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