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