apps/catalogue/models/book.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 OrderedDict
   6 import re
   7 from django.conf import settings
   8 from django.core.cache import caches
   9 from django.db import models
  10 from django.db.models import permalink
  11 import django.dispatch
  12 from django.contrib.contenttypes.fields import GenericRelation
  13 from django.core.urlresolvers import reverse
  14 from django.utils.translation import ugettext_lazy as _
  15 import jsonfield
  16 from fnpdjango.storage import BofhFileSystemStorage
  17 from catalogue import constants
  18 from catalogue.fields import EbookField
  19 from catalogue.models import Tag, Fragment, BookMedia
  20 from catalogue.utils import create_zip, split_tags, related_tag_name
  21 from catalogue import app_settings
  22 from catalogue import tasks
  23 from newtagging import managers
  24
  25 bofh_storage = BofhFileSystemStorage()
  26
  27 permanent_cache = caches['permanent']
  28
  29
  30 def _cover_upload_to(i, n):
  31     return 'book/cover/%s.jpg' % i.slug
  32
  33 def _cover_thumb_upload_to(i, n):
  34     return 'book/cover_thumb/%s.jpg' % i.slug,
  35
  36 def _ebook_upload_to(upload_path):
  37     def _upload_to(i, n):
  38         return upload_path % i.slug
  39     return _upload_to
  40
  41
  42 class Book(models.Model):
  43     """Represents a book imported from WL-XML."""
  44     title         = models.CharField(_('title'), max_length=120)
  45     sort_key = models.CharField(_('sort key'), max_length=120, db_index=True, editable=False)
  46     sort_key_author = models.CharField(_('sort key by author'), max_length=120, db_index=True, editable=False, default=u'')
  47     slug = models.SlugField(_('slug'), max_length=120, db_index=True,
  48             unique=True)
  49     common_slug = models.SlugField(_('slug'), max_length=120, db_index=True)
  50     language = models.CharField(_('language code'), max_length=3, db_index=True,
  51                     default=app_settings.DEFAULT_LANGUAGE)
  52     description   = models.TextField(_('description'), blank=True)
  53     created_at    = models.DateTimeField(_('creation date'), auto_now_add=True, db_index=True)
  54     changed_at    = models.DateTimeField(_('creation date'), auto_now=True, db_index=True)
  55     parent_number = models.IntegerField(_('parent number'), default=0)
  56     extra_info    = jsonfield.JSONField(_('extra information'), default={})
  57     gazeta_link   = models.CharField(blank=True, max_length=240)
  58     wiki_link     = models.CharField(blank=True, max_length=240)
  59     # files generated during publication
  60
  61     cover = EbookField('cover', _('cover'),
  62             null=True, blank=True,
  63             upload_to=_cover_upload_to,
  64             storage=bofh_storage, max_length=255)
  65     # Cleaner version of cover for thumbs
  66     cover_thumb = EbookField('cover_thumb', _('cover thumbnail'),
  67             null=True, blank=True,
  68             upload_to=_cover_thumb_upload_to,
  69             max_length=255)
  70     ebook_formats = constants.EBOOK_FORMATS
  71     formats = ebook_formats + ['html', 'xml']
  72
  73     parent = models.ForeignKey('self', blank=True, null=True,
  74         related_name='children')
  75
  76     _related_info = jsonfield.JSONField(blank=True, null=True, editable=False)
  77
  78     objects  = models.Manager()
  79     tagged   = managers.ModelTaggedItemManager(Tag)
  80     tags     = managers.TagDescriptor(Tag)
  81     tag_relations = GenericRelation(Tag.intermediary_table_model)
  82
  83     html_built = django.dispatch.Signal()
  84     published = django.dispatch.Signal()
  85
  86     class AlreadyExists(Exception):
  87         pass
  88
  89     class Meta:
  90         ordering = ('sort_key',)
  91         verbose_name = _('book')
  92         verbose_name_plural = _('books')
  93         app_label = 'catalogue'
  94
  95     def __unicode__(self):
  96         return self.title
  97
  98     def save(self, force_insert=False, force_update=False, reset_short_html=True, **kwargs):
  99         from sortify import sortify
 100
 101         self.sort_key = sortify(self.title)
 102         self.title = unicode(self.title) # ???
 103
 104         ret = super(Book, self).save(force_insert, force_update, **kwargs)
 105
 106         if reset_short_html:
 107             self.reset_short_html()
 108
 109         return ret
 110
 111     @permalink
 112     def get_absolute_url(self):
 113         return ('catalogue.views.book_detail', [self.slug])
 114
 115     @staticmethod
 116     @permalink
 117     def create_url(slug):
 118         return ('catalogue.views.book_detail', [slug])
 119
 120     @property
 121     def name(self):
 122         return self.title
 123
 124     def language_code(self):
 125         return constants.LANGUAGES_3TO2.get(self.language, self.language)
 126
 127     def language_name(self):
 128         return dict(settings.LANGUAGES).get(self.language_code(), "")
 129
 130     def book_tag_slug(self):
 131         return ('l-' + self.slug)[:120]
 132
 133     def book_tag(self):
 134         slug = self.book_tag_slug()
 135         book_tag, created = Tag.objects.get_or_create(slug=slug, category='book')
 136         if created:
 137             book_tag.name = self.title[:50]
 138             book_tag.sort_key = self.title.lower()
 139             book_tag.save()
 140         return book_tag
 141
 142     def has_media(self, type_):
 143         if type_ in Book.formats:
 144             return bool(getattr(self, "%s_file" % type_))
 145         else:
 146             return self.media.filter(type=type_).exists()
 147
 148     def get_media(self, type_):
 149         if self.has_media(type_):
 150             if type_ in Book.formats:
 151                 return getattr(self, "%s_file" % type_)
 152             else:
 153                 return self.media.filter(type=type_)
 154         else:
 155             return None
 156
 157     def get_mp3(self):
 158         return self.get_media("mp3")
 159     def get_odt(self):
 160         return self.get_media("odt")
 161     def get_ogg(self):
 162         return self.get_media("ogg")
 163     def get_daisy(self):
 164         return self.get_media("daisy")
 165
 166     def reset_short_html(self):
 167         if self.id is None:
 168             return
 169
 170         type(self).objects.filter(pk=self.pk).update(_related_info=None)
 171         # Fragment.short_html relies on book's tags, so reset it here too
 172         for fragm in self.fragments.all().iterator():
 173             fragm.reset_short_html()
 174
 175         try:
 176             author = self.tags.filter(category='author')[0].sort_key
 177         except IndexError:
 178             author = u''
 179         type(self).objects.filter(pk=self.pk).update(sort_key_author=author)
 180
 181
 182
 183     def has_description(self):
 184         return len(self.description) > 0
 185     has_description.short_description = _('description')
 186     has_description.boolean = True
 187
 188     # ugly ugly ugly
 189     def has_mp3_file(self):
 190         return bool(self.has_media("mp3"))
 191     has_mp3_file.short_description = 'MP3'
 192     has_mp3_file.boolean = True
 193
 194     def has_ogg_file(self):
 195         return bool(self.has_media("ogg"))
 196     has_ogg_file.short_description = 'OGG'
 197     has_ogg_file.boolean = True
 198
 199     def has_daisy_file(self):
 200         return bool(self.has_media("daisy"))
 201     has_daisy_file.short_description = 'DAISY'
 202     has_daisy_file.boolean = True
 203
 204     def wldocument(self, parse_dublincore=True, inherit=True):
 205         from catalogue.import_utils import ORMDocProvider
 206         from librarian.parser import WLDocument
 207
 208         if inherit and self.parent:
 209             meta_fallbacks = self.parent.cover_info()
 210         else:
 211             meta_fallbacks = None
 212
 213         return WLDocument.from_file(self.xml_file.path,
 214                 provider=ORMDocProvider(self),
 215                 parse_dublincore=parse_dublincore,
 216                 meta_fallbacks=meta_fallbacks)
 217
 218     @staticmethod
 219     def zip_format(format_):
 220         def pretty_file_name(book):
 221             return "%s/%s.%s" % (
 222                 book.extra_info['author'],
 223                 book.slug,
 224                 format_)
 225
 226         field_name = "%s_file" % format_
 227         books = Book.objects.filter(parent=None).exclude(**{field_name: ""})
 228         paths = [(pretty_file_name(b), getattr(b, field_name).path)
 229                     for b in books.iterator()]
 230         return create_zip(paths, app_settings.FORMAT_ZIPS[format_])
 231
 232     def zip_audiobooks(self, format_):
 233         bm = BookMedia.objects.filter(book=self, type=format_)
 234         paths = map(lambda bm: (None, bm.file.path), bm)
 235         return create_zip(paths, "%s_%s" % (self.slug, format_))
 236
 237     def search_index(self, book_info=None, index=None, index_tags=True, commit=True):
 238         if index is None:
 239             from search.index import Index
 240             index = Index()
 241         try:
 242             index.index_book(self, book_info)
 243             if index_tags:
 244                 index.index_tags()
 245             if commit:
 246                 index.index.commit()
 247         except Exception, e:
 248             index.index.rollback()
 249             raise e
 250
 251
 252     @classmethod
 253     def from_xml_file(cls, xml_file, **kwargs):
 254         from django.core.files import File
 255         from librarian import dcparser
 256
 257         # use librarian to parse meta-data
 258         book_info = dcparser.parse(xml_file)
 259
 260         if not isinstance(xml_file, File):
 261             xml_file = File(open(xml_file))
 262
 263         try:
 264             return cls.from_text_and_meta(xml_file, book_info, **kwargs)
 265         finally:
 266             xml_file.close()
 267
 268     @classmethod
 269     def from_text_and_meta(cls, raw_file, book_info, overwrite=False,
 270             dont_build=None, search_index=True,
 271             search_index_tags=True):
 272         if dont_build is None:
 273             dont_build = set()
 274         dont_build = set.union(set(dont_build), set(app_settings.DONT_BUILD))
 275
 276         # check for parts before we do anything
 277         children = []
 278         if hasattr(book_info, 'parts'):
 279             for part_url in book_info.parts:
 280                 try:
 281                     children.append(Book.objects.get(slug=part_url.slug))
 282                 except Book.DoesNotExist:
 283                     raise Book.DoesNotExist(_('Book "%s" does not exist.') %
 284                             part_url.slug)
 285
 286         # Read book metadata
 287         book_slug = book_info.url.slug
 288         if re.search(r'[^a-z0-9-]', book_slug):
 289             raise ValueError('Invalid characters in slug')
 290         book, created = Book.objects.get_or_create(slug=book_slug)
 291
 292         if created:
 293             book_shelves = []
 294             old_cover = None
 295         else:
 296             if not overwrite:
 297                 raise Book.AlreadyExists(_('Book %s already exists') % (
 298                         book_slug))
 299             # Save shelves for this book
 300             book_shelves = list(book.tags.filter(category='set'))
 301             old_cover = book.cover_info()
 302
 303         # Save XML file
 304         book.xml_file.save('%s.xml' % book.slug, raw_file, save=False)
 305
 306         book.language = book_info.language
 307         book.title = book_info.title
 308         if book_info.variant_of:
 309             book.common_slug = book_info.variant_of.slug
 310         else:
 311             book.common_slug = book.slug
 312         book.extra_info = book_info.to_dict()
 313         book.save()
 314
 315         meta_tags = Tag.tags_from_info(book_info)
 316
 317         book.tags = set(meta_tags + book_shelves)
 318
 319         cover_changed = old_cover != book.cover_info()
 320         obsolete_children = set(b for b in book.children.all()
 321                                 if b not in children)
 322         notify_cover_changed = []
 323         for n, child_book in enumerate(children):
 324             new_child = child_book.parent != book
 325             child_book.parent = book
 326             child_book.parent_number = n
 327             child_book.save()
 328             if new_child or cover_changed:
 329                 notify_cover_changed.append(child_book)
 330         # Disown unfaithful children and let them cope on their own.
 331         for child in obsolete_children:
 332             child.parent = None
 333             child.parent_number = 0
 334             child.save()
 335             tasks.fix_tree_tags.delay(child)
 336             if old_cover:
 337                 notify_cover_changed.append(child)
 338
 339         # No saves beyond this point.
 340
 341         # Build cover.
 342         if 'cover' not in dont_build:
 343             book.cover.build_delay()
 344             book.cover_thumb.build_delay()
 345
 346         # Build HTML and ebooks.
 347         book.html_file.build_delay()
 348         if not children:
 349             for format_ in constants.EBOOK_FORMATS_WITHOUT_CHILDREN:
 350                 if format_ not in dont_build:
 351                     getattr(book, '%s_file' % format_).build_delay()
 352         for format_ in constants.EBOOK_FORMATS_WITH_CHILDREN:
 353             if format_ not in dont_build:
 354                 getattr(book, '%s_file' % format_).build_delay()
 355
 356         if not settings.NO_SEARCH_INDEX and search_index:
 357             tasks.index_book.delay(book.id, book_info=book_info, index_tags=search_index_tags)
 358
 359         for child in notify_cover_changed:
 360             child.parent_cover_changed()
 361
 362         cls.published.send(sender=book)
 363         return book
 364
 365     def fix_tree_tags(self):
 366         """Fixes the l-tags on the book's subtree.
 367
 368         Makes sure that:
 369         * the book has its parents book-tags,
 370         * its fragments have the book's and its parents book-tags,
 371         * runs those for every child book too,
 372         * touches all relevant tags,
 373         * resets tag and theme counter on the book and its ancestry.
 374         """
 375         def fix_subtree(book, parent_tags):
 376             affected_tags = set(book.tags)
 377             book.tags = list(book.tags.exclude(category='book')) + parent_tags
 378             sub_parent_tags = parent_tags + [book.book_tag()]
 379             for frag in book.fragments.all():
 380                 affected_tags.update(frag.tags)
 381                 frag.tags = list(frag.tags.exclude(category='book')
 382                                     ) + sub_parent_tags
 383             for child in book.children.all():
 384                 affected_tags.update(fix_subtree(child, sub_parent_tags))
 385             return affected_tags
 386
 387         parent_tags = []
 388         parent = self.parent
 389         while parent is not None:
 390             parent_tags.append(parent.book_tag())
 391             parent = parent.parent
 392
 393         affected_tags = fix_subtree(self, parent_tags)
 394         for tag in affected_tags:
 395             tasks.touch_tag(tag)
 396
 397         book = self
 398         while book is not None:
 399             book.reset_tag_counter()
 400             book.reset_theme_counter()
 401             book = book.parent
 402
 403     def cover_info(self, inherit=True):
 404         """Returns a dictionary to serve as fallback for BookInfo.
 405
 406         For now, the only thing inherited is the cover image.
 407         """
 408         need = False
 409         info = {}
 410         for field in ('cover_url', 'cover_by', 'cover_source'):
 411             val = self.extra_info.get(field)
 412             if val:
 413                 info[field] = val
 414             else:
 415                 need = True
 416         if inherit and need and self.parent is not None:
 417             parent_info = self.parent.cover_info()
 418             parent_info.update(info)
 419             info = parent_info
 420         return info
 421
 422     def parent_cover_changed(self):
 423         """Called when parent book's cover image is changed."""
 424         if not self.cover_info(inherit=False):
 425             if 'cover' not in app_settings.DONT_BUILD:
 426                 self.cover.build_delay()
 427                 self.cover_thumb.build_delay()
 428             for format_ in constants.EBOOK_FORMATS_WITH_COVERS:
 429                 if format_ not in app_settings.DONT_BUILD:
 430                     getattr(self, '%s_file' % format_).build_delay()
 431             for child in self.children.all():
 432                 child.parent_cover_changed()
 433
 434     def other_versions(self):
 435         """Find other versions (i.e. in other languages) of the book."""
 436         return type(self).objects.filter(common_slug=self.common_slug).exclude(pk=self.pk)
 437
 438     def related_info(self):
 439         """Keeps info about related objects (tags, media) in cache field."""
 440         if self._related_info is not None:
 441             return self._related_info
 442         else:
 443             rel = {'tags': {}, 'media': {}}
 444
 445             tags = self.tags.filter(category__in=(
 446                     'author', 'kind', 'genre', 'epoch'))
 447             tags = split_tags(tags)
 448             for category in tags:
 449                 cat = []
 450                 for tag in tags[category]:
 451                     tag_info = {'slug': tag.slug, 'name': tag.name}
 452                     for lc, ln in settings.LANGUAGES:
 453                         tag_name = getattr(tag, "name_%s" % lc)
 454                         if tag_name:
 455                             tag_info["name_%s" % lc] = tag_name
 456                     cat.append(tag_info)
 457                 rel['tags'][category] = cat
 458
 459             for media_format in BookMedia.formats:
 460                 rel['media'][media_format] = self.has_media(media_format)
 461
 462             book = self
 463             parents = []
 464             while book.parent:
 465                 parents.append((book.parent.title, book.parent.slug))
 466                 book = book.parent
 467             parents = parents[::-1]
 468             if parents:
 469                 rel['parents'] = parents
 470
 471             if self.pk:
 472                 type(self).objects.filter(pk=self.pk).update(_related_info=rel)
 473             return rel
 474
 475     def related_themes(self):
 476         theme_counter = self.theme_counter
 477         book_themes = list(Tag.objects.filter(pk__in=theme_counter.keys()))
 478         for tag in book_themes:
 479             tag.count = theme_counter[tag.pk]
 480         return book_themes
 481
 482     def reset_tag_counter(self):
 483         if self.id is None:
 484             return
 485
 486         cache_key = "Book.tag_counter/%d" % self.id
 487         permanent_cache.delete(cache_key)
 488         if self.parent:
 489             self.parent.reset_tag_counter()
 490
 491     @property
 492     def tag_counter(self):
 493         if self.id:
 494             cache_key = "Book.tag_counter/%d" % self.id
 495             tags = permanent_cache.get(cache_key)
 496         else:
 497             tags = None
 498
 499         if tags is None:
 500             tags = {}
 501             for child in self.children.all().order_by().iterator():
 502                 for tag_pk, value in child.tag_counter.iteritems():
 503                     tags[tag_pk] = tags.get(tag_pk, 0) + value
 504             for tag in self.tags.exclude(category__in=('book', 'theme', 'set')).order_by().iterator():
 505                 tags[tag.pk] = 1
 506
 507             if self.id:
 508                 permanent_cache.set(cache_key, tags)
 509         return tags
 510
 511     def reset_theme_counter(self):
 512         if self.id is None:
 513             return
 514
 515         cache_key = "Book.theme_counter/%d" % self.id
 516         permanent_cache.delete(cache_key)
 517         if self.parent:
 518             self.parent.reset_theme_counter()
 519
 520     @property
 521     def theme_counter(self):
 522         if self.id:
 523             cache_key = "Book.theme_counter/%d" % self.id
 524             tags = permanent_cache.get(cache_key)
 525         else:
 526             tags = None
 527
 528         if tags is None:
 529             tags = {}
 530             for fragment in Fragment.tagged.with_any([self.book_tag()]).order_by().iterator():
 531                 for tag in fragment.tags.filter(category='theme').order_by().iterator():
 532                     tags[tag.pk] = tags.get(tag.pk, 0) + 1
 533
 534             if self.id:
 535                 permanent_cache.set(cache_key, tags)
 536         return tags
 537
 538     def pretty_title(self, html_links=False):
 539         book = self
 540         rel_info = book.related_info()
 541         names = [(related_tag_name(tag), Tag.create_url('author', tag['slug']))
 542                     for tag in rel_info['tags'].get('author', ())]
 543         if 'parents' in rel_info:
 544             books = [(name, Book.create_url(slug))
 545                         for name, slug in rel_info['parents']]
 546             names.extend(reversed(books))
 547         names.append((self.title, self.get_absolute_url()))
 548
 549         if html_links:
 550             names = ['<a href="%s">%s</a>' % (tag[1], tag[0]) for tag in names]
 551         else:
 552             names = [tag[0] for tag in names]
 553         return ', '.join(names)
 554
 555     @classmethod
 556     def tagged_top_level(cls, tags):
 557         """ Returns top-level books tagged with `tags`.
 558
 559         It only returns those books which don't have ancestors which are
 560         also tagged with those tags.
 561
 562         """
 563         # get relevant books and their tags
 564         objects = cls.tagged.with_all(tags)
 565         parents = objects.exclude(children=None).only('slug')
 566         # eliminate descendants
 567         l_tags = Tag.objects.filter(category='book',
 568             slug__in=[book.book_tag_slug() for book in parents.iterator()])
 569         descendants_keys = [book.pk for book in cls.tagged.with_any(l_tags).only('pk').iterator()]
 570         if descendants_keys:
 571             objects = objects.exclude(pk__in=descendants_keys)
 572
 573         return objects
 574
 575     @classmethod
 576     def book_list(cls, filter=None):
 577         """Generates a hierarchical listing of all books.
 578
 579         Books are optionally filtered with a test function.
 580
 581         """
 582
 583         books_by_parent = {}
 584         books = cls.objects.all().order_by('parent_number', 'sort_key').only(
 585                 'title', 'parent', 'slug')
 586         if filter:
 587             books = books.filter(filter).distinct()
 588
 589             book_ids = set(b['pk'] for b in books.values("pk").iterator())
 590             for book in books.iterator():
 591                 parent = book.parent_id
 592                 if parent not in book_ids:
 593                     parent = None
 594                 books_by_parent.setdefault(parent, []).append(book)
 595         else:
 596             for book in books.iterator():
 597                 books_by_parent.setdefault(book.parent_id, []).append(book)
 598
 599         orphans = []
 600         books_by_author = OrderedDict()
 601         for tag in Tag.objects.filter(category='author').iterator():
 602             books_by_author[tag] = []
 603
 604         for book in books_by_parent.get(None, ()):
 605             authors = list(book.tags.filter(category='author'))
 606             if authors:
 607                 for author in authors:
 608                     books_by_author[author].append(book)
 609             else:
 610                 orphans.append(book)
 611
 612         return books_by_author, orphans, books_by_parent
 613
 614     _audiences_pl = {
 615         "SP": (1, u"szkoła podstawowa"),
 616         "SP1": (1, u"szkoła podstawowa"),
 617         "SP2": (1, u"szkoła podstawowa"),
 618         "P": (1, u"szkoła podstawowa"),
 619         "G": (2, u"gimnazjum"),
 620         "L": (3, u"liceum"),
 621         "LP": (3, u"liceum"),
 622     }
 623     def audiences_pl(self):
 624         audiences = self.extra_info.get('audiences', [])
 625         audiences = sorted(set([self._audiences_pl.get(a, (99, a)) for a in audiences]))
 626         return [a[1] for a in audiences]
 627
 628     def stage_note(self):
 629         stage = self.extra_info.get('stage')
 630         if stage and stage < '0.4':
 631             return (_('This work needs modernisation'),
 632                     reverse('infopage', args=['wymagajace-uwspolczesnienia']))
 633         else:
 634             return None, None
 635
 636     def choose_fragment(self):
 637         tag = self.book_tag()
 638         fragments = Fragment.tagged.with_any([tag])
 639         if fragments.exists():
 640             return fragments.order_by('?')[0]
 641         elif self.parent:
 642             return self.parent.choose_fragment()
 643         else:
 644             return None
 645
 646 # add the file fields
 647 for format_ in Book.formats:
 648     field_name = "%s_file" % format_
 649     # This weird globals() assignment makes Django migrations comfortable.
 650     _upload_to = _ebook_upload_to('book/%s/%%s.%s' % (format_, format_))
 651     _upload_to.__name__ = '_%s_upload_to' % format_
 652     globals()[_upload_to.__name__] = _upload_to
 653
 654     EbookField(format_, _("%s file" % format_.upper()),
 655         upload_to=_upload_to,
 656         storage=bofh_storage,
 657         max_length=255,
 658         blank=True,
 659         default=''
 660     ).contribute_to_class(Book, field_name)