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 django.db import models
   6 from django.db.models import permalink, Q
   7 from django.utils.translation import ugettext_lazy as _
   8 from django.contrib.auth.models import User
   9 from django.core.files import File
  10 from django.template.loader import render_to_string
  11 from django.utils.safestring import mark_safe
  12 from django.utils.translation import get_language
  13 from django.core.urlresolvers import reverse
  14 from datetime import datetime
  15
  16 from newtagging.models import TagBase, tags_updated
  17 from newtagging import managers
  18 from catalogue.fields import JSONField
  19
  20 from librarian import dcparser, html, epub, NoDublinCore
  21 from mutagen import id3
  22
  23
  24 TAG_CATEGORIES = (
  25     ('author', _('author')),
  26     ('epoch', _('epoch')),
  27     ('kind', _('kind')),
  28     ('genre', _('genre')),
  29     ('theme', _('theme')),
  30     ('set', _('set')),
  31     ('book', _('book')),
  32 )
  33
  34
  35 class TagSubcategoryManager(models.Manager):
  36     def __init__(self, subcategory):
  37         super(TagSubcategoryManager, self).__init__()
  38         self.subcategory = subcategory
  39
  40     def get_query_set(self):
  41         return super(TagSubcategoryManager, self).get_query_set().filter(category=self.subcategory)
  42
  43
  44 class Tag(TagBase):
  45     name = models.CharField(_('name'), max_length=50, db_index=True)
  46     slug = models.SlugField(_('slug'), max_length=120, db_index=True)
  47     sort_key = models.SlugField(_('sort key'), max_length=120, db_index=True)
  48     category = models.CharField(_('category'), max_length=50, blank=False, null=False,
  49         db_index=True, choices=TAG_CATEGORIES)
  50     description = models.TextField(_('description'), blank=True)
  51     main_page = models.BooleanField(_('main page'), default=False, db_index=True, help_text=_('Show tag on main page'))
  52
  53     user = models.ForeignKey(User, blank=True, null=True)
  54     book_count = models.IntegerField(_('book count'), blank=False, null=True)
  55     death = models.IntegerField(_(u'year of death'), blank=True, null=True)
  56     gazeta_link = models.CharField(blank=True, max_length=240)
  57     wiki_link = models.CharField(blank=True, max_length=240)
  58
  59     categories_rev = {
  60         'autor': 'author',
  61         'epoka': 'epoch',
  62         'rodzaj': 'kind',
  63         'gatunek': 'genre',
  64         'motyw': 'theme',
  65         'polka': 'set',
  66     }
  67     categories_dict = dict((item[::-1] for item in categories_rev.iteritems()))
  68
  69     class Meta:
  70         ordering = ('sort_key',)
  71         verbose_name = _('tag')
  72         verbose_name_plural = _('tags')
  73         unique_together = (("slug", "category"),)
  74
  75     def __unicode__(self):
  76         return self.name
  77
  78     def __repr__(self):
  79         return "Tag(slug=%r)" % self.slug
  80
  81     @permalink
  82     def get_absolute_url(self):
  83         return ('catalogue.views.tagged_object_list', [self.url_chunk])
  84
  85     def has_description(self):
  86         return len(self.description) > 0
  87     has_description.short_description = _('description')
  88     has_description.boolean = True
  89
  90     def alive(self):
  91         return self.death is None
  92
  93     def in_pd(self):
  94         """ tests whether an author is in public domain """
  95         return self.death is not None and self.goes_to_pd() <= datetime.now().year
  96
  97     def goes_to_pd(self):
  98         """ calculates the year of public domain entry for an author """
  99         return self.death + 71 if self.death is not None else None
 100
 101     def get_count(self):
 102         """ returns global book count for book tags, fragment count for themes """
 103
 104         if self.book_count is None:
 105             if self.category == 'book':
 106                 # never used
 107                 objects = Book.objects.none()
 108             elif self.category == 'theme':
 109                 objects = Fragment.tagged.with_all((self,))
 110             else:
 111                 objects = Book.tagged.with_all((self,)).order_by()
 112                 if self.category != 'set':
 113                     # eliminate descendants
 114                     l_tags = Tag.objects.filter(slug__in=[book.book_tag_slug() for book in objects])
 115                     descendants_keys = [book.pk for book in Book.tagged.with_any(l_tags)]
 116                     if descendants_keys:
 117                         objects = objects.exclude(pk__in=descendants_keys)
 118             self.book_count = objects.count()
 119             self.save()
 120         return self.book_count
 121
 122     @staticmethod
 123     def get_tag_list(tags):
 124         if isinstance(tags, basestring):
 125             real_tags = []
 126             ambiguous_slugs = []
 127             category = None
 128             tags_splitted = tags.split('/')
 129             for index, name in enumerate(tags_splitted):
 130                 if name in Tag.categories_rev:
 131                     category = Tag.categories_rev[name]
 132                 else:
 133                     if category:
 134                         real_tags.append(Tag.objects.get(slug=name, category=category))
 135                         category = None
 136                     else:
 137                         try:
 138                             real_tags.append(Tag.objects.exclude(category='book').get(slug=name))
 139                         except Tag.MultipleObjectsReturned, e:
 140                             ambiguous_slugs.append(name)
 141
 142             if category:
 143                 # something strange left off
 144                 raise Tag.DoesNotExist()
 145             if ambiguous_slugs:
 146                 # some tags should be qualified
 147                 e = Tag.MultipleObjectsReturned()
 148                 e.tags = real_tags
 149                 e.ambiguous_slugs = ambiguous_slugs
 150                 raise e
 151             else:
 152                 return real_tags
 153         else:
 154             return TagBase.get_tag_list(tags)
 155
 156     @property
 157     def url_chunk(self):
 158         return '/'.join((Tag.categories_dict[self.category], self.slug))
 159
 160
 161 # TODO: why is this hard-coded ?
 162 def book_upload_path(ext):
 163     def get_dynamic_path(book, filename):
 164         return 'lektura/%s.%s' % (book.slug, ext)
 165     return get_dynamic_path
 166
 167
 168 class Book(models.Model):
 169     title = models.CharField(_('title'), max_length=120)
 170     slug = models.SlugField(_('slug'), max_length=120, unique=True, db_index=True)
 171     description = models.TextField(_('description'), blank=True)
 172     created_at = models.DateTimeField(_('creation date'), auto_now=True)
 173     _short_html = models.TextField(_('short HTML'), editable=False)
 174     parent_number = models.IntegerField(_('parent number'), default=0)
 175     extra_info = JSONField(_('extra information'))
 176     gazeta_link = models.CharField(blank=True, max_length=240)
 177     wiki_link = models.CharField(blank=True, max_length=240)
 178
 179
 180     # Formats
 181     xml_file = models.FileField(_('XML file'), upload_to=book_upload_path('xml'), blank=True)
 182     html_file = models.FileField(_('HTML file'), upload_to=book_upload_path('html'), blank=True)
 183     pdf_file = models.FileField(_('PDF file'), upload_to=book_upload_path('pdf'), blank=True)
 184     epub_file = models.FileField(_('EPUB file'), upload_to=book_upload_path('epub'), blank=True)
 185     odt_file = models.FileField(_('ODT file'), upload_to=book_upload_path('odt'), blank=True)
 186     txt_file = models.FileField(_('TXT file'), upload_to=book_upload_path('txt'), blank=True)
 187     mp3_file = models.FileField(_('MP3 file'), upload_to=book_upload_path('mp3'), blank=True)
 188     ogg_file = models.FileField(_('OGG file'), upload_to=book_upload_path('ogg'), blank=True)
 189
 190     parent = models.ForeignKey('self', blank=True, null=True, related_name='children')
 191
 192     objects = models.Manager()
 193     tagged = managers.ModelTaggedItemManager(Tag)
 194     tags = managers.TagDescriptor(Tag)
 195
 196     _tag_counter = JSONField(null=True, editable=False)
 197     _theme_counter = JSONField(null=True, editable=False)
 198
 199     class AlreadyExists(Exception):
 200         pass
 201
 202     class Meta:
 203         ordering = ('title',)
 204         verbose_name = _('book')
 205         verbose_name_plural = _('books')
 206
 207     def __unicode__(self):
 208         return self.title
 209
 210     def save(self, force_insert=False, force_update=False, reset_short_html=True, refresh_mp3=True, **kwargs):
 211         if reset_short_html:
 212             # Reset _short_html during save
 213             update = {}
 214             for key in filter(lambda x: x.startswith('_short_html'), self.__dict__):
 215                 update[key] = ''
 216                 self.__setattr__(key, '')
 217             # Fragment.short_html relies on book's tags, so reset it here too
 218             self.fragments.all().update(**update)
 219
 220         book = super(Book, self).save(force_insert, force_update)
 221
 222         if refresh_mp3 and self.mp3_file:
 223             print self.mp3_file, self.mp3_file.path
 224             extra_info = self.get_extra_info_value()
 225             extra_info.update(self.get_mp3_info())
 226             self.set_extra_info_value(extra_info)
 227             book = super(Book, self).save(force_insert, force_update)
 228
 229         return book
 230
 231     @permalink
 232     def get_absolute_url(self):
 233         return ('catalogue.views.book_detail', [self.slug])
 234
 235     @property
 236     def name(self):
 237         return self.title
 238
 239     def book_tag_slug(self):
 240         return ('l-' + self.slug)[:120]
 241
 242     def book_tag(self):
 243         slug = self.book_tag_slug()
 244         book_tag, created = Tag.objects.get_or_create(slug=slug, category='book')
 245         if created:
 246             book_tag.name = self.title[:50]
 247             book_tag.sort_key = slug
 248             book_tag.save()
 249         return book_tag
 250
 251     def short_html(self):
 252         key = '_short_html_%s' % get_language()
 253         short_html = getattr(self, key)
 254
 255         if short_html and len(short_html):
 256             return mark_safe(short_html)
 257         else:
 258             tags = self.tags.filter(~Q(category__in=('set', 'theme', 'book')))
 259             tags = [mark_safe(u'<a href="%s">%s</a>' % (tag.get_absolute_url(), tag.name)) for tag in tags]
 260
 261             formats = []
 262             if self.html_file:
 263                 formats.append(u'<a href="%s">%s</a>' % (reverse('book_text', kwargs={'slug': self.slug}), _('Read online')))
 264             if self.pdf_file:
 265                 formats.append(u'<a href="%s">PDF</a>' % self.pdf_file.url)
 266             if self.epub_file:
 267                 formats.append(u'<a href="%s">EPUB</a>' % self.epub_file.url)
 268             if self.odt_file:
 269                 formats.append(u'<a href="%s">ODT</a>' % self.odt_file.url)
 270             if self.txt_file:
 271                 formats.append(u'<a href="%s">TXT</a>' % self.txt_file.url)
 272             if self.mp3_file:
 273                 formats.append(u'<a href="%s">MP3</a>' % self.mp3_file.url)
 274             if self.ogg_file:
 275                 formats.append(u'<a href="%s">OGG</a>' % self.ogg_file.url)
 276
 277             formats = [mark_safe(format) for format in formats]
 278
 279             setattr(self, key, unicode(render_to_string('catalogue/book_short.html',
 280                 {'book': self, 'tags': tags, 'formats': formats})))
 281             self.save(reset_short_html=False)
 282             return mark_safe(getattr(self, key))
 283
 284
 285     def get_mp3_info(self):
 286         """Retrieves artist and director names from audio ID3 tags."""
 287         audio = id3.ID3(self.mp3_file.path)
 288         artist_name = ', '.join(', '.join(tag.text) for tag in audio.getall('TPE1'))
 289         director_name = ', '.join(', '.join(tag.text) for tag in audio.getall('TPE3'))
 290         return {'artist_name': artist_name, 'director_name': director_name}
 291
 292     def has_description(self):
 293         return len(self.description) > 0
 294     has_description.short_description = _('description')
 295     has_description.boolean = True
 296
 297     def has_pdf_file(self):
 298         return bool(self.pdf_file)
 299     has_pdf_file.short_description = 'PDF'
 300     has_pdf_file.boolean = True
 301
 302     def has_epub_file(self):
 303         return bool(self.epub_file)
 304     has_epub_file.short_description = 'EPUB'
 305     has_epub_file.boolean = True
 306
 307     def has_odt_file(self):
 308         return bool(self.odt_file)
 309     has_odt_file.short_description = 'ODT'
 310     has_odt_file.boolean = True
 311
 312     def has_html_file(self):
 313         return bool(self.html_file)
 314     has_html_file.short_description = 'HTML'
 315     has_html_file.boolean = True
 316
 317     @classmethod
 318     def from_xml_file(cls, xml_file, overwrite=False):
 319         # use librarian to parse meta-data
 320         book_info = dcparser.parse(xml_file)
 321
 322         if not isinstance(xml_file, File):
 323             xml_file = File(open(xml_file))
 324
 325         try:
 326             return cls.from_text_and_meta(xml_file, book_info, overwrite)
 327         finally:
 328             xml_file.close()
 329
 330     @classmethod
 331     def from_text_and_meta(cls, raw_file, book_info, overwrite=False):
 332         from tempfile import NamedTemporaryFile
 333         from slughifi import slughifi
 334         from markupstring import MarkupString
 335         from hashlib import sha1
 336         from django.core.files.base import ContentFile
 337         from django.core.files.storage import default_storage
 338         from StringIO import StringIO
 339
 340         # Read book metadata
 341         book_base, book_slug = book_info.url.rsplit('/', 1)
 342         book, created = Book.objects.get_or_create(slug=book_slug)
 343
 344         if created:
 345             book_shelves = []
 346         else:
 347             if not overwrite:
 348                 raise Book.AlreadyExists(_('Book %s already exists') % book_slug)
 349             # Save shelves for this book
 350             book_shelves = list(book.tags.filter(category='set'))
 351
 352         book.title = book_info.title
 353         book.set_extra_info_value(book_info.to_dict())
 354         book._short_html = ''
 355         book.save()
 356
 357         book_tags = []
 358         categories = (('kinds', 'kind'), ('genres', 'genre'), ('authors', 'author'), ('epochs', 'epoch'))
 359         for field_name, category in categories:
 360             try:
 361                 tag_names = getattr(book_info, field_name)
 362             except:
 363                 tag_names = [getattr(book_info, category)]
 364             for tag_name in tag_names:
 365                 tag_sort_key = tag_name
 366                 if category == 'author':
 367                     tag_sort_key = tag_name.last_name
 368                     tag_name = ' '.join(tag_name.first_names) + ' ' + tag_name.last_name
 369                 tag, created = Tag.objects.get_or_create(slug=slughifi(tag_name), category=category)
 370                 if created:
 371                     tag.name = tag_name
 372                     tag.sort_key = slughifi(tag_sort_key)
 373                     tag.save()
 374                 book_tags.append(tag)
 375
 376         book.tags = book_tags + book_shelves
 377
 378         book_tag = book.book_tag()
 379
 380         if hasattr(book_info, 'parts'):
 381             for n, part_url in enumerate(book_info.parts):
 382                 base, slug = part_url.rsplit('/', 1)
 383                 try:
 384                     child_book = Book.objects.get(slug=slug)
 385                     child_book.parent = book
 386                     child_book.parent_number = n
 387                     child_book.save()
 388                 except Book.DoesNotExist, e:
 389                     raise Book.DoesNotExist(_('Book with slug = "%s" does not exist.') % slug)
 390
 391         book_descendants = list(book.children.all())
 392         while len(book_descendants) > 0:
 393             child_book = book_descendants.pop(0)
 394             child_book.tags = list(child_book.tags) + [book_tag]
 395             child_book.save()
 396             for fragment in child_book.fragments.all():
 397                 fragment.tags = set(list(fragment.tags) + [book_tag])
 398             book_descendants += list(child_book.children.all())
 399
 400         # Save XML and HTML files
 401         book.xml_file.save('%s.xml' % book.slug, raw_file, save=False)
 402
 403         html_file = NamedTemporaryFile()
 404         if html.transform(book.xml_file.path, html_file, parse_dublincore=False):
 405             book.html_file.save('%s.html' % book.slug, File(html_file), save=False)
 406
 407             # Create EPUB
 408             epub_file = StringIO()
 409             try:
 410                 epub.transform(book.xml_file, epub_file)
 411                 book.epub_file.save('%s.epub' % book.slug, ContentFile(epub_file.getvalue()), save=False)
 412                 FileRecord(slug=book.slug, type='epub', sha1=sha1(epub_file.getvalue()).hexdigest()).save()
 413             except NoDublinCore:
 414                 pass
 415
 416             # Extract fragments
 417             closed_fragments, open_fragments = html.extract_fragments(book.html_file.path)
 418             for fragment in closed_fragments.values():
 419                 try:
 420                     theme_names = [s.strip() for s in fragment.themes.split(',')]
 421                 except AttributeError:
 422                     continue
 423                 themes = []
 424                 for theme_name in theme_names:
 425                     if not theme_name:
 426                         continue
 427                     tag, created = Tag.objects.get_or_create(slug=slughifi(theme_name), category='theme')
 428                     if created:
 429                         tag.name = theme_name
 430                         tag.sort_key = slughifi(theme_name)
 431                         tag.save()
 432                     themes.append(tag)
 433                 if not themes:
 434                     continue
 435
 436                 text = fragment.to_string()
 437                 short_text = ''
 438                 if (len(MarkupString(text)) > 240):
 439                     short_text = unicode(MarkupString(text)[:160])
 440                 new_fragment, created = Fragment.objects.get_or_create(anchor=fragment.id, book=book,
 441                     defaults={'text': text, 'short_text': short_text})
 442
 443                 new_fragment.save()
 444                 new_fragment.tags = set(book_tags + themes + [book_tag])
 445
 446         # refresh cache
 447         book.tag_counter
 448         book.theme_counter
 449
 450         book.save()
 451         return book
 452
 453
 454     def refresh_tag_counter(self):
 455         tags = {}
 456         for child in self.children.all().order_by():
 457             for tag_pk, value in child.tag_counter.iteritems():
 458                 tags[tag_pk] = tags.get(tag_pk, 0) + value
 459         for tag in self.tags.exclude(category__in=('book', 'theme', 'set')).order_by():
 460             tags[tag.pk] = 1
 461         self.set__tag_counter_value(tags)
 462         self.save(reset_short_html=False, refresh_mp3=False)
 463         return tags
 464
 465     def reset_tag_counter(self):
 466         self._tag_counter = None
 467         self.save(reset_short_html=False, refresh_mp3=False)
 468         if self.parent:
 469             self.parent.reset_tag_counter()
 470
 471     @property
 472     def tag_counter(self):
 473         if self._tag_counter is None:
 474             return self.refresh_tag_counter()
 475         return dict((int(k), v) for k, v in self.get__tag_counter_value().iteritems())
 476
 477     def refresh_theme_counter(self):
 478         tags = {}
 479         for fragment in Fragment.tagged.with_any([self.book_tag()]).order_by():
 480             for tag in fragment.tags.filter(category='theme').order_by():
 481                 tags[tag.pk] = tags.get(tag.pk, 0) + 1
 482         self.set__theme_counter_value(tags)
 483         self.save(reset_short_html=False, refresh_mp3=False)
 484         return tags
 485
 486     def reset_theme_counter(self):
 487         self._theme_counter = None
 488         self.save(reset_short_html=False, refresh_mp3=False)
 489         if self.parent:
 490             self.parent.reset_theme_counter()
 491
 492     @property
 493     def theme_counter(self):
 494         if self._theme_counter is None:
 495             return self.refresh_theme_counter()
 496         return dict((int(k), v) for k, v in self.get__theme_counter_value().iteritems())
 497
 498
 499
 500 class Fragment(models.Model):
 501     text = models.TextField()
 502     short_text = models.TextField(editable=False)
 503     _short_html = models.TextField(editable=False)
 504     anchor = models.CharField(max_length=120)
 505     book = models.ForeignKey(Book, related_name='fragments')
 506
 507     objects = models.Manager()
 508     tagged = managers.ModelTaggedItemManager(Tag)
 509     tags = managers.TagDescriptor(Tag)
 510
 511     class Meta:
 512         ordering = ('book', 'anchor',)
 513         verbose_name = _('fragment')
 514         verbose_name_plural = _('fragments')
 515
 516     def get_absolute_url(self):
 517         return '%s#m%s' % (reverse('book_text', kwargs={'slug': self.book.slug}), self.anchor)
 518
 519     def short_html(self):
 520         key = '_short_html_%s' % get_language()
 521         short_html = getattr(self, key)
 522         if short_html and len(short_html):
 523             return mark_safe(short_html)
 524         else:
 525             setattr(self, key, unicode(render_to_string('catalogue/fragment_short.html',
 526                 {'fragment': self})))
 527             self.save()
 528             return mark_safe(getattr(self, key))
 529
 530
 531 class BookStub(models.Model):
 532     title = models.CharField(_('title'), max_length=120)
 533     author = models.CharField(_('author'), max_length=120)
 534     pd = models.IntegerField(_('goes to public domain'), null=True, blank=True)
 535     slug = models.SlugField(_('slug'), max_length=120, unique=True, db_index=True)
 536     translator = models.TextField(_('translator'), blank=True)
 537     translator_death = models.TextField(_('year of translator\'s death'), blank=True)
 538
 539     class Meta:
 540         ordering = ('title',)
 541         verbose_name = _('book stub')
 542         verbose_name_plural = _('book stubs')
 543
 544     def __unicode__(self):
 545         return self.title
 546
 547     @permalink
 548     def get_absolute_url(self):
 549         return ('catalogue.views.book_detail', [self.slug])
 550
 551     def in_pd(self):
 552         return self.pd is not None and self.pd <= datetime.now().year
 553
 554     @property
 555     def name(self):
 556         return self.title
 557
 558
 559 class FileRecord(models.Model):
 560     slug = models.SlugField(_('slug'), max_length=120, db_index=True)
 561     type = models.CharField(_('type'), max_length=20, db_index=True)
 562     sha1 = models.CharField(_('sha-1 hash'), max_length=40)
 563     time = models.DateTimeField(_('time'), auto_now_add=True)
 564
 565     class Meta:
 566         ordering = ('-time','-slug', '-type')
 567         verbose_name = _('file record')
 568         verbose_name_plural = _('file records')
 569
 570     def __unicode__(self):
 571         return "%s %s.%s" % (self.sha1,  self.slug, self.type)
 572
 573
 574 def _tags_updated_handler(sender, affected_tags, **kwargs):
 575     # reset tag global counter
 576     Tag.objects.filter(pk__in=[tag.pk for tag in affected_tags]).update(book_count=None)
 577
 578     # if book tags changed, reset book tag counter
 579     if isinstance(sender, Book) and \
 580                 Tag.objects.filter(pk__in=(tag.pk for tag in affected_tags)).\
 581                     exclude(category__in=('book', 'theme', 'set')).count():
 582         sender.reset_tag_counter()
 583     # if fragment theme changed, reset book theme counter
 584     elif isinstance(sender, Fragment) and \
 585                 Tag.objects.filter(pk__in=(tag.pk for tag in affected_tags)).\
 586                     filter(category='theme').count():
 587         sender.book.reset_theme_counter()
 588 tags_updated.connect(_tags_updated_handler)
 589