4d1b12e4cd430f575b831e604a92094987334fcb
[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, transaction
8 from django.template.loader import render_to_string
9 from django.utils.translation import ugettext_lazy as _
10 from django.conf import settings
11 from slughifi import slughifi
12
13
14 import apiclient
15 from catalogue.helpers import cached_in_field, GalleryMerger
16 from catalogue.models import BookPublishRecord, ChunkPublishRecord
17 from catalogue.signals import post_publish
18 from catalogue.tasks import refresh_instance, book_content_updated
19 from catalogue.xml_tools import compile_text, split_xml
20 from cover.models import Image
21 import os
22 import shutil
23 import re
24
25 class Book(models.Model):
26     """ A document edited on the wiki """
27
28     title = models.CharField(_('title'), max_length=255, db_index=True)
29     slug = models.SlugField(_('slug'), max_length=128, unique=True, db_index=True)
30     public = models.BooleanField(_('public'), default=True, db_index=True)
31     gallery = models.CharField(_('scan gallery name'), max_length=255, blank=True)
32
33     #wl_slug = models.CharField(_('title'), max_length=255, null=True, db_index=True, editable=False)
34     parent = models.ForeignKey('self', null=True, blank=True, verbose_name=_('parent'), related_name="children", editable=False)
35     parent_number = models.IntegerField(_('parent number'), null=True, blank=True, db_index=True, editable=False)
36
37     # Cache
38     _short_html = models.TextField(null=True, blank=True, editable=False)
39     _single = models.NullBooleanField(editable=False, db_index=True)
40     _new_publishable = models.NullBooleanField(editable=False)
41     _published = models.NullBooleanField(editable=False)
42     _on_track = models.IntegerField(null=True, blank=True, db_index=True, editable=False)
43     dc_cover_image = models.ForeignKey(Image, blank=True, null=True,
44         db_index=True, on_delete=models.SET_NULL, editable=False)
45     dc_slug = models.CharField(max_length=128, null=True, blank=True,
46             editable=False, db_index=True)
47
48     class NoTextError(BaseException):
49         pass
50
51     class Meta:
52         app_label = 'catalogue'
53         ordering = ['title', 'slug']
54         verbose_name = _('book')
55         verbose_name_plural = _('books')
56
57
58     # Representing
59     # ============
60
61     def __iter__(self):
62         return iter(self.chunk_set.all())
63
64     def __getitem__(self, chunk):
65         return self.chunk_set.all()[chunk]
66
67     def __len__(self):
68         return self.chunk_set.count()
69
70     def __nonzero__(self):
71         """
72             Necessary so that __len__ isn't used for bool evaluation.
73         """
74         return True
75
76     def __unicode__(self):
77         return self.title
78
79     @models.permalink
80     def get_absolute_url(self):
81         return ("catalogue_book", [self.slug])
82
83     def correct_about(self):
84         return "http://%s%s" % (
85             Site.objects.get_current().domain,
86             self.get_absolute_url()
87         )
88
89     # Creating & manipulating
90     # =======================
91
92     def accessible(self, request):
93         return self.public or request.user.is_authenticated()
94
95     @classmethod
96     @transaction.commit_on_success
97     def create(cls, creator, text, *args, **kwargs):
98         b = cls.objects.create(*args, **kwargs)
99         b.chunk_set.all().update(creator=creator)
100         b[0].commit(text, author=creator)
101         return b
102
103     def add(self, *args, **kwargs):
104         """Add a new chunk at the end."""
105         return self.chunk_set.reverse()[0].split(*args, **kwargs)
106
107     @classmethod
108     @transaction.commit_on_success
109     def import_xml_text(cls, text=u'', previous_book=None,
110                 commit_args=None, **kwargs):
111         """Imports a book from XML, splitting it into chunks as necessary."""
112         texts = split_xml(text)
113         if previous_book:
114             instance = previous_book
115         else:
116             instance = cls(**kwargs)
117             instance.save()
118
119         # if there are more parts, set the rest to empty strings
120         book_len = len(instance)
121         for i in range(book_len - len(texts)):
122             texts.append((u'pusta część %d' % (i + 1), u''))
123
124         i = 0
125         for i, (title, text) in enumerate(texts):
126             if not title:
127                 title = u'część %d' % (i + 1)
128
129             slug = slughifi(title)
130
131             if i < book_len:
132                 chunk = instance[i]
133                 chunk.slug = slug[:50]
134                 chunk.title = title[:255]
135                 chunk.save()
136             else:
137                 chunk = instance.add(slug, title)
138
139             chunk.commit(text, **commit_args)
140
141         return instance
142
143     def make_chunk_slug(self, proposed):
144         """ 
145             Finds a chunk slug not yet used in the book.
146         """
147         slugs = set(c.slug for c in self)
148         i = 1
149         new_slug = proposed[:50]
150         while new_slug in slugs:
151             new_slug = "%s_%d" % (proposed[:45], i)
152             i += 1
153         return new_slug
154
155     @transaction.commit_on_success
156     def append(self, other, slugs=None, titles=None):
157         """Add all chunks of another book to self."""
158         assert self != other
159
160         number = self[len(self) - 1].number + 1
161         len_other = len(other)
162         single = len_other == 1
163
164         if slugs is not None:
165             assert len(slugs) == len_other
166         if titles is not None:
167             assert len(titles) == len_other
168             if slugs is None:
169                 slugs = [slughifi(t) for t in titles]
170
171         for i, chunk in enumerate(other):
172             # move chunk to new book
173             chunk.book = self
174             chunk.number = number
175
176             if titles is None:
177                 # try some title guessing
178                 if other.title.startswith(self.title):
179                     other_title_part = other.title[len(self.title):].lstrip(' /')
180                 else:
181                     other_title_part = other.title
182
183                 if single:
184                     # special treatment for appending one-parters:
185                     # just use the guessed title and original book slug
186                     chunk.title = other_title_part
187                     if other.slug.startswith(self.slug):
188                         chunk.slug = other.slug[len(self.slug):].lstrip('-_')
189                     else:
190                         chunk.slug = other.slug
191                 else:
192                     chunk.title = ("%s, %s" % (other_title_part, chunk.title))[:255]
193             else:
194                 chunk.slug = slugs[i]
195                 chunk.title = titles[i]
196
197             chunk.slug = self.make_chunk_slug(chunk.slug)
198             chunk.save()
199             number += 1
200         assert not other.chunk_set.exists()
201
202         gm = GalleryMerger(self.gallery, other.gallery)
203         self.gallery = gm.merge()
204
205         # and move the gallery starts
206         if gm.was_merged:
207                 for chunk in self[len(self) - len_other:]:
208                         old_start = chunk.gallery_start or 1
209                         chunk.gallery_start = old_start + gm.dest_size - gm.num_deleted
210                         chunk.save()
211
212         other.delete()
213
214
215     @transaction.commit_on_success
216     def prepend_history(self, other):
217         """Prepend history from all the other book's chunks to own."""
218         assert self != other
219
220         for i in range(len(self), len(other)):
221             title = u"pusta część %d" % i
222             chunk = self.add(slughifi(title), title)
223             chunk.commit('')
224
225         for i in range(len(other)):
226             self[i].prepend_history(other[0])
227
228         assert not other.chunk_set.exists()
229         other.delete()
230
231     def split(self):
232         """Splits all the chunks into separate books."""
233         self.title
234         for chunk in self:
235             book = Book.objects.create(title=chunk.title, slug=chunk.slug,
236                     public=self.public, gallery=self.gallery)
237             book[0].delete()
238             chunk.book = book
239             chunk.number = 1
240             chunk.save()
241         assert not self.chunk_set.exists()
242         self.delete()
243
244     # State & cache
245     # =============
246
247     def last_published(self):
248         try:
249             return self.publish_log.all()[0].timestamp
250         except IndexError:
251             return None
252
253     def assert_publishable(self):
254         assert self.chunk_set.exists(), _('No chunks in the book.')
255         try:
256             changes = self.get_current_changes(publishable=True)
257         except self.NoTextError:
258             raise AssertionError(_('Not all chunks have publishable revisions.'))
259
260         from librarian import NoDublinCore, ParseError, ValidationError
261
262         try:
263             bi = self.wldocument(changes=changes, strict=True).book_info
264         except ParseError, e:
265             raise AssertionError(_('Invalid XML') + ': ' + unicode(e))
266         except NoDublinCore:
267             raise AssertionError(_('No Dublin Core found.'))
268         except ValidationError, e:
269             raise AssertionError(_('Invalid Dublin Core') + ': ' + unicode(e))
270
271         valid_about = self.correct_about()
272         assert bi.about == valid_about, _("rdf:about is not") + " " + valid_about
273
274     def publishable_error(self):
275         try:
276             return self.assert_publishable()
277         except AssertionError, e:
278             return e
279         else:
280             return None
281
282     def hidden(self):
283         return self.slug.startswith('.')
284
285     def is_new_publishable(self):
286         """Checks if book is ready for publishing.
287
288         Returns True if there is a publishable version newer than the one
289         already published.
290
291         """
292         new_publishable = False
293         if not self.chunk_set.exists():
294             return False
295         for chunk in self:
296             change = chunk.publishable()
297             if not change:
298                 return False
299             if not new_publishable and not change.publish_log.exists():
300                 new_publishable = True
301         return new_publishable
302     new_publishable = cached_in_field('_new_publishable')(is_new_publishable)
303
304     def is_published(self):
305         return self.publish_log.exists()
306     published = cached_in_field('_published')(is_published)
307
308     def get_on_track(self):
309         if self.published:
310             return -1
311         stages = [ch.stage.ordering if ch.stage is not None else 0
312                     for ch in self]
313         if not len(stages):
314             return 0
315         return min(stages)
316     on_track = cached_in_field('_on_track')(get_on_track)
317
318     def is_single(self):
319         return len(self) == 1
320     single = cached_in_field('_single')(is_single)
321
322     @cached_in_field('_short_html')
323     def short_html(self):
324         return render_to_string('catalogue/book_list/book.html', {'book': self})
325
326     def book_info(self, publishable=True):
327         try:
328             book_xml = self.materialize(publishable=publishable)
329         except self.NoTextError:
330             pass
331         else:
332             from librarian.dcparser import BookInfo
333             from librarian import NoDublinCore, ParseError, ValidationError
334             try:
335                 return BookInfo.from_string(book_xml.encode('utf-8'))
336             except (self.NoTextError, ParseError, NoDublinCore, ValidationError):
337                 return None
338
339     def refresh_dc_cache(self):
340         update = {
341             'dc_slug': None,
342             'dc_cover_image': None,
343         }
344
345         info = self.book_info()
346         if info is not None:
347             update['dc_slug'] = info.url.slug
348             if info.cover_source:
349                 try:
350                     image = Image.objects.get(pk=int(info.cover_source.rstrip('/').rsplit('/', 1)[-1]))
351                 except:
352                     pass
353                 else:
354                     if info.cover_source == image.get_full_url():
355                         update['dc_cover_image'] = image
356         Book.objects.filter(pk=self.pk).update(**update)
357
358     def touch(self):
359         # this should only really be done when text or publishable status changes
360         book_content_updated.delay(self)
361
362         update = {
363             "_new_publishable": self.is_new_publishable(),
364             "_published": self.is_published(),
365             "_single": self.is_single(),
366             "_on_track": self.get_on_track(),
367             "_short_html": None,
368         }
369         Book.objects.filter(pk=self.pk).update(**update)
370         refresh_instance(self)
371
372     def refresh(self):
373         """This should be done offline."""
374         self.short_html
375         self.single
376         self.new_publishable
377         self.published
378
379     # Materializing & publishing
380     # ==========================
381
382     def get_current_changes(self, publishable=True):
383         """
384             Returns a list containing one Change for every Chunk in the Book.
385             Takes the most recent revision (publishable, if set).
386             Throws an error, if a proper revision is unavailable for a Chunk.
387         """
388         if publishable:
389             changes = [chunk.publishable() for chunk in self]
390         else:
391             changes = [chunk.head for chunk in self if chunk.head is not None]
392         if None in changes:
393             raise self.NoTextError('Some chunks have no available text.')
394         return changes
395
396     def materialize(self, publishable=False, changes=None):
397         """ 
398             Get full text of the document compiled from chunks.
399             Takes the current versions of all texts
400             or versions most recently tagged for publishing,
401             or a specified iterable changes.
402         """
403         if changes is None:
404             changes = self.get_current_changes(publishable)
405         return compile_text(change.materialize() for change in changes)
406
407     def wldocument(self, publishable=True, changes=None, 
408             parse_dublincore=True, strict=False):
409         from catalogue.ebook_utils import RedakcjaDocProvider
410         from librarian.parser import WLDocument
411
412         return WLDocument.from_string(
413                 self.materialize(publishable=publishable, changes=changes),
414                 provider=RedakcjaDocProvider(publishable=publishable),
415                 parse_dublincore=parse_dublincore,
416                 strict=strict)
417
418     def publish(self, user):
419         """
420             Publishes a book on behalf of a (local) user.
421         """
422         self.assert_publishable()
423         changes = self.get_current_changes(publishable=True)
424         book_xml = self.materialize(changes=changes)
425         apiclient.api_call(user, "books/", {"book_xml": book_xml})
426         # record the publish
427         br = BookPublishRecord.objects.create(book=self, user=user)
428         for c in changes:
429             ChunkPublishRecord.objects.create(book_record=br, change=c)
430         post_publish.send(sender=br)