1524c34566fe4c24b56f9117e6b45fe65acac3ba
[wolnelektury.git] / src / 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 from collections import OrderedDict
6 from random import randint
7 import re
8 from django.conf import settings
9 from django.db import connection, models, transaction
10 from django.db.models import permalink
11 import django.dispatch
12 from django.contrib.contenttypes.fields import GenericRelation
13 from django.core.urlresolvers import reverse
14 from django.utils.translation import ugettext_lazy as _
15 import jsonfield
16 from fnpdjango.storage import BofhFileSystemStorage
17 from ssify import flush_ssi_includes
18 from newtagging import managers
19 from catalogue import constants
20 from catalogue.fields import EbookField
21 from catalogue.models import Tag, Fragment, BookMedia
22 from catalogue.utils import create_zip
23 from catalogue import app_settings
24 from catalogue import tasks
25
26 bofh_storage = BofhFileSystemStorage()
27
28
29 def _cover_upload_to(i, n):
30     return 'book/cover/%s.jpg' % i.slug
31
32 def _cover_thumb_upload_to(i, n):
33     return 'book/cover_thumb/%s.jpg' % i.slug
34
35 def _ebook_upload_to(upload_path):
36     def _upload_to(i, n):
37         return upload_path % i.slug
38     return _upload_to
39
40
41 class Book(models.Model):
42     """Represents a book imported from WL-XML."""
43     title         = models.CharField(_('title'), max_length=32767)
44     sort_key = models.CharField(_('sort key'), max_length=120, db_index=True, editable=False)
45     sort_key_author = models.CharField(_('sort key by author'), max_length=120, db_index=True, editable=False, default=u'')
46     slug = models.SlugField(_('slug'), max_length=120, db_index=True,
47             unique=True)
48     common_slug = models.SlugField(_('slug'), max_length=120, db_index=True)
49     language = models.CharField(_('language code'), max_length=3, db_index=True,
50                     default=app_settings.DEFAULT_LANGUAGE)
51     description   = models.TextField(_('description'), blank=True)
52     created_at    = models.DateTimeField(_('creation date'), auto_now_add=True, db_index=True)
53     changed_at    = models.DateTimeField(_('creation date'), auto_now=True, db_index=True)
54     parent_number = models.IntegerField(_('parent number'), default=0)
55     extra_info    = jsonfield.JSONField(_('extra information'), default={})
56     gazeta_link   = models.CharField(blank=True, max_length=240)
57     wiki_link     = models.CharField(blank=True, max_length=240)
58
59     # files generated during publication
60     cover = EbookField('cover', _('cover'),
61             null=True, blank=True,
62             upload_to=_cover_upload_to,
63             storage=bofh_storage, max_length=255)
64     # Cleaner version of cover for thumbs
65     cover_thumb = EbookField('cover_thumb', _('cover thumbnail'),
66             null=True, blank=True,
67             upload_to=_cover_thumb_upload_to,
68             max_length=255)
69     ebook_formats = constants.EBOOK_FORMATS
70     formats = ebook_formats + ['html', 'xml']
71
72     parent = models.ForeignKey('self', blank=True, null=True,
73         related_name='children')
74     ancestor = models.ManyToManyField('self', blank=True,
75         editable=False, related_name='descendant', symmetrical=False)
76
77     objects  = models.Manager()
78     tagged   = managers.ModelTaggedItemManager(Tag)
79     tags     = managers.TagDescriptor(Tag)
80     tag_relations = GenericRelation(Tag.intermediary_table_model)
81
82     html_built = django.dispatch.Signal()
83     published = django.dispatch.Signal()
84
85     short_html_url_name = 'catalogue_book_short'
86
87     class AlreadyExists(Exception):
88         pass
89
90     class Meta:
91         ordering = ('sort_key',)
92         verbose_name = _('book')
93         verbose_name_plural = _('books')
94         app_label = 'catalogue'
95
96     def __unicode__(self):
97         return self.title
98
99     def get_initial(self):
100         try:
101             return re.search(r'\w', self.title, re.U).group(0)
102         except AttributeError:
103             return None
104
105     def author_str(self):
106         return ", ".join(str(t) for t in self.tags.filter(category='author'))
107
108     def save(self, force_insert=False, force_update=False, **kwargs):
109         from sortify import sortify
110
111         self.sort_key = sortify(self.title)[:120]
112         self.title = unicode(self.title) # ???
113
114         try:
115             author = self.tags.filter(category='author')[0].sort_key
116         except IndexError:
117             author = u''
118         self.sort_key_author = author
119
120         ret = super(Book, self).save(force_insert, force_update, **kwargs)
121
122         return ret
123
124     @permalink
125     def get_absolute_url(self):
126         return ('catalogue.views.book_detail', [self.slug])
127
128     @staticmethod
129     @permalink
130     def create_url(slug):
131         return ('catalogue.views.book_detail', [slug])
132
133     @property
134     def name(self):
135         return self.title
136
137     def language_code(self):
138         return constants.LANGUAGES_3TO2.get(self.language, self.language)
139
140     def language_name(self):
141         return dict(settings.LANGUAGES).get(self.language_code(), "")
142
143     def has_media(self, type_):
144         if type_ in Book.formats:
145             return bool(getattr(self, "%s_file" % type_))
146         else:
147             return self.media.filter(type=type_).exists()
148
149     def get_media(self, type_):
150         if self.has_media(type_):
151             if type_ in Book.formats:
152                 return getattr(self, "%s_file" % type_)
153             else:
154                 return self.media.filter(type=type_)
155         else:
156             return None
157
158     def get_mp3(self):
159         return self.get_media("mp3")
160     def get_odt(self):
161         return self.get_media("odt")
162     def get_ogg(self):
163         return self.get_media("ogg")
164     def get_daisy(self):
165         return self.get_media("daisy")
166
167     def has_description(self):
168         return len(self.description) > 0
169     has_description.short_description = _('description')
170     has_description.boolean = True
171
172     # ugly ugly ugly
173     def has_mp3_file(self):
174         return bool(self.has_media("mp3"))
175     has_mp3_file.short_description = 'MP3'
176     has_mp3_file.boolean = True
177
178     def has_ogg_file(self):
179         return bool(self.has_media("ogg"))
180     has_ogg_file.short_description = 'OGG'
181     has_ogg_file.boolean = True
182
183     def has_daisy_file(self):
184         return bool(self.has_media("daisy"))
185     has_daisy_file.short_description = 'DAISY'
186     has_daisy_file.boolean = True
187
188     def wldocument(self, parse_dublincore=True, inherit=True):
189         from catalogue.import_utils import ORMDocProvider
190         from librarian.parser import WLDocument
191
192         if inherit and self.parent:
193             meta_fallbacks = self.parent.cover_info()
194         else:
195             meta_fallbacks = None
196
197         return WLDocument.from_file(self.xml_file.path,
198                 provider=ORMDocProvider(self),
199                 parse_dublincore=parse_dublincore,
200                 meta_fallbacks=meta_fallbacks)
201
202     @staticmethod
203     def zip_format(format_):
204         def pretty_file_name(book):
205             return "%s/%s.%s" % (
206                 book.extra_info['author'],
207                 book.slug,
208                 format_)
209
210         field_name = "%s_file" % format_
211         books = Book.objects.filter(parent=None).exclude(**{field_name: ""})
212         paths = [(pretty_file_name(b), getattr(b, field_name).path)
213                     for b in books.iterator()]
214         return create_zip(paths, app_settings.FORMAT_ZIPS[format_])
215
216     def zip_audiobooks(self, format_):
217         bm = BookMedia.objects.filter(book=self, type=format_)
218         paths = map(lambda bm: (None, bm.file.path), bm)
219         return create_zip(paths, "%s_%s" % (self.slug, format_))
220
221     def search_index(self, book_info=None, index=None, index_tags=True, commit=True):
222         if index is None:
223             from search.index import Index
224             index = Index()
225         try:
226             index.index_book(self, book_info)
227             if index_tags:
228                 index.index_tags()
229             if commit:
230                 index.index.commit()
231         except Exception, e:
232             index.index.rollback()
233             raise e
234
235
236     @classmethod
237     def from_xml_file(cls, xml_file, **kwargs):
238         from django.core.files import File
239         from librarian import dcparser
240
241         # use librarian to parse meta-data
242         book_info = dcparser.parse(xml_file)
243
244         if not isinstance(xml_file, File):
245             xml_file = File(open(xml_file))
246
247         try:
248             return cls.from_text_and_meta(xml_file, book_info, **kwargs)
249         finally:
250             xml_file.close()
251
252     @classmethod
253     def from_text_and_meta(cls, raw_file, book_info, overwrite=False,
254             dont_build=None, search_index=True,
255             search_index_tags=True):
256         if dont_build is None:
257             dont_build = set()
258         dont_build = set.union(set(dont_build), set(app_settings.DONT_BUILD))
259
260         # check for parts before we do anything
261         children = []
262         if hasattr(book_info, 'parts'):
263             for part_url in book_info.parts:
264                 try:
265                     children.append(Book.objects.get(slug=part_url.slug))
266                 except Book.DoesNotExist:
267                     raise Book.DoesNotExist(_('Book "%s" does not exist.') %
268                             part_url.slug)
269
270         # Read book metadata
271         book_slug = book_info.url.slug
272         if re.search(r'[^a-z0-9-]', book_slug):
273             raise ValueError('Invalid characters in slug')
274         book, created = Book.objects.get_or_create(slug=book_slug)
275
276         if created:
277             book_shelves = []
278             old_cover = None
279         else:
280             if not overwrite:
281                 raise Book.AlreadyExists(_('Book %s already exists') % (
282                         book_slug))
283             # Save shelves for this book
284             book_shelves = list(book.tags.filter(category='set'))
285             old_cover = book.cover_info()
286
287         # Save XML file
288         book.xml_file.save('%s.xml' % book.slug, raw_file, save=False)
289
290         book.language = book_info.language
291         book.title = book_info.title
292         if book_info.variant_of:
293             book.common_slug = book_info.variant_of.slug
294         else:
295             book.common_slug = book.slug
296         book.extra_info = book_info.to_dict()
297         book.save()
298
299         meta_tags = Tag.tags_from_info(book_info)
300
301         book.tags = set(meta_tags + book_shelves)
302
303         cover_changed = old_cover != book.cover_info()
304         obsolete_children = set(b for b in book.children.all()
305                                 if b not in children)
306         notify_cover_changed = []
307         for n, child_book in enumerate(children):
308             new_child = child_book.parent != book
309             child_book.parent = book
310             child_book.parent_number = n
311             child_book.save()
312             if new_child or cover_changed:
313                 notify_cover_changed.append(child_book)
314         # Disown unfaithful children and let them cope on their own.
315         for child in obsolete_children:
316             child.parent = None
317             child.parent_number = 0
318             child.save()
319             if old_cover:
320                 notify_cover_changed.append(child)
321
322         cls.repopulate_ancestors()
323         tasks.update_counters.delay()
324
325         # No saves beyond this point.
326
327         # Build cover.
328         if 'cover' not in dont_build:
329             book.cover.build_delay()
330             book.cover_thumb.build_delay()
331
332         # Build HTML and ebooks.
333         book.html_file.build_delay()
334         if not children:
335             for format_ in constants.EBOOK_FORMATS_WITHOUT_CHILDREN:
336                 if format_ not in dont_build:
337                     getattr(book, '%s_file' % format_).build_delay()
338         for format_ in constants.EBOOK_FORMATS_WITH_CHILDREN:
339             if format_ not in dont_build:
340                 getattr(book, '%s_file' % format_).build_delay()
341
342         if not settings.NO_SEARCH_INDEX and search_index:
343             tasks.index_book.delay(book.id, book_info=book_info, index_tags=search_index_tags)
344
345         for child in notify_cover_changed:
346             child.parent_cover_changed()
347
348         cls.published.send(sender=cls, instance=book)
349         return book
350
351     @classmethod
352     def repopulate_ancestors(cls):
353         """Fixes the ancestry cache."""
354         # TODO: table names
355         with transaction.atomic():
356             cursor = connection.cursor()
357             if connection.vendor == 'postgres':
358                 cursor.execute("TRUNCATE catalogue_book_ancestor")
359                 cursor.execute("""
360                     WITH RECURSIVE ancestry AS (
361                         SELECT book.id, book.parent_id
362                         FROM catalogue_book AS book
363                         WHERE book.parent_id IS NOT NULL
364                         UNION
365                         SELECT ancestor.id, book.parent_id
366                         FROM ancestry AS ancestor, catalogue_book AS book
367                         WHERE ancestor.parent_id = book.id
368                             AND book.parent_id IS NOT NULL
369                         )
370                     INSERT INTO catalogue_book_ancestor
371                         (from_book_id, to_book_id)
372                         SELECT id, parent_id
373                         FROM ancestry
374                         ORDER BY id;
375                     """)
376             else:
377                 cursor.execute("DELETE FROM catalogue_book_ancestor")
378                 for b in cls.objects.exclude(parent=None):
379                     parent = b.parent
380                     while parent is not None:
381                         b.ancestor.add(parent)
382                         parent = parent.parent
383
384     def flush_includes(self, languages=True):
385         if not languages:
386             return
387         if languages is True:
388             languages = [lc for (lc, _ln) in settings.LANGUAGES]
389         flush_ssi_includes([
390             template % (self.pk, lang)
391             for template in [
392                 '/katalog/b/%d/mini.%s.html',
393                 '/katalog/b/%d/mini_nolink.%s.html',
394                 '/katalog/b/%d/short.%s.html',
395                 '/katalog/b/%d/wide.%s.html',
396                 '/api/include/book/%d.%s.json',
397                 '/api/include/book/%d.%s.xml',
398                 ]
399             for lang in languages
400             ])
401
402     def cover_info(self, inherit=True):
403         """Returns a dictionary to serve as fallback for BookInfo.
404
405         For now, the only thing inherited is the cover image.
406         """
407         need = False
408         info = {}
409         for field in ('cover_url', 'cover_by', 'cover_source'):
410             val = self.extra_info.get(field)
411             if val:
412                 info[field] = val
413             else:
414                 need = True
415         if inherit and need and self.parent is not None:
416             parent_info = self.parent.cover_info()
417             parent_info.update(info)
418             info = parent_info
419         return info
420
421     def related_themes(self):
422         return Tag.objects.usage_for_queryset(
423             Fragment.objects.filter(models.Q(book=self) | models.Q(book__ancestor=self)),
424             counts=True).filter(category='theme')
425
426     def parent_cover_changed(self):
427         """Called when parent book's cover image is changed."""
428         if not self.cover_info(inherit=False):
429             if 'cover' not in app_settings.DONT_BUILD:
430                 self.cover.build_delay()
431                 self.cover_thumb.build_delay()
432             for format_ in constants.EBOOK_FORMATS_WITH_COVERS:
433                 if format_ not in app_settings.DONT_BUILD:
434                     getattr(self, '%s_file' % format_).build_delay()
435             for child in self.children.all():
436                 child.parent_cover_changed()
437
438     def other_versions(self):
439         """Find other versions (i.e. in other languages) of the book."""
440         return type(self).objects.filter(common_slug=self.common_slug).exclude(pk=self.pk)
441
442     def parents(self):
443         books = []
444         parent = self.parent
445         while parent is not None:
446             books.insert(0, parent)
447             parent = parent.parent
448         return books
449
450     def pretty_title(self, html_links=False):
451         names = [(tag.name, tag.get_absolute_url())
452             for tag in self.tags.filter(category='author')]
453         books = self.parents() + [self]
454         names.extend([(b.title, b.get_absolute_url()) for b in books])
455
456         if html_links:
457             names = ['<a href="%s">%s</a>' % (tag[1], tag[0]) for tag in names]
458         else:
459             names = [tag[0] for tag in names]
460         return ', '.join(names)
461
462     @classmethod
463     def tagged_top_level(cls, tags):
464         """ Returns top-level books tagged with `tags`.
465
466         It only returns those books which don't have ancestors which are
467         also tagged with those tags.
468
469         """
470         objects = cls.tagged.with_all(tags)
471         return objects.exclude(ancestor__in=objects)
472
473     @classmethod
474     def book_list(cls, filter=None):
475         """Generates a hierarchical listing of all books.
476
477         Books are optionally filtered with a test function.
478
479         """
480
481         books_by_parent = {}
482         books = cls.objects.all().order_by('parent_number', 'sort_key').only(
483                 'title', 'parent', 'slug')
484         if filter:
485             books = books.filter(filter).distinct()
486
487             book_ids = set(b['pk'] for b in books.values("pk").iterator())
488             for book in books.iterator():
489                 parent = book.parent_id
490                 if parent not in book_ids:
491                     parent = None
492                 books_by_parent.setdefault(parent, []).append(book)
493         else:
494             for book in books.iterator():
495                 books_by_parent.setdefault(book.parent_id, []).append(book)
496
497         orphans = []
498         books_by_author = OrderedDict()
499         for tag in Tag.objects.filter(category='author').iterator():
500             books_by_author[tag] = []
501
502         for book in books_by_parent.get(None, ()):
503             authors = list(book.tags.filter(category='author'))
504             if authors:
505                 for author in authors:
506                     books_by_author[author].append(book)
507             else:
508                 orphans.append(book)
509
510         return books_by_author, orphans, books_by_parent
511
512     _audiences_pl = {
513         "SP": (1, u"szkoła podstawowa"),
514         "SP1": (1, u"szkoła podstawowa"),
515         "SP2": (1, u"szkoła podstawowa"),
516         "P": (1, u"szkoła podstawowa"),
517         "G": (2, u"gimnazjum"),
518         "L": (3, u"liceum"),
519         "LP": (3, u"liceum"),
520     }
521     def audiences_pl(self):
522         audiences = self.extra_info.get('audiences', [])
523         audiences = sorted(set([self._audiences_pl.get(a, (99, a)) for a in audiences]))
524         return [a[1] for a in audiences]
525
526     def stage_note(self):
527         stage = self.extra_info.get('stage')
528         if stage and stage < '0.4':
529             return (_('This work needs modernisation'),
530                     reverse('infopage', args=['wymagajace-uwspolczesnienia']))
531         else:
532             return None, None
533
534     def choose_fragment(self):
535         fragments = self.fragments.order_by()
536         fragments_count = fragments.count()
537         if not fragments_count and self.children.exists():
538             fragments = Fragment.objects.filter(book__ancestor=self).order_by()
539             fragments_count = fragments.count()
540         if fragments_count:
541             return fragments[randint(0, fragments_count - 1)]
542         elif self.parent:
543             return self.parent.choose_fragment()
544         else:
545             return None
546
547
548 # add the file fields
549 for format_ in Book.formats:
550     field_name = "%s_file" % format_
551     # This weird globals() assignment makes Django migrations comfortable.
552     _upload_to = _ebook_upload_to('book/%s/%%s.%s' % (format_, format_))
553     _upload_to.__name__ = '_%s_upload_to' % format_
554     globals()[_upload_to.__name__] = _upload_to
555
556     EbookField(format_, _("%s file" % format_.upper()),
557         upload_to=_upload_to,
558         storage=bofh_storage,
559         max_length=255,
560         blank=True,
561         default=''
562     ).contribute_to_class(Book, field_name)