random fixes
[wolnelektury.git] / apps / catalogue / models.py
index 930f98d..cace1a5 100644 (file)
@@ -2,35 +2,38 @@
 # This file is part of Wolnelektury, licensed under GNU Affero GPLv3 or later.
 # Copyright © Fundacja Nowoczesna Polska. See NOTICE for more information.
 #
 # 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
 import django.dispatch
 
 from django.db import models
 from django.db.models import permalink, Q
 import django.dispatch
-from django.core.cache import cache
+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.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.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 django.utils.safestring import mark_safe
 from django.utils.translation import get_language
 from django.core.urlresolvers import reverse
-from django.db.models.signals import post_save, m2m_changed, pre_delete
+from django.db.models.signals import post_save, m2m_changed, pre_delete, post_delete
+import jsonfield
 
 from django.conf import settings
 
 from newtagging.models import TagBase, tags_updated
 from newtagging import managers
 from catalogue.fields import JSONField, OverwritingFileField
 
 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, 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
 
 
-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
+import search
 
 
+# Those are hard-coded here so that makemessages sees them.
 TAG_CATEGORIES = (
     ('author', _('author')),
     ('epoch', _('epoch')),
 TAG_CATEGORIES = (
     ('author', _('author')),
     ('epoch', _('epoch')),
@@ -41,15 +44,8 @@ TAG_CATEGORIES = (
     ('book', _('book')),
 )
 
     ('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
+permanent_cache = get_cache('permanent')
 
 
 class TagSubcategoryManager(models.Manager):
 
 
 class TagSubcategoryManager(models.Manager):
@@ -68,7 +64,6 @@ 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)
     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)
 
     user = models.ForeignKey(User, blank=True, null=True)
     book_count = models.IntegerField(_('book count'), blank=True, null=True)
@@ -113,25 +108,22 @@ class Tag(TagBase):
     has_description.boolean = True
 
     def get_count(self):
     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):
 
     @staticmethod
     def get_tag_list(tags):
@@ -175,26 +167,89 @@ class Tag(TagBase):
     def url_chunk(self):
         return '/'.join((Tag.categories_dict[self.category], self.slug))
 
     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=None, maxlen=100):
 
 # 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.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):
 
 
 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)
     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)
@@ -211,16 +266,25 @@ class BookMedia(models.Model):
         verbose_name_plural = _('book media')
 
     def save(self, *args, **kwargs):
         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:
         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):
                 self.file.save(None, ExistingFile(self.file.path), save=False, leave=True)
 
         super(BookMedia, self).save(*args, **kwargs)
         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)
         extra_info = self.get_extra_info_value()
         extra_info.update(self.read_meta())
         self.set_extra_info_value(extra_info)
@@ -231,6 +295,8 @@ class BookMedia(models.Model):
         """
             Reads some metadata from the audiobook.
         """
         """
             Reads some metadata from the audiobook.
         """
+        import mutagen
+        from mutagen import id3
 
         artist_name = director_name = project = funded_by = ''
         if self.type == 'mp3':
 
         artist_name = director_name = project = funded_by = ''
         if self.type == 'mp3':
@@ -263,6 +329,8 @@ class BookMedia(models.Model):
         """
             Reads source file SHA1 from audiobok metadata.
         """
         """
             Reads source file SHA1 from audiobok metadata.
         """
+        import mutagen
+        from mutagen import id3
 
         if filetype == 'mp3':
             try:
 
         if filetype == 'mp3':
             try:
@@ -283,8 +351,12 @@ class BookMedia(models.Model):
 
 class Book(models.Model):
     title         = models.CharField(_('title'), max_length=120)
 
 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,
+            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)
     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)
@@ -293,18 +365,22 @@ 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
     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)
-    
+
+    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')
     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)
 
     html_built = django.dispatch.Signal()
     objects  = models.Manager()
     tagged   = managers.ModelTaggedItemManager(Tag)
     tags     = managers.TagDescriptor(Tag)
 
     html_built = django.dispatch.Signal()
+    published = django.dispatch.Signal()
 
     class AlreadyExists(Exception):
         pass
 
     class AlreadyExists(Exception):
         pass
@@ -318,6 +394,8 @@ class Book(models.Model):
         return self.title
 
     def save(self, force_insert=False, force_update=False, reset_short_html=True, **kwargs):
         return self.title
 
     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)
         self.sort_key = sortify(self.title)
 
         ret = super(Book, self).save(force_insert, force_update)
@@ -348,49 +426,15 @@ class Book(models.Model):
         return book_tag
 
     def has_media(self, type):
         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.formats:
+            return bool(getattr(self, "%s_file" % type))
         else:
         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):
 
     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.formats:
+                return getattr(self, "%s_file" % type)
             else:                                             
                 return self.media.filter(type=type)
         else:
             else:                                             
                 return self.media.filter(type=type)
         else:
@@ -409,92 +453,17 @@ class Book(models.Model):
         if self.id is None:
             return
 
         if self.id is None:
             return
 
-        cache_key = "Book.short_html/%d/%s"
-        for lang, langname in settings.LANGUAGES:
-            cache.delete(cache_key % (self.id, lang))
+        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()
 
         # Fragment.short_html relies on book's tags, so reset it here too
         for fragm in self.fragments.all():
             fragm.reset_short_html()
 
-    def short_html(self):
-        if self.id:
-            cache_key = "Book.short_html/%d/%s" % (self.id, get_language())
-            short_html = cache.get(cache_key)
-        else:
-            short_html = None
-
-        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'<a href="%s">%s</a>' % (tag.get_absolute_url(), tag.name)) for tag in tags]
-
-            formats = []
-            # files generated during publication               
-            if self.has_media("html"):
-                formats.append(u'<a href="%s">%s</a>' % (reverse('book_text', kwargs={'slug': self.slug}), _('Read online')))
-            if self.has_media("pdf"):
-                formats.append(u'<a href="%s">PDF</a>' % self.get_media('pdf').url)
-            if self.root_ancestor.has_media("epub"):
-                formats.append(u'<a href="%s">EPUB</a>' % self.root_ancestor.get_media('epub').url)
-            if self.has_media("txt"):
-                formats.append(u'<a href="%s">TXT</a>' % self.get_media('txt').url)
-            # other files
-            for m in self.media.order_by('type'):
-                formats.append(u'<a href="%s">%s</a>' % (m.file.url, m.type.upper()))
-
-            formats = [mark_safe(format) for format in formats]
-
-            short_html = unicode(render_to_string('catalogue/book_short.html',
-                {'book': self, 'tags': tags, 'formats': formats}))
-
-            if self.id:
-                cache.set(cache_key, short_html, CACHE_FOREVER)
-            return mark_safe(short_html)
-
-    @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 has_description(self):
         return len(self.description) > 0
     has_description.short_description = _('description')
     has_description.boolean = True
 
     # ugly ugly ugly
     def has_description(self):
         return len(self.description) > 0
     has_description.short_description = _('description')
     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'
-    has_odt_file.boolean = True
-
     def has_mp3_file(self):
         return bool(self.has_media("mp3"))
     has_mp3_file.short_description = 'MP3'
     def has_mp3_file(self):
         return bool(self.has_media("mp3"))
     has_mp3_file.short_description = 'MP3'
@@ -510,86 +479,104 @@ class Book(models.Model):
     has_daisy_file.short_description = 'DAISY'
     has_daisy_file.boolean = True
 
     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.import_utils import ORMDocProvider
+        from librarian.parser import WLDocument
 
 
-        """
-        from librarian import pdf, ParseError
-        from tempfile import NamedTemporaryFile
-        import os
+        return WLDocument.from_file(self.xml_file.path,
+                provider=ORMDocProvider(self),
+                parse_dublincore=parse_dublincore)
 
 
-        try:
-            path, fname = os.path.realpath(self.xml_file.path).rsplit('/', 1)
-            try:
-                pdf_file = NamedTemporaryFile(delete=False)
-
-                pdf.transform(BookImportDocProvider(self),
-                          file_path=str(self.xml_file.path),
-                          output_file=pdf_file,
-                          )
-
-                self.pdf_file.save('%s.pdf' % self.slug, File(open(pdf_file.name)))
-            finally:
-                unlink(pdf_file.name)
-
-        except ParseError, e:
-            print '%(file)s:%(name)s:%(message)s; use -v to see more output' % {
-                'file': self.xml_file.path,
-                'name': e.__class__.__name__,
-                'message': e
-                }
-
-    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.
-        """
+    def build_cover(self, book_info=None):
+        """(Re)builds the cover image."""
         from StringIO import StringIO
         from StringIO import StringIO
-        from hashlib import sha1
         from django.core.files.base import ContentFile
         from django.core.files.base import ContentFile
+        from librarian.cover import WLCover
 
 
-        if self.parent:
-            # don't need an epub
-            return
+        if book_info is None:
+            book_info = self.wldocument().book_info
 
 
-        epub_file = StringIO()
-        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()
-        except NoDublinCore:
-            pass
+        cover = WLCover(book_info).image()
+        imgstr = StringIO()
+        cover.save(imgstr, 'png')
+        self.cover.save(None, ContentFile(imgstr.getvalue()))
 
 
-        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())
+    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
+
+        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)
+        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):
 
     def build_txt(self):
-        from StringIO import StringIO
         from django.core.files.base import ContentFile
         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):
 
 
     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()
 
 
         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.slug,
+                    ContentFile(html_output.get_string()))
 
             # get ancestor l-tags for adding to new fragments
             ancestor_tags = []
 
             # get ancestor l-tags for adding to new fragments
             ancestor_tags = []
@@ -621,9 +608,9 @@ class Book(models.Model):
                     continue
 
                 text = fragment.to_string()
                     continue
 
                 text = fragment.to_string()
-                short_text = ''
-                if (len(MarkupString(text)) > 240):
-                    short_text = unicode(MarkupString(text)[:160])
+                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 = Fragment.objects.create(anchor=fragment.id, book=self,
                     text=text, short_text=short_text)
 
@@ -634,9 +621,47 @@ class Book(models.Model):
             return True
         return False
 
             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, index_tags=True):
+        if reuse_index:
+            idx = search.ReusableIndex()
+        else:
+            idx = search.Index()
+            
+        idx.open()
+        try:
+            idx.index_book(self, book_info)
+            if index_tags:
+                idx.index_tags()
+        finally:
+            idx.close()
 
     @classmethod
     def from_xml_file(cls, xml_file, **kwargs):
 
     @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)
 
         # use librarian to parse meta-data
         book_info = dcparser.parse(xml_file)
 
@@ -649,23 +674,26 @@ class Book(models.Model):
             xml_file.close()
 
     @classmethod
             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,
+            search_index=True, search_index_tags=True, search_index_reuse=False):
         import re
         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:
 
         # 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:
                 try:
-                    children.append(Book.objects.get(slug=slug))
+                    children.append(Book.objects.get(slug=part_url.slug))
                 except Book.DoesNotExist, e:
                 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
 
 
         # 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)
 
             raise ValueError('Invalid characters in slug')
         book, created = Book.objects.get_or_create(slug=book_slug)
 
@@ -673,32 +701,21 @@ class Book(models.Model):
             book_shelves = []
         else:
             if not overwrite:
             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'))
 
             # Save shelves for this book
             book_shelves = list(book.tags.filter(category='set'))
 
+        book.language = book_info.language
         book.title = book_info.title
         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()
 
         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)
 
 
         book.tags = set(meta_tags + book_shelves)
 
@@ -719,37 +736,81 @@ class Book(models.Model):
             if not settings.NO_BUILD_TXT and build_txt:
                 book.build_txt()
 
             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:
         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:
 
         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:
+            book.search_index(index_tags=search_index_tags, reuse_index=search_index_reuse)
+            #index_book.delay(book.id, book_info)
 
         book_descendants = list(book.children.all())
 
         book_descendants = list(book.children.all())
+        descendants_tags = set()
         # add l-tag to descendants and their fragments
         # add l-tag to descendants and their fragments
-        # delete unnecessary EPUB files
         while len(book_descendants) > 0:
             child_book = book_descendants.pop(0)
         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())
 
             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(tag)
+
         book.save()
 
         # refresh cache
         book.reset_tag_counter()
         book.reset_theme_counter()
 
         book.save()
 
         # refresh cache
         book.reset_tag_counter()
         book.reset_theme_counter()
 
+        cls.published.send(sender=book)
         return 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': {}}
+
+            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):
         if self.id is None:
             return
 
         cache_key = "Book.tag_counter/%d" % self.id
     def reset_tag_counter(self):
         if self.id is None:
             return
 
         cache_key = "Book.tag_counter/%d" % self.id
-        cache.delete(cache_key)
+        permanent_cache.delete(cache_key)
         if self.parent:
             self.parent.reset_tag_counter()
 
         if self.parent:
             self.parent.reset_tag_counter()
 
@@ -757,7 +818,7 @@ class Book(models.Model):
     def tag_counter(self):
         if self.id:
             cache_key = "Book.tag_counter/%d" % self.id
     def tag_counter(self):
         if self.id:
             cache_key = "Book.tag_counter/%d" % self.id
-            tags = cache.get(cache_key)
+            tags = permanent_cache.get(cache_key)
         else:
             tags = None
 
         else:
             tags = None
 
@@ -770,7 +831,7 @@ class Book(models.Model):
                 tags[tag.pk] = 1
 
             if self.id:
                 tags[tag.pk] = 1
 
             if self.id:
-                cache.set(cache_key, tags, CACHE_FOREVER)
+                permanent_cache.set(cache_key, tags)
         return tags
 
     def reset_theme_counter(self):
         return tags
 
     def reset_theme_counter(self):
@@ -778,7 +839,7 @@ class Book(models.Model):
             return
 
         cache_key = "Book.theme_counter/%d" % self.id
             return
 
         cache_key = "Book.theme_counter/%d" % self.id
-        cache.delete(cache_key)
+        permanent_cache.delete(cache_key)
         if self.parent:
             self.parent.reset_theme_counter()
 
         if self.parent:
             self.parent.reset_theme_counter()
 
@@ -786,7 +847,7 @@ class Book(models.Model):
     def theme_counter(self):
         if self.id:
             cache_key = "Book.theme_counter/%d" % self.id
     def theme_counter(self):
         if self.id:
             cache_key = "Book.theme_counter/%d" % self.id
-            tags = cache.get(cache_key)
+            tags = permanent_cache.get(cache_key)
         else:
             tags = None
 
         else:
             tags = None
 
@@ -797,7 +858,7 @@ class Book(models.Model):
                     tags[tag.pk] = tags.get(tag.pk, 0) + 1
 
             if self.id:
                     tags[tag.pk] = tags.get(tag.pk, 0) + 1
 
             if self.id:
-                cache.set(cache_key, tags, CACHE_FOREVER)
+                permanent_cache.set(cache_key, tags)
         return tags
 
     def pretty_title(self, html_links=False):
         return tags
 
     def pretty_title(self, html_links=False):
@@ -835,6 +896,85 @@ class Book(models.Model):
 
         return objects
 
 
         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()
 
 class Fragment(models.Model):
     text = models.TextField()
@@ -852,7 +992,7 @@ class Fragment(models.Model):
         verbose_name_plural = _('fragments')
 
     def get_absolute_url(self):
         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:
 
     def reset_short_html(self):
         if self.id is None:
@@ -860,12 +1000,16 @@ class Fragment(models.Model):
 
         cache_key = "Fragment.short_html/%d/%s"
         for lang, langname in settings.LANGUAGES:
 
         cache_key = "Fragment.short_html/%d/%s"
         for lang, langname in settings.LANGUAGES:
-            cache.delete(cache_key % (self.id, lang))
+            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):
         if self.id:
             cache_key = "Fragment.short_html/%d/%s" % (self.id, get_language())
 
     def short_html(self):
         if self.id:
             cache_key = "Fragment.short_html/%d/%s" % (self.id, get_language())
-            short_html = cache.get(cache_key)
+            short_html = permanent_cache.get(cache_key)
         else:
             short_html = None
 
         else:
             short_html = None
 
@@ -875,23 +1019,27 @@ class Fragment(models.Model):
             short_html = unicode(render_to_string('catalogue/fragment_short.html',
                 {'fragment': self}))
             if self.id:
             short_html = unicode(render_to_string('catalogue/fragment_short.html',
                 {'fragment': self}))
             if self.id:
-                cache.set(cache_key, short_html, CACHE_FOREVER)
+                permanent_cache.set(cache_key, short_html)
             return mark_safe(short_html)
 
 
             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 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)
+
+    models.SlugField(_('slug'), max_length=120, unique=True, db_index=True)
+    book_slugs = models.TextField(_('book slugs'))
 
     class Meta:
 
     class Meta:
-        ordering = ('-time','-slug', '-type')
-        verbose_name = _('file record')
-        verbose_name_plural = _('file records')
+        ordering = ('title',)
+        verbose_name = _('collection')
+        verbose_name_plural = _('collections')
 
     def __unicode__(self):
 
     def __unicode__(self):
-        return "%s %s.%s" % (self.sha1,  self.slug, self.type)
+        return self.title
+
 
 ###########
 #
 
 ###########
 #
@@ -903,7 +1051,8 @@ 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
 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(tag)
 
     # if book tags changed, reset book tag counter
     if isinstance(sender, Book) and \
 
     # if book tags changed, reset book tag counter
     if isinstance(sender, Book) and \
@@ -924,8 +1073,20 @@ def _pre_delete_handler(sender, instance, **kwargs):
         instance.book.save()
 pre_delete.connect(_pre_delete_handler)
 
         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)
 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)
+
+
+@django.dispatch.receiver(post_delete, sender=Book)
+def _remove_book_from_index_handler(sender, instance, **kwargs):
+    """ remove the book from search index, when it is deleted."""
+    idx = search.Index()
+    idx.open(timeout=10000)  # 10 seconds timeout.
+    try:
+        idx.remove_book(instance)
+    finally:
+        idx.close()