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