X-Git-Url: https://git.mdrn.pl/wolnelektury.git/blobdiff_plain/5a605e04abaa18a80e391c5cc26c48e0125d6f3c..b07520a11332374de645ed22c74959afa8a4b7ed:/apps/catalogue/models.py diff --git a/apps/catalogue/models.py b/apps/catalogue/models.py index b82879ba5..78e6b0b3d 100644 --- a/apps/catalogue/models.py +++ b/apps/catalogue/models.py @@ -2,7 +2,7 @@ # This file is part of Wolnelektury, licensed under GNU Affero GPLv3 or later. # Copyright © Fundacja Nowoczesna Polska. See NOTICE for more information. # -from datetime import datetime +from collections import namedtuple from django.db import models from django.db.models import permalink, Q @@ -23,13 +23,16 @@ from django.conf import settings from newtagging.models import TagBase, tags_updated from newtagging import managers from catalogue.fields import JSONField, OverwritingFileField -from catalogue.utils import create_zip +from catalogue.utils import create_zip, split_tags +from catalogue.tasks import touch_tag, index_book from shutil import copy from glob import glob import re from os import path +import search + TAG_CATEGORIES = ( ('author', _('author')), ('epoch', _('epoch')), @@ -40,13 +43,6 @@ TAG_CATEGORIES = ( ('book', _('book')), ) -MEDIA_FORMATS = ( - ('odt', _('ODT file')), - ('mp3', _('MP3 file')), - ('ogg', _('OGG file')), - ('daisy', _('DAISY file')), -) - # not quite, but Django wants you to set a timeout CACHE_FOREVER = 2419200 # 28 days @@ -67,7 +63,6 @@ class Tag(TagBase): 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=True, null=True) @@ -112,25 +107,22 @@ class Tag(TagBase): has_description.boolean = True 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): @@ -174,16 +166,43 @@ 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: - if media.type == 'daisy': - ext = 'daisy.zip' - else: - ext = media.type + # BookMedia case + ext = media.formats[media.type].ext if media is None or not media.name: name = slughifi(filename.split(".")[0]) else: @@ -202,8 +221,10 @@ def get_customized_pdf_path(book, customizations): """ customizations.sort() h = hash(tuple(customizations)) + pdf_name = '%s-custom-%s' % (book.slug, h) - pdf_file = models.get_dynamic_path(None, pdf_name, ext='pdf') + pdf_file = get_dynamic_path(None, pdf_name, ext='pdf') + return pdf_file @@ -218,7 +239,16 @@ def get_existing_customized_pdf(book): class BookMedia(models.Model): - type = models.CharField(_('type'), choices=MEDIA_FORMATS, max_length="100") + 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) @@ -241,7 +271,7 @@ class BookMedia(models.Model): try: old = BookMedia.objects.get(pk=self.pk) except BookMedia.DoesNotExist, e: - pass + old = None else: # if name changed, change the file name, too if slughifi(self.name) != slughifi(old.name): @@ -250,7 +280,9 @@ class BookMedia(models.Model): super(BookMedia, self).save(*args, **kwargs) # remove the zip package for book with modified media - remove_zip(self.book.slug) + 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()) @@ -319,7 +351,11 @@ class BookMedia(models.Model): class Book(models.Model): 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, unique=True, db_index=True) + 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) @@ -329,8 +365,11 @@ class Book(models.Model): wiki_link = models.CharField(blank=True, max_length=240) # files generated during publication - file_types = ['epub', 'html', 'mobi', 'pdf', 'txt', 'xml'] - + 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'] + parent = models.ForeignKey('self', blank=True, null=True, related_name='children') objects = models.Manager() tagged = managers.ModelTaggedItemManager(Tag) @@ -383,14 +422,14 @@ class Book(models.Model): return book_tag def has_media(self, type): - if type in Book.file_types: + if type in Book.formats: return bool(getattr(self, "%s_file" % type)) else: return self.media.filter(type=type).exists() def get_media(self, type): if self.has_media(type): - if type in Book.file_types: + if type in Book.formats: return getattr(self, "%s_file" % type) else: return self.media.filter(type=type) @@ -413,6 +452,7 @@ class Book(models.Model): cache_key = "Book.short_html/%d/%s" for lang, langname in settings.LANGUAGES: cache.delete(cache_key % (self.id, lang)) + cache.delete("Book.mini_box/%d" % (self.id, )) # Fragment.short_html relies on book's tags, so reset it here too for fragm in self.fragments.all(): fragm.reset_short_html() @@ -427,26 +467,15 @@ class Book(models.Model): if short_html is not None: return mark_safe(short_html) 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] + tags = self.tags.filter(category__in=('author', 'kind', 'genre', 'epoch')) + tags = split_tags(tags) - formats = [] + formats = {} # files generated during publication - if self.has_media("html"): - formats.append(u'%s' % (reverse('book_text', kwargs={'slug': self.slug}), _('Read online'))) - if self.has_media("pdf"): - formats.append(u'PDF' % self.get_media('pdf').url) - if self.has_media("mobi"): - formats.append(u'MOBI' % self.get_media('mobi').url) - if self.root_ancestor.has_media("epub"): - formats.append(u'EPUB' % self.root_ancestor.get_media('epub').url) - if self.has_media("txt"): - formats.append(u'TXT' % self.get_media('txt').url) - # other files - for m in self.media.order_by('type'): - formats.append(u'%s' % (m.file.url, m.type.upper())) - - formats = [mark_safe(format) for format in formats] + for ebook_format in self.ebook_formats: + if self.has_media(ebook_format): + formats[ebook_format] = self.get_media(ebook_format) + short_html = unicode(render_to_string('catalogue/book_short.html', {'book': self, 'tags': tags, 'formats': formats})) @@ -455,17 +484,22 @@ class Book(models.Model): cache.set(cache_key, short_html, CACHE_FOREVER) return mark_safe(short_html) - @property - def root_ancestor(self): - """ returns the oldest ancestor """ + def mini_box(self): + if self.id: + cache_key = "Book.mini_box/%d" % (self.id, ) + short_html = cache.get(cache_key) + else: + short_html = None - if not hasattr(self, '_root_ancestor'): - book = self - while book.parent: - book = book.parent - self._root_ancestor = book - return self._root_ancestor + if short_html is None: + authors = self.tags.filter(category='author') + short_html = unicode(render_to_string('catalogue/book_mini_box.html', + {'book': self, 'authors': authors, 'STATIC_URL': settings.STATIC_URL})) + + if self.id: + cache.set(cache_key, short_html, CACHE_FOREVER) + return mark_safe(short_html) def has_description(self): return len(self.description) > 0 @@ -473,11 +507,6 @@ class Book(models.Model): has_description.boolean = True # ugly ugly ugly - def has_odt_file(self): - return bool(self.has_media("odt")) - has_odt_file.short_description = 'ODT' - has_odt_file.boolean = True - def has_mp3_file(self): return bool(self.has_media("mp3")) has_mp3_file.short_description = 'MP3' @@ -493,116 +522,94 @@ class Book(models.Model): 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 django.core.files.base import ContentFile + from librarian.cover import WLCover + + if book_info is None: + book_info = self.wldocument().book_info + + cover = WLCover(book_info).image() + imgstr = StringIO() + cover.save(imgstr, 'png') + self.cover.save(None, ContentFile(imgstr.getvalue())) + 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 tempfile import NamedTemporaryFile from os import unlink from django.core.files import File - from librarian import pdf - from catalogue.utils import ORMDocProvider, remove_zip + from catalogue.utils import remove_zip - try: - pdf_file = NamedTemporaryFile(delete=False) - pdf.transform(ORMDocProvider(self), - file_path=str(self.xml_file.path), - output_file=pdf_file, - customizations=customizations - ) - - 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_file.name))) - self.pdf_file = current_self.pdf_file - else: - print "safing %s" % file_name - print "to: %s" % DefaultStorage().path(file_name) - DefaultStorage().save(file_name, File(open(pdf_file.name))) - finally: - unlink(pdf_file.name) + pdf = self.wldocument().as_pdf(customizations=customizations) + + 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 - # remove cached downloadables - remove_zip(settings.ALL_PDF_ZIP) - for customized_pdf in get_existing_customized_pdf(self): - unlink(customized_pdf) + # 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 tempfile import NamedTemporaryFile - from os import unlink from django.core.files import File - from librarian import mobi - from catalogue.utils import ORMDocProvider, remove_zip + from catalogue.utils import remove_zip - try: - mobi_file = NamedTemporaryFile(suffix='.mobi', delete=False) - mobi.transform(ORMDocProvider(self), verbose=1, - file_path=str(self.xml_file.path), - output_file=mobi_file.name, - ) + mobi = self.wldocument().as_mobi() - self.mobi_file.save('%s.mobi' % self.slug, File(open(mobi_file.name))) - finally: - unlink(mobi_file.name) + 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, 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 epub, NoDublinCore - from catalogue.utils import ORMDocProvider, remove_zip - - if self.parent: - # don't need an epub - return + def build_epub(self): + """(Re)builds the epub file.""" + from django.core.files import File + from catalogue.utils import remove_zip - epub_file = StringIO() - try: - epub.transform(ORMDocProvider(self), self.slug, output_file=epub_file) - self.epub_file.save('%s.epub' % self.slug, ContentFile(epub_file.getvalue())) - FileRecord(slug=self.slug, type='epub', sha1=sha1(epub_file.getvalue()).hexdigest()).save() - except NoDublinCore: - pass + epub = self.wldocument().as_epub() - 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()) + 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 StringIO import StringIO from django.core.files.base import ContentFile - from librarian import text - out = StringIO() - text.transform(open(self.xml_file.path), out) - self.txt_file.save('%s.txt' % self.slug, ContentFile(out.getvalue())) + text = self.wldocument().as_text() + self.txt_file.save('%s.txt' % self.slug, ContentFile(text.get_string())) def build_html(self): - from tempfile import NamedTemporaryFile from markupstring import MarkupString - from django.core.files import File + from django.core.files.base import ContentFile from slughifi import slughifi from librarian import html @@ -610,9 +617,10 @@ class Book(models.Model): category__in=('author', 'epoch', 'genre', 'kind'))) book_tag = self.book_tag() - html_file = NamedTemporaryFile() - if html.transform(self.xml_file.path, html_file, parse_dublincore=False): - self.html_file.save('%s.html' % self.slug, File(html_file)) + 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 = [] @@ -673,12 +681,25 @@ class Book(models.Model): getattr(settings, "ALL_%s_ZIP" % format_.upper())) return result.wait() - def zip_audiobooks(self): - bm = BookMedia.objects.filter(book=self, type='mp3') + 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, self.slug) + result = create_zip.delay(paths, "%s_%s" % (self.slug, format_)) return result.wait() + def search_index(self, book_info=None): + if settings.CELERY_ALWAYS_EAGER: + 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, **kwargs): from django.core.files import File @@ -697,25 +718,25 @@ class Book(models.Model): @classmethod def from_text_and_meta(cls, raw_file, book_info, overwrite=False, - build_epub=True, build_txt=True, build_pdf=True, build_mobi=True): + build_epub=True, build_txt=True, build_pdf=True, build_mobi=True, + search_index=True): import re - from slughifi import slughifi from sortify import sortify # check for parts before we do anything children = [] if hasattr(book_info, 'parts'): for part_url in book_info.parts: - base, slug = part_url.rsplit('/', 1) try: - children.append(Book.objects.get(slug=slug)) + children.append(Book.objects.get(slug=part_url.slug)) except Book.DoesNotExist, e: - raise Book.DoesNotExist(_('Book with slug = "%s" does not exist.') % slug) + raise Book.DoesNotExist(_('Book "%s" does not exist.') % + part_url.slug) # Read book metadata - book_base, book_slug = book_info.url.rsplit('/', 1) - if re.search(r'[^a-zA-Z0-9-]', book_slug): + 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) @@ -723,32 +744,21 @@ class Book(models.Model): 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.save() - meta_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 = sortify(tag_sort_key.lower()) - tag.save() - meta_tags.append(tag) + meta_tags = Tag.tags_from_info(book_info) book.tags = set(meta_tags + book_shelves) @@ -769,26 +779,35 @@ class Book(models.Model): if not settings.NO_BUILD_TXT and build_txt: book.build_txt() + book.build_cover(book_info) + if not settings.NO_BUILD_EPUB and build_epub: - book.root_ancestor.build_epub() + book.build_epub() if not settings.NO_BUILD_PDF and build_pdf: - book.root_ancestor.build_pdf() + book.build_pdf() if not settings.NO_BUILD_MOBI and build_mobi: book.build_mobi() + if not settings.NO_SEARCH_INDEX and 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] 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.delay(tag) + book.save() # refresh cache @@ -898,7 +917,8 @@ class Book(models.Model): """ books_by_parent = {} - books = cls.objects.all().order_by('parent_number', 'sort_key').only('title', 'parent', 'slug') + 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)) @@ -949,7 +969,7 @@ def _has_factory(ftype): # add the file fields -for t in Book.file_types: +for t in Book.formats: field_name = "%s_file" % t models.FileField(_("%s file" % t.upper()), upload_to=book_upload_path(t), @@ -974,7 +994,7 @@ 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: @@ -1001,20 +1021,6 @@ class Fragment(models.Model): return mark_safe(short_html) -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 @@ -1025,7 +1031,8 @@ class FileRecord(models.Model): def _tags_updated_handler(sender, affected_tags, **kwargs): # reset tag global counter # we want Tag.changed_at updated for API to know the tag was touched - Tag.objects.filter(pk__in=[tag.pk for tag in affected_tags]).update(book_count=None, changed_at=datetime.now()) + for tag in affected_tags: + touch_tag.delay(tag) # if book tags changed, reset book tag counter if isinstance(sender, Book) and \