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