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