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