#880: hide copyrighted stuff,
[redakcja.git] / apps / catalogue / models / book.py
1 # -*- coding: utf-8 -*-
2 #
3 # This file is part of FNP-Redakcja, licensed under GNU Affero GPLv3 or later.
4 # Copyright © Fundacja Nowoczesna Polska. See NOTICE for more information.
5 #
6 from django.contrib.sites.models import Site
7 from django.db import models
8 from django.template.loader import render_to_string
9 from django.utils.translation import ugettext_lazy as _
10 from slughifi import slughifi
11 from librarian import NoDublinCore, ParseError, ValidationError
12 from librarian.dcparser import BookInfo
13
14 import apiclient
15 from catalogue.helpers import cached_in_field
16 from catalogue.models import BookPublishRecord, ChunkPublishRecord
17 from catalogue.signals import post_publish
18 from catalogue.tasks import refresh_instance
19 from catalogue.xml_tools import compile_text, split_xml
20
21
22 class Book(models.Model):
23     """ A document edited on the wiki """
24
25     title = models.CharField(_('title'), max_length=255, db_index=True)
26     slug = models.SlugField(_('slug'), max_length=128, unique=True, db_index=True)
27     public = models.BooleanField(_('public'), default=True, db_index=True)
28     gallery = models.CharField(_('scan gallery name'), max_length=255, blank=True)
29
30     #wl_slug = models.CharField(_('title'), max_length=255, null=True, db_index=True, editable=False)
31     parent = models.ForeignKey('self', null=True, blank=True, verbose_name=_('parent'), related_name="children", editable=False)
32     parent_number = models.IntegerField(_('parent number'), null=True, blank=True, db_index=True, editable=False)
33
34     # Cache
35     _short_html = models.TextField(null=True, blank=True, editable=False)
36     _single = models.NullBooleanField(editable=False, db_index=True)
37     _new_publishable = models.NullBooleanField(editable=False)
38     _published = models.NullBooleanField(editable=False)
39
40     class NoTextError(BaseException):
41         pass
42
43     class Meta:
44         app_label = 'catalogue'
45         ordering = ['title', 'slug']
46         verbose_name = _('book')
47         verbose_name_plural = _('books')
48
49
50     # Representing
51     # ============
52
53     def __iter__(self):
54         return iter(self.chunk_set.all())
55
56     def __getitem__(self, chunk):
57         return self.chunk_set.all()[chunk]
58
59     def __len__(self):
60         return self.chunk_set.count()
61
62     def __nonzero__(self):
63         """
64             Necessary so that __len__ isn't used for bool evaluation.
65         """
66         return True
67
68     def __unicode__(self):
69         return self.title
70
71     @models.permalink
72     def get_absolute_url(self):
73         return ("catalogue_book", [self.slug])
74
75
76     # Creating & manipulating
77     # =======================
78
79     def accessible(self, request):
80         return self.public or request.user.is_authenticated()
81
82     @classmethod
83     def create(cls, creator, text, *args, **kwargs):
84         b = cls.objects.create(*args, **kwargs)
85         b.chunk_set.all().update(creator=creator)
86         b[0].commit(text, author=creator)
87         return b
88
89     def add(self, *args, **kwargs):
90         """Add a new chunk at the end."""
91         return self.chunk_set.reverse()[0].split(*args, **kwargs)
92
93     @classmethod
94     def import_xml_text(cls, text=u'', previous_book=None,
95                 commit_args=None, **kwargs):
96         """Imports a book from XML, splitting it into chunks as necessary."""
97         texts = split_xml(text)
98         if previous_book:
99             instance = previous_book
100         else:
101             instance = cls(**kwargs)
102             instance.save()
103
104         # if there are more parts, set the rest to empty strings
105         book_len = len(instance)
106         for i in range(book_len - len(texts)):
107             texts.append((u'pusta część %d' % (i + 1), u''))
108
109         i = 0
110         for i, (title, text) in enumerate(texts):
111             if not title:
112                 title = u'część %d' % (i + 1)
113
114             slug = slughifi(title)
115
116             if i < book_len:
117                 chunk = instance[i]
118                 chunk.slug = slug[:50]
119                 chunk.title = title[:255]
120                 chunk.save()
121             else:
122                 chunk = instance.add(slug, title)
123
124             chunk.commit(text, **commit_args)
125
126         return instance
127
128     def make_chunk_slug(self, proposed):
129         """ 
130             Finds a chunk slug not yet used in the book.
131         """
132         slugs = set(c.slug for c in self)
133         i = 1
134         new_slug = proposed[:50]
135         while new_slug in slugs:
136             new_slug = "%s_%d" % (proposed[:45], i)
137             i += 1
138         return new_slug
139
140     def append(self, other, slugs=None, titles=None):
141         """Add all chunks of another book to self."""
142         assert self != other
143
144         number = self[len(self) - 1].number + 1
145         len_other = len(other)
146         single = len_other == 1
147
148         if slugs is not None:
149             assert len(slugs) == len_other
150         if titles is not None:
151             assert len(titles) == len_other
152             if slugs is None:
153                 slugs = [slughifi(t) for t in titles]
154
155         for i, chunk in enumerate(other):
156             # move chunk to new book
157             chunk.book = self
158             chunk.number = number
159
160             if titles is None:
161                 # try some title guessing
162                 if other.title.startswith(self.title):
163                     other_title_part = other.title[len(self.title):].lstrip(' /')
164                 else:
165                     other_title_part = other.title
166
167                 if single:
168                     # special treatment for appending one-parters:
169                     # just use the guessed title and original book slug
170                     chunk.title = other_title_part
171                     if other.slug.startswith(self.slug):
172                         chunk.slug = other.slug[len(self.slug):].lstrip('-_')
173                     else:
174                         chunk.slug = other.slug
175                 else:
176                     chunk.title = ("%s, %s" % (other_title_part, chunk.title))[:255]
177             else:
178                 chunk.slug = slugs[i]
179                 chunk.title = titles[i]
180
181             chunk.slug = self.make_chunk_slug(chunk.slug)
182             chunk.save()
183             number += 1
184         assert not other.chunk_set.exists()
185         other.delete()
186
187
188     # State & cache
189     # =============
190
191     def last_published(self):
192         try:
193             return self.publish_log.all()[0].timestamp
194         except IndexError:
195             return None
196
197     def assert_publishable(self):
198         assert self.chunk_set.exists(), _('No chunks in the book.')
199         try:
200             changes = self.get_current_changes(publishable=True)
201         except self.NoTextError:
202             raise AssertionError(_('Not all chunks have publishable revisions.'))
203         book_xml = self.materialize(changes=changes)
204
205         try:
206             bi = BookInfo.from_string(book_xml.encode('utf-8'))
207         except ParseError, e:
208             raise AssertionError(_('Invalid XML') + ': ' + str(e))
209         except NoDublinCore:
210             raise AssertionError(_('No Dublin Core found.'))
211         except ValidationError, e:
212             raise AssertionError(_('Invalid Dublin Core') + ': ' + str(e))
213
214         valid_about = "http://%s%s" % (Site.objects.get_current().domain, self.get_absolute_url())
215         assert bi.about == valid_about, _("rdf:about is not") + " " + valid_about
216
217     def hidden(self):
218         return self.slug.startswith('.')
219
220     def is_new_publishable(self):
221         """Checks if book is ready for publishing.
222
223         Returns True if there is a publishable version newer than the one
224         already published.
225
226         """
227         new_publishable = False
228         if not self.chunk_set.exists():
229             return False
230         for chunk in self:
231             change = chunk.publishable()
232             if not change:
233                 return False
234             if not new_publishable and not change.publish_log.exists():
235                 new_publishable = True
236         return new_publishable
237     new_publishable = cached_in_field('_new_publishable')(is_new_publishable)
238
239     def is_published(self):
240         return self.publish_log.exists()
241     published = cached_in_field('_published')(is_published)
242
243     def is_single(self):
244         return len(self) == 1
245     single = cached_in_field('_single')(is_single)
246
247     @cached_in_field('_short_html')
248     def short_html(self):
249         return render_to_string('catalogue/book_list/book.html', {'book': self})
250
251     def touch(self):
252         update = {
253             "_new_publishable": self.is_new_publishable(),
254             "_published": self.is_published(),
255             "_single": self.is_single(),
256             "_short_html": None,
257         }
258         Book.objects.filter(pk=self.pk).update(**update)
259         refresh_instance(self)
260
261     def refresh(self):
262         """This should be done offline."""
263         self.short_html
264         self.single
265         self.new_publishable
266         self.published
267
268     # Materializing & publishing
269     # ==========================
270
271     def get_current_changes(self, publishable=True):
272         """
273             Returns a list containing one Change for every Chunk in the Book.
274             Takes the most recent revision (publishable, if set).
275             Throws an error, if a proper revision is unavailable for a Chunk.
276         """
277         if publishable:
278             changes = [chunk.publishable() for chunk in self]
279         else:
280             changes = [chunk.head for chunk in self if chunk.head is not None]
281         if None in changes:
282             raise self.NoTextError('Some chunks have no available text.')
283         return changes
284
285     def materialize(self, publishable=False, changes=None):
286         """ 
287             Get full text of the document compiled from chunks.
288             Takes the current versions of all texts
289             or versions most recently tagged for publishing,
290             or a specified iterable changes.
291         """
292         if changes is None:
293             changes = self.get_current_changes(publishable)
294         return compile_text(change.materialize() for change in changes)
295
296     def publish(self, user):
297         """
298             Publishes a book on behalf of a (local) user.
299         """
300         self.assert_publishable()
301         changes = self.get_current_changes(publishable=True)
302         book_xml = self.materialize(changes=changes)
303         apiclient.api_call(user, "books/", {"book_xml": book_xml})
304         # record the publish
305         br = BookPublishRecord.objects.create(book=self, user=user)
306         for c in changes:
307             ChunkPublishRecord.objects.create(book_record=br, change=c)
308         post_publish.send(sender=br)