X-Git-Url: https://git.mdrn.pl/wolnelektury.git/blobdiff_plain/73e3cc6c4b80f1afbfc78fc7ccd6f6cfc90b719c..2f0409de9ac65bcd54064bf925539a1fe89225bb:/apps/catalogue/models.py?ds=inline diff --git a/apps/catalogue/models.py b/apps/catalogue/models.py index 0273e47c7..22cd0ea76 100644 --- a/apps/catalogue/models.py +++ b/apps/catalogue/models.py @@ -8,10 +8,11 @@ from django.db import models from django.db.models import permalink, Q import django.dispatch from django.core.cache import 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 @@ -22,14 +23,12 @@ 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 ExistingFile, BookImportDocProvider +from catalogue.utils import create_zip +from shutil import copy +from glob import glob +import re +from os import path -from librarian import dcparser, html, epub, NoDublinCore -import mutagen -from mutagen import id3 -from slughifi import slughifi -from sortify import sortify -from os import unlink TAG_CATEGORIES = ( ('author', _('author')), @@ -175,22 +174,74 @@ 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: + tag_names = [getattr(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) + 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 + 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=None, maxlen=100): - def get_dynamic_path(media, filename, ext=ext): - # how to put related book's slug here? - if not ext: - if media.type == 'daisy': - ext = 'daisy.zip' - else: - ext = media.type - if 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) - return get_dynamic_path + 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.fileid(), 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.fileid(),) + 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): @@ -211,6 +262,9 @@ class BookMedia(models.Model): 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: @@ -221,6 +275,10 @@ class BookMedia(models.Model): 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 + remove_zip(self.book.fileid()) + extra_info = self.get_extra_info_value() extra_info.update(self.read_meta()) self.set_extra_info_value(extra_info) @@ -231,6 +289,8 @@ class BookMedia(models.Model): """ Reads some metadata from the audiobook. """ + import mutagen + from mutagen import id3 artist_name = director_name = project = funded_by = '' if self.type == 'mp3': @@ -263,6 +323,8 @@ class BookMedia(models.Model): """ Reads source file SHA1 from audiobok metadata. """ + import mutagen + from mutagen import id3 if filetype == 'mp3': try: @@ -283,8 +345,10 @@ 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) + sort_key = models.CharField(_('sort key'), max_length=120, db_index=True, editable=False) + 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) @@ -293,11 +357,8 @@ class Book(models.Model): gazeta_link = models.CharField(blank=True, max_length=240) wiki_link = models.CharField(blank=True, max_length=240) # files generated during publication - 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) - txt_file = models.FileField(_('TXT file'), upload_to=book_upload_path('txt'), blank=True) + + file_types = ['epub', 'html', 'mobi', 'pdf', 'txt', 'xml'] parent = models.ForeignKey('self', blank=True, null=True, related_name='children') objects = models.Manager() @@ -305,11 +366,16 @@ class Book(models.Model): tags = managers.TagDescriptor(Tag) html_built = django.dispatch.Signal() + published = django.dispatch.Signal() + + URLID_RE = r'[a-z0-9-]+(?:/[a-z]{3})?' + FILEID_RE = r'[a-z0-9-]+(?:_[a-z]{3})?' class AlreadyExists(Exception): pass class Meta: + unique_together = [['slug', 'language']] ordering = ('sort_key',) verbose_name = _('book') verbose_name_plural = _('books') @@ -317,7 +383,45 @@ class Book(models.Model): def __unicode__(self): return self.title + def urlid(self, sep='/'): + stem = self.slug + if self.language != settings.CATALOGUE_DEFAULT_LANGUAGE: + stem += sep + self.language + return stem + + def fileid(self): + return self.urlid('_') + + @staticmethod + def split_urlid(urlid, sep='/', default_lang=settings.CATALOGUE_DEFAULT_LANGUAGE): + """Splits a URL book id into slug and language code. + + Returns a dictionary usable i.e. for object lookup, or None. + + >>> Book.split_urlid("a-slug/pol", default_lang="eng") + {'slug': 'a-slug', 'language': 'pol'} + >>> Book.split_urlid("a-slug", default_lang="eng") + {'slug': 'a-slug', 'language': 'eng'} + >>> Book.split_urlid("a-slug_pol", "_", default_lang="eng") + {'slug': 'a-slug', 'language': 'pol'} + >>> Book.split_urlid("a-slug/eng", default_lang="eng") + + """ + parts = urlid.rsplit(sep, 1) + if len(parts) == 2: + if parts[1] == default_lang: + return None + return {'slug': parts[0], 'language': parts[1]} + else: + return {'slug': urlid, 'language': default_lang} + + @classmethod + def split_fileid(cls, fileid): + return cls.split_urlid(fileid, '_') + 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) @@ -329,14 +433,18 @@ class Book(models.Model): @permalink def get_absolute_url(self): - return ('catalogue.views.book_detail', [self.slug]) + return ('catalogue.views.book_detail', [self.urlid()]) @property def name(self): return self.title def book_tag_slug(self): - return ('l-' + self.slug)[:120] + stem = 'l-' + self.slug + if self.language != settings.CATALOGUE_DEFAULT_LANGUAGE: + return stem[:116] + ' ' + self.language + else: + return stem[:120] def book_tag(self): slug = self.book_tag_slug() @@ -348,49 +456,15 @@ class Book(models.Model): return book_tag def has_media(self, type): - if type == 'xml': - if self.xml_file: - return True - else: - return False - elif type == 'html': - if self.html_file: - return True - else: - return False - elif type == 'txt': - if self.txt_file: - return True - else: - return False - elif type == 'pdf': - if self.pdf_file: - return True - else: - return False - elif type == 'epub': - if self.epub_file: - return True - else: - return False + if type in Book.file_types: + return bool(getattr(self, "%s_file" % type)) else: - if self.media.filter(type=type).exists(): - return True - else: - return False + return self.media.filter(type=type).exists() def get_media(self, type): if self.has_media(type): - if type == "xml": - return self.xml_file - elif type == "html": - return self.html_file - elif type == "epub": - return self.epub_file - elif type == "txt": - return self.txt_file - elif type == "pdf": - return self.pdf_file + if type in Book.file_types: + return getattr(self, "%s_file" % type) else: return self.media.filter(type=type) else: @@ -430,11 +504,13 @@ class Book(models.Model): tags = [mark_safe(u'%s' % (tag.get_absolute_url(), tag.name)) for tag in tags] formats = [] - # files generated during publication + # files generated during publication if self.has_media("html"): - formats.append(u'%s' % (reverse('book_text', kwargs={'slug': self.slug}), _('Read online'))) + formats.append(u'%s' % (reverse('book_text', args=[self.fileid()]), _('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"): @@ -470,26 +546,6 @@ class Book(models.Model): has_description.boolean = True # ugly ugly ugly - 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_txt_file(self): - return bool(self.txt_file) - has_txt_file.short_description = 'HTML' - has_txt_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 has_odt_file(self): return bool(self.has_media("odt")) has_odt_file.short_description = 'ODT' @@ -510,46 +566,74 @@ class Book(models.Model): has_daisy_file.short_description = 'DAISY' has_daisy_file.boolean = True - def build_pdf(self): - """ (Re)builds the pdf file. + def wldocument(self, parse_dublincore=True): + from catalogue.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_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 librarian import pdf - from tempfile import NamedTemporaryFile - import os + from os import unlink + from django.core.files import File + from catalogue.utils import remove_zip - path, fname = os.path.realpath(self.xml_file.path).rsplit('/', 1) - try: - pdf_file = NamedTemporaryFile(delete=False) + pdf = self.wldocument().as_pdf(customizations=customizations) - pdf.transform(BookImportDocProvider(self), - file_path=str(self.xml_file.path), - output_file=pdf_file, - ) + 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.fileid(), + File(open(pdf.get_filename()))) + self.pdf_file = current_self.pdf_file - self.pdf_file.save('%s.pdf' % self.slug, File(open(pdf_file.name))) - finally: - unlink(pdf_file.name) + # 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.fileid(), 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 django.core.files import File + from catalogue.utils import remove_zip if self.parent: # don't need an epub return - epub_file = StringIO() + epub = self.wldocument().as_epub() + try: - epub.transform(BookImportDocProvider(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() + epub.transform(ORMDocProvider(self), self.fileid(), output_file=epub_file) + self.epub_file.save('%s.epub' % self.fileid(), File(open(epub.get_filename()))) except NoDublinCore: pass @@ -562,27 +646,30 @@ class Book(models.Model): child_book.save() book_descendants += list(child_book.children.all()) + # 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.fileid(), ContentFile(text.get_string())) def build_html(self): - from tempfile import NamedTemporaryFile from markupstring import MarkupString + 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_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.fileid(), + ContentFile(html_output.get_string())) # get ancestor l-tags for adding to new fragments ancestor_tags = [] @@ -627,9 +714,33 @@ class Book(models.Model): 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.fileid(), + 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): + bm = BookMedia.objects.filter(book=self, type='mp3') + paths = map(lambda bm: (None, bm.file.path), bm) + result = create_zip.delay(paths, self.fileid()) + return result.wait() @classmethod 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) @@ -642,31 +753,36 @@ class Book(models.Model): xml_file.close() @classmethod - def from_text_and_meta(cls, raw_file, book_info, overwrite=False, build_epub=True, build_txt=True, build_pdf=True): + def from_text_and_meta(cls, raw_file, book_info, overwrite=False, + build_epub=True, build_txt=True, build_pdf=True, build_mobi=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: - base, slug = part_url.rsplit('/', 1) try: - children.append(Book.objects.get(slug=slug)) + children.append(Book.objects.get( + slug=part_url.slug, language=part_url.language)) except Book.DoesNotExist, e: - raise Book.DoesNotExist(_('Book with slug = "%s" does not exist.') % slug) + raise Book.DoesNotExist(_('Book "%s/%s" does not exist.') % + (part_url.slug, part_url.language)) # Read book metadata - book_base, book_slug = book_info.url.rsplit('/', 1) + book_slug = book_info.url.slug + language = book_info.language if re.search(r'[^a-zA-Z0-9-]', book_slug): raise ValueError('Invalid characters in slug') - book, created = Book.objects.get_or_create(slug=book_slug) + book, created = Book.objects.get_or_create(slug=book_slug, language=language) if created: book_shelves = [] else: if not overwrite: - raise Book.AlreadyExists(_('Book %s already exists') % book_slug) + raise Book.AlreadyExists(_('Book %s/%s already exists') % ( + book_slug, language)) # Save shelves for this book book_shelves = list(book.tags.filter(category='set')) @@ -674,24 +790,7 @@ class Book(models.Model): 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) @@ -718,6 +817,9 @@ class Book(models.Model): if not settings.NO_BUILD_PDF and build_pdf: book.root_ancestor.build_pdf() + if not settings.NO_BUILD_MOBI and build_mobi: + book.build_mobi() + book_descendants = list(book.children.all()) # add l-tag to descendants and their fragments # delete unnecessary EPUB files @@ -735,6 +837,7 @@ class Book(models.Model): book.reset_tag_counter() book.reset_theme_counter() + cls.published.send(sender=book) return book def reset_tag_counter(self): @@ -828,6 +931,75 @@ class Book(models.Model): 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', 'language') + 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 _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.file_types: + 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() @@ -845,7 +1017,7 @@ 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' % (self.book.get_html_url(), self.anchor) def reset_short_html(self): if self.id is None: @@ -872,20 +1044,6 @@ 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