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