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