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     @staticmethod
 178     def tags_from_info(info):
 179         from slughifi import slughifi
 180         from sortify import sortify
 181         meta_tags = []
 182         categories = (('kinds', 'kind'), ('genres', 'genre'), ('authors', 'author'), ('epochs', 'epoch'))
 183         for field_name, category in categories:
 184             try:
 185                 tag_names = getattr(info, field_name)
 186             except:
 187                 tag_names = [getattr(info, category)]
 188             for tag_name in tag_names:
 189                 tag_sort_key = tag_name
 190                 if category == 'author':
 191                     tag_sort_key = tag_name.last_name
 192                     tag_name = ' '.join(tag_name.first_names) + ' ' + tag_name.last_name
 193                 tag, created = Tag.objects.get_or_create(slug=slughifi(tag_name), category=category)
 194                 if created:
 195                     tag.name = tag_name
 196                     tag.sort_key = sortify(tag_sort_key.lower())
 197                     tag.save()
 198                 meta_tags.append(tag)
 199         return meta_tags
 200
 201
 202
 203 def get_dynamic_path(media, filename, ext=None, maxlen=100):
 204     from slughifi import slughifi
 205
 206     # how to put related book's slug here?
 207     if not ext:
 208         if media.type == 'daisy':
 209             ext = 'daisy.zip'
 210         else:
 211             ext = media.type
 212     if media is None or not media.name:
 213         name = slughifi(filename.split(".")[0])
 214     else:
 215         name = slughifi(media.name)
 216     return 'book/%s/%s.%s' % (ext, name[:maxlen-len('book/%s/.%s' % (ext, ext))-4], ext)
 217
 218
 219 # TODO: why is this hard-coded ?
 220 def book_upload_path(ext=None, maxlen=100):
 221     return lambda *args: get_dynamic_path(*args, ext=ext, maxlen=maxlen)
 222
 223
 224 def get_customized_pdf_path(book, customizations):
 225     """
 226     Returns a MEDIA_ROOT relative path for a customized pdf. The name will contain a hash of customization options.
 227     """
 228     customizations.sort()
 229     h = hash(tuple(customizations))
 230
 231     pdf_name = '%s-custom-%s' % (book.fileid(), h)
 232     pdf_file = get_dynamic_path(None, pdf_name, ext='pdf')
 233
 234     return pdf_file
 235
 236
 237 def get_existing_customized_pdf(book):
 238     """
 239     Returns a list of paths to generated customized pdf of a book
 240     """
 241     pdf_glob = '%s-custom-' % (book.fileid(),)
 242     pdf_glob = get_dynamic_path(None, pdf_glob, ext='pdf')
 243     pdf_glob = re.sub(r"[.]([a-z0-9]+)$", "*.\\1", pdf_glob)
 244     return glob(path.join(settings.MEDIA_ROOT, pdf_glob))
 245
 246
 247 class BookMedia(models.Model):
 248     type        = models.CharField(_('type'), choices=MEDIA_FORMATS, max_length="100")
 249     name        = models.CharField(_('name'), max_length="100")
 250     file        = OverwritingFileField(_('file'), upload_to=book_upload_path())
 251     uploaded_at = models.DateTimeField(_('creation date'), auto_now_add=True, editable=False)
 252     extra_info  = JSONField(_('extra information'), default='{}', editable=False)
 253     book = models.ForeignKey('Book', related_name='media')
 254     source_sha1 = models.CharField(null=True, blank=True, max_length=40, editable=False)
 255
 256     def __unicode__(self):
 257         return "%s (%s)" % (self.name, self.file.name.split("/")[-1])
 258
 259     class Meta:
 260         ordering            = ('type', 'name')
 261         verbose_name        = _('book media')
 262         verbose_name_plural = _('book media')
 263
 264     def save(self, *args, **kwargs):
 265         from slughifi import slughifi
 266         from catalogue.utils import ExistingFile, remove_zip
 267
 268         try:
 269             old = BookMedia.objects.get(pk=self.pk)
 270         except BookMedia.DoesNotExist, e:
 271             pass
 272         else:
 273             # if name changed, change the file name, too
 274             if slughifi(self.name) != slughifi(old.name):
 275                 self.file.save(None, ExistingFile(self.file.path), save=False, leave=True)
 276
 277         super(BookMedia, self).save(*args, **kwargs)
 278
 279         # remove the zip package for book with modified media
 280         remove_zip(self.book.fileid())
 281
 282         extra_info = self.get_extra_info_value()
 283         extra_info.update(self.read_meta())
 284         self.set_extra_info_value(extra_info)
 285         self.source_sha1 = self.read_source_sha1(self.file.path, self.type)
 286         return super(BookMedia, self).save(*args, **kwargs)
 287
 288     def read_meta(self):
 289         """
 290             Reads some metadata from the audiobook.
 291         """
 292         import mutagen
 293         from mutagen import id3
 294
 295         artist_name = director_name = project = funded_by = ''
 296         if self.type == 'mp3':
 297             try:
 298                 audio = id3.ID3(self.file.path)
 299                 artist_name = ', '.join(', '.join(tag.text) for tag in audio.getall('TPE1'))
 300                 director_name = ', '.join(', '.join(tag.text) for tag in audio.getall('TPE3'))
 301                 project = ", ".join([t.data for t in audio.getall('PRIV')
 302                         if t.owner=='wolnelektury.pl?project'])
 303                 funded_by = ", ".join([t.data for t in audio.getall('PRIV')
 304                         if t.owner=='wolnelektury.pl?funded_by'])
 305             except:
 306                 pass
 307         elif self.type == 'ogg':
 308             try:
 309                 audio = mutagen.File(self.file.path)
 310                 artist_name = ', '.join(audio.get('artist', []))
 311                 director_name = ', '.join(audio.get('conductor', []))
 312                 project = ", ".join(audio.get('project', []))
 313                 funded_by = ", ".join(audio.get('funded_by', []))
 314             except:
 315                 pass
 316         else:
 317             return {}
 318         return {'artist_name': artist_name, 'director_name': director_name,
 319                 'project': project, 'funded_by': funded_by}
 320
 321     @staticmethod
 322     def read_source_sha1(filepath, filetype):
 323         """
 324             Reads source file SHA1 from audiobok metadata.
 325         """
 326         import mutagen
 327         from mutagen import id3
 328
 329         if filetype == 'mp3':
 330             try:
 331                 audio = id3.ID3(filepath)
 332                 return [t.data for t in audio.getall('PRIV')
 333                         if t.owner=='wolnelektury.pl?flac_sha1'][0]
 334             except:
 335                 return None
 336         elif filetype == 'ogg':
 337             try:
 338                 audio = mutagen.File(filepath)
 339                 return audio.get('flac_sha1', [None])[0]
 340             except:
 341                 return None
 342         else:
 343             return None
 344
 345
 346 class Book(models.Model):
 347     title         = models.CharField(_('title'), max_length=120)
 348     sort_key = models.CharField(_('sort key'), max_length=120, db_index=True, editable=False)
 349     slug          = models.SlugField(_('slug'), max_length=120, db_index=True)
 350     language = models.CharField(_('language code'), max_length=3, db_index=True,
 351                     default=settings.CATALOGUE_DEFAULT_LANGUAGE)
 352     description   = models.TextField(_('description'), blank=True)
 353     created_at    = models.DateTimeField(_('creation date'), auto_now_add=True, db_index=True)
 354     changed_at    = models.DateTimeField(_('creation date'), auto_now=True, db_index=True)
 355     parent_number = models.IntegerField(_('parent number'), default=0)
 356     extra_info    = JSONField(_('extra information'), default='{}')
 357     gazeta_link   = models.CharField(blank=True, max_length=240)
 358     wiki_link     = models.CharField(blank=True, max_length=240)
 359     # files generated during publication
 360
 361     file_types = ['epub', 'html', 'mobi', 'pdf', 'txt', 'xml']
 362
 363     parent        = models.ForeignKey('self', blank=True, null=True, related_name='children')
 364     objects  = models.Manager()
 365     tagged   = managers.ModelTaggedItemManager(Tag)
 366     tags     = managers.TagDescriptor(Tag)
 367
 368     html_built = django.dispatch.Signal()
 369     published = django.dispatch.Signal()
 370
 371     URLID_RE = r'[a-z0-9-]+(?:/[a-z]{3})?'
 372     FILEID_RE = r'[a-z0-9-]+(?:_[a-z]{3})?'
 373
 374     class AlreadyExists(Exception):
 375         pass
 376
 377     class Meta:
 378         unique_together = [['slug', 'language']]
 379         ordering = ('sort_key',)
 380         verbose_name = _('book')
 381         verbose_name_plural = _('books')
 382
 383     def __unicode__(self):
 384         return self.title
 385
 386     def urlid(self, sep='/'):
 387         stem = self.slug
 388         if self.language != settings.CATALOGUE_DEFAULT_LANGUAGE:
 389             stem += sep + self.language
 390         return stem
 391
 392     def fileid(self):
 393         return self.urlid('_')
 394
 395     @staticmethod
 396     def split_urlid(urlid, sep='/', default_lang=settings.CATALOGUE_DEFAULT_LANGUAGE):
 397         """Splits a URL book id into slug and language code.
 398
 399         Returns a dictionary usable i.e. for object lookup, or None.
 400
 401         >>> Book.split_urlid("a-slug/pol", default_lang="eng")
 402         {'slug': 'a-slug', 'language': 'pol'}
 403         >>> Book.split_urlid("a-slug", default_lang="eng")
 404         {'slug': 'a-slug', 'language': 'eng'}
 405         >>> Book.split_urlid("a-slug_pol", "_", default_lang="eng")
 406         {'slug': 'a-slug', 'language': 'pol'}
 407         >>> Book.split_urlid("a-slug/eng", default_lang="eng")
 408
 409         """
 410         parts = urlid.rsplit(sep, 1)
 411         if len(parts) == 2:
 412             if parts[1] == default_lang:
 413                 return None
 414             return {'slug': parts[0], 'language': parts[1]}
 415         else:
 416             return {'slug': urlid, 'language': default_lang}
 417
 418     @classmethod
 419     def split_fileid(cls, fileid):
 420         return cls.split_urlid(fileid, '_')
 421
 422     def save(self, force_insert=False, force_update=False, reset_short_html=True, **kwargs):
 423         from sortify import sortify
 424
 425         self.sort_key = sortify(self.title)
 426
 427         ret = super(Book, self).save(force_insert, force_update)
 428
 429         if reset_short_html:
 430             self.reset_short_html()
 431
 432         return ret
 433
 434     @permalink
 435     def get_absolute_url(self):
 436         return ('catalogue.views.book_detail', [self.urlid()])
 437
 438     @property
 439     def name(self):
 440         return self.title
 441
 442     def book_tag_slug(self):
 443         stem = 'l-' + self.slug
 444         if self.language != settings.CATALOGUE_DEFAULT_LANGUAGE:
 445             return stem[:116] + ' ' + self.language
 446         else:
 447             return stem[:120]
 448
 449     def book_tag(self):
 450         slug = self.book_tag_slug()
 451         book_tag, created = Tag.objects.get_or_create(slug=slug, category='book')
 452         if created:
 453             book_tag.name = self.title[:50]
 454             book_tag.sort_key = self.title.lower()
 455             book_tag.save()
 456         return book_tag
 457
 458     def has_media(self, type):
 459         if type in Book.file_types:
 460             return bool(getattr(self, "%s_file" % type))
 461         else:
 462             return self.media.filter(type=type).exists()
 463
 464     def get_media(self, type):
 465         if self.has_media(type):
 466             if type in Book.file_types:
 467                 return getattr(self, "%s_file" % type)
 468             else:
 469                 return self.media.filter(type=type)
 470         else:
 471             return None
 472
 473     def get_mp3(self):
 474         return self.get_media("mp3")
 475     def get_odt(self):
 476         return self.get_media("odt")
 477     def get_ogg(self):
 478         return self.get_media("ogg")
 479     def get_daisy(self):
 480         return self.get_media("daisy")
 481
 482     def reset_short_html(self):
 483         if self.id is None:
 484             return
 485
 486         cache_key = "Book.short_html/%d/%s"
 487         for lang, langname in settings.LANGUAGES:
 488             cache.delete(cache_key % (self.id, lang))
 489         # Fragment.short_html relies on book's tags, so reset it here too
 490         for fragm in self.fragments.all():
 491             fragm.reset_short_html()
 492
 493     def short_html(self):
 494         if self.id:
 495             cache_key = "Book.short_html/%d/%s" % (self.id, get_language())
 496             short_html = cache.get(cache_key)
 497         else:
 498             short_html = None
 499
 500         if short_html is not None:
 501             return mark_safe(short_html)
 502         else:
 503             tags = self.tags.filter(~Q(category__in=('set', 'theme', 'book')))
 504             tags = [mark_safe(u'<a href="%s">%s</a>' % (tag.get_absolute_url(), tag.name)) for tag in tags]
 505
 506             formats = []
 507             # files generated during publication
 508             if self.has_media("html"):
 509                 formats.append(u'<a href="%s">%s</a>' % (reverse('book_text', args=[self.fileid()]), _('Read online')))
 510             if self.has_media("pdf"):
 511                 formats.append(u'<a href="%s">PDF</a>' % self.get_media('pdf').url)
 512             if self.has_media("mobi"):
 513                 formats.append(u'<a href="%s">MOBI</a>' % self.get_media('mobi').url)
 514             if self.root_ancestor.has_media("epub"):
 515                 formats.append(u'<a href="%s">EPUB</a>' % self.root_ancestor.get_media('epub').url)
 516             if self.has_media("txt"):
 517                 formats.append(u'<a href="%s">TXT</a>' % self.get_media('txt').url)
 518             # other files
 519             for m in self.media.order_by('type'):
 520                 formats.append(u'<a href="%s">%s</a>' % (m.file.url, m.type.upper()))
 521
 522             formats = [mark_safe(format) for format in formats]
 523
 524             short_html = unicode(render_to_string('catalogue/book_short.html',
 525                 {'book': self, 'tags': tags, 'formats': formats}))
 526
 527             if self.id:
 528                 cache.set(cache_key, short_html, CACHE_FOREVER)
 529             return mark_safe(short_html)
 530
 531     @property
 532     def root_ancestor(self):
 533         """ returns the oldest ancestor """
 534
 535         if not hasattr(self, '_root_ancestor'):
 536             book = self
 537             while book.parent:
 538                 book = book.parent
 539             self._root_ancestor = book
 540         return self._root_ancestor
 541
 542
 543     def has_description(self):
 544         return len(self.description) > 0
 545     has_description.short_description = _('description')
 546     has_description.boolean = True
 547
 548     # ugly ugly ugly
 549     def has_odt_file(self):
 550         return bool(self.has_media("odt"))
 551     has_odt_file.short_description = 'ODT'
 552     has_odt_file.boolean = True
 553
 554     def has_mp3_file(self):
 555         return bool(self.has_media("mp3"))
 556     has_mp3_file.short_description = 'MP3'
 557     has_mp3_file.boolean = True
 558
 559     def has_ogg_file(self):
 560         return bool(self.has_media("ogg"))
 561     has_ogg_file.short_description = 'OGG'
 562     has_ogg_file.boolean = True
 563
 564     def has_daisy_file(self):
 565         return bool(self.has_media("daisy"))
 566     has_daisy_file.short_description = 'DAISY'
 567     has_daisy_file.boolean = True
 568
 569     def wldocument(self, parse_dublincore=True):
 570         from catalogue.utils import ORMDocProvider
 571         from librarian.parser import WLDocument
 572
 573         return WLDocument.from_file(self.xml_file.path,
 574                 provider=ORMDocProvider(self),
 575                 parse_dublincore=parse_dublincore)
 576
 577     def build_pdf(self, customizations=None, file_name=None):
 578         """ (Re)builds the pdf file.
 579         customizations - customizations which are passed to LaTeX class file.
 580         file_name - save the pdf file under a different name and DO NOT save it in db.
 581         """
 582         from os import unlink
 583         from django.core.files import File
 584         from catalogue.utils import remove_zip
 585
 586         pdf = self.wldocument().as_pdf(customizations=customizations)
 587
 588         if file_name is None:
 589             # we'd like to be sure not to overwrite changes happening while
 590             # (timely) pdf generation is taking place (async celery scenario)
 591             current_self = Book.objects.get(id=self.id)
 592             current_self.pdf_file.save('%s.pdf' % self.fileid(),
 593                     File(open(pdf.get_filename())))
 594             self.pdf_file = current_self.pdf_file
 595
 596             # remove cached downloadables
 597             remove_zip(settings.ALL_PDF_ZIP)
 598
 599             for customized_pdf in get_existing_customized_pdf(self):
 600                 unlink(customized_pdf)
 601         else:
 602             print "saving %s" % file_name
 603             print "to: %s" % DefaultStorage().path(file_name)
 604             DefaultStorage().save(file_name, File(open(pdf.get_filename())))
 605
 606     def build_mobi(self):
 607         """ (Re)builds the MOBI file.
 608
 609         """
 610         from django.core.files import File
 611         from catalogue.utils import remove_zip
 612
 613         mobi = self.wldocument().as_mobi()
 614
 615         self.mobi_file.save('%s.mobi' % self.fileid(), File(open(mobi.get_filename())))
 616
 617         # remove zip with all mobi files
 618         remove_zip(settings.ALL_MOBI_ZIP)
 619
 620     def build_epub(self, remove_descendants=True):
 621         """ (Re)builds the epub file.
 622             If book has a parent, does nothing.
 623             Unless remove_descendants is False, descendants' epubs are removed.
 624         """
 625         from django.core.files import File
 626         from catalogue.utils import remove_zip
 627
 628         if self.parent:
 629             # don't need an epub
 630             return
 631
 632         epub = self.wldocument().as_epub()
 633
 634         try:
 635             epub.transform(ORMDocProvider(self), self.fileid(), output_file=epub_file)
 636             self.epub_file.save('%s.epub' % self.fileid(), File(open(epub.get_filename())))
 637         except NoDublinCore:
 638             pass
 639
 640         book_descendants = list(self.children.all())
 641         while len(book_descendants) > 0:
 642             child_book = book_descendants.pop(0)
 643             if remove_descendants and child_book.has_epub_file():
 644                 child_book.epub_file.delete()
 645             # save anyway, to refresh short_html
 646             child_book.save()
 647             book_descendants += list(child_book.children.all())
 648
 649         # remove zip package with all epub files
 650         remove_zip(settings.ALL_EPUB_ZIP)
 651
 652     def build_txt(self):
 653         from django.core.files.base import ContentFile
 654
 655         text = self.wldocument().as_text()
 656         self.txt_file.save('%s.txt' % self.fileid(), ContentFile(text.get_string()))
 657
 658
 659     def build_html(self):
 660         from markupstring import MarkupString
 661         from django.core.files.base import ContentFile
 662         from slughifi import slughifi
 663         from librarian import html
 664
 665         meta_tags = list(self.tags.filter(
 666             category__in=('author', 'epoch', 'genre', 'kind')))
 667         book_tag = self.book_tag()
 668
 669         html_output = self.wldocument(parse_dublincore=False).as_html()
 670         if html_output:
 671             self.html_file.save('%s.html' % self.fileid(),
 672                     ContentFile(html_output.get_string()))
 673
 674             # get ancestor l-tags for adding to new fragments
 675             ancestor_tags = []
 676             p = self.parent
 677             while p:
 678                 ancestor_tags.append(p.book_tag())
 679                 p = p.parent
 680
 681             # Delete old fragments and create them from scratch
 682             self.fragments.all().delete()
 683             # Extract fragments
 684             closed_fragments, open_fragments = html.extract_fragments(self.html_file.path)
 685             for fragment in closed_fragments.values():
 686                 try:
 687                     theme_names = [s.strip() for s in fragment.themes.split(',')]
 688                 except AttributeError:
 689                     continue
 690                 themes = []
 691                 for theme_name in theme_names:
 692                     if not theme_name:
 693                         continue
 694                     tag, created = Tag.objects.get_or_create(slug=slughifi(theme_name), category='theme')
 695                     if created:
 696                         tag.name = theme_name
 697                         tag.sort_key = theme_name.lower()
 698                         tag.save()
 699                     themes.append(tag)
 700                 if not themes:
 701                     continue
 702
 703                 text = fragment.to_string()
 704                 short_text = ''
 705                 if (len(MarkupString(text)) > 240):
 706                     short_text = unicode(MarkupString(text)[:160])
 707                 new_fragment = Fragment.objects.create(anchor=fragment.id, book=self,
 708                     text=text, short_text=short_text)
 709
 710                 new_fragment.save()
 711                 new_fragment.tags = set(meta_tags + themes + [book_tag] + ancestor_tags)
 712             self.save()
 713             self.html_built.send(sender=self)
 714             return True
 715         return False
 716
 717     @staticmethod
 718     def zip_format(format_):
 719         def pretty_file_name(book):
 720             return "%s/%s.%s" % (
 721                 b.get_extra_info_value()['author'],
 722                 b.fileid(),
 723                 format_)
 724
 725         field_name = "%s_file" % format_
 726         books = Book.objects.filter(parent=None).exclude(**{field_name: ""})
 727         paths = [(pretty_file_name(b), getattr(b, field_name).path)
 728                     for b in books]
 729         result = create_zip.delay(paths,
 730                     getattr(settings, "ALL_%s_ZIP" % format_.upper()))
 731         return result.wait()
 732
 733     def zip_audiobooks(self):
 734         bm = BookMedia.objects.filter(book=self, type='mp3')
 735         paths = map(lambda bm: (None, bm.file.path), bm)
 736         result = create_zip.delay(paths, self.fileid())
 737         return result.wait()
 738
 739     @classmethod
 740     def from_xml_file(cls, xml_file, **kwargs):
 741         from django.core.files import File
 742         from librarian import dcparser
 743
 744         # use librarian to parse meta-data
 745         book_info = dcparser.parse(xml_file)
 746
 747         if not isinstance(xml_file, File):
 748             xml_file = File(open(xml_file))
 749
 750         try:
 751             return cls.from_text_and_meta(xml_file, book_info, **kwargs)
 752         finally:
 753             xml_file.close()
 754
 755     @classmethod
 756     def from_text_and_meta(cls, raw_file, book_info, overwrite=False,
 757             build_epub=True, build_txt=True, build_pdf=True, build_mobi=True):
 758         import re
 759         from sortify import sortify
 760
 761         # check for parts before we do anything
 762         children = []
 763         if hasattr(book_info, 'parts'):
 764             for part_url in book_info.parts:
 765                 try:
 766                     children.append(Book.objects.get(
 767                         slug=part_url.slug, language=part_url.language))
 768                 except Book.DoesNotExist, e:
 769                     raise Book.DoesNotExist(_('Book "%s/%s" does not exist.') %
 770                             (part_url.slug, part_url.language))
 771
 772
 773         # Read book metadata
 774         book_slug = book_info.url.slug
 775         language = book_info.language
 776         if re.search(r'[^a-zA-Z0-9-]', book_slug):
 777             raise ValueError('Invalid characters in slug')
 778         book, created = Book.objects.get_or_create(slug=book_slug, language=language)
 779
 780         if created:
 781             book_shelves = []
 782         else:
 783             if not overwrite:
 784                 raise Book.AlreadyExists(_('Book %s/%s already exists') % (
 785                         book_slug, language))
 786             # Save shelves for this book
 787             book_shelves = list(book.tags.filter(category='set'))
 788
 789         book.title = book_info.title
 790         book.set_extra_info_value(book_info.to_dict())
 791         book.save()
 792
 793         meta_tags = Tag.tags_from_info(book_info)
 794
 795         book.tags = set(meta_tags + book_shelves)
 796
 797         book_tag = book.book_tag()
 798
 799         for n, child_book in enumerate(children):
 800             child_book.parent = book
 801             child_book.parent_number = n
 802             child_book.save()
 803
 804         # Save XML and HTML files
 805         book.xml_file.save('%s.xml' % book.slug, raw_file, save=False)
 806
 807         # delete old fragments when overwriting
 808         book.fragments.all().delete()
 809
 810         if book.build_html():
 811             if not settings.NO_BUILD_TXT and build_txt:
 812                 book.build_txt()
 813
 814         if not settings.NO_BUILD_EPUB and build_epub:
 815             book.root_ancestor.build_epub()
 816
 817         if not settings.NO_BUILD_PDF and build_pdf:
 818             book.root_ancestor.build_pdf()
 819
 820         if not settings.NO_BUILD_MOBI and build_mobi:
 821             book.build_mobi()
 822
 823         book_descendants = list(book.children.all())
 824         # add l-tag to descendants and their fragments
 825         # delete unnecessary EPUB files
 826         while len(book_descendants) > 0:
 827             child_book = book_descendants.pop(0)
 828             child_book.tags = list(child_book.tags) + [book_tag]
 829             child_book.save()
 830             for fragment in child_book.fragments.all():
 831                 fragment.tags = set(list(fragment.tags) + [book_tag])
 832             book_descendants += list(child_book.children.all())
 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.file_types:
 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     Tag.objects.filter(pk__in=[tag.pk for tag in affected_tags]).update(book_count=None, changed_at=datetime.now())
1058
1059     # if book tags changed, reset book tag counter
1060     if isinstance(sender, Book) and \
1061                 Tag.objects.filter(pk__in=(tag.pk for tag in affected_tags)).\
1062                     exclude(category__in=('book', 'theme', 'set')).count():
1063         sender.reset_tag_counter()
1064     # if fragment theme changed, reset book theme counter
1065     elif isinstance(sender, Fragment) and \
1066                 Tag.objects.filter(pk__in=(tag.pk for tag in affected_tags)).\
1067                     filter(category='theme').count():
1068         sender.book.reset_theme_counter()
1069 tags_updated.connect(_tags_updated_handler)
1070
1071
1072 def _pre_delete_handler(sender, instance, **kwargs):
1073     """ refresh Book on BookMedia delete """
1074     if sender == BookMedia:
1075         instance.book.save()
1076 pre_delete.connect(_pre_delete_handler)
1077
1078 def _post_save_handler(sender, instance, **kwargs):
1079     """ refresh all the short_html stuff on BookMedia update """
1080     if sender == BookMedia:
1081         instance.book.save()
1082 post_save.connect(_post_save_handler)