fbae111eaf89e98a9a4eeb8e8269d2bdb6af04df
[wolnelektury.git] / apps / catalogue / models.py
1 # -*- coding: utf-8 -*-
2 # This file is part of Wolnelektury, licensed under GNU Affero GPLv3 or later.
3 # Copyright © Fundacja Nowoczesna Polska. See NOTICE for more information.
4 #
5 from collections import namedtuple
6
7 from django.db import models
8 from django.db.models import permalink, Q
9 import django.dispatch
10 from django.core.cache import cache
11 from django.core.files.storage import DefaultStorage
12 from django.utils.translation import ugettext_lazy as _
13 from django.contrib.auth.models import User
14 from django.template.loader import render_to_string
15 from django.utils.datastructures import SortedDict
16 from django.utils.safestring import mark_safe
17 from django.utils.translation import get_language
18 from django.core.urlresolvers import reverse
19 from django.db.models.signals import post_save, m2m_changed, pre_delete
20
21 from django.conf import settings
22
23 from newtagging.models import TagBase, tags_updated
24 from newtagging import managers
25 from catalogue.fields import JSONField, OverwritingFileField
26 from catalogue.utils import create_zip, split_tags
27 from catalogue.tasks import touch_tag, index_book
28 from shutil import copy
29 from glob import glob
30 import re
31 from os import path
32
33
34 import search
35
36 # Those are hard-coded here so that makemessages sees them.
37 TAG_CATEGORIES = (
38     ('author', _('author')),
39     ('epoch', _('epoch')),
40     ('kind', _('kind')),
41     ('genre', _('genre')),
42     ('theme', _('theme')),
43     ('set', _('set')),
44     ('book', _('book')),
45 )
46
47 # not quite, but Django wants you to set a timeout
48 CACHE_FOREVER = 2419200  # 28 days
49
50
51 class TagSubcategoryManager(models.Manager):
52     def __init__(self, subcategory):
53         super(TagSubcategoryManager, self).__init__()
54         self.subcategory = subcategory
55
56     def get_query_set(self):
57         return super(TagSubcategoryManager, self).get_query_set().filter(category=self.subcategory)
58
59
60 class Tag(TagBase):
61     name = models.CharField(_('name'), max_length=50, db_index=True)
62     slug = models.SlugField(_('slug'), max_length=120, db_index=True)
63     sort_key = models.CharField(_('sort key'), max_length=120, db_index=True)
64     category = models.CharField(_('category'), max_length=50, blank=False, null=False,
65         db_index=True, choices=TAG_CATEGORIES)
66     description = models.TextField(_('description'), blank=True)
67
68     user = models.ForeignKey(User, blank=True, null=True)
69     book_count = models.IntegerField(_('book count'), blank=True, null=True)
70     gazeta_link = models.CharField(blank=True, max_length=240)
71     wiki_link = models.CharField(blank=True, max_length=240)
72
73     created_at    = models.DateTimeField(_('creation date'), auto_now_add=True, db_index=True)
74     changed_at    = models.DateTimeField(_('creation date'), auto_now=True, db_index=True)
75
76     class UrlDeprecationWarning(DeprecationWarning):
77         pass
78
79     categories_rev = {
80         'autor': 'author',
81         'epoka': 'epoch',
82         'rodzaj': 'kind',
83         'gatunek': 'genre',
84         'motyw': 'theme',
85         'polka': 'set',
86     }
87     categories_dict = dict((item[::-1] for item in categories_rev.iteritems()))
88
89     class Meta:
90         ordering = ('sort_key',)
91         verbose_name = _('tag')
92         verbose_name_plural = _('tags')
93         unique_together = (("slug", "category"),)
94
95     def __unicode__(self):
96         return self.name
97
98     def __repr__(self):
99         return "Tag(slug=%r)" % self.slug
100
101     @permalink
102     def get_absolute_url(self):
103         return ('catalogue.views.tagged_object_list', [self.url_chunk])
104
105     def has_description(self):
106         return len(self.description) > 0
107     has_description.short_description = _('description')
108     has_description.boolean = True
109
110     def get_count(self):
111         """Returns global book count for book tags, fragment count for themes."""
112
113         if self.category == 'book':
114             # never used
115             objects = Book.objects.none()
116         elif self.category == 'theme':
117             objects = Fragment.tagged.with_all((self,))
118         else:
119             objects = Book.tagged.with_all((self,)).order_by()
120             if self.category != 'set':
121                 # eliminate descendants
122                 l_tags = Tag.objects.filter(slug__in=[book.book_tag_slug() for book in objects])
123                 descendants_keys = [book.pk for book in Book.tagged.with_any(l_tags)]
124                 if descendants_keys:
125                     objects = objects.exclude(pk__in=descendants_keys)
126         return objects.count()
127
128     @staticmethod
129     def get_tag_list(tags):
130         if isinstance(tags, basestring):
131             real_tags = []
132             ambiguous_slugs = []
133             category = None
134             deprecated = False
135             tags_splitted = tags.split('/')
136             for name in tags_splitted:
137                 if category:
138                     real_tags.append(Tag.objects.get(slug=name, category=category))
139                     category = None
140                 elif name in Tag.categories_rev:
141                     category = Tag.categories_rev[name]
142                 else:
143                     try:
144                         real_tags.append(Tag.objects.exclude(category='book').get(slug=name))
145                         deprecated = True 
146                     except Tag.MultipleObjectsReturned, e:
147                         ambiguous_slugs.append(name)
148
149             if category:
150                 # something strange left off
151                 raise Tag.DoesNotExist()
152             if ambiguous_slugs:
153                 # some tags should be qualified
154                 e = Tag.MultipleObjectsReturned()
155                 e.tags = real_tags
156                 e.ambiguous_slugs = ambiguous_slugs
157                 raise e
158             if deprecated:
159                 e = Tag.UrlDeprecationWarning()
160                 e.tags = real_tags
161                 raise e
162             return real_tags
163         else:
164             return TagBase.get_tag_list(tags)
165
166     @property
167     def url_chunk(self):
168         return '/'.join((Tag.categories_dict[self.category], self.slug))
169
170     @staticmethod
171     def tags_from_info(info):
172         from slughifi import slughifi
173         from sortify import sortify
174         meta_tags = []
175         categories = (('kinds', 'kind'), ('genres', 'genre'), ('authors', 'author'), ('epochs', 'epoch'))
176         for field_name, category in categories:
177             try:
178                 tag_names = getattr(info, field_name)
179             except:
180                 try:
181                     tag_names = [getattr(info, category)]
182                 except:
183                     # For instance, Pictures do not have 'genre' field.
184                     continue
185             for tag_name in tag_names:
186                 tag_sort_key = tag_name
187                 if category == 'author':
188                     tag_sort_key = tag_name.last_name
189                     tag_name = tag_name.readable()
190                 tag, created = Tag.objects.get_or_create(slug=slughifi(tag_name), category=category)
191                 if created:
192                     tag.name = tag_name
193                     tag.sort_key = sortify(tag_sort_key.lower())
194                     tag.save()
195                 meta_tags.append(tag)
196         return meta_tags
197
198
199
200 def get_dynamic_path(media, filename, ext=None, maxlen=100):
201     from slughifi import slughifi
202
203     # how to put related book's slug here?
204     if not ext:
205         # BookMedia case
206         ext = media.formats[media.type].ext
207     if media is None or not media.name:
208         name = slughifi(filename.split(".")[0])
209     else:
210         name = slughifi(media.name)
211     return 'book/%s/%s.%s' % (ext, name[:maxlen-len('book/%s/.%s' % (ext, ext))-4], ext)
212
213
214 # TODO: why is this hard-coded ?
215 def book_upload_path(ext=None, maxlen=100):
216     return lambda *args: get_dynamic_path(*args, ext=ext, maxlen=maxlen)
217
218
219 def get_customized_pdf_path(book, customizations):
220     """
221     Returns a MEDIA_ROOT relative path for a customized pdf. The name will contain a hash of customization options.
222     """
223     customizations.sort()
224     h = hash(tuple(customizations))
225
226     pdf_name = '%s-custom-%s' % (book.slug, h)
227     pdf_file = get_dynamic_path(None, pdf_name, ext='pdf')
228
229     return pdf_file
230
231
232 def get_existing_customized_pdf(book):
233     """
234     Returns a list of paths to generated customized pdf of a book
235     """
236     pdf_glob = '%s-custom-' % (book.slug,)
237     pdf_glob = get_dynamic_path(None, pdf_glob, ext='pdf')
238     pdf_glob = re.sub(r"[.]([a-z0-9]+)$", "*.\\1", pdf_glob)
239     return glob(path.join(settings.MEDIA_ROOT, pdf_glob))
240
241
242 class BookMedia(models.Model):
243     FileFormat = namedtuple("FileFormat", "name ext")
244     formats = SortedDict([
245         ('mp3', FileFormat(name='MP3', ext='mp3')),
246         ('ogg', FileFormat(name='Ogg Vorbis', ext='ogg')),
247         ('daisy', FileFormat(name='DAISY', ext='daisy.zip')),
248     ])
249     format_choices = [(k, _('%s file') % t.name)
250             for k, t in formats.items()]
251
252     type        = models.CharField(_('type'), choices=format_choices, max_length="100")
253     name        = models.CharField(_('name'), max_length="100")
254     file        = OverwritingFileField(_('file'), upload_to=book_upload_path())
255     uploaded_at = models.DateTimeField(_('creation date'), auto_now_add=True, editable=False)
256     extra_info  = JSONField(_('extra information'), default='{}', editable=False)
257     book = models.ForeignKey('Book', related_name='media')
258     source_sha1 = models.CharField(null=True, blank=True, max_length=40, editable=False)
259
260     def __unicode__(self):
261         return "%s (%s)" % (self.name, self.file.name.split("/")[-1])
262
263     class Meta:
264         ordering            = ('type', 'name')
265         verbose_name        = _('book media')
266         verbose_name_plural = _('book media')
267
268     def save(self, *args, **kwargs):
269         from slughifi import slughifi
270         from catalogue.utils import ExistingFile, remove_zip
271
272         try:
273             old = BookMedia.objects.get(pk=self.pk)
274         except BookMedia.DoesNotExist, e:
275             old = None
276         else:
277             # if name changed, change the file name, too
278             if slughifi(self.name) != slughifi(old.name):
279                 self.file.save(None, ExistingFile(self.file.path), save=False, leave=True)
280
281         super(BookMedia, self).save(*args, **kwargs)
282
283         # remove the zip package for book with modified media
284         if old:
285             remove_zip("%s_%s" % (old.book.slug, old.type))
286         remove_zip("%s_%s" % (self.book.slug, self.type))
287
288         extra_info = self.get_extra_info_value()
289         extra_info.update(self.read_meta())
290         self.set_extra_info_value(extra_info)
291         self.source_sha1 = self.read_source_sha1(self.file.path, self.type)
292         return super(BookMedia, self).save(*args, **kwargs)
293
294     def read_meta(self):
295         """
296             Reads some metadata from the audiobook.
297         """
298         import mutagen
299         from mutagen import id3
300
301         artist_name = director_name = project = funded_by = ''
302         if self.type == 'mp3':
303             try:
304                 audio = id3.ID3(self.file.path)
305                 artist_name = ', '.join(', '.join(tag.text) for tag in audio.getall('TPE1'))
306                 director_name = ', '.join(', '.join(tag.text) for tag in audio.getall('TPE3'))
307                 project = ", ".join([t.data for t in audio.getall('PRIV') 
308                         if t.owner=='wolnelektury.pl?project'])
309                 funded_by = ", ".join([t.data for t in audio.getall('PRIV') 
310                         if t.owner=='wolnelektury.pl?funded_by'])
311             except:
312                 pass
313         elif self.type == 'ogg':
314             try:
315                 audio = mutagen.File(self.file.path)
316                 artist_name = ', '.join(audio.get('artist', []))
317                 director_name = ', '.join(audio.get('conductor', []))
318                 project = ", ".join(audio.get('project', []))
319                 funded_by = ", ".join(audio.get('funded_by', []))
320             except:
321                 pass
322         else:
323             return {}
324         return {'artist_name': artist_name, 'director_name': director_name,
325                 'project': project, 'funded_by': funded_by}
326
327     @staticmethod
328     def read_source_sha1(filepath, filetype):
329         """
330             Reads source file SHA1 from audiobok metadata.
331         """
332         import mutagen
333         from mutagen import id3
334
335         if filetype == 'mp3':
336             try:
337                 audio = id3.ID3(filepath)
338                 return [t.data for t in audio.getall('PRIV') 
339                         if t.owner=='wolnelektury.pl?flac_sha1'][0]
340             except:
341                 return None
342         elif filetype == 'ogg':
343             try:
344                 audio = mutagen.File(filepath)
345                 return audio.get('flac_sha1', [None])[0] 
346             except:
347                 return None
348         else:
349             return None
350
351
352 class Book(models.Model):
353     title         = models.CharField(_('title'), max_length=120)
354     sort_key = models.CharField(_('sort key'), max_length=120, db_index=True, editable=False)
355     slug = models.SlugField(_('slug'), max_length=120, db_index=True,
356             unique=True)
357     common_slug = models.SlugField(_('slug'), max_length=120, db_index=True)
358     language = models.CharField(_('language code'), max_length=3, db_index=True,
359                     default=settings.CATALOGUE_DEFAULT_LANGUAGE)
360     description   = models.TextField(_('description'), blank=True)
361     created_at    = models.DateTimeField(_('creation date'), auto_now_add=True, db_index=True)
362     changed_at    = models.DateTimeField(_('creation date'), auto_now=True, db_index=True)
363     parent_number = models.IntegerField(_('parent number'), default=0)
364     extra_info    = JSONField(_('extra information'), default='{}')
365     gazeta_link   = models.CharField(blank=True, max_length=240)
366     wiki_link     = models.CharField(blank=True, max_length=240)
367     # files generated during publication
368
369     cover = models.FileField(_('cover'), upload_to=book_upload_path('png'),
370                 null=True, blank=True)
371     ebook_formats = ['pdf', 'epub', 'mobi', 'txt']
372     formats = ebook_formats + ['html', 'xml']
373
374     parent        = models.ForeignKey('self', blank=True, null=True, related_name='children')
375     objects  = models.Manager()
376     tagged   = managers.ModelTaggedItemManager(Tag)
377     tags     = managers.TagDescriptor(Tag)
378
379     html_built = django.dispatch.Signal()
380     published = django.dispatch.Signal()
381
382     class AlreadyExists(Exception):
383         pass
384
385     class Meta:
386         ordering = ('sort_key',)
387         verbose_name = _('book')
388         verbose_name_plural = _('books')
389
390     def __unicode__(self):
391         return self.title
392
393     def save(self, force_insert=False, force_update=False, reset_short_html=True, **kwargs):
394         from sortify import sortify
395
396         self.sort_key = sortify(self.title)
397
398         ret = super(Book, self).save(force_insert, force_update)
399
400         if reset_short_html:
401             self.reset_short_html()
402
403         return ret
404
405     @permalink
406     def get_absolute_url(self):
407         return ('catalogue.views.book_detail', [self.slug])
408
409     @property
410     def name(self):
411         return self.title
412
413     def book_tag_slug(self):
414         return ('l-' + self.slug)[:120]
415
416     def book_tag(self):
417         slug = self.book_tag_slug()
418         book_tag, created = Tag.objects.get_or_create(slug=slug, category='book')
419         if created:
420             book_tag.name = self.title[:50]
421             book_tag.sort_key = self.title.lower()
422             book_tag.save()
423         return book_tag
424
425     def has_media(self, type):
426         if type in Book.formats:
427             return bool(getattr(self, "%s_file" % type))
428         else:
429             return self.media.filter(type=type).exists()
430
431     def get_media(self, type):
432         if self.has_media(type):
433             if type in Book.formats:
434                 return getattr(self, "%s_file" % type)
435             else:                                             
436                 return self.media.filter(type=type)
437         else:
438             return None
439
440     def get_mp3(self):
441         return self.get_media("mp3")
442     def get_odt(self):
443         return self.get_media("odt")
444     def get_ogg(self):
445         return self.get_media("ogg")
446     def get_daisy(self):
447         return self.get_media("daisy")                       
448
449     def reset_short_html(self):
450         if self.id is None:
451             return
452
453         cache_key = "Book.short_html/%d/%s"
454         for lang, langname in settings.LANGUAGES:
455             cache.delete(cache_key % (self.id, lang))
456         cache.delete("Book.mini_box/%d" % (self.id, ))
457         # Fragment.short_html relies on book's tags, so reset it here too
458         for fragm in self.fragments.all():
459             fragm.reset_short_html()
460
461     def short_html(self):
462         if self.id:
463             cache_key = "Book.short_html/%d/%s" % (self.id, get_language())
464             short_html = cache.get(cache_key)
465         else:
466             short_html = None
467
468         if short_html is not None:
469             return mark_safe(short_html)
470         else:
471             tags = self.tags.filter(category__in=('author', 'kind', 'genre', 'epoch'))
472             tags = split_tags(tags)
473
474             formats = {}
475             # files generated during publication
476             for ebook_format in self.ebook_formats:
477                 if self.has_media(ebook_format):
478                     formats[ebook_format] = self.get_media(ebook_format)
479
480
481             short_html = unicode(render_to_string('catalogue/book_short.html',
482                 {'book': self, 'tags': tags, 'formats': formats}))
483
484             if self.id:
485                 cache.set(cache_key, short_html, CACHE_FOREVER)
486             return mark_safe(short_html)
487
488     def mini_box(self):
489         if self.id:
490             cache_key = "Book.mini_box/%d" % (self.id, )
491             short_html = cache.get(cache_key)
492         else:
493             short_html = None
494
495         if short_html is None:
496             authors = self.tags.filter(category='author')
497
498             short_html = unicode(render_to_string('catalogue/book_mini_box.html',
499                 {'book': self, 'authors': authors, 'STATIC_URL': settings.STATIC_URL}))
500
501             if self.id:
502                 cache.set(cache_key, short_html, CACHE_FOREVER)
503         return mark_safe(short_html)
504
505     def has_description(self):
506         return len(self.description) > 0
507     has_description.short_description = _('description')
508     has_description.boolean = True
509
510     # ugly ugly ugly
511     def has_mp3_file(self):
512         return bool(self.has_media("mp3"))
513     has_mp3_file.short_description = 'MP3'
514     has_mp3_file.boolean = True
515
516     def has_ogg_file(self):
517         return bool(self.has_media("ogg"))
518     has_ogg_file.short_description = 'OGG'
519     has_ogg_file.boolean = True
520
521     def has_daisy_file(self):
522         return bool(self.has_media("daisy"))
523     has_daisy_file.short_description = 'DAISY'
524     has_daisy_file.boolean = True
525
526     def wldocument(self, parse_dublincore=True):
527         from catalogue.import_utils import ORMDocProvider
528         from librarian.parser import WLDocument
529
530         return WLDocument.from_file(self.xml_file.path,
531                 provider=ORMDocProvider(self),
532                 parse_dublincore=parse_dublincore)
533
534     def build_cover(self, book_info=None):
535         """(Re)builds the cover image."""
536         from StringIO import StringIO
537         from django.core.files.base import ContentFile
538         from librarian.cover import WLCover
539
540         if book_info is None:
541             book_info = self.wldocument().book_info
542
543         cover = WLCover(book_info).image()
544         imgstr = StringIO()
545         cover.save(imgstr, 'png')
546         self.cover.save(None, ContentFile(imgstr.getvalue()))
547
548     def build_pdf(self, customizations=None, file_name=None):
549         """ (Re)builds the pdf file.
550         customizations - customizations which are passed to LaTeX class file.
551         file_name - save the pdf file under a different name and DO NOT save it in db.
552         """
553         from os import unlink
554         from django.core.files import File
555         from catalogue.utils import remove_zip
556
557         pdf = self.wldocument().as_pdf(customizations=customizations)
558
559         if file_name is None:
560             # we'd like to be sure not to overwrite changes happening while
561             # (timely) pdf generation is taking place (async celery scenario)
562             current_self = Book.objects.get(id=self.id)
563             current_self.pdf_file.save('%s.pdf' % self.slug,
564                     File(open(pdf.get_filename())))
565             self.pdf_file = current_self.pdf_file
566
567             # remove cached downloadables
568             remove_zip(settings.ALL_PDF_ZIP)
569
570             for customized_pdf in get_existing_customized_pdf(self):
571                 unlink(customized_pdf)
572         else:
573             print "saving %s" % file_name
574             print "to: %s" % DefaultStorage().path(file_name)
575             DefaultStorage().save(file_name, File(open(pdf.get_filename())))
576
577     def build_mobi(self):
578         """ (Re)builds the MOBI file.
579
580         """
581         from django.core.files import File
582         from catalogue.utils import remove_zip
583
584         mobi = self.wldocument().as_mobi()
585
586         self.mobi_file.save('%s.mobi' % self.slug, File(open(mobi.get_filename())))
587
588         # remove zip with all mobi files
589         remove_zip(settings.ALL_MOBI_ZIP)
590
591     def build_epub(self):
592         """(Re)builds the epub file."""
593         from django.core.files import File
594         from catalogue.utils import remove_zip
595
596         epub = self.wldocument().as_epub()
597
598         self.epub_file.save('%s.epub' % self.slug,
599                 File(open(epub.get_filename())))
600
601         # remove zip package with all epub files
602         remove_zip(settings.ALL_EPUB_ZIP)
603
604     def build_txt(self):
605         from django.core.files.base import ContentFile
606
607         text = self.wldocument().as_text()
608         self.txt_file.save('%s.txt' % self.slug, ContentFile(text.get_string()))
609
610
611     def build_html(self):
612         from markupstring import MarkupString
613         from django.core.files.base import ContentFile
614         from slughifi import slughifi
615         from librarian import html
616
617         meta_tags = list(self.tags.filter(
618             category__in=('author', 'epoch', 'genre', 'kind')))
619         book_tag = self.book_tag()
620
621         html_output = self.wldocument(parse_dublincore=False).as_html()
622         if html_output:
623             self.html_file.save('%s.html' % self.slug,
624                     ContentFile(html_output.get_string()))
625
626             # get ancestor l-tags for adding to new fragments
627             ancestor_tags = []
628             p = self.parent
629             while p:
630                 ancestor_tags.append(p.book_tag())
631                 p = p.parent
632
633             # Delete old fragments and create them from scratch
634             self.fragments.all().delete()
635             # Extract fragments
636             closed_fragments, open_fragments = html.extract_fragments(self.html_file.path)
637             for fragment in closed_fragments.values():
638                 try:
639                     theme_names = [s.strip() for s in fragment.themes.split(',')]
640                 except AttributeError:
641                     continue
642                 themes = []
643                 for theme_name in theme_names:
644                     if not theme_name:
645                         continue
646                     tag, created = Tag.objects.get_or_create(slug=slughifi(theme_name), category='theme')
647                     if created:
648                         tag.name = theme_name
649                         tag.sort_key = theme_name.lower()
650                         tag.save()
651                     themes.append(tag)
652                 if not themes:
653                     continue
654
655                 text = fragment.to_string()
656                 short_text = ''
657                 if (len(MarkupString(text)) > 240):
658                     short_text = unicode(MarkupString(text)[:160])
659                 new_fragment = Fragment.objects.create(anchor=fragment.id, book=self,
660                     text=text, short_text=short_text)
661
662                 new_fragment.save()
663                 new_fragment.tags = set(meta_tags + themes + [book_tag] + ancestor_tags)
664             self.save()
665             self.html_built.send(sender=self)
666             return True
667         return False
668
669     @staticmethod
670     def zip_format(format_):
671         def pretty_file_name(book):
672             return "%s/%s.%s" % (
673                 b.get_extra_info_value()['author'],
674                 b.slug,
675                 format_)
676
677         field_name = "%s_file" % format_
678         books = Book.objects.filter(parent=None).exclude(**{field_name: ""})
679         paths = [(pretty_file_name(b), getattr(b, field_name).path)
680                     for b in books]
681         result = create_zip.delay(paths,
682                     getattr(settings, "ALL_%s_ZIP" % format_.upper()))
683         return result.wait()
684
685     def zip_audiobooks(self, format_):
686         bm = BookMedia.objects.filter(book=self, type=format_)
687         paths = map(lambda bm: (None, bm.file.path), bm)
688         result = create_zip.delay(paths, "%s_%s" % (self.slug, format_))
689         return result.wait()
690
691     def search_index(self, book_info=None, reuse_index=False):
692         if reuse_index:
693             idx = search.ReusableIndex()
694         else:
695             idx = search.Index()
696             
697         idx.open()
698         try:
699             idx.index_book(self, book_info)
700             idx.index_tags()
701         finally:
702             idx.close()
703
704     @classmethod
705     def from_xml_file(cls, xml_file, **kwargs):
706         from django.core.files import File
707         from librarian import dcparser
708
709         # use librarian to parse meta-data
710         book_info = dcparser.parse(xml_file)
711
712         if not isinstance(xml_file, File):
713             xml_file = File(open(xml_file))
714
715         try:
716             return cls.from_text_and_meta(xml_file, book_info, **kwargs)
717         finally:
718             xml_file.close()
719
720     @classmethod
721     def from_text_and_meta(cls, raw_file, book_info, overwrite=False,
722             build_epub=True, build_txt=True, build_pdf=True, build_mobi=True,
723             search_index=True):
724         import re
725         from sortify import sortify
726
727         # check for parts before we do anything
728         children = []
729         if hasattr(book_info, 'parts'):
730             for part_url in book_info.parts:
731                 try:
732                     children.append(Book.objects.get(slug=part_url.slug))
733                 except Book.DoesNotExist, e:
734                     raise Book.DoesNotExist(_('Book "%s" does not exist.') %
735                             part_url.slug)
736
737
738         # Read book metadata
739         book_slug = book_info.url.slug
740         if re.search(r'[^a-z0-9-]', book_slug):
741             raise ValueError('Invalid characters in slug')
742         book, created = Book.objects.get_or_create(slug=book_slug)
743
744         if created:
745             book_shelves = []
746         else:
747             if not overwrite:
748                 raise Book.AlreadyExists(_('Book %s already exists') % (
749                         book_slug))
750             # Save shelves for this book
751             book_shelves = list(book.tags.filter(category='set'))
752
753         book.language = book_info.language
754         book.title = book_info.title
755         if book_info.variant_of:
756             book.common_slug = book_info.variant_of.slug
757         else:
758             book.common_slug = book.slug
759         book.set_extra_info_value(book_info.to_dict())
760         book.save()
761
762         meta_tags = Tag.tags_from_info(book_info)
763
764         book.tags = set(meta_tags + book_shelves)
765
766         book_tag = book.book_tag()
767
768         for n, child_book in enumerate(children):
769             child_book.parent = book
770             child_book.parent_number = n
771             child_book.save()
772
773         # Save XML and HTML files
774         book.xml_file.save('%s.xml' % book.slug, raw_file, save=False)
775
776         # delete old fragments when overwriting
777         book.fragments.all().delete()
778
779         if book.build_html():
780             if not settings.NO_BUILD_TXT and build_txt:
781                 book.build_txt()
782
783         book.build_cover(book_info)
784
785         if not settings.NO_BUILD_EPUB and build_epub:
786             book.build_epub()
787
788         if not settings.NO_BUILD_PDF and build_pdf:
789             book.build_pdf()
790
791         if not settings.NO_BUILD_MOBI and build_mobi:
792             book.build_mobi()
793
794         if not settings.NO_SEARCH_INDEX and search_index:
795             index_book.delay(book.id, book_info)
796
797         book_descendants = list(book.children.all())
798         descendants_tags = set()
799         # add l-tag to descendants and their fragments
800         while len(book_descendants) > 0:
801             child_book = book_descendants.pop(0)
802             descendants_tags.update(child_book.tags)
803             child_book.tags = list(child_book.tags) + [book_tag]
804             child_book.save()
805             for fragment in child_book.fragments.all():
806                 fragment.tags = set(list(fragment.tags) + [book_tag])
807             book_descendants += list(child_book.children.all())
808
809         for tag in descendants_tags:
810             touch_tag.delay(tag)
811
812         book.save()
813
814         # refresh cache
815         book.reset_tag_counter()
816         book.reset_theme_counter()
817
818         cls.published.send(sender=book)
819         return book
820
821     def reset_tag_counter(self):
822         if self.id is None:
823             return
824
825         cache_key = "Book.tag_counter/%d" % self.id
826         cache.delete(cache_key)
827         if self.parent:
828             self.parent.reset_tag_counter()
829
830     @property
831     def tag_counter(self):
832         if self.id:
833             cache_key = "Book.tag_counter/%d" % self.id
834             tags = cache.get(cache_key)
835         else:
836             tags = None
837
838         if tags is None:
839             tags = {}
840             for child in self.children.all().order_by():
841                 for tag_pk, value in child.tag_counter.iteritems():
842                     tags[tag_pk] = tags.get(tag_pk, 0) + value
843             for tag in self.tags.exclude(category__in=('book', 'theme', 'set')).order_by():
844                 tags[tag.pk] = 1
845
846             if self.id:
847                 cache.set(cache_key, tags, CACHE_FOREVER)
848         return tags
849
850     def reset_theme_counter(self):
851         if self.id is None:
852             return
853
854         cache_key = "Book.theme_counter/%d" % self.id
855         cache.delete(cache_key)
856         if self.parent:
857             self.parent.reset_theme_counter()
858
859     @property
860     def theme_counter(self):
861         if self.id:
862             cache_key = "Book.theme_counter/%d" % self.id
863             tags = cache.get(cache_key)
864         else:
865             tags = None
866
867         if tags is None:
868             tags = {}
869             for fragment in Fragment.tagged.with_any([self.book_tag()]).order_by():
870                 for tag in fragment.tags.filter(category='theme').order_by():
871                     tags[tag.pk] = tags.get(tag.pk, 0) + 1
872
873             if self.id:
874                 cache.set(cache_key, tags, CACHE_FOREVER)
875         return tags
876
877     def pretty_title(self, html_links=False):
878         book = self
879         names = list(book.tags.filter(category='author'))
880
881         books = []
882         while book:
883             books.append(book)
884             book = book.parent
885         names.extend(reversed(books))
886
887         if html_links:
888             names = ['<a href="%s">%s</a>' % (tag.get_absolute_url(), tag.name) for tag in names]
889         else:
890             names = [tag.name for tag in names]
891
892         return ', '.join(names)
893
894     @classmethod
895     def tagged_top_level(cls, tags):
896         """ Returns top-level books tagged with `tags'.
897
898         It only returns those books which don't have ancestors which are
899         also tagged with those tags.
900
901         """
902         # get relevant books and their tags
903         objects = cls.tagged.with_all(tags)
904         # eliminate descendants
905         l_tags = Tag.objects.filter(category='book', slug__in=[book.book_tag_slug() for book in objects])
906         descendants_keys = [book.pk for book in cls.tagged.with_any(l_tags)]
907         if descendants_keys:
908             objects = objects.exclude(pk__in=descendants_keys)
909
910         return objects
911
912     @classmethod
913     def book_list(cls, filter=None):
914         """Generates a hierarchical listing of all books.
915
916         Books are optionally filtered with a test function.
917
918         """
919
920         books_by_parent = {}
921         books = cls.objects.all().order_by('parent_number', 'sort_key').only(
922                 'title', 'parent', 'slug')
923         if filter:
924             books = books.filter(filter).distinct()
925             book_ids = set((book.pk for book in books))
926             for book in books:
927                 parent = book.parent_id
928                 if parent not in book_ids:
929                     parent = None
930                 books_by_parent.setdefault(parent, []).append(book)
931         else:
932             for book in books:
933                 books_by_parent.setdefault(book.parent_id, []).append(book)
934
935         orphans = []
936         books_by_author = SortedDict()
937         for tag in Tag.objects.filter(category='author'):
938             books_by_author[tag] = []
939
940         for book in books_by_parent.get(None,()):
941             authors = list(book.tags.filter(category='author'))
942             if authors:
943                 for author in authors:
944                     books_by_author[author].append(book)
945             else:
946                 orphans.append(book)
947
948         return books_by_author, orphans, books_by_parent
949
950     _audiences_pl = {
951         "SP1": (1, u"szkoła podstawowa"),
952         "SP2": (1, u"szkoła podstawowa"),
953         "P": (1, u"szkoła podstawowa"),
954         "G": (2, u"gimnazjum"),
955         "L": (3, u"liceum"),
956         "LP": (3, u"liceum"),
957     }
958     def audiences_pl(self):
959         audiences = self.get_extra_info_value().get('audiences', [])
960         audiences = sorted(set([self._audiences_pl[a] for a in audiences]))
961         return [a[1] for a in audiences]
962
963
964 def _has_factory(ftype):
965     has = lambda self: bool(getattr(self, "%s_file" % ftype))
966     has.short_description = t.upper()
967     has.boolean = True
968     has.__name__ = "has_%s_file" % ftype
969     return has
970
971     
972 # add the file fields
973 for t in Book.formats:
974     field_name = "%s_file" % t
975     models.FileField(_("%s file" % t.upper()),
976             upload_to=book_upload_path(t),
977             blank=True).contribute_to_class(Book, field_name)
978
979     setattr(Book, "has_%s_file" % t, _has_factory(t))
980
981
982 class Fragment(models.Model):
983     text = models.TextField()
984     short_text = models.TextField(editable=False)
985     anchor = models.CharField(max_length=120)
986     book = models.ForeignKey(Book, related_name='fragments')
987
988     objects = models.Manager()
989     tagged = managers.ModelTaggedItemManager(Tag)
990     tags = managers.TagDescriptor(Tag)
991
992     class Meta:
993         ordering = ('book', 'anchor',)
994         verbose_name = _('fragment')
995         verbose_name_plural = _('fragments')
996
997     def get_absolute_url(self):
998         return '%s#m%s' % (reverse('book_text', args=[self.book.slug]), self.anchor)
999
1000     def reset_short_html(self):
1001         if self.id is None:
1002             return
1003
1004         cache_key = "Fragment.short_html/%d/%s"
1005         for lang, langname in settings.LANGUAGES:
1006             cache.delete(cache_key % (self.id, lang))
1007
1008     def short_html(self):
1009         if self.id:
1010             cache_key = "Fragment.short_html/%d/%s" % (self.id, get_language())
1011             short_html = cache.get(cache_key)
1012         else:
1013             short_html = None
1014
1015         if short_html is not None:
1016             return mark_safe(short_html)
1017         else:
1018             short_html = unicode(render_to_string('catalogue/fragment_short.html',
1019                 {'fragment': self}))
1020             if self.id:
1021                 cache.set(cache_key, short_html, CACHE_FOREVER)
1022             return mark_safe(short_html)
1023
1024
1025 ###########
1026 #
1027 # SIGNALS
1028 #
1029 ###########
1030
1031
1032 def _tags_updated_handler(sender, affected_tags, **kwargs):
1033     # reset tag global counter
1034     # we want Tag.changed_at updated for API to know the tag was touched
1035     for tag in affected_tags:
1036         touch_tag.delay(tag)
1037
1038     # if book tags changed, reset book tag counter
1039     if isinstance(sender, Book) and \
1040                 Tag.objects.filter(pk__in=(tag.pk for tag in affected_tags)).\
1041                     exclude(category__in=('book', 'theme', 'set')).count():
1042         sender.reset_tag_counter()
1043     # if fragment theme changed, reset book theme counter
1044     elif isinstance(sender, Fragment) and \
1045                 Tag.objects.filter(pk__in=(tag.pk for tag in affected_tags)).\
1046                     filter(category='theme').count():
1047         sender.book.reset_theme_counter()
1048 tags_updated.connect(_tags_updated_handler)
1049
1050
1051 def _pre_delete_handler(sender, instance, **kwargs):
1052     """ refresh Book on BookMedia delete """
1053     if sender == BookMedia:
1054         instance.book.save()
1055 pre_delete.connect(_pre_delete_handler)
1056
1057 def _post_save_handler(sender, instance, **kwargs):
1058     """ refresh all the short_html stuff on BookMedia update """
1059     if sender == BookMedia:
1060         instance.book.save()
1061 post_save.connect(_post_save_handler)