apps/catalogue/models.py

   1 # -*- coding: utf-8 -*-
   2 # This file is part of Wolnelektury, licensed under GNU Affero GPLv3 or later.
   3 # Copyright © Fundacja Nowoczesna Polska. See NOTICE for more information.
   4 #
   5 from collections import namedtuple
   6
   7 from django.db import models
   8 from django.db.models import permalink, Q
   9 import django.dispatch
  10 from django.core.cache import get_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, post_delete
  20 import jsonfield
  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, split_tags, truncate_html_words
  28 from catalogue.tasks import touch_tag, index_book
  29 from shutil import copy
  30 from glob import glob
  31 import re
  32 from os import path
  33
  34 import search
  35
  36 # Those are hard-coded here so that makemessages sees them.
  37 TAG_CATEGORIES = (
  38     ('author', _('author')),
  39     ('epoch', _('epoch')),
  40     ('kind', _('kind')),
  41     ('genre', _('genre')),
  42     ('theme', _('theme')),
  43     ('set', _('set')),
  44     ('book', _('book')),
  45 )
  46
  47
  48 permanent_cache = get_cache('permanent')
  49
  50
  51 class TagSubcategoryManager(models.Manager):
  52     def __init__(self, subcategory):
  53         super(TagSubcategoryManager, self).__init__()
  54         self.subcategory = subcategory
  55
  56     def get_query_set(self):
  57         return super(TagSubcategoryManager, self).get_query_set().filter(category=self.subcategory)
  58
  59
  60 class Tag(TagBase):
  61     name = models.CharField(_('name'), max_length=50, db_index=True)
  62     slug = models.SlugField(_('slug'), max_length=120, db_index=True)
  63     sort_key = models.CharField(_('sort key'), max_length=120, db_index=True)
  64     category = models.CharField(_('category'), max_length=50, blank=False, null=False,
  65         db_index=True, choices=TAG_CATEGORIES)
  66     description = models.TextField(_('description'), blank=True)
  67
  68     user = models.ForeignKey(User, blank=True, null=True)
  69     book_count = models.IntegerField(_('book count'), blank=True, null=True)
  70     gazeta_link = models.CharField(blank=True, max_length=240)
  71     wiki_link = models.CharField(blank=True, max_length=240)
  72
  73     created_at    = models.DateTimeField(_('creation date'), auto_now_add=True, db_index=True)
  74     changed_at    = models.DateTimeField(_('creation date'), auto_now=True, db_index=True)
  75
  76     class UrlDeprecationWarning(DeprecationWarning):
  77         pass
  78
  79     categories_rev = {
  80         'autor': 'author',
  81         'epoka': 'epoch',
  82         'rodzaj': 'kind',
  83         'gatunek': 'genre',
  84         'motyw': 'theme',
  85         'polka': 'set',
  86     }
  87     categories_dict = dict((item[::-1] for item in categories_rev.iteritems()))
  88
  89     class Meta:
  90         ordering = ('sort_key',)
  91         verbose_name = _('tag')
  92         verbose_name_plural = _('tags')
  93         unique_together = (("slug", "category"),)
  94
  95     def __unicode__(self):
  96         return self.name
  97
  98     def __repr__(self):
  99         return "Tag(slug=%r)" % self.slug
 100
 101     @permalink
 102     def get_absolute_url(self):
 103         return ('catalogue.views.tagged_object_list', [self.url_chunk])
 104
 105     def has_description(self):
 106         return len(self.description) > 0
 107     has_description.short_description = _('description')
 108     has_description.boolean = True
 109
 110     def get_count(self):
 111         """Returns global book count for book tags, fragment count for themes."""
 112
 113         if self.category == 'book':
 114             # never used
 115             objects = Book.objects.none()
 116         elif self.category == 'theme':
 117             objects = Fragment.tagged.with_all((self,))
 118         else:
 119             objects = Book.tagged.with_all((self,)).order_by()
 120             if self.category != 'set':
 121                 # eliminate descendants
 122                 l_tags = Tag.objects.filter(slug__in=[book.book_tag_slug() for book in objects])
 123                 descendants_keys = [book.pk for book in Book.tagged.with_any(l_tags)]
 124                 if descendants_keys:
 125                     objects = objects.exclude(pk__in=descendants_keys)
 126         return objects.count()
 127
 128     @staticmethod
 129     def get_tag_list(tags):
 130         if isinstance(tags, basestring):
 131             real_tags = []
 132             ambiguous_slugs = []
 133             category = None
 134             deprecated = False
 135             tags_splitted = tags.split('/')
 136             for name in tags_splitted:
 137                 if category:
 138                     real_tags.append(Tag.objects.get(slug=name, category=category))
 139                     category = None
 140                 elif name in Tag.categories_rev:
 141                     category = Tag.categories_rev[name]
 142                 else:
 143                     try:
 144                         real_tags.append(Tag.objects.exclude(category='book').get(slug=name))
 145                         deprecated = True
 146                     except Tag.MultipleObjectsReturned, e:
 147                         ambiguous_slugs.append(name)
 148
 149             if category:
 150                 # something strange left off
 151                 raise Tag.DoesNotExist()
 152             if ambiguous_slugs:
 153                 # some tags should be qualified
 154                 e = Tag.MultipleObjectsReturned()
 155                 e.tags = real_tags
 156                 e.ambiguous_slugs = ambiguous_slugs
 157                 raise e
 158             if deprecated:
 159                 e = Tag.UrlDeprecationWarning()
 160                 e.tags = real_tags
 161                 raise e
 162             return real_tags
 163         else:
 164             return TagBase.get_tag_list(tags)
 165
 166     @property
 167     def url_chunk(self):
 168         return '/'.join((Tag.categories_dict[self.category], self.slug))
 169
 170     @staticmethod
 171     def tags_from_info(info):
 172         from slughifi import slughifi
 173         from sortify import sortify
 174         meta_tags = []
 175         categories = (('kinds', 'kind'), ('genres', 'genre'), ('authors', 'author'), ('epochs', 'epoch'))
 176         for field_name, category in categories:
 177             try:
 178                 tag_names = getattr(info, field_name)
 179             except:
 180                 try:
 181                     tag_names = [getattr(info, category)]
 182                 except:
 183                     # For instance, Pictures do not have 'genre' field.
 184                     continue
 185             for tag_name in tag_names:
 186                 tag_sort_key = tag_name
 187                 if category == 'author':
 188                     tag_sort_key = tag_name.last_name
 189                     tag_name = tag_name.readable()
 190                 tag, created = Tag.objects.get_or_create(slug=slughifi(tag_name), category=category)
 191                 if created:
 192                     tag.name = tag_name
 193                     tag.sort_key = sortify(tag_sort_key.lower())
 194                     tag.save()
 195                 meta_tags.append(tag)
 196         return meta_tags
 197
 198
 199
 200 def get_dynamic_path(media, filename, ext=None, maxlen=100):
 201     from slughifi import slughifi
 202
 203     # how to put related book's slug here?
 204     if not ext:
 205         # BookMedia case
 206         ext = media.formats[media.type].ext
 207     if media is None or not media.name:
 208         name = slughifi(filename.split(".")[0])
 209     else:
 210         name = slughifi(media.name)
 211     return 'book/%s/%s.%s' % (ext, name[:maxlen-len('book/%s/.%s' % (ext, ext))-4], ext)
 212
 213
 214 # TODO: why is this hard-coded ?
 215 def book_upload_path(ext=None, maxlen=100):
 216     return lambda *args: get_dynamic_path(*args, ext=ext, maxlen=maxlen)
 217
 218
 219 def customizations_hash(customizations):
 220     customizations.sort()
 221     return hash(tuple(customizations))
 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     h = customizations_hash(customizations)
 229     pdf_name = '%s-custom-%s' % (book.slug, h)
 230     pdf_file = get_dynamic_path(None, pdf_name, ext='pdf')
 231
 232     return pdf_file
 233
 234
 235 def get_existing_customized_pdf(book):
 236     """
 237     Returns a list of paths to generated customized pdf of a book
 238     """
 239     pdf_glob = '%s-custom-' % (book.slug,)
 240     pdf_glob = get_dynamic_path(None, pdf_glob, ext='pdf')
 241     pdf_glob = re.sub(r"[.]([a-z0-9]+)$", "*.\\1", pdf_glob)
 242     return glob(path.join(settings.MEDIA_ROOT, pdf_glob))
 243
 244
 245 class BookMedia(models.Model):
 246     FileFormat = namedtuple("FileFormat", "name ext")
 247     formats = SortedDict([
 248         ('mp3', FileFormat(name='MP3', ext='mp3')),
 249         ('ogg', FileFormat(name='Ogg Vorbis', ext='ogg')),
 250         ('daisy', FileFormat(name='DAISY', ext='daisy.zip')),
 251     ])
 252     format_choices = [(k, _('%s file') % t.name)
 253             for k, t in formats.items()]
 254
 255     type        = models.CharField(_('type'), choices=format_choices, max_length="100")
 256     name        = models.CharField(_('name'), max_length="100")
 257     file        = OverwritingFileField(_('file'), upload_to=book_upload_path())
 258     uploaded_at = models.DateTimeField(_('creation date'), auto_now_add=True, editable=False)
 259     extra_info  = JSONField(_('extra information'), default='{}', editable=False)
 260     book = models.ForeignKey('Book', related_name='media')
 261     source_sha1 = models.CharField(null=True, blank=True, max_length=40, editable=False)
 262
 263     def __unicode__(self):
 264         return "%s (%s)" % (self.name, self.file.name.split("/")[-1])
 265
 266     class Meta:
 267         ordering            = ('type', 'name')
 268         verbose_name        = _('book media')
 269         verbose_name_plural = _('book media')
 270
 271     def save(self, *args, **kwargs):
 272         from slughifi import slughifi
 273         from catalogue.utils import ExistingFile, remove_zip
 274
 275         try:
 276             old = BookMedia.objects.get(pk=self.pk)
 277         except BookMedia.DoesNotExist, e:
 278             old = None
 279         else:
 280             # if name changed, change the file name, too
 281             if slughifi(self.name) != slughifi(old.name):
 282                 self.file.save(None, ExistingFile(self.file.path), save=False, leave=True)
 283
 284         super(BookMedia, self).save(*args, **kwargs)
 285
 286         # remove the zip package for book with modified media
 287         if old:
 288             remove_zip("%s_%s" % (old.book.slug, old.type))
 289         remove_zip("%s_%s" % (self.book.slug, self.type))
 290
 291         extra_info = self.get_extra_info_value()
 292         extra_info.update(self.read_meta())
 293         self.set_extra_info_value(extra_info)
 294         self.source_sha1 = self.read_source_sha1(self.file.path, self.type)
 295         return super(BookMedia, self).save(*args, **kwargs)
 296
 297     def read_meta(self):
 298         """
 299             Reads some metadata from the audiobook.
 300         """
 301         import mutagen
 302         from mutagen import id3
 303
 304         artist_name = director_name = project = funded_by = ''
 305         if self.type == 'mp3':
 306             try:
 307                 audio = id3.ID3(self.file.path)
 308                 artist_name = ', '.join(', '.join(tag.text) for tag in audio.getall('TPE1'))
 309                 director_name = ', '.join(', '.join(tag.text) for tag in audio.getall('TPE3'))
 310                 project = ", ".join([t.data for t in audio.getall('PRIV')
 311                         if t.owner=='wolnelektury.pl?project'])
 312                 funded_by = ", ".join([t.data for t in audio.getall('PRIV')
 313                         if t.owner=='wolnelektury.pl?funded_by'])
 314             except:
 315                 pass
 316         elif self.type == 'ogg':
 317             try:
 318                 audio = mutagen.File(self.file.path)
 319                 artist_name = ', '.join(audio.get('artist', []))
 320                 director_name = ', '.join(audio.get('conductor', []))
 321                 project = ", ".join(audio.get('project', []))
 322                 funded_by = ", ".join(audio.get('funded_by', []))
 323             except:
 324                 pass
 325         else:
 326             return {}
 327         return {'artist_name': artist_name, 'director_name': director_name,
 328                 'project': project, 'funded_by': funded_by}
 329
 330     @staticmethod
 331     def read_source_sha1(filepath, filetype):
 332         """
 333             Reads source file SHA1 from audiobok metadata.
 334         """
 335         import mutagen
 336         from mutagen import id3
 337
 338         if filetype == 'mp3':
 339             try:
 340                 audio = id3.ID3(filepath)
 341                 return [t.data for t in audio.getall('PRIV')
 342                         if t.owner=='wolnelektury.pl?flac_sha1'][0]
 343             except:
 344                 return None
 345         elif filetype == 'ogg':
 346             try:
 347                 audio = mutagen.File(filepath)
 348                 return audio.get('flac_sha1', [None])[0]
 349             except:
 350                 return None
 351         else:
 352             return None
 353
 354
 355 class Book(models.Model):
 356     title         = models.CharField(_('title'), max_length=120)
 357     sort_key = models.CharField(_('sort key'), max_length=120, db_index=True, editable=False)
 358     slug = models.SlugField(_('slug'), max_length=120, db_index=True,
 359             unique=True)
 360     common_slug = models.SlugField(_('slug'), max_length=120, db_index=True)
 361     language = models.CharField(_('language code'), max_length=3, db_index=True,
 362                     default=settings.CATALOGUE_DEFAULT_LANGUAGE)
 363     description   = models.TextField(_('description'), blank=True)
 364     created_at    = models.DateTimeField(_('creation date'), auto_now_add=True, db_index=True)
 365     changed_at    = models.DateTimeField(_('creation date'), auto_now=True, db_index=True)
 366     parent_number = models.IntegerField(_('parent number'), default=0)
 367     extra_info    = JSONField(_('extra information'), default='{}')
 368     gazeta_link   = models.CharField(blank=True, max_length=240)
 369     wiki_link     = models.CharField(blank=True, max_length=240)
 370     # files generated during publication
 371
 372     cover = models.FileField(_('cover'), upload_to=book_upload_path('png'),
 373                 null=True, blank=True)
 374     ebook_formats = ['pdf', 'epub', 'mobi', 'txt']
 375     formats = ebook_formats + ['html', 'xml']
 376
 377     parent        = models.ForeignKey('self', blank=True, null=True, related_name='children')
 378
 379     _related_info = jsonfield.JSONField(blank=True, null=True, editable=False)
 380
 381     objects  = models.Manager()
 382     tagged   = managers.ModelTaggedItemManager(Tag)
 383     tags     = managers.TagDescriptor(Tag)
 384
 385     html_built = django.dispatch.Signal()
 386     published = django.dispatch.Signal()
 387
 388     class AlreadyExists(Exception):
 389         pass
 390
 391     class Meta:
 392         ordering = ('sort_key',)
 393         verbose_name = _('book')
 394         verbose_name_plural = _('books')
 395
 396     def __unicode__(self):
 397         return self.title
 398
 399     def save(self, force_insert=False, force_update=False, reset_short_html=True, **kwargs):
 400         from sortify import sortify
 401
 402         self.sort_key = sortify(self.title)
 403
 404         ret = super(Book, self).save(force_insert, force_update)
 405
 406         if reset_short_html:
 407             self.reset_short_html()
 408
 409         return ret
 410
 411     @permalink
 412     def get_absolute_url(self):
 413         return ('catalogue.views.book_detail', [self.slug])
 414
 415     @property
 416     def name(self):
 417         return self.title
 418
 419     def book_tag_slug(self):
 420         return ('l-' + self.slug)[: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.formats:
 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.formats:
 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         type(self).objects.filter(pk=self.pk).update(_related_info=None)
 460         # Fragment.short_html relies on book's tags, so reset it here too
 461         for fragm in self.fragments.all():
 462             fragm.reset_short_html()
 463
 464     def has_description(self):
 465         return len(self.description) > 0
 466     has_description.short_description = _('description')
 467     has_description.boolean = True
 468
 469     # ugly ugly ugly
 470     def has_mp3_file(self):
 471         return bool(self.has_media("mp3"))
 472     has_mp3_file.short_description = 'MP3'
 473     has_mp3_file.boolean = True
 474
 475     def has_ogg_file(self):
 476         return bool(self.has_media("ogg"))
 477     has_ogg_file.short_description = 'OGG'
 478     has_ogg_file.boolean = True
 479
 480     def has_daisy_file(self):
 481         return bool(self.has_media("daisy"))
 482     has_daisy_file.short_description = 'DAISY'
 483     has_daisy_file.boolean = True
 484
 485     def wldocument(self, parse_dublincore=True):
 486         from catalogue.import_utils import ORMDocProvider
 487         from librarian.parser import WLDocument
 488
 489         return WLDocument.from_file(self.xml_file.path,
 490                 provider=ORMDocProvider(self),
 491                 parse_dublincore=parse_dublincore)
 492
 493     def build_cover(self, book_info=None):
 494         """(Re)builds the cover image."""
 495         from StringIO import StringIO
 496         from django.core.files.base import ContentFile
 497         from librarian.cover import WLCover
 498
 499         if book_info is None:
 500             book_info = self.wldocument().book_info
 501
 502         cover = WLCover(book_info).image()
 503         imgstr = StringIO()
 504         cover.save(imgstr, 'png')
 505         self.cover.save(None, ContentFile(imgstr.getvalue()))
 506
 507     def build_pdf(self, customizations=None, file_name=None):
 508         """ (Re)builds the pdf file.
 509         customizations - customizations which are passed to LaTeX class file.
 510         file_name - save the pdf file under a different name and DO NOT save it in db.
 511         """
 512         from os import unlink
 513         from django.core.files import File
 514         from catalogue.utils import remove_zip
 515
 516         pdf = self.wldocument().as_pdf(customizations=customizations)
 517
 518         if file_name is None:
 519             # we'd like to be sure not to overwrite changes happening while
 520             # (timely) pdf generation is taking place (async celery scenario)
 521             current_self = Book.objects.get(id=self.id)
 522             current_self.pdf_file.save('%s.pdf' % self.slug,
 523                     File(open(pdf.get_filename())))
 524             self.pdf_file = current_self.pdf_file
 525
 526             # remove cached downloadables
 527             remove_zip(settings.ALL_PDF_ZIP)
 528
 529             for customized_pdf in get_existing_customized_pdf(self):
 530                 unlink(customized_pdf)
 531         else:
 532             print "saving %s" % file_name
 533             print "to: %s" % DefaultStorage().path(file_name)
 534             DefaultStorage().save(file_name, File(open(pdf.get_filename())))
 535
 536     def build_mobi(self):
 537         """ (Re)builds the MOBI file.
 538
 539         """
 540         from django.core.files import File
 541         from catalogue.utils import remove_zip
 542
 543         mobi = self.wldocument().as_mobi()
 544
 545         self.mobi_file.save('%s.mobi' % self.slug, File(open(mobi.get_filename())))
 546
 547         # remove zip with all mobi files
 548         remove_zip(settings.ALL_MOBI_ZIP)
 549
 550     def build_epub(self):
 551         """(Re)builds the epub file."""
 552         from django.core.files import File
 553         from catalogue.utils import remove_zip
 554
 555         epub = self.wldocument().as_epub()
 556
 557         self.epub_file.save('%s.epub' % self.slug,
 558                 File(open(epub.get_filename())))
 559
 560         # remove zip package with all epub files
 561         remove_zip(settings.ALL_EPUB_ZIP)
 562
 563     def build_txt(self):
 564         from django.core.files.base import ContentFile
 565
 566         text = self.wldocument().as_text()
 567         self.txt_file.save('%s.txt' % self.slug, ContentFile(text.get_string()))
 568
 569
 570     def build_html(self):
 571         from django.core.files.base import ContentFile
 572         from slughifi import slughifi
 573         from librarian import html
 574
 575         meta_tags = list(self.tags.filter(
 576             category__in=('author', 'epoch', 'genre', 'kind')))
 577         book_tag = self.book_tag()
 578
 579         html_output = self.wldocument(parse_dublincore=False).as_html()
 580         if html_output:
 581             self.html_file.save('%s.html' % self.slug,
 582                     ContentFile(html_output.get_string()))
 583
 584             # get ancestor l-tags for adding to new fragments
 585             ancestor_tags = []
 586             p = self.parent
 587             while p:
 588                 ancestor_tags.append(p.book_tag())
 589                 p = p.parent
 590
 591             # Delete old fragments and create them from scratch
 592             self.fragments.all().delete()
 593             # Extract fragments
 594             closed_fragments, open_fragments = html.extract_fragments(self.html_file.path)
 595             for fragment in closed_fragments.values():
 596                 try:
 597                     theme_names = [s.strip() for s in fragment.themes.split(',')]
 598                 except AttributeError:
 599                     continue
 600                 themes = []
 601                 for theme_name in theme_names:
 602                     if not theme_name:
 603                         continue
 604                     tag, created = Tag.objects.get_or_create(slug=slughifi(theme_name), category='theme')
 605                     if created:
 606                         tag.name = theme_name
 607                         tag.sort_key = theme_name.lower()
 608                         tag.save()
 609                     themes.append(tag)
 610                 if not themes:
 611                     continue
 612
 613                 text = fragment.to_string()
 614                 short_text = truncate_html_words(text, 15)
 615                 if text == short_text:
 616                     short_text = ''
 617                 new_fragment = Fragment.objects.create(anchor=fragment.id, book=self,
 618                     text=text, short_text=short_text)
 619
 620                 new_fragment.save()
 621                 new_fragment.tags = set(meta_tags + themes + [book_tag] + ancestor_tags)
 622             self.save()
 623             self.html_built.send(sender=self)
 624             return True
 625         return False
 626
 627     @staticmethod
 628     def zip_format(format_):
 629         def pretty_file_name(book):
 630             return "%s/%s.%s" % (
 631                 b.get_extra_info_value()['author'],
 632                 b.slug,
 633                 format_)
 634
 635         field_name = "%s_file" % format_
 636         books = Book.objects.filter(parent=None).exclude(**{field_name: ""})
 637         paths = [(pretty_file_name(b), getattr(b, field_name).path)
 638                     for b in books]
 639         result = create_zip.delay(paths,
 640                     getattr(settings, "ALL_%s_ZIP" % format_.upper()))
 641         return result.wait()
 642
 643     def zip_audiobooks(self, format_):
 644         bm = BookMedia.objects.filter(book=self, type=format_)
 645         paths = map(lambda bm: (None, bm.file.path), bm)
 646         result = create_zip.delay(paths, "%s_%s" % (self.slug, format_))
 647         return result.wait()
 648
 649     def search_index(self, book_info=None, reuse_index=False, index_tags=True):
 650         if reuse_index:
 651             idx = search.ReusableIndex()
 652         else:
 653             idx = search.Index()
 654
 655         idx.open()
 656         try:
 657             idx.index_book(self, book_info)
 658             if index_tags:
 659                 idx.index_tags()
 660         finally:
 661             idx.close()
 662
 663     @classmethod
 664     def from_xml_file(cls, xml_file, **kwargs):
 665         from django.core.files import File
 666         from librarian import dcparser
 667
 668         # use librarian to parse meta-data
 669         book_info = dcparser.parse(xml_file)
 670
 671         if not isinstance(xml_file, File):
 672             xml_file = File(open(xml_file))
 673
 674         try:
 675             return cls.from_text_and_meta(xml_file, book_info, **kwargs)
 676         finally:
 677             xml_file.close()
 678
 679     @classmethod
 680     def from_text_and_meta(cls, raw_file, book_info, overwrite=False,
 681             build_epub=True, build_txt=True, build_pdf=True, build_mobi=True,
 682             search_index=True, search_index_tags=True, search_index_reuse=False):
 683         import re
 684         from sortify import sortify
 685
 686         # check for parts before we do anything
 687         children = []
 688         if hasattr(book_info, 'parts'):
 689             for part_url in book_info.parts:
 690                 try:
 691                     children.append(Book.objects.get(slug=part_url.slug))
 692                 except Book.DoesNotExist, e:
 693                     raise Book.DoesNotExist(_('Book "%s" does not exist.') %
 694                             part_url.slug)
 695
 696
 697         # Read book metadata
 698         book_slug = book_info.url.slug
 699         if re.search(r'[^a-z0-9-]', book_slug):
 700             raise ValueError('Invalid characters in slug')
 701         book, created = Book.objects.get_or_create(slug=book_slug)
 702
 703         if created:
 704             book_shelves = []
 705         else:
 706             if not overwrite:
 707                 raise Book.AlreadyExists(_('Book %s already exists') % (
 708                         book_slug))
 709             # Save shelves for this book
 710             book_shelves = list(book.tags.filter(category='set'))
 711
 712         book.language = book_info.language
 713         book.title = book_info.title
 714         if book_info.variant_of:
 715             book.common_slug = book_info.variant_of.slug
 716         else:
 717             book.common_slug = book.slug
 718         book.set_extra_info_value(book_info.to_dict())
 719         book.save()
 720
 721         meta_tags = Tag.tags_from_info(book_info)
 722
 723         book.tags = set(meta_tags + book_shelves)
 724
 725         book_tag = book.book_tag()
 726
 727         for n, child_book in enumerate(children):
 728             child_book.parent = book
 729             child_book.parent_number = n
 730             child_book.save()
 731
 732         # Save XML and HTML files
 733         book.xml_file.save('%s.xml' % book.slug, raw_file, save=False)
 734
 735         # delete old fragments when overwriting
 736         book.fragments.all().delete()
 737
 738         if book.build_html():
 739             if not settings.NO_BUILD_TXT and build_txt:
 740                 book.build_txt()
 741
 742         book.build_cover(book_info)
 743
 744         if not settings.NO_BUILD_EPUB and build_epub:
 745             book.build_epub()
 746
 747         if not settings.NO_BUILD_PDF and build_pdf:
 748             book.build_pdf()
 749
 750         if not settings.NO_BUILD_MOBI and build_mobi:
 751             book.build_mobi()
 752
 753         if not settings.NO_SEARCH_INDEX and search_index:
 754             book.search_index(index_tags=search_index_tags, reuse_index=search_index_reuse)
 755             #index_book.delay(book.id, book_info)
 756
 757         book_descendants = list(book.children.all())
 758         descendants_tags = set()
 759         # add l-tag to descendants and their fragments
 760         while len(book_descendants) > 0:
 761             child_book = book_descendants.pop(0)
 762             descendants_tags.update(child_book.tags)
 763             child_book.tags = list(child_book.tags) + [book_tag]
 764             child_book.save()
 765             for fragment in child_book.fragments.all():
 766                 fragment.tags = set(list(fragment.tags) + [book_tag])
 767             book_descendants += list(child_book.children.all())
 768
 769         for tag in descendants_tags:
 770             touch_tag(tag)
 771
 772         book.save()
 773
 774         # refresh cache
 775         book.reset_tag_counter()
 776         book.reset_theme_counter()
 777
 778         cls.published.send(sender=book)
 779         return book
 780
 781     def related_info(self):
 782         """Keeps info about related objects (tags, media) in cache field."""
 783         if self._related_info is not None:
 784             return self._related_info
 785         else:
 786             rel = {'tags': {}, 'media': {}}
 787
 788             tags = self.tags.filter(category__in=(
 789                     'author', 'kind', 'genre', 'epoch'))
 790             tags = split_tags(tags)
 791             for category in tags:
 792                 rel['tags'][category] = [
 793                         (t.name, t.slug) for t in tags[category]]
 794
 795             for media_format in BookMedia.formats:
 796                 rel['media'][media_format] = self.has_media(media_format)
 797
 798             book = self
 799             parents = []
 800             while book.parent:
 801                 parents.append((book.parent.title, book.parent.slug))
 802                 book = book.parent
 803             parents = parents[::-1]
 804             if parents:
 805                 rel['parents'] = parents
 806
 807             if self.pk:
 808                 type(self).objects.filter(pk=self.pk).update(_related_info=rel)
 809             return rel
 810
 811     def reset_tag_counter(self):
 812         if self.id is None:
 813             return
 814
 815         cache_key = "Book.tag_counter/%d" % self.id
 816         permanent_cache.delete(cache_key)
 817         if self.parent:
 818             self.parent.reset_tag_counter()
 819
 820     @property
 821     def tag_counter(self):
 822         if self.id:
 823             cache_key = "Book.tag_counter/%d" % self.id
 824             tags = permanent_cache.get(cache_key)
 825         else:
 826             tags = None
 827
 828         if tags is None:
 829             tags = {}
 830             for child in self.children.all().order_by():
 831                 for tag_pk, value in child.tag_counter.iteritems():
 832                     tags[tag_pk] = tags.get(tag_pk, 0) + value
 833             for tag in self.tags.exclude(category__in=('book', 'theme', 'set')).order_by():
 834                 tags[tag.pk] = 1
 835
 836             if self.id:
 837                 permanent_cache.set(cache_key, tags)
 838         return tags
 839
 840     def reset_theme_counter(self):
 841         if self.id is None:
 842             return
 843
 844         cache_key = "Book.theme_counter/%d" % self.id
 845         permanent_cache.delete(cache_key)
 846         if self.parent:
 847             self.parent.reset_theme_counter()
 848
 849     @property
 850     def theme_counter(self):
 851         if self.id:
 852             cache_key = "Book.theme_counter/%d" % self.id
 853             tags = permanent_cache.get(cache_key)
 854         else:
 855             tags = None
 856
 857         if tags is None:
 858             tags = {}
 859             for fragment in Fragment.tagged.with_any([self.book_tag()]).order_by():
 860                 for tag in fragment.tags.filter(category='theme').order_by():
 861                     tags[tag.pk] = tags.get(tag.pk, 0) + 1
 862
 863             if self.id:
 864                 permanent_cache.set(cache_key, tags)
 865         return tags
 866
 867     def pretty_title(self, html_links=False):
 868         book = self
 869         names = list(book.tags.filter(category='author'))
 870
 871         books = []
 872         while book:
 873             books.append(book)
 874             book = book.parent
 875         names.extend(reversed(books))
 876
 877         if html_links:
 878             names = ['<a href="%s">%s</a>' % (tag.get_absolute_url(), tag.name) for tag in names]
 879         else:
 880             names = [tag.name for tag in names]
 881
 882         return ', '.join(names)
 883
 884     @classmethod
 885     def tagged_top_level(cls, tags):
 886         """ Returns top-level books tagged with `tags'.
 887
 888         It only returns those books which don't have ancestors which are
 889         also tagged with those tags.
 890
 891         """
 892         # get relevant books and their tags
 893         objects = cls.tagged.with_all(tags)
 894         # eliminate descendants
 895         l_tags = Tag.objects.filter(category='book', slug__in=[book.book_tag_slug() for book in objects])
 896         descendants_keys = [book.pk for book in cls.tagged.with_any(l_tags)]
 897         if descendants_keys:
 898             objects = objects.exclude(pk__in=descendants_keys)
 899
 900         return objects
 901
 902     @classmethod
 903     def book_list(cls, filter=None):
 904         """Generates a hierarchical listing of all books.
 905
 906         Books are optionally filtered with a test function.
 907
 908         """
 909
 910         books_by_parent = {}
 911         books = cls.objects.all().order_by('parent_number', 'sort_key').only(
 912                 'title', 'parent', 'slug')
 913         if filter:
 914             books = books.filter(filter).distinct()
 915             book_ids = set((book.pk for book in books))
 916             for book in books:
 917                 parent = book.parent_id
 918                 if parent not in book_ids:
 919                     parent = None
 920                 books_by_parent.setdefault(parent, []).append(book)
 921         else:
 922             for book in books:
 923                 books_by_parent.setdefault(book.parent_id, []).append(book)
 924
 925         orphans = []
 926         books_by_author = SortedDict()
 927         for tag in Tag.objects.filter(category='author'):
 928             books_by_author[tag] = []
 929
 930         for book in books_by_parent.get(None,()):
 931             authors = list(book.tags.filter(category='author'))
 932             if authors:
 933                 for author in authors:
 934                     books_by_author[author].append(book)
 935             else:
 936                 orphans.append(book)
 937
 938         return books_by_author, orphans, books_by_parent
 939
 940     _audiences_pl = {
 941         "SP1": (1, u"szkoła podstawowa"),
 942         "SP2": (1, u"szkoła podstawowa"),
 943         "P": (1, u"szkoła podstawowa"),
 944         "G": (2, u"gimnazjum"),
 945         "L": (3, u"liceum"),
 946         "LP": (3, u"liceum"),
 947     }
 948     def audiences_pl(self):
 949         audiences = self.get_extra_info_value().get('audiences', [])
 950         audiences = sorted(set([self._audiences_pl[a] for a in audiences]))
 951         return [a[1] for a in audiences]
 952
 953     def choose_fragment(self):
 954         tag = self.book_tag()
 955         fragments = Fragment.tagged.with_any([tag])
 956         if fragments.exists():
 957             return fragments.order_by('?')[0]
 958         elif self.parent:
 959             return self.parent.choose_fragment()
 960         else:
 961             return None
 962
 963
 964 def _has_factory(ftype):
 965     has = lambda self: bool(getattr(self, "%s_file" % ftype))
 966     has.short_description = t.upper()
 967     has.boolean = True
 968     has.__name__ = "has_%s_file" % ftype
 969     return has
 970
 971
 972 # add the file fields
 973 for t in Book.formats:
 974     field_name = "%s_file" % t
 975     models.FileField(_("%s file" % t.upper()),
 976             upload_to=book_upload_path(t),
 977             blank=True).contribute_to_class(Book, field_name)
 978
 979     setattr(Book, "has_%s_file" % t, _has_factory(t))
 980
 981
 982 class Fragment(models.Model):
 983     text = models.TextField()
 984     short_text = models.TextField(editable=False)
 985     anchor = models.CharField(max_length=120)
 986     book = models.ForeignKey(Book, related_name='fragments')
 987
 988     objects = models.Manager()
 989     tagged = managers.ModelTaggedItemManager(Tag)
 990     tags = managers.TagDescriptor(Tag)
 991
 992     class Meta:
 993         ordering = ('book', 'anchor',)
 994         verbose_name = _('fragment')
 995         verbose_name_plural = _('fragments')
 996
 997     def get_absolute_url(self):
 998         return '%s#m%s' % (reverse('book_text', args=[self.book.slug]), self.anchor)
 999
1000     def reset_short_html(self):
1001         if self.id is None:
1002             return
1003
1004         cache_key = "Fragment.short_html/%d/%s"
1005         for lang, langname in settings.LANGUAGES:
1006             permanent_cache.delete(cache_key % (self.id, lang))
1007
1008     def get_short_text(self):
1009         """Returns short version of the fragment."""
1010         return self.short_text if self.short_text else self.text
1011
1012     def short_html(self):
1013         if self.id:
1014             cache_key = "Fragment.short_html/%d/%s" % (self.id, get_language())
1015             short_html = permanent_cache.get(cache_key)
1016         else:
1017             short_html = None
1018
1019         if short_html is not None:
1020             return mark_safe(short_html)
1021         else:
1022             short_html = unicode(render_to_string('catalogue/fragment_short.html',
1023                 {'fragment': self}))
1024             if self.id:
1025                 permanent_cache.set(cache_key, short_html)
1026             return mark_safe(short_html)
1027
1028
1029 class Collection(models.Model):
1030     """A collection of books, which might be defined before publishing them."""
1031     title = models.CharField(_('title'), max_length=120, db_index=True)
1032     slug = models.SlugField(_('slug'), max_length=120, primary_key=True)
1033     description = models.TextField(_('description'), null=True, blank=True)
1034
1035     models.SlugField(_('slug'), max_length=120, unique=True, db_index=True)
1036     book_slugs = models.TextField(_('book slugs'))
1037
1038     class Meta:
1039         ordering = ('title',)
1040         verbose_name = _('collection')
1041         verbose_name_plural = _('collections')
1042
1043     def __unicode__(self):
1044         return self.title
1045
1046
1047 ###########
1048 #
1049 # SIGNALS
1050 #
1051 ###########
1052
1053
1054 def _tags_updated_handler(sender, affected_tags, **kwargs):
1055     # reset tag global counter
1056     # we want Tag.changed_at updated for API to know the tag was touched
1057     for tag in affected_tags:
1058         touch_tag(tag)
1059
1060     # if book tags changed, reset book tag counter
1061     if isinstance(sender, Book) and \
1062                 Tag.objects.filter(pk__in=(tag.pk for tag in affected_tags)).\
1063                     exclude(category__in=('book', 'theme', 'set')).count():
1064         sender.reset_tag_counter()
1065     # if fragment theme changed, reset book theme counter
1066     elif isinstance(sender, Fragment) and \
1067                 Tag.objects.filter(pk__in=(tag.pk for tag in affected_tags)).\
1068                     filter(category='theme').count():
1069         sender.book.reset_theme_counter()
1070 tags_updated.connect(_tags_updated_handler)
1071
1072
1073 def _pre_delete_handler(sender, instance, **kwargs):
1074     """ refresh Book on BookMedia delete """
1075     if sender == BookMedia:
1076         instance.book.save()
1077 pre_delete.connect(_pre_delete_handler)
1078
1079
1080 def _post_save_handler(sender, instance, **kwargs):
1081     """ refresh all the short_html stuff on BookMedia update """
1082     if sender == BookMedia:
1083         instance.book.save()
1084 post_save.connect(_post_save_handler)
1085
1086
1087 @django.dispatch.receiver(post_delete, sender=Book)
1088 def _remove_book_from_index_handler(sender, instance, **kwargs):
1089     """ remove the book from search index, when it is deleted."""
1090     search.JVM.attachCurrentThread()
1091     idx = search.Index()
1092     idx.open(timeout=10000)  # 10 seconds timeout.
1093     try:
1094         idx.remove_book(instance)
1095         idx.index_tags()
1096     finally:
1097         idx.close()