X-Git-Url: https://git.mdrn.pl/wolnelektury.git/blobdiff_plain/e3f24620b9a69f1eb0652a9ffec2c2c3d6a10b8e..f2c1566914d2ee615546134b9e080fc32c1a1094:/apps/catalogue/models.py?ds=sidebyside diff --git a/apps/catalogue/models.py b/apps/catalogue/models.py index 7f6268fdb..6863b2a40 100644 --- a/apps/catalogue/models.py +++ b/apps/catalogue/models.py @@ -1,4 +1,7 @@ # -*- coding: utf-8 -*- +# This file is part of Wolnelektury, licensed under GNU Affero GPLv3 or later. +# Copyright © Fundacja Nowoczesna Polska. See NOTICE for more information. +# from django.db import models from django.db.models import permalink, Q from django.utils.translation import ugettext_lazy as _ @@ -6,12 +9,16 @@ from django.contrib.auth.models import User from django.core.files import File from django.template.loader import render_to_string from django.utils.safestring import mark_safe +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 TAG_CATEGORIES = ( @@ -21,6 +28,7 @@ TAG_CATEGORIES = ( ('genre', _('genre')), ('theme', _('theme')), ('set', _('set')), + ('book', _('book')), ) @@ -28,48 +36,129 @@ class TagSubcategoryManager(models.Manager): def __init__(self, subcategory): super(TagSubcategoryManager, self).__init__() self.subcategory = subcategory - + def get_query_set(self): return super(TagSubcategoryManager, self).get_query_set().filter(category=self.subcategory) 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, + category = models.CharField(_('category'), max_length=50, blank=False, null=False, db_index=True, choices=TAG_CATEGORIES) description = models.TextField(_('description'), blank=True) 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) - - def has_description(self): - return len(self.description) > 0 - has_description.short_description = _('description') - has_description.boolean = True + 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())) - @permalink - def get_absolute_url(self): - return ('catalogue.views.tagged_object_list', [self.slug]) - class Meta: ordering = ('sort_key',) verbose_name = _('tag') verbose_name_plural = _('tags') - + unique_together = (("slug", "category"),) + def __unicode__(self): return self.name + def __repr__(self): + return "Tag(slug=%r)" % self.slug + + @permalink + def get_absolute_url(self): + return ('catalogue.views.tagged_object_list', [self.url_chunk]) + + def has_description(self): + return len(self.description) > 0 + has_description.short_description = _('description') + has_description.boolean = True + + def alive(self): + return self.death is None + + def in_pd(self): + """ tests whether an author is in public domain """ + return self.death is not None and self.goes_to_pd() <= datetime.now().year + + def goes_to_pd(self): + """ 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 ? def book_upload_path(ext): def get_dynamic_path(book, filename): return 'lektura/%s.%s' % (book.slug, ext) @@ -83,145 +172,384 @@ class Book(models.Model): created_at = models.DateTimeField(_('creation date'), auto_now=True) _short_html = models.TextField(_('short HTML'), editable=False) parent_number = models.IntegerField(_('parent number'), default=0) - + extra_info = JSONField(_('extra information')) + gazeta_link = models.CharField(blank=True, max_length=240) + wiki_link = models.CharField(blank=True, max_length=240) + + # Formats xml_file = models.FileField(_('XML file'), upload_to=book_upload_path('xml'), blank=True) html_file = models.FileField(_('HTML file'), upload_to=book_upload_path('html'), blank=True) pdf_file = models.FileField(_('PDF file'), upload_to=book_upload_path('pdf'), blank=True) + epub_file = models.FileField(_('EPUB file'), upload_to=book_upload_path('epub'), blank=True) odt_file = models.FileField(_('ODT file'), upload_to=book_upload_path('odt'), blank=True) 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) + + class AlreadyExists(Exception): + pass + + class Meta: + ordering = ('title',) + verbose_name = _('book') + verbose_name_plural = _('books') + + def __unicode__(self): + return self.title + + 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) + + if refresh_mp3 and self.mp3_file: + print self.mp3_file, self.mp3_file.path + extra_info = self.get_extra_info_value() + extra_info.update(self.get_mp3_info()) + self.set_extra_info_value(extra_info) + book = super(Book, self).save(force_insert, force_update) + + return book + + @permalink + def get_absolute_url(self): + return ('catalogue.views.book_detail', [self.slug]) + + @property + def name(self): + return self.title + + def book_tag_slug(self): + return ('l-' + self.slug)[:120] + + def book_tag(self): + 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.save() + return book_tag + def short_html(self): - if len(self._short_html): - return mark_safe(self._short_html) + key = '_short_html_%s' % get_language() + short_html = getattr(self, key) + + if short_html and len(short_html): + return mark_safe(short_html) else: - tags = self.tags.filter(~Q(category__in=('set', 'theme'))) - tags = [u'%s' % (tag.get_absolute_url(), tag.name) for tag in tags] + tags = self.tags.filter(~Q(category__in=('set', 'theme', 'book'))) + tags = [mark_safe(u'%s' % (tag.get_absolute_url(), tag.name)) for tag in tags] formats = [] if self.html_file: - formats.append(u'Czytaj online' % reverse('book_text', kwargs={'slug': self.slug})) + formats.append(u'%s' % (reverse('book_text', kwargs={'slug': self.slug}), _('Read online'))) if self.pdf_file: - formats.append(u'Plik PDF' % self.pdf_file.url) + formats.append(u'PDF' % self.pdf_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'Plik ODT' % self.odt_file.url) - - self._short_html = unicode(render_to_string('catalogue/book_short.html', - {'book': self, 'tags': tags, 'formats': formats})) - self.save() - return mark_safe(self._short_html) - + formats.append(u'ODT' % self.odt_file.url) + if self.txt_file: + formats.append(u'TXT' % self.txt_file.url) + if self.mp3_file: + 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] + + setattr(self, key, unicode(render_to_string('catalogue/book_short.html', + {'book': self, 'tags': tags, 'formats': formats}))) + self.save(reset_short_html=False) + 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) + artist_name = ', '.join(', '.join(tag.text) for tag in audio.getall('TPE1')) + director_name = ', '.join(', '.join(tag.text) for tag in audio.getall('TPE3')) + return {'artist_name': artist_name, 'director_name': director_name} + def has_description(self): return len(self.description) > 0 has_description.short_description = _('description') has_description.boolean = True - + def has_pdf_file(self): return bool(self.pdf_file) has_pdf_file.short_description = 'PDF' has_pdf_file.boolean = True - + + def has_epub_file(self): + return bool(self.epub_file) + has_epub_file.short_description = 'EPUB' + has_epub_file.boolean = True + def has_odt_file(self): return bool(self.odt_file) has_odt_file.short_description = 'ODT' has_odt_file.boolean = True - + def has_html_file(self): return bool(self.html_file) has_html_file.short_description = 'HTML' has_html_file.boolean = True - @staticmethod - def from_xml_file(xml_file): + 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(open(xml_file)) + + try: + return cls.from_text_and_meta(xml_file, book_info, overwrite) + finally: + xml_file.close() + + @classmethod + def from_text_and_meta(cls, raw_file, book_info, overwrite=False): from tempfile import NamedTemporaryFile from slughifi import slughifi from markupstring import MarkupString - + from django.core.files.storage import default_storage + # Read book metadata - book_info = dcparser.parse(xml_file) book_base, book_slug = book_info.url.rsplit('/', 1) - book = Book(title=book_info.title, slug=book_slug) + book, created = Book.objects.get_or_create(slug=book_slug) + + if created: + book_shelves = [] + else: + if not overwrite: + raise Book.AlreadyExists(_('Book %s already exists') % book_slug) + # Save shelves for this book + book_shelves = list(book.tags.filter(category='set')) + + book.title = book_info.title + book.set_extra_info_value(book_info.to_dict()) + book._short_html = '' 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() + if hasattr(book_info, 'parts'): for n, part_url in enumerate(book_info.parts): base, slug = part_url.rsplit('/', 1) - child_book = Book.objects.get(slug=slug) - child_book.parent = book - child_book.parent_number = n - child_book.save() - + try: + child_book = Book.objects.get(slug=slug) + child_book.parent = book + child_book.parent_number = n + child_book.save() + except Book.DoesNotExist, e: + raise Book.DoesNotExist(_('Book with slug = "%s" does not exist.') % slug) + # Save XML and HTML files - book.xml_file.save('%s.xml' % book.slug, File(file(xml_file)), save=False) - + book.xml_file.save('%s.xml' % book.slug, raw_file, save=False) + html_file = NamedTemporaryFile() - html.transform(book.xml_file.path, html_file) - book.html_file.save('%s.html' % book.slug, File(html_file), save=False) - - # 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 = Fragment(text=text, short_text=short_text, anchor=fragment.id, book=book) - - 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 created: - tag.name = theme_name - tag.sort_key = slughifi(theme_name) - tag.category = 'theme' - tag.save() - themes.append(tag) - new_fragment.save() - new_fragment.tags = list(book.tags) + themes - book_themes += themes - - book_themes = set(book_themes) - book.tags = list(book.tags) + list(book_themes) - return book.save() - - @permalink - def get_absolute_url(self): - return ('catalogue.views.book_detail', [self.slug]) - - class Meta: - ordering = ('title',) - verbose_name = _('book') - verbose_name_plural = _('books') + 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) + + # Extract fragments + closed_fragments, open_fragments = html.extract_fragments(book.html_file.path) + for fragment in closed_fragments.values(): + try: + theme_names = [s.strip() for s in fragment.themes.split(',')] + except AttributeError: + continue + themes = [] + for theme_name in theme_names: + 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.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(book_tags + themes + [book_tag]) + + book.build_epub(remove_descendants=False) + + 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.tag_counter + book.theme_counter + + book.save() + return book + + + def refresh_tag_counter(self): + tags = {} + for child in self.children.all().order_by(): + for tag_pk, value in child.tag_counter.iteritems(): + tags[tag_pk] = tags.get(tag_pk, 0) + value + for tag in self.tags.exclude(category__in=('book', 'theme', 'set')).order_by(): + tags[tag.pk] = 1 + 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: + return self.refresh_tag_counter() + return dict((int(k), v) for k, v in self.get__tag_counter_value().iteritems()) + + def refresh_theme_counter(self): + tags = {} + for fragment in Fragment.tagged.with_any([self.book_tag()]).order_by(): + for tag in fragment.tags.filter(category='theme').order_by(): + tags[tag.pk] = tags.get(tag.pk, 0) + 1 + 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()) - def __unicode__(self): - return self.title class Fragment(models.Model): @@ -234,24 +562,83 @@ class Fragment(models.Model): objects = models.Manager() tagged = managers.ModelTaggedItemManager(Tag) tags = managers.TagDescriptor(Tag) - + + class Meta: + ordering = ('book', 'anchor',) + verbose_name = _('fragment') + verbose_name_plural = _('fragments') + + def get_absolute_url(self): + return '%s#m%s' % (reverse('book_text', kwargs={'slug': self.book.slug}), self.anchor) + def short_html(self): - if len(self._short_html): - return mark_safe(self._short_html) + key = '_short_html_%s' % get_language() + short_html = getattr(self, key) + if short_html and len(short_html): + return mark_safe(short_html) else: - book_authors = [u'%s' % (tag.get_absolute_url(), tag.name) - for tag in self.book.tags if tag.category == 'author'] - - self._short_html = unicode(render_to_string('catalogue/fragment_short.html', - {'fragment': self, 'book': self.book, 'book_authors': book_authors})) + setattr(self, key, unicode(render_to_string('catalogue/fragment_short.html', + {'fragment': self}))) self.save() - return mark_safe(self._short_html) - + return mark_safe(getattr(self, key)) + + +class BookStub(models.Model): + title = models.CharField(_('title'), max_length=120) + author = models.CharField(_('author'), max_length=120) + pd = models.IntegerField(_('goes to public domain'), null=True, blank=True) + slug = models.SlugField(_('slug'), max_length=120, unique=True, db_index=True) + translator = models.TextField(_('translator'), blank=True) + translator_death = models.TextField(_('year of translator\'s death'), blank=True) + + class Meta: + ordering = ('title',) + verbose_name = _('book stub') + verbose_name_plural = _('book stubs') + + def __unicode__(self): + return self.title + + @permalink def get_absolute_url(self): - return '%s#m%s' % (reverse('book_text', kwargs={'slug': self.book.slug}), self.anchor) - + return ('catalogue.views.book_detail', [self.slug]) + + def in_pd(self): + return self.pd is not None and self.pd <= datetime.now().year + + @property + def name(self): + 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 = ('book', 'anchor',) - verbose_name = _('fragment') - verbose_name_plural = _('fragments') + 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)