X-Git-Url: https://git.mdrn.pl/wolnelektury.git/blobdiff_plain/ed20fe9aabacabc81232e58bed2372124e538afa..6e374f03fd5a0b9c7725f8268dd9c7d57d49ee89:/apps/catalogue/models.py?ds=inline diff --git a/apps/catalogue/models.py b/apps/catalogue/models.py index e692ad8c5..091e66148 100644 --- a/apps/catalogue/models.py +++ b/apps/catalogue/models.py @@ -2,25 +2,38 @@ # This file is part of Wolnelektury, licensed under GNU Affero GPLv3 or later. # Copyright © Fundacja Nowoczesna Polska. See NOTICE for more information. # +from collections import namedtuple + from django.db import models from django.db.models import permalink, Q +import django.dispatch +from django.core.cache import get_cache +from django.core.files.storage import DefaultStorage from django.utils.translation import ugettext_lazy as _ from django.contrib.auth.models import User -from django.core.files import File from django.template.loader import render_to_string +from django.utils.datastructures import SortedDict 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 django.db.models.signals import post_save, m2m_changed, pre_delete +import jsonfield + +from django.conf import settings from newtagging.models import TagBase, tags_updated from newtagging import managers -from catalogue.fields import JSONField - -from librarian import dcparser, html, epub, NoDublinCore -from mutagen import id3 +from catalogue.fields import JSONField, OverwritingFileField +from catalogue.utils import create_zip, split_tags, truncate_html_words +from catalogue.tasks import touch_tag, index_book +from shutil import copy +from glob import glob +import re +from os import path +import search +# Those are hard-coded here so that makemessages sees them. TAG_CATEGORIES = ( ('author', _('author')), ('epoch', _('epoch')), @@ -32,6 +45,9 @@ TAG_CATEGORIES = ( ) +permanent_cache = get_cache('permanent') + + class TagSubcategoryManager(models.Manager): def __init__(self, subcategory): super(TagSubcategoryManager, self).__init__() @@ -44,18 +60,22 @@ class TagSubcategoryManager(models.Manager): class Tag(TagBase): name = models.CharField(_('name'), max_length=50, 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) + sort_key = models.CharField(_('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) 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) - book_count = models.IntegerField(_('book count'), blank=False, null=True) - death = models.IntegerField(_(u'year of death'), blank=True, null=True) + book_count = models.IntegerField(_('book count'), blank=True, null=True) gazeta_link = models.CharField(blank=True, max_length=240) wiki_link = models.CharField(blank=True, max_length=240) + created_at = models.DateTimeField(_('creation date'), auto_now_add=True, db_index=True) + changed_at = models.DateTimeField(_('creation date'), auto_now=True, db_index=True) + + class UrlDeprecationWarning(DeprecationWarning): + pass + categories_rev = { 'autor': 'author', 'epoka': 'epoch', @@ -87,37 +107,23 @@ class Tag(TagBase): 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 + """Returns global book count for book tags, fragment count for themes.""" + + 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) + return objects.count() @staticmethod def get_tag_list(tags): @@ -125,19 +131,20 @@ class Tag(TagBase): real_tags = [] ambiguous_slugs = [] category = None + deprecated = False tags_splitted = tags.split('/') - for index, name in enumerate(tags_splitted): - if name in Tag.categories_rev: + for name in tags_splitted: + if category: + real_tags.append(Tag.objects.get(slug=name, category=category)) + category = None + elif 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) + try: + real_tags.append(Tag.objects.exclude(category='book').get(slug=name)) + deprecated = True + except Tag.MultipleObjectsReturned, e: + ambiguous_slugs.append(name) if category: # something strange left off @@ -148,8 +155,11 @@ class Tag(TagBase): e.tags = real_tags e.ambiguous_slugs = ambiguous_slugs raise e - else: - return real_tags + if deprecated: + e = Tag.UrlDeprecationWarning() + e.tags = real_tags + raise e + return real_tags else: return TagBase.get_tag_list(tags) @@ -157,77 +167,243 @@ class Tag(TagBase): def url_chunk(self): return '/'.join((Tag.categories_dict[self.category], self.slug)) + @staticmethod + def tags_from_info(info): + from slughifi import slughifi + from sortify import sortify + meta_tags = [] + categories = (('kinds', 'kind'), ('genres', 'genre'), ('authors', 'author'), ('epochs', 'epoch')) + for field_name, category in categories: + try: + tag_names = getattr(info, field_name) + except: + try: + tag_names = [getattr(info, category)] + except: + # For instance, Pictures do not have 'genre' field. + continue + for tag_name in tag_names: + tag_sort_key = tag_name + if category == 'author': + tag_sort_key = tag_name.last_name + tag_name = tag_name.readable() + tag, created = Tag.objects.get_or_create(slug=slughifi(tag_name), category=category) + if created: + tag.name = tag_name + tag.sort_key = sortify(tag_sort_key.lower()) + tag.save() + meta_tags.append(tag) + return meta_tags + + + +def get_dynamic_path(media, filename, ext=None, maxlen=100): + from slughifi import slughifi + + # how to put related book's slug here? + if not ext: + # BookMedia case + ext = media.formats[media.type].ext + if media is None or not media.name: + name = slughifi(filename.split(".")[0]) + else: + name = slughifi(media.name) + return 'book/%s/%s.%s' % (ext, name[:maxlen-len('book/%s/.%s' % (ext, ext))-4], ext) + # TODO: why is this hard-coded ? -def book_upload_path(ext): - def get_dynamic_path(book, filename): - return 'lektura/%s.%s' % (book.slug, ext) - return get_dynamic_path +def book_upload_path(ext=None, maxlen=100): + return lambda *args: get_dynamic_path(*args, ext=ext, maxlen=maxlen) + + +def get_customized_pdf_path(book, customizations): + """ + Returns a MEDIA_ROOT relative path for a customized pdf. The name will contain a hash of customization options. + """ + customizations.sort() + h = hash(tuple(customizations)) + + pdf_name = '%s-custom-%s' % (book.slug, h) + pdf_file = get_dynamic_path(None, pdf_name, ext='pdf') + + return pdf_file + + +def get_existing_customized_pdf(book): + """ + Returns a list of paths to generated customized pdf of a book + """ + pdf_glob = '%s-custom-' % (book.slug,) + pdf_glob = get_dynamic_path(None, pdf_glob, ext='pdf') + pdf_glob = re.sub(r"[.]([a-z0-9]+)$", "*.\\1", pdf_glob) + return glob(path.join(settings.MEDIA_ROOT, pdf_glob)) + + +class BookMedia(models.Model): + FileFormat = namedtuple("FileFormat", "name ext") + formats = SortedDict([ + ('mp3', FileFormat(name='MP3', ext='mp3')), + ('ogg', FileFormat(name='Ogg Vorbis', ext='ogg')), + ('daisy', FileFormat(name='DAISY', ext='daisy.zip')), + ]) + format_choices = [(k, _('%s file') % t.name) + for k, t in formats.items()] + + type = models.CharField(_('type'), choices=format_choices, max_length="100") + name = models.CharField(_('name'), max_length="100") + file = OverwritingFileField(_('file'), upload_to=book_upload_path()) + uploaded_at = models.DateTimeField(_('creation date'), auto_now_add=True, editable=False) + extra_info = JSONField(_('extra information'), default='{}', editable=False) + book = models.ForeignKey('Book', related_name='media') + source_sha1 = models.CharField(null=True, blank=True, max_length=40, editable=False) + + def __unicode__(self): + return "%s (%s)" % (self.name, self.file.name.split("/")[-1]) + + class Meta: + ordering = ('type', 'name') + verbose_name = _('book media') + verbose_name_plural = _('book media') + + def save(self, *args, **kwargs): + from slughifi import slughifi + from catalogue.utils import ExistingFile, remove_zip + + try: + old = BookMedia.objects.get(pk=self.pk) + except BookMedia.DoesNotExist, e: + old = None + else: + # if name changed, change the file name, too + if slughifi(self.name) != slughifi(old.name): + self.file.save(None, ExistingFile(self.file.path), save=False, leave=True) + + super(BookMedia, self).save(*args, **kwargs) + + # remove the zip package for book with modified media + if old: + remove_zip("%s_%s" % (old.book.slug, old.type)) + remove_zip("%s_%s" % (self.book.slug, self.type)) + + extra_info = self.get_extra_info_value() + extra_info.update(self.read_meta()) + self.set_extra_info_value(extra_info) + self.source_sha1 = self.read_source_sha1(self.file.path, self.type) + return super(BookMedia, self).save(*args, **kwargs) + + def read_meta(self): + """ + Reads some metadata from the audiobook. + """ + import mutagen + from mutagen import id3 + + artist_name = director_name = project = funded_by = '' + if self.type == 'mp3': + try: + audio = id3.ID3(self.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')) + project = ", ".join([t.data for t in audio.getall('PRIV') + if t.owner=='wolnelektury.pl?project']) + funded_by = ", ".join([t.data for t in audio.getall('PRIV') + if t.owner=='wolnelektury.pl?funded_by']) + except: + pass + elif self.type == 'ogg': + try: + audio = mutagen.File(self.file.path) + artist_name = ', '.join(audio.get('artist', [])) + director_name = ', '.join(audio.get('conductor', [])) + project = ", ".join(audio.get('project', [])) + funded_by = ", ".join(audio.get('funded_by', [])) + except: + pass + else: + return {} + return {'artist_name': artist_name, 'director_name': director_name, + 'project': project, 'funded_by': funded_by} + + @staticmethod + def read_source_sha1(filepath, filetype): + """ + Reads source file SHA1 from audiobok metadata. + """ + import mutagen + from mutagen import id3 + + if filetype == 'mp3': + try: + audio = id3.ID3(filepath) + return [t.data for t in audio.getall('PRIV') + if t.owner=='wolnelektury.pl?flac_sha1'][0] + except: + return None + elif filetype == 'ogg': + try: + audio = mutagen.File(filepath) + return audio.get('flac_sha1', [None])[0] + except: + return None + else: + return None class Book(models.Model): - title = models.CharField(_('title'), max_length=120) - slug = models.SlugField(_('slug'), max_length=120, unique=True, db_index=True) - description = models.TextField(_('description'), blank=True) - created_at = models.DateTimeField(_('creation date'), auto_now=True) - _short_html = models.TextField(_('short HTML'), editable=False) + title = models.CharField(_('title'), max_length=120) + sort_key = models.CharField(_('sort key'), max_length=120, db_index=True, editable=False) + slug = models.SlugField(_('slug'), max_length=120, db_index=True, + unique=True) + common_slug = models.SlugField(_('slug'), max_length=120, db_index=True) + language = models.CharField(_('language code'), max_length=3, db_index=True, + default=settings.CATALOGUE_DEFAULT_LANGUAGE) + description = models.TextField(_('description'), blank=True) + created_at = models.DateTimeField(_('creation date'), auto_now_add=True, db_index=True) + changed_at = models.DateTimeField(_('creation date'), auto_now=True, db_index=True) 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) + extra_info = JSONField(_('extra information'), default='{}') + gazeta_link = models.CharField(blank=True, max_length=240) + wiki_link = models.CharField(blank=True, max_length=240) + # files generated during publication + cover = models.FileField(_('cover'), upload_to=book_upload_path('png'), + null=True, blank=True) + ebook_formats = ['pdf', 'epub', 'mobi', 'txt'] + formats = ebook_formats + ['html', 'xml'] - # 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') - parent = models.ForeignKey('self', blank=True, null=True, related_name='children') + _related_info = jsonfield.JSONField(blank=True, null=True, editable=False) - objects = models.Manager() - tagged = managers.ModelTaggedItemManager(Tag) - tags = managers.TagDescriptor(Tag) + 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) + html_built = django.dispatch.Signal() + published = django.dispatch.Signal() class AlreadyExists(Exception): pass class Meta: - ordering = ('title',) + ordering = ('sort_key',) 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): + def save(self, force_insert=False, force_update=False, reset_short_html=True, **kwargs): + from sortify import sortify + + self.sort_key = sortify(self.title) + + ret = super(Book, self).save(force_insert, force_update) + 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) + self.reset_short_html() - return book + return ret @permalink def get_absolute_url(self): @@ -245,138 +421,246 @@ class Book(models.Model): 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.sort_key = self.title.lower() book_tag.save() return book_tag - def short_html(self): - key = '_short_html_%s' % get_language() - short_html = getattr(self, key) - - if short_html and len(short_html): - return mark_safe(short_html) + def has_media(self, type): + if type in Book.formats: + return bool(getattr(self, "%s_file" % type)) else: - 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'%s' % (reverse('book_text', kwargs={'slug': self.slug}), _('Read online'))) - if self.pdf_file: - 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'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 - + return self.media.filter(type=type).exists() + + def get_media(self, type): + if self.has_media(type): + if type in Book.formats: + return getattr(self, "%s_file" % type) + else: + return self.media.filter(type=type) + else: + return None + + def get_mp3(self): + return self.get_media("mp3") + def get_odt(self): + return self.get_media("odt") + def get_ogg(self): + return self.get_media("ogg") + def get_daisy(self): + return self.get_media("daisy") + + def reset_short_html(self): + if self.id is None: + return - 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} + type(self).objects.filter(pk=self.pk).update(_related_info=None) + # Fragment.short_html relies on book's tags, so reset it here too + for fragm in self.fragments.all(): + fragm.reset_short_html() 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 - - 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. - """ - + # ugly ugly ugly + def has_mp3_file(self): + return bool(self.has_media("mp3")) + has_mp3_file.short_description = 'MP3' + has_mp3_file.boolean = True + + def has_ogg_file(self): + return bool(self.has_media("ogg")) + has_ogg_file.short_description = 'OGG' + has_ogg_file.boolean = True + + def has_daisy_file(self): + return bool(self.has_media("daisy")) + has_daisy_file.short_description = 'DAISY' + has_daisy_file.boolean = True + + def wldocument(self, parse_dublincore=True): + from catalogue.import_utils import ORMDocProvider + from librarian.parser import WLDocument + + return WLDocument.from_file(self.xml_file.path, + provider=ORMDocProvider(self), + parse_dublincore=parse_dublincore) + + def build_cover(self, book_info=None): + """(Re)builds the cover image.""" from StringIO import StringIO - from hashlib import sha1 from django.core.files.base import ContentFile - from librarian import DocProvider + from librarian.cover import WLCover - class BookImportDocProvider(DocProvider): - """ used for joined EPUBs """ + if book_info is None: + book_info = self.wldocument().book_info - def __init__(self, book): - self.book = book + cover = WLCover(book_info).image() + imgstr = StringIO() + cover.save(imgstr, 'png') + self.cover.save(None, ContentFile(imgstr.getvalue())) - def by_slug(self, slug): - if slug == self.book.slug: - return self.book.xml_file - else: - return Book.objects.get(slug=slug).xml_file + def build_pdf(self, customizations=None, file_name=None): + """ (Re)builds the pdf file. + customizations - customizations which are passed to LaTeX class file. + file_name - save the pdf file under a different name and DO NOT save it in db. + """ + from os import unlink + from django.core.files import File + from catalogue.utils import remove_zip - if self.parent: - # don't need an epub - return + pdf = self.wldocument().as_pdf(customizations=customizations) - 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 + if file_name is None: + # we'd like to be sure not to overwrite changes happening while + # (timely) pdf generation is taking place (async celery scenario) + current_self = Book.objects.get(id=self.id) + current_self.pdf_file.save('%s.pdf' % self.slug, + File(open(pdf.get_filename()))) + self.pdf_file = current_self.pdf_file - 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()) + # remove cached downloadables + remove_zip(settings.ALL_PDF_ZIP) + + for customized_pdf in get_existing_customized_pdf(self): + unlink(customized_pdf) + else: + print "saving %s" % file_name + print "to: %s" % DefaultStorage().path(file_name) + DefaultStorage().save(file_name, File(open(pdf.get_filename()))) + + def build_mobi(self): + """ (Re)builds the MOBI file. + + """ + from django.core.files import File + from catalogue.utils import remove_zip + + mobi = self.wldocument().as_mobi() + + self.mobi_file.save('%s.mobi' % self.slug, File(open(mobi.get_filename()))) + + # remove zip with all mobi files + remove_zip(settings.ALL_MOBI_ZIP) + + def build_epub(self): + """(Re)builds the epub file.""" + from django.core.files import File + from catalogue.utils import remove_zip + + epub = self.wldocument().as_epub() + + self.epub_file.save('%s.epub' % self.slug, + File(open(epub.get_filename()))) + + # remove zip package with all epub files + remove_zip(settings.ALL_EPUB_ZIP) + + def build_txt(self): + from django.core.files.base import ContentFile + + text = self.wldocument().as_text() + self.txt_file.save('%s.txt' % self.slug, ContentFile(text.get_string())) + + + def build_html(self): + from django.core.files.base import ContentFile + from slughifi import slughifi + from librarian import html + + meta_tags = list(self.tags.filter( + category__in=('author', 'epoch', 'genre', 'kind'))) + book_tag = self.book_tag() + + html_output = self.wldocument(parse_dublincore=False).as_html() + if html_output: + self.html_file.save('%s.html' % self.slug, + ContentFile(html_output.get_string())) + + # get ancestor l-tags for adding to new fragments + ancestor_tags = [] + p = self.parent + while p: + ancestor_tags.append(p.book_tag()) + p = p.parent + + # Delete old fragments and create them from scratch + self.fragments.all().delete() + # Extract fragments + closed_fragments, open_fragments = html.extract_fragments(self.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 = theme_name.lower() + tag.save() + themes.append(tag) + if not themes: + continue + + text = fragment.to_string() + short_text = truncate_html_words(text, 15) + if text == short_text: + short_text = '' + new_fragment = Fragment.objects.create(anchor=fragment.id, book=self, + text=text, short_text=short_text) + + new_fragment.save() + new_fragment.tags = set(meta_tags + themes + [book_tag] + ancestor_tags) + self.save() + self.html_built.send(sender=self) + return True + return False + @staticmethod + def zip_format(format_): + def pretty_file_name(book): + return "%s/%s.%s" % ( + b.get_extra_info_value()['author'], + b.slug, + format_) + + field_name = "%s_file" % format_ + books = Book.objects.filter(parent=None).exclude(**{field_name: ""}) + paths = [(pretty_file_name(b), getattr(b, field_name).path) + for b in books] + result = create_zip.delay(paths, + getattr(settings, "ALL_%s_ZIP" % format_.upper())) + return result.wait() + + def zip_audiobooks(self, format_): + bm = BookMedia.objects.filter(book=self, type=format_) + paths = map(lambda bm: (None, bm.file.path), bm) + result = create_zip.delay(paths, "%s_%s" % (self.slug, format_)) + return result.wait() + + def search_index(self, book_info=None, reuse_index=False): + if reuse_index: + idx = search.ReusableIndex() + else: + idx = search.Index() + + idx.open() + try: + idx.index_book(self, book_info) + idx.index_tags() + finally: + idx.close() @classmethod - def from_xml_file(cls, xml_file, overwrite=False): + def from_xml_file(cls, xml_file, **kwargs): + from django.core.files import File + from librarian import dcparser + # use librarian to parse meta-data book_info = dcparser.parse(xml_file) @@ -384,67 +668,62 @@ class Book(models.Model): xml_file = File(open(xml_file)) try: - return cls.from_text_and_meta(xml_file, book_info, overwrite) + return cls.from_text_and_meta(xml_file, book_info, **kwargs) 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 + def from_text_and_meta(cls, raw_file, book_info, overwrite=False, + build_epub=True, build_txt=True, build_pdf=True, build_mobi=True, + search_index=True): + import re + from sortify import sortify + + # check for parts before we do anything + children = [] + if hasattr(book_info, 'parts'): + for part_url in book_info.parts: + try: + children.append(Book.objects.get(slug=part_url.slug)) + except Book.DoesNotExist, e: + raise Book.DoesNotExist(_('Book "%s" does not exist.') % + part_url.slug) + # Read book metadata - book_base, book_slug = book_info.url.rsplit('/', 1) + book_slug = book_info.url.slug + if re.search(r'[^a-z0-9-]', book_slug): + raise ValueError('Invalid characters in 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) + raise Book.AlreadyExists(_('Book %s already exists') % ( + book_slug)) # Save shelves for this book book_shelves = list(book.tags.filter(category='set')) + book.language = book_info.language book.title = book_info.title + if book_info.variant_of: + book.common_slug = book_info.variant_of.slug + else: + book.common_slug = book.slug book.set_extra_info_value(book_info.to_dict()) - book._short_html = '' book.save() - 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) + meta_tags = Tag.tags_from_info(book_info) - book.tags = book_tags + book_shelves + book.tags = set(meta_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) - 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) + for n, child_book in enumerate(children): + child_book.parent = book + child_book.parent_number = n + child_book.save() # Save XML and HTML files book.xml_file.save('%s.xml' % book.slug, raw_file, save=False) @@ -452,121 +731,253 @@ class Book(models.Model): # 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) + if book.build_html(): + if not settings.NO_BUILD_TXT and build_txt: + book.build_txt() - # 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 + book.build_cover(book_info) - # 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 + if not settings.NO_BUILD_EPUB and build_epub: + book.build_epub() - 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}) + if not settings.NO_BUILD_PDF and build_pdf: + book.build_pdf() - new_fragment.save() - new_fragment.tags = set(book_tags + themes + [book_tag] + ancestor_tags) + if not settings.NO_BUILD_MOBI and build_mobi: + book.build_mobi() - if not book.parent: - book.build_epub(remove_descendants=False) + if not settings.NO_SEARCH_INDEX and search_index: + book.search_index() + #index_book.delay(book.id, book_info) book_descendants = list(book.children.all()) + descendants_tags = set() # add l-tag to descendants and their fragments - # delete unnecessary EPUB files while len(book_descendants) > 0: child_book = book_descendants.pop(0) + descendants_tags.update(child_book.tags) 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()) + for tag in descendants_tags: + touch_tag(tag) + + book.save() + # refresh cache book.reset_tag_counter() book.reset_theme_counter() - book.save() + cls.published.send(sender=book) return book + def related_info(self): + """Keeps info about related objects (tags, media) in cache field.""" + if self._related_info is not None: + return self._related_info + else: + rel = {'tags': {}, 'media': {}} - 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 + tags = self.tags.filter(category__in=( + 'author', 'kind', 'genre', 'epoch')) + tags = split_tags(tags) + for category in tags: + rel['tags'][category] = [ + (t.name, t.slug) for t in tags[category]] + + for media_format in BookMedia.formats: + rel['media'][media_format] = self.has_media(media_format) + + book = self + parents = [] + while book.parent: + parents.append((book.parent.title, book.parent.slug)) + book = book.parent + parents = parents[::-1] + if parents: + rel['parents'] = parents + + if self.pk: + type(self).objects.filter(pk=self.pk).update(_related_info=rel) + return rel def reset_tag_counter(self): - self._tag_counter = None - self.save(reset_short_html=False, refresh_mp3=False) + if self.id is None: + return + + cache_key = "Book.tag_counter/%d" % self.id + permanent_cache.delete(cache_key) 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) + if self.id: + cache_key = "Book.tag_counter/%d" % self.id + tags = permanent_cache.get(cache_key) + else: + tags = None + + if tags is None: + 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 + + if self.id: + permanent_cache.set(cache_key, tags) return tags def reset_theme_counter(self): - self._theme_counter = None - self.save(reset_short_html=False, refresh_mp3=False) + if self.id is None: + return + + cache_key = "Book.theme_counter/%d" % self.id + permanent_cache.delete(cache_key) 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()) + if self.id: + cache_key = "Book.theme_counter/%d" % self.id + tags = permanent_cache.get(cache_key) + else: + tags = None + + if tags is None: + 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 + + if self.id: + permanent_cache.set(cache_key, tags) + return tags + + def pretty_title(self, html_links=False): + book = self + names = list(book.tags.filter(category='author')) + + books = [] + while book: + books.append(book) + book = book.parent + names.extend(reversed(books)) + + if html_links: + names = ['%s' % (tag.get_absolute_url(), tag.name) for tag in names] + else: + names = [tag.name for tag in names] + + return ', '.join(names) + + @classmethod + def tagged_top_level(cls, tags): + """ Returns top-level books tagged with `tags'. + + It only returns those books which don't have ancestors which are + also tagged with those tags. + + """ + # get relevant books and their tags + objects = cls.tagged.with_all(tags) + # eliminate descendants + l_tags = Tag.objects.filter(category='book', slug__in=[book.book_tag_slug() for book in objects]) + descendants_keys = [book.pk for book in cls.tagged.with_any(l_tags)] + if descendants_keys: + objects = objects.exclude(pk__in=descendants_keys) + + return objects + + @classmethod + def book_list(cls, filter=None): + """Generates a hierarchical listing of all books. + + Books are optionally filtered with a test function. + + """ + + books_by_parent = {} + books = cls.objects.all().order_by('parent_number', 'sort_key').only( + 'title', 'parent', 'slug') + if filter: + books = books.filter(filter).distinct() + book_ids = set((book.pk for book in books)) + for book in books: + parent = book.parent_id + if parent not in book_ids: + parent = None + books_by_parent.setdefault(parent, []).append(book) + else: + for book in books: + books_by_parent.setdefault(book.parent_id, []).append(book) + + orphans = [] + books_by_author = SortedDict() + for tag in Tag.objects.filter(category='author'): + books_by_author[tag] = [] + + for book in books_by_parent.get(None,()): + authors = list(book.tags.filter(category='author')) + if authors: + for author in authors: + books_by_author[author].append(book) + else: + orphans.append(book) + + return books_by_author, orphans, books_by_parent + + _audiences_pl = { + "SP1": (1, u"szkoła podstawowa"), + "SP2": (1, u"szkoła podstawowa"), + "P": (1, u"szkoła podstawowa"), + "G": (2, u"gimnazjum"), + "L": (3, u"liceum"), + "LP": (3, u"liceum"), + } + def audiences_pl(self): + audiences = self.get_extra_info_value().get('audiences', []) + audiences = sorted(set([self._audiences_pl[a] for a in audiences])) + return [a[1] for a in audiences] + + def choose_fragment(self): + tag = self.book_tag() + fragments = Fragment.tagged.with_any([tag]) + if fragments.exists(): + return fragments.order_by('?')[0] + elif self.parent: + return self.parent.choose_fragment() + else: + return None + + +def _has_factory(ftype): + has = lambda self: bool(getattr(self, "%s_file" % ftype)) + has.short_description = t.upper() + has.boolean = True + has.__name__ = "has_%s_file" % ftype + return has + + +# add the file fields +for t in Book.formats: + field_name = "%s_file" % t + models.FileField(_("%s file" % t.upper()), + upload_to=book_upload_path(t), + blank=True).contribute_to_class(Book, field_name) + setattr(Book, "has_%s_file" % t, _has_factory(t)) class Fragment(models.Model): text = models.TextField() short_text = models.TextField(editable=False) - _short_html = models.TextField(editable=False) anchor = models.CharField(max_length=120) book = models.ForeignKey(Book, related_name='fragments') @@ -580,66 +991,67 @@ class Fragment(models.Model): verbose_name_plural = _('fragments') def get_absolute_url(self): - return '%s#m%s' % (reverse('book_text', kwargs={'slug': self.book.slug}), self.anchor) + return '%s#m%s' % (reverse('book_text', args=[self.book.slug]), self.anchor) + + def reset_short_html(self): + if self.id is None: + return + + cache_key = "Fragment.short_html/%d/%s" + for lang, langname in settings.LANGUAGES: + permanent_cache.delete(cache_key % (self.id, lang)) + + def get_short_text(self): + """Returns short version of the fragment.""" + return self.short_text if self.short_text else self.text def short_html(self): - key = '_short_html_%s' % get_language() - short_html = getattr(self, key) - if short_html and len(short_html): + if self.id: + cache_key = "Fragment.short_html/%d/%s" % (self.id, get_language()) + short_html = permanent_cache.get(cache_key) + else: + short_html = None + + if short_html is not None: return mark_safe(short_html) else: - setattr(self, key, unicode(render_to_string('catalogue/fragment_short.html', - {'fragment': self}))) - self.save() - return mark_safe(getattr(self, key)) + short_html = unicode(render_to_string('catalogue/fragment_short.html', + {'fragment': self})) + if self.id: + permanent_cache.set(cache_key, short_html) + return mark_safe(short_html) + +class Collection(models.Model): + """A collection of books, which might be defined before publishing them.""" + title = models.CharField(_('title'), max_length=120, db_index=True) + slug = models.SlugField(_('slug'), max_length=120, primary_key=True) + description = models.TextField(_('description'), null=True, blank=True) -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) + models.SlugField(_('slug'), max_length=120, unique=True, db_index=True) + book_slugs = models.TextField(_('book slugs')) class Meta: ordering = ('title',) - verbose_name = _('book stub') - verbose_name_plural = _('book stubs') + verbose_name = _('collection') + verbose_name_plural = _('collections') def __unicode__(self): return self.title - @permalink - def get_absolute_url(self): - 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 = ('-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) +########### +# +# SIGNALS +# +########### 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) + # we want Tag.changed_at updated for API to know the tag was touched + for tag in affected_tags: + touch_tag(tag) # if book tags changed, reset book tag counter if isinstance(sender, Book) and \ @@ -653,3 +1065,15 @@ def _tags_updated_handler(sender, affected_tags, **kwargs): sender.book.reset_theme_counter() tags_updated.connect(_tags_updated_handler) + +def _pre_delete_handler(sender, instance, **kwargs): + """ refresh Book on BookMedia delete """ + if sender == BookMedia: + instance.book.save() +pre_delete.connect(_pre_delete_handler) + +def _post_save_handler(sender, instance, **kwargs): + """ refresh all the short_html stuff on BookMedia update """ + if sender == BookMedia: + instance.book.save() +post_save.connect(_post_save_handler)