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.core.urlresolvers import reverse
  12 from django.utils.datastructures import SortedDict
  13 from django.utils.translation import ugettext_lazy as _
  14 import jsonfield
  15 from fnpdjango.storage import BofhFileSystemStorage
  16 from catalogue import constants
  17 from catalogue.fields import EbookField
  18 from catalogue.models import Tag, Fragment, BookMedia
  19 from catalogue.utils import create_zip, split_tags, related_tag_name
  20 from catalogue import app_settings
  21 from catalogue import tasks
  22 from newtagging import managers
  23
  24 bofh_storage = BofhFileSystemStorage()
  25
  26 permanent_cache = get_cache('permanent')
  27
  28
  29 def _cover_upload_to(i, n):
  30     return 'book/cover/%s.jpg' % i.slug
  31 def _cover_thumb_upload_to(i, n):
  32     return 'book/cover_thumb/%s.jpg' % i.slug,
  33 def _ebook_upload_to(upload_path):
  34     def _upload_to(i, n):
  35         return upload_path % i.slug
  36     return _upload_to
  37
  38
  39 class Book(models.Model):
  40     """Represents a book imported from WL-XML."""
  41     title         = models.CharField(_('title'), max_length=120)
  42     sort_key = models.CharField(_('sort key'), max_length=120, db_index=True, editable=False)
  43     sort_key_author = models.CharField(_('sort key by author'), max_length=120, db_index=True, editable=False, default=u'')
  44     slug = models.SlugField(_('slug'), max_length=120, db_index=True,
  45             unique=True)
  46     common_slug = models.SlugField(_('slug'), max_length=120, db_index=True)
  47     language = models.CharField(_('language code'), max_length=3, db_index=True,
  48                     default=app_settings.DEFAULT_LANGUAGE)
  49     description   = models.TextField(_('description'), blank=True)
  50     created_at    = models.DateTimeField(_('creation date'), auto_now_add=True, db_index=True)
  51     changed_at    = models.DateTimeField(_('creation date'), auto_now=True, db_index=True)
  52     parent_number = models.IntegerField(_('parent number'), default=0)
  53     extra_info    = jsonfield.JSONField(_('extra information'), default={})
  54     gazeta_link   = models.CharField(blank=True, max_length=240)
  55     wiki_link     = models.CharField(blank=True, max_length=240)
  56     # files generated during publication
  57
  58     cover = EbookField('cover', _('cover'),
  59             null=True, blank=True,
  60             upload_to=_cover_upload_to,
  61             storage=bofh_storage, max_length=255)
  62     # Cleaner version of cover for thumbs
  63     cover_thumb = EbookField('cover_thumb', _('cover thumbnail'),
  64             null=True, blank=True,
  65             upload_to=_cover_thumb_upload_to,
  66             max_length=255)
  67     ebook_formats = constants.EBOOK_FORMATS
  68     formats = ebook_formats + ['html', 'xml']
  69
  70     parent = models.ForeignKey('self', blank=True, null=True,
  71         related_name='children')
  72
  73     _related_info = jsonfield.JSONField(blank=True, null=True, editable=False)
  74
  75     objects  = models.Manager()
  76     tagged   = managers.ModelTaggedItemManager(Tag)
  77     tags     = managers.TagDescriptor(Tag)
  78
  79     html_built = django.dispatch.Signal()
  80     published = django.dispatch.Signal()
  81
  82     class AlreadyExists(Exception):
  83         pass
  84
  85     class Meta:
  86         ordering = ('sort_key',)
  87         verbose_name = _('book')
  88         verbose_name_plural = _('books')
  89         app_label = 'catalogue'
  90
  91     def __unicode__(self):
  92         return self.title
  93
  94     def save(self, force_insert=False, force_update=False, reset_short_html=True, **kwargs):
  95         from sortify import sortify
  96
  97         self.sort_key = sortify(self.title)
  98         self.title = unicode(self.title) # ???
  99
 100         ret = super(Book, self).save(force_insert, force_update, **kwargs)
 101
 102         if reset_short_html:
 103             self.reset_short_html()
 104
 105         return ret
 106
 107     @permalink
 108     def get_absolute_url(self):
 109         return ('catalogue.views.book_detail', [self.slug])
 110
 111     @staticmethod
 112     @permalink
 113     def create_url(slug):
 114         return ('catalogue.views.book_detail', [slug])
 115
 116     @property
 117     def name(self):
 118         return self.title
 119
 120     def language_code(self):
 121         return constants.LANGUAGES_3TO2.get(self.language, self.language)
 122
 123     def language_name(self):
 124         return dict(settings.LANGUAGES).get(self.language_code(), "")
 125
 126     def book_tag_slug(self):
 127         return ('l-' + self.slug)[:120]
 128
 129     def book_tag(self):
 130         slug = self.book_tag_slug()
 131         book_tag, created = Tag.objects.get_or_create(slug=slug, category='book')
 132         if created:
 133             book_tag.name = self.title[:50]
 134             book_tag.sort_key = self.title.lower()
 135             book_tag.save()
 136         return book_tag
 137
 138     def has_media(self, type_):
 139         if type_ in Book.formats:
 140             return bool(getattr(self, "%s_file" % type_))
 141         else:
 142             return self.media.filter(type=type_).exists()
 143
 144     def get_media(self, type_):
 145         if self.has_media(type_):
 146             if type_ in Book.formats:
 147                 return getattr(self, "%s_file" % type_)
 148             else:
 149                 return self.media.filter(type=type_)
 150         else:
 151             return None
 152
 153     def get_mp3(self):
 154         return self.get_media("mp3")
 155     def get_odt(self):
 156         return self.get_media("odt")
 157     def get_ogg(self):
 158         return self.get_media("ogg")
 159     def get_daisy(self):
 160         return self.get_media("daisy")
 161
 162     def reset_short_html(self):
 163         if self.id is None:
 164             return
 165
 166         type(self).objects.filter(pk=self.pk).update(_related_info=None)
 167         # Fragment.short_html relies on book's tags, so reset it here too
 168         for fragm in self.fragments.all().iterator():
 169             fragm.reset_short_html()
 170
 171         try:
 172             author = self.tags.filter(category='author')[0].sort_key
 173         except IndexError:
 174             author = u''
 175         type(self).objects.filter(pk=self.pk).update(sort_key_author=author)
 176
 177
 178
 179     def has_description(self):
 180         return len(self.description) > 0
 181     has_description.short_description = _('description')
 182     has_description.boolean = True
 183
 184     # ugly ugly ugly
 185     def has_mp3_file(self):
 186         return bool(self.has_media("mp3"))
 187     has_mp3_file.short_description = 'MP3'
 188     has_mp3_file.boolean = True
 189
 190     def has_ogg_file(self):
 191         return bool(self.has_media("ogg"))
 192     has_ogg_file.short_description = 'OGG'
 193     has_ogg_file.boolean = True
 194
 195     def has_daisy_file(self):
 196         return bool(self.has_media("daisy"))
 197     has_daisy_file.short_description = 'DAISY'
 198     has_daisy_file.boolean = True
 199
 200     def wldocument(self, parse_dublincore=True, inherit=True):
 201         from catalogue.import_utils import ORMDocProvider
 202         from librarian.parser import WLDocument
 203
 204         if inherit and self.parent:
 205             meta_fallbacks = self.parent.cover_info()
 206         else:
 207             meta_fallbacks = None
 208
 209         return WLDocument.from_file(self.xml_file.path,
 210                 provider=ORMDocProvider(self),
 211                 parse_dublincore=parse_dublincore,
 212                 meta_fallbacks=meta_fallbacks)
 213
 214     @staticmethod
 215     def zip_format(format_):
 216         def pretty_file_name(book):
 217             return "%s/%s.%s" % (
 218                 book.extra_info['author'],
 219                 book.slug,
 220                 format_)
 221
 222         field_name = "%s_file" % format_
 223         books = Book.objects.filter(parent=None).exclude(**{field_name: ""})
 224         paths = [(pretty_file_name(b), getattr(b, field_name).path)
 225                     for b in books.iterator()]
 226         return create_zip(paths, app_settings.FORMAT_ZIPS[format_])
 227
 228     def zip_audiobooks(self, format_):
 229         bm = BookMedia.objects.filter(book=self, type=format_)
 230         paths = map(lambda bm: (None, bm.file.path), bm)
 231         return create_zip(paths, "%s_%s" % (self.slug, format_))
 232
 233     def search_index(self, book_info=None, index=None, index_tags=True, commit=True):
 234         import search
 235         if index is None:
 236             index = search.Index()
 237         try:
 238             index.index_book(self, book_info)
 239             if index_tags:
 240                 index.index_tags()
 241             if commit:
 242                 index.index.commit()
 243         except Exception, e:
 244             index.index.rollback()
 245             raise e
 246
 247
 248     @classmethod
 249     def from_xml_file(cls, xml_file, **kwargs):
 250         from django.core.files import File
 251         from librarian import dcparser
 252
 253         # use librarian to parse meta-data
 254         book_info = dcparser.parse(xml_file)
 255
 256         if not isinstance(xml_file, File):
 257             xml_file = File(open(xml_file))
 258
 259         try:
 260             return cls.from_text_and_meta(xml_file, book_info, **kwargs)
 261         finally:
 262             xml_file.close()
 263
 264     @classmethod
 265     def from_text_and_meta(cls, raw_file, book_info, overwrite=False,
 266             dont_build=None, search_index=True,
 267             search_index_tags=True):
 268         if dont_build is None:
 269             dont_build = set()
 270         dont_build = set.union(set(dont_build), set(app_settings.DONT_BUILD))
 271
 272         # check for parts before we do anything
 273         children = []
 274         if hasattr(book_info, 'parts'):
 275             for part_url in book_info.parts:
 276                 try:
 277                     children.append(Book.objects.get(slug=part_url.slug))
 278                 except Book.DoesNotExist:
 279                     raise Book.DoesNotExist(_('Book "%s" does not exist.') %
 280                             part_url.slug)
 281
 282         # Read book metadata
 283         book_slug = book_info.url.slug
 284         if re.search(r'[^a-z0-9-]', book_slug):
 285             raise ValueError('Invalid characters in slug')
 286         book, created = Book.objects.get_or_create(slug=book_slug)
 287
 288         if created:
 289             book_shelves = []
 290             old_cover = None
 291         else:
 292             if not overwrite:
 293                 raise Book.AlreadyExists(_('Book %s already exists') % (
 294                         book_slug))
 295             # Save shelves for this book
 296             book_shelves = list(book.tags.filter(category='set'))
 297             old_cover = book.cover_info()
 298
 299         # Save XML file
 300         book.xml_file.save('%s.xml' % book.slug, raw_file, save=False)
 301
 302         book.language = book_info.language
 303         book.title = book_info.title
 304         if book_info.variant_of:
 305             book.common_slug = book_info.variant_of.slug
 306         else:
 307             book.common_slug = book.slug
 308         book.extra_info = book_info.to_dict()
 309         book.save()
 310
 311         meta_tags = Tag.tags_from_info(book_info)
 312
 313         book.tags = set(meta_tags + book_shelves)
 314
 315         cover_changed = old_cover != book.cover_info()
 316         obsolete_children = set(b for b in book.children.all()
 317                                 if b not in children)
 318         notify_cover_changed = []
 319         for n, child_book in enumerate(children):
 320             new_child = child_book.parent != book
 321             child_book.parent = book
 322             child_book.parent_number = n
 323             child_book.save()
 324             if new_child or cover_changed:
 325                 notify_cover_changed.append(child_book)
 326         # Disown unfaithful children and let them cope on their own.
 327         for child in obsolete_children:
 328             child.parent = None
 329             child.parent_number = 0
 330             child.save()
 331             tasks.fix_tree_tags.delay(child)
 332             if old_cover:
 333                 notify_cover_changed.append(child)
 334
 335         # No saves beyond this point.
 336
 337         # Build cover.
 338         if 'cover' not in dont_build:
 339             book.cover.build_delay()
 340             book.cover_thumb.build_delay()
 341
 342         # Build HTML and ebooks.
 343         book.html_file.build_delay()
 344         if not children:
 345             for format_ in constants.EBOOK_FORMATS_WITHOUT_CHILDREN:
 346                 if format_ not in dont_build:
 347                     getattr(book, '%s_file' % format_).build_delay()
 348         for format_ in constants.EBOOK_FORMATS_WITH_CHILDREN:
 349             if format_ not in dont_build:
 350                 getattr(book, '%s_file' % format_).build_delay()
 351
 352         if not settings.NO_SEARCH_INDEX and search_index:
 353             tasks.index_book.delay(book.id, book_info=book_info, index_tags=search_index_tags)
 354
 355         for child in notify_cover_changed:
 356             child.parent_cover_changed()
 357
 358         cls.published.send(sender=book)
 359         return book
 360
 361     def fix_tree_tags(self):
 362         """Fixes the l-tags on the book's subtree.
 363
 364         Makes sure that:
 365         * the book has its parents book-tags,
 366         * its fragments have the book's and its parents book-tags,
 367         * runs those for every child book too,
 368         * touches all relevant tags,
 369         * resets tag and theme counter on the book and its ancestry.
 370         """
 371         def fix_subtree(book, parent_tags):
 372             affected_tags = set(book.tags)
 373             book.tags = list(book.tags.exclude(category='book')) + parent_tags
 374             sub_parent_tags = parent_tags + [book.book_tag()]
 375             for frag in book.fragments.all():
 376                 affected_tags.update(frag.tags)
 377                 frag.tags = list(frag.tags.exclude(category='book')
 378                                     ) + sub_parent_tags
 379             for child in book.children.all():
 380                 affected_tags.update(fix_subtree(child, sub_parent_tags))
 381             return affected_tags
 382
 383         parent_tags = []
 384         parent = self.parent
 385         while parent is not None:
 386             parent_tags.append(parent.book_tag())
 387             parent = parent.parent
 388
 389         affected_tags = fix_subtree(self, parent_tags)
 390         for tag in affected_tags:
 391             tasks.touch_tag(tag)
 392
 393         book = self
 394         while book is not None:
 395             book.reset_tag_counter()
 396             book.reset_theme_counter()
 397             book = book.parent
 398
 399     def cover_info(self, inherit=True):
 400         """Returns a dictionary to serve as fallback for BookInfo.
 401
 402         For now, the only thing inherited is the cover image.
 403         """
 404         need = False
 405         info = {}
 406         for field in ('cover_url', 'cover_by', 'cover_source'):
 407             val = self.extra_info.get(field)
 408             if val:
 409                 info[field] = val
 410             else:
 411                 need = True
 412         if inherit and need and self.parent is not None:
 413             parent_info = self.parent.cover_info()
 414             parent_info.update(info)
 415             info = parent_info
 416         return info
 417
 418     def parent_cover_changed(self):
 419         """Called when parent book's cover image is changed."""
 420         if not self.cover_info(inherit=False):
 421             if 'cover' not in app_settings.DONT_BUILD:
 422                 self.cover.build_delay()
 423                 self.cover_thumb.build_delay()
 424             for format_ in constants.EBOOK_FORMATS_WITH_COVERS:
 425                 if format_ not in app_settings.DONT_BUILD:
 426                     getattr(self, '%s_file' % format_).build_delay()
 427             for child in self.children.all():
 428                 child.parent_cover_changed()
 429
 430     def other_versions(self):
 431         """Find other versions (i.e. in other languages) of the book."""
 432         return type(self).objects.filter(common_slug=self.common_slug).exclude(pk=self.pk)
 433
 434     def related_info(self):
 435         """Keeps info about related objects (tags, media) in cache field."""
 436         if self._related_info is not None:
 437             return self._related_info
 438         else:
 439             rel = {'tags': {}, 'media': {}}
 440
 441             tags = self.tags.filter(category__in=(
 442                     'author', 'kind', 'genre', 'epoch'))
 443             tags = split_tags(tags)
 444             for category in tags:
 445                 cat = []
 446                 for tag in tags[category]:
 447                     tag_info = {'slug': tag.slug, 'name': tag.name}
 448                     for lc, ln in settings.LANGUAGES:
 449                         tag_name = getattr(tag, "name_%s" % lc)
 450                         if tag_name:
 451                             tag_info["name_%s" % lc] = tag_name
 452                     cat.append(tag_info)
 453                 rel['tags'][category] = cat
 454
 455             for media_format in BookMedia.formats:
 456                 rel['media'][media_format] = self.has_media(media_format)
 457
 458             book = self
 459             parents = []
 460             while book.parent:
 461                 parents.append((book.parent.title, book.parent.slug))
 462                 book = book.parent
 463             parents = parents[::-1]
 464             if parents:
 465                 rel['parents'] = parents
 466
 467             if self.pk:
 468                 type(self).objects.filter(pk=self.pk).update(_related_info=rel)
 469             return rel
 470
 471     def related_themes(self):
 472         theme_counter = self.theme_counter
 473         book_themes = list(Tag.objects.filter(pk__in=theme_counter.keys()))
 474         for tag in book_themes:
 475             tag.count = theme_counter[tag.pk]
 476         return book_themes
 477
 478     def reset_tag_counter(self):
 479         if self.id is None:
 480             return
 481
 482         cache_key = "Book.tag_counter/%d" % self.id
 483         permanent_cache.delete(cache_key)
 484         if self.parent:
 485             self.parent.reset_tag_counter()
 486
 487     @property
 488     def tag_counter(self):
 489         if self.id:
 490             cache_key = "Book.tag_counter/%d" % self.id
 491             tags = permanent_cache.get(cache_key)
 492         else:
 493             tags = None
 494
 495         if tags is None:
 496             tags = {}
 497             for child in self.children.all().order_by().iterator():
 498                 for tag_pk, value in child.tag_counter.iteritems():
 499                     tags[tag_pk] = tags.get(tag_pk, 0) + value
 500             for tag in self.tags.exclude(category__in=('book', 'theme', 'set')).order_by().iterator():
 501                 tags[tag.pk] = 1
 502
 503             if self.id:
 504                 permanent_cache.set(cache_key, tags)
 505         return tags
 506
 507     def reset_theme_counter(self):
 508         if self.id is None:
 509             return
 510
 511         cache_key = "Book.theme_counter/%d" % self.id
 512         permanent_cache.delete(cache_key)
 513         if self.parent:
 514             self.parent.reset_theme_counter()
 515
 516     @property
 517     def theme_counter(self):
 518         if self.id:
 519             cache_key = "Book.theme_counter/%d" % self.id
 520             tags = permanent_cache.get(cache_key)
 521         else:
 522             tags = None
 523
 524         if tags is None:
 525             tags = {}
 526             for fragment in Fragment.tagged.with_any([self.book_tag()]).order_by().iterator():
 527                 for tag in fragment.tags.filter(category='theme').order_by().iterator():
 528                     tags[tag.pk] = tags.get(tag.pk, 0) + 1
 529
 530             if self.id:
 531                 permanent_cache.set(cache_key, tags)
 532         return tags
 533
 534     def pretty_title(self, html_links=False):
 535         book = self
 536         rel_info = book.related_info()
 537         names = [(related_tag_name(tag), Tag.create_url('author', tag['slug']))
 538                     for tag in rel_info['tags'].get('author', ())]
 539         if 'parents' in rel_info:
 540             books = [(name, Book.create_url(slug))
 541                         for name, slug in rel_info['parents']]
 542             names.extend(reversed(books))
 543         names.append((self.title, self.get_absolute_url()))
 544
 545         if html_links:
 546             names = ['<a href="%s">%s</a>' % (tag[1], tag[0]) for tag in names]
 547         else:
 548             names = [tag[0] for tag in names]
 549         return ', '.join(names)
 550
 551     @classmethod
 552     def tagged_top_level(cls, tags):
 553         """ Returns top-level books tagged with `tags`.
 554
 555         It only returns those books which don't have ancestors which are
 556         also tagged with those tags.
 557
 558         """
 559         # get relevant books and their tags
 560         objects = cls.tagged.with_all(tags)
 561         parents = objects.exclude(children=None).only('slug')
 562         # eliminate descendants
 563         l_tags = Tag.objects.filter(category='book',
 564             slug__in=[book.book_tag_slug() for book in parents.iterator()])
 565         descendants_keys = [book.pk for book in cls.tagged.with_any(l_tags).only('pk').iterator()]
 566         if descendants_keys:
 567             objects = objects.exclude(pk__in=descendants_keys)
 568
 569         return objects
 570
 571     @classmethod
 572     def book_list(cls, filter=None):
 573         """Generates a hierarchical listing of all books.
 574
 575         Books are optionally filtered with a test function.
 576
 577         """
 578
 579         books_by_parent = {}
 580         books = cls.objects.all().order_by('parent_number', 'sort_key').only(
 581                 'title', 'parent', 'slug')
 582         if filter:
 583             books = books.filter(filter).distinct()
 584
 585             book_ids = set(b['pk'] for b in books.values("pk").iterator())
 586             for book in books.iterator():
 587                 parent = book.parent_id
 588                 if parent not in book_ids:
 589                     parent = None
 590                 books_by_parent.setdefault(parent, []).append(book)
 591         else:
 592             for book in books.iterator():
 593                 books_by_parent.setdefault(book.parent_id, []).append(book)
 594
 595         orphans = []
 596         books_by_author = SortedDict()
 597         for tag in Tag.objects.filter(category='author').iterator():
 598             books_by_author[tag] = []
 599
 600         for book in books_by_parent.get(None, ()):
 601             authors = list(book.tags.filter(category='author'))
 602             if authors:
 603                 for author in authors:
 604                     books_by_author[author].append(book)
 605             else:
 606                 orphans.append(book)
 607
 608         return books_by_author, orphans, books_by_parent
 609
 610     _audiences_pl = {
 611         "SP": (1, u"szkoła podstawowa"),
 612         "SP1": (1, u"szkoła podstawowa"),
 613         "SP2": (1, u"szkoła podstawowa"),
 614         "P": (1, u"szkoła podstawowa"),
 615         "G": (2, u"gimnazjum"),
 616         "L": (3, u"liceum"),
 617         "LP": (3, u"liceum"),
 618     }
 619     def audiences_pl(self):
 620         audiences = self.extra_info.get('audiences', [])
 621         audiences = sorted(set([self._audiences_pl.get(a, (99, a)) for a in audiences]))
 622         return [a[1] for a in audiences]
 623
 624     def stage_note(self):
 625         stage = self.extra_info.get('stage')
 626         if stage and stage < '0.4':
 627             return (_('This work needs modernisation'),
 628                     reverse('infopage', args=['wymagajace-uwspolczesnienia']))
 629         else:
 630             return None, None
 631
 632     def choose_fragment(self):
 633         tag = self.book_tag()
 634         fragments = Fragment.tagged.with_any([tag])
 635         if fragments.exists():
 636             return fragments.order_by('?')[0]
 637         elif self.parent:
 638             return self.parent.choose_fragment()
 639         else:
 640             return None
 641
 642
 643 # add the file fields
 644 for format_ in Book.formats:
 645     field_name = "%s_file" % format_
 646     upload_to = _ebook_upload_to('book/%s/%%s.%s' % (format_, format_))
 647     EbookField(format_, _("%s file" % format_.upper()),
 648         upload_to=upload_to,
 649         storage=bofh_storage,
 650         max_length=255,
 651         blank=True,
 652         default=''
 653     ).contribute_to_class(Book, field_name)