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                         chunk.gallery_start += gm.dest_size - gm.num_deleted
 209                         chunk.save()
 210
 211         other.delete()
 212
 213
 214     @transaction.commit_on_success
 215     def prepend_history(self, other):
 216         """Prepend history from all the other book's chunks to own."""
 217         assert self != other
 218
 219         for i in range(len(self), len(other)):
 220             title = u"pusta część %d" % i
 221             chunk = self.add(slughifi(title), title)
 222             chunk.commit('')
 223
 224         for i in range(len(other)):
 225             self[i].prepend_history(other[0])
 226
 227         assert not other.chunk_set.exists()
 228         other.delete()
 229
 230     def split(self):
 231         """Splits all the chunks into separate books."""
 232         self.title
 233         for chunk in self:
 234             book = Book.objects.create(title=chunk.title, slug=chunk.slug,
 235                     public=self.public, gallery=self.gallery)
 236             book[0].delete()
 237             chunk.book = book
 238             chunk.number = 1
 239             chunk.save()
 240         assert not self.chunk_set.exists()
 241         self.delete()
 242
 243     # State & cache
 244     # =============
 245
 246     def last_published(self):
 247         try:
 248             return self.publish_log.all()[0].timestamp
 249         except IndexError:
 250             return None
 251
 252     def assert_publishable(self):
 253         assert self.chunk_set.exists(), _('No chunks in the book.')
 254         try:
 255             changes = self.get_current_changes(publishable=True)
 256         except self.NoTextError:
 257             raise AssertionError(_('Not all chunks have publishable revisions.'))
 258         book_xml = self.materialize(changes=changes)
 259
 260         from librarian.dcparser import BookInfo
 261         from librarian import NoDublinCore, ParseError, ValidationError
 262
 263         try:
 264             bi = BookInfo.from_string(book_xml.encode('utf-8'), strict=True)
 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 hidden(self):
 276         return self.slug.startswith('.')
 277
 278     def is_new_publishable(self):
 279         """Checks if book is ready for publishing.
 280
 281         Returns True if there is a publishable version newer than the one
 282         already published.
 283
 284         """
 285         new_publishable = False
 286         if not self.chunk_set.exists():
 287             return False
 288         for chunk in self:
 289             change = chunk.publishable()
 290             if not change:
 291                 return False
 292             if not new_publishable and not change.publish_log.exists():
 293                 new_publishable = True
 294         return new_publishable
 295     new_publishable = cached_in_field('_new_publishable')(is_new_publishable)
 296
 297     def is_published(self):
 298         return self.publish_log.exists()
 299     published = cached_in_field('_published')(is_published)
 300
 301     def get_on_track(self):
 302         if self.published:
 303             return -1
 304         stages = [ch.stage.ordering if ch.stage is not None else 0
 305                     for ch in self]
 306         if not len(stages):
 307             return 0
 308         return min(stages)
 309     on_track = cached_in_field('_on_track')(get_on_track)
 310
 311     def is_single(self):
 312         return len(self) == 1
 313     single = cached_in_field('_single')(is_single)
 314
 315     @cached_in_field('_short_html')
 316     def short_html(self):
 317         return render_to_string('catalogue/book_list/book.html', {'book': self})
 318
 319     def book_info(self, publishable=True):
 320         try:
 321             book_xml = self.materialize(publishable=publishable)
 322         except self.NoTextError:
 323             pass
 324         else:
 325             from librarian.dcparser import BookInfo
 326             from librarian import NoDublinCore, ParseError, ValidationError
 327             try:
 328                 return BookInfo.from_string(book_xml.encode('utf-8'))
 329             except (self.NoTextError, ParseError, NoDublinCore, ValidationError):
 330                 return None
 331
 332     def refresh_dc_cache(self):
 333         update = {
 334             'dc_slug': None,
 335             'dc_cover_image': None,
 336         }
 337
 338         info = self.book_info()
 339         if info is not None:
 340             update['dc_slug'] = info.url.slug
 341             if info.cover_source:
 342                 try:
 343                     image = Image.objects.get(pk=int(info.cover_source.rstrip('/').rsplit('/', 1)[-1]))
 344                 except:
 345                     pass
 346                 else:
 347                     if info.cover_source == image.get_full_url():
 348                         update['dc_cover_image'] = image
 349         Book.objects.filter(pk=self.pk).update(**update)
 350
 351     def touch(self):
 352         # this should only really be done when text or publishable status changes
 353         book_content_updated.delay(self)
 354
 355         update = {
 356             "_new_publishable": self.is_new_publishable(),
 357             "_published": self.is_published(),
 358             "_single": self.is_single(),
 359             "_on_track": self.get_on_track(),
 360             "_short_html": None,
 361         }
 362         Book.objects.filter(pk=self.pk).update(**update)
 363         refresh_instance(self)
 364
 365     def refresh(self):
 366         """This should be done offline."""
 367         self.short_html
 368         self.single
 369         self.new_publishable
 370         self.published
 371
 372     # Materializing & publishing
 373     # ==========================
 374
 375     def get_current_changes(self, publishable=True):
 376         """
 377             Returns a list containing one Change for every Chunk in the Book.
 378             Takes the most recent revision (publishable, if set).
 379             Throws an error, if a proper revision is unavailable for a Chunk.
 380         """
 381         if publishable:
 382             changes = [chunk.publishable() for chunk in self]
 383         else:
 384             changes = [chunk.head for chunk in self if chunk.head is not None]
 385         if None in changes:
 386             raise self.NoTextError('Some chunks have no available text.')
 387         return changes
 388
 389     def materialize(self, publishable=False, changes=None):
 390         """
 391             Get full text of the document compiled from chunks.
 392             Takes the current versions of all texts
 393             or versions most recently tagged for publishing,
 394             or a specified iterable changes.
 395         """
 396         if changes is None:
 397             changes = self.get_current_changes(publishable)
 398         return compile_text(change.materialize() for change in changes)
 399
 400     def wldocument(self, publishable=True, changes=None, parse_dublincore=True):
 401         from catalogue.ebook_utils import RedakcjaDocProvider
 402         from librarian.parser import WLDocument
 403
 404         return WLDocument.from_string(
 405                 self.materialize(publishable=publishable, changes=changes),
 406                 provider=RedakcjaDocProvider(publishable=publishable),
 407                 parse_dublincore=parse_dublincore)
 408
 409     def publish(self, user):
 410         """
 411             Publishes a book on behalf of a (local) user.
 412         """
 413         self.assert_publishable()
 414         changes = self.get_current_changes(publishable=True)
 415         book_xml = self.materialize(changes=changes)
 416         apiclient.api_call(user, "books/", {"book_xml": book_xml})
 417         # record the publish
 418         br = BookPublishRecord.objects.create(book=self, user=user)
 419         for c in changes:
 420             ChunkPublishRecord.objects.create(book_record=br, change=c)
 421         post_publish.send(sender=br)