_url.language))
 741                 except Book.DoesNotExist, e:
 742                     raise Book.DoesNotExist(_('Book "%s/%s" does not exist.') %
 743                             (part_url.slug, part_url.language))
 744
 745
 746         # Read book metadata
 747         book_slug = book_info.url.slug
 748         language = book_info.language
 749         if re.search(r'[^a-zA-Z0-9-]', book_slug):
 750             raise ValueError('Invalid characters in slug')
 751         book, created = Book.objects.get_or_create(slug=book_slug, language=language)
 752
 753         if created:
 754             book_shelves = []
 755         else:
 756             if not overwrite:
 757                 raise Book.AlreadyExists(_('Book %s/%s already exists') % (
 758                         book_slug, language))
 759             # Save shelves for this book
 760             book_shelves = list(book.tags.filter(category='set'))
 761
 762         book.title = book_info.title
 763         book.set_extra_info_value(book_info.to_dict())
 764         book.save()
 765
 766         meta_tags = []
 767         categories = (('kinds', 'kind'), ('genres', 'genre'), ('authors', 'author'), ('epochs', 'epoch'))
 768         for field_name, category in categories:
 769             try:
 770                 tag_names = getattr(book_info, field_name)
 771             except:
 772                 tag_names = [getattr(book_info, category)]
 773             for tag_name in tag_names:
 774                 tag_sort_key = tag_name
 775                 if category == 'author':
 776                     tag_sort_key = tag_name.last_name
 777                     tag_name = ' '.join(tag_name.first_names) + ' ' + tag_name.last_name
 778                 tag, created = Tag.objects.get_or_create(slug=slughifi(tag_name), category=category)
 779                 if created:
 780                     tag.name = tag_name
 781                     tag.sort_key = sortify(tag_sort_key.lower())
 782                     tag.save()
 783                 meta_tags.append(tag)
 784
 785         book.tags = set(meta_tags + book_shelves)
 786
 787         book_tag = book.book_tag()
 788
 789         for n, child_book in enumerate(children):
 790             child_book.parent = book
 791             child_book.parent_number = n
 792             child_book.save()
 793
 794         # Save XML and HTML files
 795         book.xml_file.save('%s.xml' % book.slug, raw_file, save=False)
 796
 797         # delete old fragments when overwriting
 798         book.fragments.all().delete()
 799
 800         if book.build_html():
 801             if not settings.NO_BUILD_TXT and build_txt:
 802                 book.build_txt()
 803
 804         if not settings.NO_BUILD_EPUB and build_epub:
 805             book.root_ancestor.build_epub()
 806
 807         if not settings.NO_BUILD_PDF and build_pdf:
 808             book.root_ancestor.build_pdf()
 809
 810         if not settings.NO_BUILD_MOBI and build_mobi:
 811             book.build_mobi()
 812
 813         book_descendants = list(book.children.all())
 814         # add l-tag to descendants and their fragments
 815         # delete unnecessary EPUB files
 816         while len(book_descendants) > 0:
 817             child_book = book_descendants.pop(0)
 818             child_book.tags = list(child_book.tags) + [book_tag]
 819             child_book.save()
 820             for fragment in child_book.fragments.all():
 821                 fragment.tags = set(list(fragment.tags) + [book_tag])
 822             book_descendants += list(child_book.children.all())
 823
 824         book.save()
 825
 826         # refresh cache
 827         book.reset_tag_counter()
 828         book.reset_theme_counter()
 829
 830         cls.published.send(sender=book)
 831         return book
 832
 833     def reset_tag_counter(self):
 834         if self.id is None:
 835             return
 836
 837         cache_key = "Book.tag_counter/%d" % self.id
 838         cache.delete(cache_key)
 839         if self.parent:
 840             self.parent.reset_tag_counter()
 841
 842     @property
 843     def tag_counter(self):
 844         if self.id:
 845             cache_key = "Book.tag_counter/%d" % self.id
 846             tags = cache.get(cache_key)
 847         else:
 848             tags = None
 849
 850         if tags is None:
 851             tags = {}
 852             for child in self.children.all().order_by():
 853                 for tag_pk, value in child.tag_counter.iteritems():
 854                     tags[tag_pk] = tags.get(tag_pk, 0) + value
 855             for tag in self.tags.exclude(category__in=('book', 'theme', 'set')).order_by():
 856                 tags[tag.pk] = 1
 857
 858             if self.id:
 859                 cache.set(cache_key, tags, CACHE_FOREVER)
 860         return tags
 861
 862     def reset_theme_counter(self):
 863         if self.id is None:
 864             return
 865
 866         cache_key = "Book.theme_counter/%d" % self.id
 867         cache.delete(cache_key)
 868         if self.parent:
 869             self.parent.reset_theme_counter()
 870
 871     @property
 872     def theme_counter(self):
 873         if self.id:
 874             cache_key = "Book.theme_counter/%d" % self.id
 875             tags = cache.get(cache_key)
 876         else:
 877             tags = None
 878
 879         if tags is None:
 880             tags = {}
 881             for fragment in Fragment.tagged.with_any([self.book_tag()]).order_by():
 882                 for tag in fragment.tags.filter(category='theme').order_by():
 883                     tags[tag.pk] = tags.get(tag.pk, 0) + 1
 884
 885             if self.id:
 886                 cache.set(cache_key, tags, CACHE_FOREVER)
 887         return tags
 888
 889     def pretty_title(self, html_links=False):
 890         book = self
 891         names = list(book.tags.filter(category='author'))
 892
 893         books = []
 894         while book:
 895             books.append(book)
 896             book = book.parent
 897         names.extend(reversed(books))
 898
 899         if html_links:
 900             names = ['<a href="%s">%s</a>' % (tag.get_absolute_url(), tag.name) for tag in names]
 901         else:
 902             names = [tag.name for tag in names]
 903
 904         return ', '.join(names)
 905
 906     @classmethod
 907     def tagged_top_level(cls, tags):
 908         """ Returns top-level books tagged with `tags'.
 909
 910         It only returns those books which don't have ancestors which are
 911         also tagged with those tags.
 912
 913         """
 914         # get relevant books and their tags
 915         objects = cls.tagged.with_all(tags)
 916         # eliminate descendants
 917         l_tags = Tag.objects.filter(category='book', slug__in=[book.book_tag_slug() for book in objects])
 918         descendants_keys = [book.pk for book in cls.tagged.with_any(l_tags)]
 919         if descendants_keys:
 920             objects = objects.exclude(pk__in=descendants_keys)
 921
 922         return objects
 923
 924     @classmethod
 925     def book_list(cls, filter=None):
 926         """Generates a hierarchical listing of all books.
 927
 928         Books are optionally filtered with a test function.
 929
 930         """
 931
 932         books_by_parent = {}
 933         books = cls.objects.all().order_by('parent_number', 'sort_key').only(
 934                 'title', 'parent', 'slug', 'language')
 935         if filter:
 936             books = books.filter(filter).distinct()
 937             book_ids = set((book.pk for book in books))
 938             for book in books:
 939                 parent = book.parent_id
 940                 if parent not in book_ids:
 941                     parent = None
 942                 books_by_parent.setdefault(parent, []).append(book)
 943         else:
 944             for book in books:
 945                 books_by_parent.setdefault(book.parent_id, []).append(book)
 946
 947         orphans = []
 948         books_by_author = SortedDict()
 949         for tag in Tag.objects.filter(category='author'):
 950             books_by_author[tag] = []
 951
 952         for book in books_by_parent.get(None,()):
 953             authors = list(book.tags.filter(category='author'))
 954             if authors:
 955                 for author in authors:
 956                     books_by_author[author].append(book)
 957             else:
 958                 orphans.append(book)
 959
 960         return books_by_author, orphans, books_by_parent
 961
 962     _audiences_pl = {
 963         "SP1": (1, u"szkoła podstawowa"),
 964         "SP2": (1, u"szkoła podstawowa"),
 965         "P": (1, u"szkoła podstawowa"),
 966         "G": (2, u"gimnazjum"),
 967         "L": (3, u"liceum"),
 968         "LP": (3, u"liceum"),
 969     }
 970     def audiences_pl(self):
 971         audiences = self.get_extra_info_value().get('audiences', [])
 972         audiences = sorted(set([self._audiences_pl[a] for a in audiences]))
 973         return [a[1] for a in audiences]
 974
 975
 976 def _has_factory(ftype):
 977     has = lambda self: bool(getattr(self, "%s_file" % ftype))
 978     has.short_description = t.upper()
 979     has.boolean = True
 980     has.__name__ = "has_%s_file" % ftype
 981     return has
 982
 983
 984 # add the file fields
 985 for t in Book.file_types:
 986     field_name = "%s_file" % t
 987     models.FileField(_("%s file" % t.upper()),
 988             upload_to=book_upload_path(t),
 989             blank=True).contribute_to_class(Book, field_name)
 990
 991     setattr(Book, "has_%s_file" % t, _has_factory(t))
 992
 993
 994 class Fragment(models.Model):
 995     text = models.TextField()
 996     short_text = models.TextField(editable=False)
 997     anchor = models.CharField(max_length=120)
 998     book = models.ForeignKey(Book, related_name='fragments')
 999
1000     objects = models.Manager()
1001     tagged = managers.ModelTaggedItemManager(Tag)
1002     tags = managers.TagDescriptor(Tag)
1003
1004     class Meta:
1005         ordering = ('book', 'anchor',)
1006         verbose_name = _('fragment')
1007         verbose_name_plural = _('fragments')
1008
1009     def get_absolute_url(self):
1010         return '%s#m%s' % (self.book.get_html_url(), self.anchor)
1011
1012     def reset_short_html(self):
1013         if self.id is None:
1014             return
1015
1016         cache_key = "Fragment.short_html/%d/%s"
1017         for lang, langname in settings.LANGUAGES:
1018             cache.delete(cache_key % (self.id, lang))
1019
1020     def short_html(self):
1021         if self.id:
1022             cache_key = "Fragment.short_html/%d/%s" % (self.id, get_language())
1023             short_html = cache.get(cache_key)
1024         else:
1025             short_html = None
1026
1027         if short_html is not None:
1028             return mark_safe(short_html)
1029         else:
1030             short_html = unicode(render_to_string('catalogue/fragment_short.html',
1031                 {'fragment': self}))
1032             if self.id:
1033                 cache.set(cache_key, short_html, CACHE_FOREVER)
1034             return mark_safe(short_html)
1035
1036
1037 ###########
1038 #
1039 # SIGNALS
1040 #
1041 ###########
1042
1043
1044 def _tags_updated_handler(sender, affected_tags, **kwargs):
1045     # reset tag global counter
1046     # we want Tag.changed_at updated for API to know the tag was touched
1047     Tag.objects.filter(pk__in=[tag.pk for tag in affected_tags]).update(book_count=None, changed_at=datetime.now())
1048
1049     # if book tags changed, reset book tag counter
1050     if isinstance(sender, Book) and \
1051                 Tag.objects.filter(pk__in=(tag.pk for tag in affected_tags)).\
1052                     exclude(category__in=('book', 'theme', 'set')).count():
1053         sender.reset_tag_counter()
1054     # if fragment theme changed, reset book theme counter
1055     elif isinstance(sender, Fragment) and \
1056                 Tag.objects.filter(pk__in=(tag.pk for tag in affected_tags)).\
1057                     filter(category='theme').count():
1058         sender.book.reset_theme_counter()
1059 tags_updated.connect(_tags_updated_handler)
1060
1061
1062 def _pre_delete_handler(sender, instance, **kwargs):
1063     """ refresh Book on BookMedia delete """
1064     if sender == BookMedia:
1065         instance.book.save()
1066 pre_delete.connect(_pre_delete_handler)
1067
1068 def _post_save_handler(sender, instance, **kwargs):
1069     """ refresh all the short_html stuff on BookMedia update """
1070     if sender == BookMedia:
1071         instance.book.save()
1072 post_save.connect(_post_save_handler)