X-Git-Url: https://git.mdrn.pl/wolnelektury.git/blobdiff_plain/eb11e65717bdbddbc3ce08fa437739fd7bf5c279..b3b6cbc7a545af2f3fa6f04f9281fd994b2adea7:/apps/catalogue/models.py?ds=inline diff --git a/apps/catalogue/models.py b/apps/catalogue/models.py index 37e6c2e5e..e692ad8c5 100644 --- a/apps/catalogue/models.py +++ b/apps/catalogue/models.py @@ -13,11 +13,11 @@ from django.utils.translation import get_language from django.core.urlresolvers import reverse from datetime import datetime -from newtagging.models import TagBase +from newtagging.models import TagBase, tags_updated from newtagging import managers from catalogue.fields import JSONField -from librarian import html, dcparser +from librarian import dcparser, html, epub, NoDublinCore from mutagen import id3 @@ -43,7 +43,7 @@ class TagSubcategoryManager(models.Manager): class Tag(TagBase): name = models.CharField(_('name'), max_length=50, db_index=True) - slug = models.SlugField(_('slug'), max_length=120, unique=True, db_index=True) + slug = models.SlugField(_('slug'), max_length=120, db_index=True) sort_key = models.SlugField(_('sort key'), max_length=120, db_index=True) category = models.CharField(_('category'), max_length=50, blank=False, null=False, db_index=True, choices=TAG_CATEGORIES) @@ -51,15 +51,26 @@ class Tag(TagBase): main_page = models.BooleanField(_('main page'), default=False, db_index=True, help_text=_('Show tag on main page')) user = models.ForeignKey(User, blank=True, null=True) - book_count = models.IntegerField(_('book count'), default=0, blank=False, null=False) + book_count = models.IntegerField(_('book count'), blank=False, null=True) death = models.IntegerField(_(u'year of death'), blank=True, null=True) gazeta_link = models.CharField(blank=True, max_length=240) wiki_link = models.CharField(blank=True, max_length=240) + categories_rev = { + 'autor': 'author', + 'epoka': 'epoch', + 'rodzaj': 'kind', + 'gatunek': 'genre', + 'motyw': 'theme', + 'polka': 'set', + } + categories_dict = dict((item[::-1] for item in categories_rev.iteritems())) + class Meta: ordering = ('sort_key',) verbose_name = _('tag') verbose_name_plural = _('tags') + unique_together = (("slug", "category"),) def __unicode__(self): return self.name @@ -69,7 +80,7 @@ class Tag(TagBase): @permalink def get_absolute_url(self): - return ('catalogue.views.tagged_object_list', [self.slug]) + return ('catalogue.views.tagged_object_list', [self.url_chunk]) def has_description(self): return len(self.description) > 0 @@ -87,16 +98,67 @@ class Tag(TagBase): """ calculates the year of public domain entry for an author """ return self.death + 71 if self.death is not None else None + def get_count(self): + """ returns global book count for book tags, fragment count for themes """ + + if self.book_count is None: + if self.category == 'book': + # never used + objects = Book.objects.none() + elif self.category == 'theme': + objects = Fragment.tagged.with_all((self,)) + else: + objects = Book.tagged.with_all((self,)).order_by() + if self.category != 'set': + # eliminate descendants + l_tags = Tag.objects.filter(slug__in=[book.book_tag_slug() for book in objects]) + descendants_keys = [book.pk for book in Book.tagged.with_any(l_tags)] + if descendants_keys: + objects = objects.exclude(pk__in=descendants_keys) + self.book_count = objects.count() + self.save() + return self.book_count + @staticmethod def get_tag_list(tags): if isinstance(tags, basestring): - tag_slugs = tags.split('/') - return [Tag.objects.get(slug=slug) for slug in tag_slugs] + real_tags = [] + ambiguous_slugs = [] + category = None + tags_splitted = tags.split('/') + for index, name in enumerate(tags_splitted): + if name in Tag.categories_rev: + category = Tag.categories_rev[name] + else: + if category: + real_tags.append(Tag.objects.get(slug=name, category=category)) + category = None + else: + try: + real_tags.append(Tag.objects.exclude(category='book').get(slug=name)) + except Tag.MultipleObjectsReturned, e: + ambiguous_slugs.append(name) + + if category: + # something strange left off + raise Tag.DoesNotExist() + if ambiguous_slugs: + # some tags should be qualified + e = Tag.MultipleObjectsReturned() + e.tags = real_tags + e.ambiguous_slugs = ambiguous_slugs + raise e + else: + return real_tags else: return TagBase.get_tag_list(tags) + @property + def url_chunk(self): + return '/'.join((Tag.categories_dict[self.category], self.slug)) + -# TODO: why is this hard-coded ? +# TODO: why is this hard-coded ? def book_upload_path(ext): def get_dynamic_path(book, filename): return 'lektura/%s.%s' % (book.slug, ext) @@ -124,13 +186,14 @@ class Book(models.Model): txt_file = models.FileField(_('TXT file'), upload_to=book_upload_path('txt'), blank=True) mp3_file = models.FileField(_('MP3 file'), upload_to=book_upload_path('mp3'), blank=True) ogg_file = models.FileField(_('OGG file'), upload_to=book_upload_path('ogg'), blank=True) + daisy_file = models.FileField(_('DAISY file'), upload_to=book_upload_path('daisy.zip'), blank=True) parent = models.ForeignKey('self', blank=True, null=True, related_name='children') objects = models.Manager() tagged = managers.ModelTaggedItemManager(Tag) tags = managers.TagDescriptor(Tag) - + _tag_counter = JSONField(null=True, editable=False) _theme_counter = JSONField(null=True, editable=False) @@ -145,11 +208,15 @@ class Book(models.Model): def __unicode__(self): return self.title - def save(self, force_insert=False, force_update=False, reset_short_html=True, refresh_mp3=True): + def save(self, force_insert=False, force_update=False, reset_short_html=True, refresh_mp3=True, **kwargs): if reset_short_html: # Reset _short_html during save + update = {} for key in filter(lambda x: x.startswith('_short_html'), self.__dict__): + update[key] = '' self.__setattr__(key, '') + # Fragment.short_html relies on book's tags, so reset it here too + self.fragments.all().update(**update) book = super(Book, self).save(force_insert, force_update) @@ -169,14 +236,16 @@ class Book(models.Model): @property def name(self): return self.title - + + def book_tag_slug(self): + return ('l-' + self.slug)[:120] + def book_tag(self): - slug = ('l-' + self.slug)[:120] - book_tag, created = Tag.objects.get_or_create(slug=slug) + slug = self.book_tag_slug() + book_tag, created = Tag.objects.get_or_create(slug=slug, category='book') if created: book_tag.name = self.title[:50] book_tag.sort_key = slug - book_tag.category = 'book' book_tag.save() return book_tag @@ -195,8 +264,8 @@ class Book(models.Model): formats.append(u'%s' % (reverse('book_text', kwargs={'slug': self.slug}), _('Read online'))) if self.pdf_file: formats.append(u'PDF' % self.pdf_file.url) - if self.epub_file: - formats.append(u'EPUB' % self.epub_file.url) + if self.root_ancestor.epub_file: + formats.append(u'EPUB' % self.root_ancestor.epub_file.url) if self.odt_file: formats.append(u'ODT' % self.odt_file.url) if self.txt_file: @@ -205,6 +274,8 @@ class Book(models.Model): formats.append(u'MP3' % self.mp3_file.url) if self.ogg_file: formats.append(u'OGG' % self.ogg_file.url) + if self.daisy_file: + formats.append(u'DAISY' % self.daisy_file.url) formats = [mark_safe(format) for format in formats] @@ -214,6 +285,18 @@ class Book(models.Model): return mark_safe(getattr(self, key)) + @property + def root_ancestor(self): + """ returns the oldest ancestor """ + + if not hasattr(self, '_root_ancestor'): + book = self + while book.parent: + book = book.parent + self._root_ancestor = book + return self._root_ancestor + + def get_mp3_info(self): """Retrieves artist and director names from audio ID3 tags.""" audio = id3.ID3(self.mp3_file.path) @@ -246,13 +329,59 @@ class Book(models.Model): has_html_file.short_description = 'HTML' has_html_file.boolean = True + def build_epub(self, remove_descendants=True): + """ (Re)builds the epub file. + If book has a parent, does nothing. + Unless remove_descendants is False, descendants' epubs are removed. + """ + + from StringIO import StringIO + from hashlib import sha1 + from django.core.files.base import ContentFile + from librarian import DocProvider + + class BookImportDocProvider(DocProvider): + """ used for joined EPUBs """ + + def __init__(self, book): + self.book = book + + def by_slug(self, slug): + if slug == self.book.slug: + return self.book.xml_file + else: + return Book.objects.get(slug=slug).xml_file + + if self.parent: + # don't need an epub + return + + epub_file = StringIO() + try: + epub.transform(BookImportDocProvider(self), self.slug, epub_file) + self.epub_file.save('%s.epub' % self.slug, ContentFile(epub_file.getvalue()), save=False) + self.save() + FileRecord(slug=self.slug, type='epub', sha1=sha1(epub_file.getvalue()).hexdigest()).save() + except NoDublinCore: + pass + + book_descendants = list(self.children.all()) + while len(book_descendants) > 0: + child_book = book_descendants.pop(0) + if remove_descendants and child_book.has_epub_file(): + child_book.epub_file.delete() + # save anyway, to refresh short_html + child_book.save() + book_descendants += list(child_book.children.all()) + + @classmethod def from_xml_file(cls, xml_file, overwrite=False): # use librarian to parse meta-data book_info = dcparser.parse(xml_file) if not isinstance(xml_file, File): - xml_file = File(xml_file) + xml_file = File(open(xml_file)) try: return cls.from_text_and_meta(xml_file, book_info, overwrite) @@ -264,6 +393,7 @@ class Book(models.Model): from tempfile import NamedTemporaryFile from slughifi import slughifi from markupstring import MarkupString + from django.core.files.storage import default_storage # Read book metadata book_base, book_slug = book_info.url.rsplit('/', 1) @@ -283,21 +413,25 @@ class Book(models.Model): book.save() book_tags = [] - for category in ('kind', 'genre', 'author', 'epoch'): - tag_name = getattr(book_info, category) - tag_sort_key = tag_name - if category == 'author': - tag_sort_key = tag_name.last_name - tag_name = ' '.join(tag_name.first_names) + ' ' + tag_name.last_name - tag, created = Tag.objects.get_or_create(slug=slughifi(tag_name)) - if created: - tag.name = tag_name - tag.sort_key = slughifi(tag_sort_key) - tag.category = category - tag.save() - book_tags.append(tag) - - book.tags = book_tags + categories = (('kinds', 'kind'), ('genres', 'genre'), ('authors', 'author'), ('epochs', 'epoch')) + for field_name, category in categories: + try: + tag_names = getattr(book_info, field_name) + except: + tag_names = [getattr(book_info, category)] + for tag_name in tag_names: + tag_sort_key = tag_name + if category == 'author': + tag_sort_key = tag_name.last_name + tag_name = ' '.join(tag_name.first_names) + ' ' + tag_name.last_name + tag, created = Tag.objects.get_or_create(slug=slughifi(tag_name), category=category) + if created: + tag.name = tag_name + tag.sort_key = slughifi(tag_sort_key) + tag.save() + book_tags.append(tag) + + book.tags = book_tags + book_shelves book_tag = book.book_tag() @@ -312,57 +446,77 @@ class Book(models.Model): except Book.DoesNotExist, e: raise Book.DoesNotExist(_('Book with slug = "%s" does not exist.') % slug) - book_descendants = list(book.children.all()) - while len(book_descendants) > 0: - child_book = book_descendants.pop(0) - child_book.tags = list(child_book.tags) + [book_tag] - child_book.save() - for fragment in child_book.fragments.all(): - fragment.tags = set(list(fragment.tags) + [book_tag]) - book_descendants += list(child_book.children.all()) - # Save XML and HTML files book.xml_file.save('%s.xml' % book.slug, raw_file, save=False) + # delete old fragments when overwriting + book.fragments.all().delete() + html_file = NamedTemporaryFile() if html.transform(book.xml_file.path, html_file, parse_dublincore=False): book.html_file.save('%s.html' % book.slug, File(html_file), save=False) + # get ancestor l-tags for adding to new fragments + ancestor_tags = [] + p = book.parent + while p: + ancestor_tags.append(p.book_tag()) + p = p.parent + # Extract fragments closed_fragments, open_fragments = html.extract_fragments(book.html_file.path) - book_themes = [] for fragment in closed_fragments.values(): - text = fragment.to_string() - short_text = '' - if (len(MarkupString(text)) > 240): - short_text = unicode(MarkupString(text)[:160]) - new_fragment, created = Fragment.objects.get_or_create(anchor=fragment.id, book=book, - defaults={'text': text, 'short_text': short_text}) - try: theme_names = [s.strip() for s in fragment.themes.split(',')] except AttributeError: continue themes = [] for theme_name in theme_names: - tag, created = Tag.objects.get_or_create(slug=slughifi(theme_name)) + if not theme_name: + continue + tag, created = Tag.objects.get_or_create(slug=slughifi(theme_name), category='theme') if created: tag.name = theme_name tag.sort_key = slughifi(theme_name) - tag.category = 'theme' tag.save() themes.append(tag) + if not themes: + continue + + text = fragment.to_string() + short_text = '' + if (len(MarkupString(text)) > 240): + short_text = unicode(MarkupString(text)[:160]) + new_fragment, created = Fragment.objects.get_or_create(anchor=fragment.id, book=book, + defaults={'text': text, 'short_text': short_text}) + new_fragment.save() - new_fragment.tags = set(list(book.tags) + themes + [book_tag]) - book_themes += themes + new_fragment.tags = set(book_tags + themes + [book_tag] + ancestor_tags) + + if not book.parent: + book.build_epub(remove_descendants=False) - book_themes = set(book_themes) - book.tags = list(book.tags) + list(book_themes) + book_shelves + book_descendants = list(book.children.all()) + # add l-tag to descendants and their fragments + # delete unnecessary EPUB files + while len(book_descendants) > 0: + child_book = book_descendants.pop(0) + child_book.tags = list(child_book.tags) + [book_tag] + if child_book.has_epub_file(): + child_book.epub_file.delete() + child_book.save() + for fragment in child_book.fragments.all(): + fragment.tags = set(list(fragment.tags) + [book_tag]) + book_descendants += list(child_book.children.all()) + + # refresh cache + book.reset_tag_counter() + book.reset_theme_counter() book.save() return book - - + + def refresh_tag_counter(self): tags = {} for child in self.children.all().order_by(): @@ -373,7 +527,13 @@ class Book(models.Model): self.set__tag_counter_value(tags) self.save(reset_short_html=False, refresh_mp3=False) return tags - + + def reset_tag_counter(self): + self._tag_counter = None + self.save(reset_short_html=False, refresh_mp3=False) + if self.parent: + self.parent.reset_tag_counter() + @property def tag_counter(self): if self._tag_counter is None: @@ -388,13 +548,19 @@ class Book(models.Model): self.set__theme_counter_value(tags) self.save(reset_short_html=False, refresh_mp3=False) return tags - + + def reset_theme_counter(self): + self._theme_counter = None + self.save(reset_short_html=False, refresh_mp3=False) + if self.parent: + self.parent.reset_theme_counter() + @property def theme_counter(self): if self._theme_counter is None: return self.refresh_theme_counter() return dict((int(k), v) for k, v in self.get__theme_counter_value().iteritems()) - + class Fragment(models.Model): @@ -422,11 +588,8 @@ class Fragment(models.Model): if short_html and len(short_html): return mark_safe(short_html) else: - book_authors = [mark_safe(u'%s' % (tag.get_absolute_url(), tag.name)) - for tag in self.book.tags if tag.category == 'author'] - setattr(self, key, unicode(render_to_string('catalogue/fragment_short.html', - {'fragment': self, 'book': self.book, 'book_authors': book_authors}))) + {'fragment': self}))) self.save() return mark_safe(getattr(self, key)) @@ -459,3 +622,34 @@ class BookStub(models.Model): return self.title +class FileRecord(models.Model): + slug = models.SlugField(_('slug'), max_length=120, db_index=True) + type = models.CharField(_('type'), max_length=20, db_index=True) + sha1 = models.CharField(_('sha-1 hash'), max_length=40) + time = models.DateTimeField(_('time'), auto_now_add=True) + + class Meta: + ordering = ('-time','-slug', '-type') + verbose_name = _('file record') + verbose_name_plural = _('file records') + + def __unicode__(self): + return "%s %s.%s" % (self.sha1, self.slug, self.type) + + +def _tags_updated_handler(sender, affected_tags, **kwargs): + # reset tag global counter + Tag.objects.filter(pk__in=[tag.pk for tag in affected_tags]).update(book_count=None) + + # if book tags changed, reset book tag counter + if isinstance(sender, Book) and \ + Tag.objects.filter(pk__in=(tag.pk for tag in affected_tags)).\ + exclude(category__in=('book', 'theme', 'set')).count(): + sender.reset_tag_counter() + # if fragment theme changed, reset book theme counter + elif isinstance(sender, Fragment) and \ + Tag.objects.filter(pk__in=(tag.pk for tag in affected_tags)).\ + filter(category='theme').count(): + sender.book.reset_theme_counter() +tags_updated.connect(_tags_updated_handler) +