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