6a31f4bff8b89be65fa792f321a08407e9760e6d
[wolnelektury.git] / 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 as 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
18 from catalogue import app_settings
19 from catalogue import tasks
20 from newtagging import managers
21
22
23 permanent_cache = get_cache('permanent')
24
25
26 class Book(models.Model):
27     """Represents a book imported from WL-XML."""
28     title         = models.CharField(_('title'), max_length=120)
29     sort_key = models.CharField(_('sort key'), max_length=120, db_index=True, editable=False)
30     slug = models.SlugField(_('slug'), max_length=120, db_index=True,
31             unique=True)
32     common_slug = models.SlugField(_('slug'), max_length=120, db_index=True)
33     language = models.CharField(_('language code'), max_length=3, db_index=True,
34                     default=app_settings.DEFAULT_LANGUAGE)
35     description   = models.TextField(_('description'), blank=True)
36     created_at    = models.DateTimeField(_('creation date'), auto_now_add=True, db_index=True)
37     changed_at    = models.DateTimeField(_('creation date'), auto_now=True, db_index=True)
38     parent_number = models.IntegerField(_('parent number'), default=0)
39     extra_info    = jsonfield.JSONField(_('extra information'), default='{}')
40     gazeta_link   = models.CharField(blank=True, max_length=240)
41     wiki_link     = models.CharField(blank=True, max_length=240)
42     # files generated during publication
43
44     cover = EbookField('cover', _('cover'),
45                 upload_to=book_upload_path('jpg'), null=True, blank=True)
46     ebook_formats = constants.EBOOK_FORMATS
47     formats = ebook_formats + ['html', 'xml']
48
49     parent = models.ForeignKey('self', blank=True, null=True,
50         related_name='children')
51
52     _related_info = jsonfield.JSONField(blank=True, null=True, editable=False)
53
54     objects  = models.Manager()
55     tagged   = managers.ModelTaggedItemManager(Tag)
56     tags     = managers.TagDescriptor(Tag)
57
58     html_built = django.dispatch.Signal()
59     published = django.dispatch.Signal()
60
61     class AlreadyExists(Exception):
62         pass
63
64     class Meta:
65         ordering = ('sort_key',)
66         verbose_name = _('book')
67         verbose_name_plural = _('books')
68         app_label = 'catalogue'
69
70     def __unicode__(self):
71         return self.title
72
73     def save(self, force_insert=False, force_update=False, reset_short_html=True, **kwargs):
74         from sortify import sortify
75
76         self.sort_key = sortify(self.title)
77
78         ret = super(Book, self).save(force_insert, force_update)
79
80         if reset_short_html:
81             self.reset_short_html()
82
83         return ret
84
85     @permalink
86     def get_absolute_url(self):
87         return ('catalogue.views.book_detail', [self.slug])
88
89     @property
90     def name(self):
91         return self.title
92
93     def book_tag_slug(self):
94         return ('l-' + self.slug)[:120]
95
96     def book_tag(self):
97         slug = self.book_tag_slug()
98         book_tag, created = Tag.objects.get_or_create(slug=slug, category='book')
99         if created:
100             book_tag.name = self.title[:50]
101             book_tag.sort_key = self.title.lower()
102             book_tag.save()
103         return book_tag
104
105     def has_media(self, type_):
106         if type_ in Book.formats:
107             return bool(getattr(self, "%s_file" % type_))
108         else:
109             return self.media.filter(type=type_).exists()
110
111     def get_media(self, type_):
112         if self.has_media(type_):
113             if type_ in Book.formats:
114                 return getattr(self, "%s_file" % type_)
115             else:                                             
116                 return self.media.filter(type=type_)
117         else:
118             return None
119
120     def get_mp3(self):
121         return self.get_media("mp3")
122     def get_odt(self):
123         return self.get_media("odt")
124     def get_ogg(self):
125         return self.get_media("ogg")
126     def get_daisy(self):
127         return self.get_media("daisy")                       
128
129     def reset_short_html(self):
130         if self.id is None:
131             return
132
133         type(self).objects.filter(pk=self.pk).update(_related_info=None)
134         # Fragment.short_html relies on book's tags, so reset it here too
135         for fragm in self.fragments.all().iterator():
136             fragm.reset_short_html()
137
138     def has_description(self):
139         return len(self.description) > 0
140     has_description.short_description = _('description')
141     has_description.boolean = True
142
143     # ugly ugly ugly
144     def has_mp3_file(self):
145         return bool(self.has_media("mp3"))
146     has_mp3_file.short_description = 'MP3'
147     has_mp3_file.boolean = True
148
149     def has_ogg_file(self):
150         return bool(self.has_media("ogg"))
151     has_ogg_file.short_description = 'OGG'
152     has_ogg_file.boolean = True
153
154     def has_daisy_file(self):
155         return bool(self.has_media("daisy"))
156     has_daisy_file.short_description = 'DAISY'
157     has_daisy_file.boolean = True
158
159     def wldocument(self, parse_dublincore=True, inherit=True):
160         from catalogue.import_utils import ORMDocProvider
161         from librarian.parser import WLDocument
162
163         if inherit and self.parent:
164             meta_fallbacks = self.parent.cover_info()
165         else:
166             meta_fallbacks = None
167
168         return WLDocument.from_file(self.xml_file.path,
169                 provider=ORMDocProvider(self),
170                 parse_dublincore=parse_dublincore,
171                 meta_fallbacks=meta_fallbacks)
172
173     @staticmethod
174     def zip_format(format_):
175         def pretty_file_name(book):
176             return "%s/%s.%s" % (
177                 book.extra_info['author'],
178                 book.slug,
179                 format_)
180
181         field_name = "%s_file" % format_
182         books = Book.objects.filter(parent=None).exclude(**{field_name: ""})
183         paths = [(pretty_file_name(b), getattr(b, field_name).path)
184                     for b in books.iterator()]
185         return create_zip(paths, app_settings.FORMAT_ZIPS[format_])
186
187     def zip_audiobooks(self, format_):
188         bm = BookMedia.objects.filter(book=self, type=format_)
189         paths = map(lambda bm: (None, bm.file.path), bm)
190         return create_zip(paths, "%s_%s" % (self.slug, format_))
191
192     def search_index(self, book_info=None, reuse_index=False, index_tags=True):
193         import search
194         if reuse_index:
195             idx = search.ReusableIndex()
196         else:
197             idx = search.Index()
198             
199         idx.open()
200         try:
201             idx.index_book(self, book_info)
202             if index_tags:
203                 idx.index_tags()
204         finally:
205             idx.close()
206
207     @classmethod
208     def from_xml_file(cls, xml_file, **kwargs):
209         from django.core.files import File
210         from librarian import dcparser
211
212         # use librarian to parse meta-data
213         book_info = dcparser.parse(xml_file)
214
215         if not isinstance(xml_file, File):
216             xml_file = File(open(xml_file))
217
218         try:
219             return cls.from_text_and_meta(xml_file, book_info, **kwargs)
220         finally:
221             xml_file.close()
222
223     @classmethod
224     def from_text_and_meta(cls, raw_file, book_info, overwrite=False,
225             dont_build=None, search_index=True,
226             search_index_tags=True, search_index_reuse=False):
227         if dont_build is None:
228             dont_build = set()
229         dont_build = set.union(set(dont_build), set(app_settings.DONT_BUILD))
230
231         # check for parts before we do anything
232         children = []
233         if hasattr(book_info, 'parts'):
234             for part_url in book_info.parts:
235                 try:
236                     children.append(Book.objects.get(slug=part_url.slug))
237                 except Book.DoesNotExist:
238                     raise Book.DoesNotExist(_('Book "%s" does not exist.') %
239                             part_url.slug)
240
241         # Read book metadata
242         book_slug = book_info.url.slug
243         if re.search(r'[^a-z0-9-]', book_slug):
244             raise ValueError('Invalid characters in slug')
245         book, created = Book.objects.get_or_create(slug=book_slug)
246
247         if created:
248             book_shelves = []
249             old_cover = None
250         else:
251             if not overwrite:
252                 raise Book.AlreadyExists(_('Book %s already exists') % (
253                         book_slug))
254             # Save shelves for this book
255             book_shelves = list(book.tags.filter(category='set'))
256             old_cover = book.cover_info()
257
258         # Save XML file
259         book.xml_file.save('%s.xml' % book.slug, raw_file, save=False)
260
261         book.language = book_info.language
262         book.title = book_info.title
263         if book_info.variant_of:
264             book.common_slug = book_info.variant_of.slug
265         else:
266             book.common_slug = book.slug
267         book.extra_info = book_info.to_dict()
268         book.save()
269
270         meta_tags = Tag.tags_from_info(book_info)
271
272         book.tags = set(meta_tags + book_shelves)
273
274         cover_changed = old_cover != book.cover_info()
275         obsolete_children = set(b for b in book.children.all()
276                                 if b not in children)
277         notify_cover_changed = []
278         for n, child_book in enumerate(children):
279             new_child = child_book.parent != book
280             child_book.parent = book
281             child_book.parent_number = n
282             child_book.save()
283             if new_child or cover_changed:
284                 notify_cover_changed.append(child_book)
285         # Disown unfaithful children and let them cope on their own.
286         for child in obsolete_children:
287             child.parent = None
288             child.parent_number = 0
289             child.save()
290             tasks.fix_tree_tags.delay(child)
291             if old_cover:
292                 notify_cover_changed.append(child)
293
294         # delete old fragments when overwriting
295         book.fragments.all().delete()
296         # Build HTML, fix the tree tags, build cover.
297         has_own_text = bool(book.html_file.build())
298         tasks.fix_tree_tags.delay(book)
299         if 'cover' not in dont_build:
300             book.cover.build_delay()
301         
302         # No saves behind this point.
303
304         if has_own_text:
305             for format_ in constants.EBOOK_FORMATS_WITHOUT_CHILDREN:
306                 if format_ not in dont_build:
307                     getattr(book, '%s_file' % format_).build_delay()
308         for format_ in constants.EBOOK_FORMATS_WITH_CHILDREN:
309             if format_ not in dont_build:
310                 getattr(book, '%s_file' % format_).build_delay()
311
312         if not settings.NO_SEARCH_INDEX and search_index:
313             book.search_index(index_tags=search_index_tags, reuse_index=search_index_reuse)
314             #index_book.delay(book.id, book_info)
315
316         for child in notify_cover_changed:
317             child.parent_cover_changed()
318
319         cls.published.send(sender=book)
320         return book
321
322     def fix_tree_tags(self):
323         """Fixes the l-tags on the book's subtree.
324
325         Makes sure that:
326         * the book has its parents book-tags,
327         * its fragments have the book's and its parents book-tags,
328         * runs those for every child book too,
329         * touches all relevant tags,
330         * resets tag and theme counter on the book and its ancestry.
331         """
332         def fix_subtree(book, parent_tags):
333             affected_tags = set(book.tags)
334             book.tags = list(book.tags.exclude(category='book')) + parent_tags
335             sub_parent_tags = parent_tags + [book.book_tag()]
336             for frag in book.fragments.all():
337                 affected_tags.update(frag.tags)
338                 frag.tags = list(frag.tags.exclude(category='book')
339                                     ) + sub_parent_tags
340             for child in book.children.all():
341                 affected_tags.update(fix_subtree(child, sub_parent_tags))
342             return affected_tags
343
344         parent_tags = []
345         parent = self.parent
346         while parent is not None:
347             parent_tags.append(parent.book_tag())
348             parent = parent.parent
349
350         affected_tags = fix_subtree(self, parent_tags)
351         for tag in affected_tags:
352             tasks.touch_tag(tag)
353
354         book = self
355         while book is not None:
356             book.reset_tag_counter()
357             book.reset_theme_counter()
358             book = book.parent
359
360     def cover_info(self, inherit=True):
361         """Returns a dictionary to serve as fallback for BookInfo.
362
363         For now, the only thing inherited is the cover image.
364         """
365         need = False
366         info = {}
367         for field in ('cover_url', 'cover_by', 'cover_source'):
368             val = self.extra_info.get(field)
369             if val:
370                 info[field] = val
371             else:
372                 need = True
373         if inherit and need and self.parent is not None:
374             parent_info = self.parent.cover_info()
375             parent_info.update(info)
376             info = parent_info
377         return info
378
379     def parent_cover_changed(self):
380         """Called when parent book's cover image is changed."""
381         if not self.cover_info(inherit=False):
382             if 'cover' not in app_settings.DONT_BUILD:
383                 self.cover.build_delay()
384             for format_ in constants.EBOOK_FORMATS_WITH_COVERS:
385                 if format_ not in app_settings.DONT_BUILD:
386                     getattr(self, '%s_file' % format_).build_delay()
387             for child in self.children.all():
388                 child.parent_cover_changed()
389
390     def related_info(self):
391         """Keeps info about related objects (tags, media) in cache field."""
392         if self._related_info is not None:
393             return self._related_info
394         else:
395             rel = {'tags': {}, 'media': {}}
396
397             tags = self.tags.filter(category__in=(
398                     'author', 'kind', 'genre', 'epoch'))
399             tags = split_tags(tags)
400             for category in tags:
401                 rel['tags'][category] = [
402                         (t.name, t.slug) for t in tags[category]]
403
404             for media_format in BookMedia.formats:
405                 rel['media'][media_format] = self.has_media(media_format)
406
407             book = self
408             parents = []
409             while book.parent:
410                 parents.append((book.parent.title, book.parent.slug))
411                 book = book.parent
412             parents = parents[::-1]
413             if parents:
414                 rel['parents'] = parents
415
416             if self.pk:
417                 type(self).objects.filter(pk=self.pk).update(_related_info=rel)
418             return rel
419
420     def related_themes(self):
421         theme_counter = self.theme_counter
422         book_themes = list(Tag.objects.filter(pk__in=theme_counter.keys()))
423         for tag in book_themes:
424             tag.count = theme_counter[tag.pk]
425         return book_themes
426
427     def reset_tag_counter(self):
428         if self.id is None:
429             return
430
431         cache_key = "Book.tag_counter/%d" % self.id
432         permanent_cache.delete(cache_key)
433         if self.parent:
434             self.parent.reset_tag_counter()
435
436     @property
437     def tag_counter(self):
438         if self.id:
439             cache_key = "Book.tag_counter/%d" % self.id
440             tags = permanent_cache.get(cache_key)
441         else:
442             tags = None
443
444         if tags is None:
445             tags = {}
446             for child in self.children.all().order_by().iterator():
447                 for tag_pk, value in child.tag_counter.iteritems():
448                     tags[tag_pk] = tags.get(tag_pk, 0) + value
449             for tag in self.tags.exclude(category__in=('book', 'theme', 'set')).order_by().iterator():
450                 tags[tag.pk] = 1
451
452             if self.id:
453                 permanent_cache.set(cache_key, tags)
454         return tags
455
456     def reset_theme_counter(self):
457         if self.id is None:
458             return
459
460         cache_key = "Book.theme_counter/%d" % self.id
461         permanent_cache.delete(cache_key)
462         if self.parent:
463             self.parent.reset_theme_counter()
464
465     @property
466     def theme_counter(self):
467         if self.id:
468             cache_key = "Book.theme_counter/%d" % self.id
469             tags = permanent_cache.get(cache_key)
470         else:
471             tags = None
472
473         if tags is None:
474             tags = {}
475             for fragment in Fragment.tagged.with_any([self.book_tag()]).order_by().iterator():
476                 for tag in fragment.tags.filter(category='theme').order_by().iterator():
477                     tags[tag.pk] = tags.get(tag.pk, 0) + 1
478
479             if self.id:
480                 permanent_cache.set(cache_key, tags)
481         return tags
482
483     def pretty_title(self, html_links=False):
484         book = self
485         names = list(book.tags.filter(category='author'))
486
487         books = []
488         while book:
489             books.append(book)
490             book = book.parent
491         names.extend(reversed(books))
492
493         if html_links:
494             names = ['<a href="%s">%s</a>' % (tag.get_absolute_url(), tag.name) for tag in names]
495         else:
496             names = [tag.name for tag in names]
497
498         return ', '.join(names)
499
500     @classmethod
501     def tagged_top_level(cls, tags):
502         """ Returns top-level books tagged with `tags`.
503
504         It only returns those books which don't have ancestors which are
505         also tagged with those tags.
506
507         """
508         # get relevant books and their tags
509         objects = cls.tagged.with_all(tags)
510         # eliminate descendants
511         l_tags = Tag.objects.filter(category='book',
512             slug__in=[book.book_tag_slug() for book in objects.iterator()])
513         descendants_keys = [book.pk for book in cls.tagged.with_any(l_tags).iterator()]
514         if descendants_keys:
515             objects = objects.exclude(pk__in=descendants_keys)
516
517         return objects
518
519     @classmethod
520     def book_list(cls, filter=None):
521         """Generates a hierarchical listing of all books.
522
523         Books are optionally filtered with a test function.
524
525         """
526
527         books_by_parent = {}
528         books = cls.objects.all().order_by('parent_number', 'sort_key').only(
529                 'title', 'parent', 'slug')
530         if filter:
531             books = books.filter(filter).distinct()
532             
533             book_ids = set(b['pk'] for b in books.values("pk").iterator())
534             for book in books.iterator():
535                 parent = book.parent_id
536                 if parent not in book_ids:
537                     parent = None
538                 books_by_parent.setdefault(parent, []).append(book)
539         else:
540             for book in books.iterator():
541                 books_by_parent.setdefault(book.parent_id, []).append(book)
542
543         orphans = []
544         books_by_author = SortedDict()
545         for tag in Tag.objects.filter(category='author').iterator():
546             books_by_author[tag] = []
547
548         for book in books_by_parent.get(None,()):
549             authors = list(book.tags.filter(category='author'))
550             if authors:
551                 for author in authors:
552                     books_by_author[author].append(book)
553             else:
554                 orphans.append(book)
555
556         return books_by_author, orphans, books_by_parent
557
558     _audiences_pl = {
559         "SP1": (1, u"szkoła podstawowa"),
560         "SP2": (1, u"szkoła podstawowa"),
561         "P": (1, u"szkoła podstawowa"),
562         "G": (2, u"gimnazjum"),
563         "L": (3, u"liceum"),
564         "LP": (3, u"liceum"),
565     }
566     def audiences_pl(self):
567         audiences = self.extra_info.get('audiences', [])
568         audiences = sorted(set([self._audiences_pl[a] for a in audiences]))
569         return [a[1] for a in audiences]
570
571     def choose_fragment(self):
572         tag = self.book_tag()
573         fragments = Fragment.tagged.with_any([tag])
574         if fragments.exists():
575             return fragments.order_by('?')[0]
576         elif self.parent:
577             return self.parent.choose_fragment()
578         else:
579             return None
580
581
582 # add the file fields
583 for format_ in Book.formats:
584     field_name = "%s_file" % format_
585     EbookField(format_, _("%s file" % format_.upper()),
586             upload_to=book_upload_path(format_),
587             blank=True, default='').contribute_to_class(Book, field_name)