librarian fix
[wolnelektury.git] / apps / catalogue / models.py
index 624211e..e692ad8 100644 (file)
@@ -186,6 +186,7 @@ class Book(models.Model):
     txt_file = models.FileField(_('TXT file'), upload_to=book_upload_path('txt'), blank=True)
     mp3_file = models.FileField(_('MP3 file'), upload_to=book_upload_path('mp3'), blank=True)
     ogg_file = models.FileField(_('OGG file'), upload_to=book_upload_path('ogg'), blank=True)
+    daisy_file = models.FileField(_('DAISY file'), upload_to=book_upload_path('daisy.zip'), blank=True)
 
     parent = models.ForeignKey('self', blank=True, null=True, related_name='children')
 
@@ -273,6 +274,8 @@ class Book(models.Model):
                 formats.append(u'<a href="%s">MP3</a>' % self.mp3_file.url)
             if self.ogg_file:
                 formats.append(u'<a href="%s">OGG</a>' % self.ogg_file.url)
+            if self.daisy_file:
+                formats.append(u'<a href="%s">DAISY</a>' % self.daisy_file.url)
 
             formats = [mark_safe(format) for format in formats]
 
@@ -326,6 +329,52 @@ class Book(models.Model):
     has_html_file.short_description = 'HTML'
     has_html_file.boolean = True
 
+    def build_epub(self, remove_descendants=True):
+        """ (Re)builds the epub file.
+            If book has a parent, does nothing.
+            Unless remove_descendants is False, descendants' epubs are removed.
+        """
+    
+        from StringIO import StringIO
+        from hashlib import sha1
+        from django.core.files.base import ContentFile
+        from librarian import DocProvider
+
+        class BookImportDocProvider(DocProvider):
+            """ used for joined EPUBs """
+
+            def __init__(self, book):
+                self.book = book
+
+            def by_slug(self, slug):
+                if slug == self.book.slug:
+                    return self.book.xml_file
+                else:
+                    return Book.objects.get(slug=slug).xml_file
+
+        if self.parent:
+            # don't need an epub
+            return
+
+        epub_file = StringIO()
+        try:
+            epub.transform(BookImportDocProvider(self), self.slug, epub_file)
+            self.epub_file.save('%s.epub' % self.slug, ContentFile(epub_file.getvalue()), save=False)
+            self.save()
+            FileRecord(slug=self.slug, type='epub', sha1=sha1(epub_file.getvalue()).hexdigest()).save()
+        except NoDublinCore:
+            pass
+
+        book_descendants = list(self.children.all())
+        while len(book_descendants) > 0:
+            child_book = book_descendants.pop(0)
+            if remove_descendants and child_book.has_epub_file():
+                child_book.epub_file.delete()
+            # save anyway, to refresh short_html
+            child_book.save()
+            book_descendants += list(child_book.children.all())
+
+
     @classmethod
     def from_xml_file(cls, xml_file, overwrite=False):
         # use librarian to parse meta-data
@@ -344,24 +393,7 @@ class Book(models.Model):
         from tempfile import NamedTemporaryFile
         from slughifi import slughifi
         from markupstring import MarkupString
-        from hashlib import sha1
-        from django.core.files.base import ContentFile
         from django.core.files.storage import default_storage
-        from StringIO import StringIO
-
-        from librarian import DocProvider
-
-        class BookImportDocProvider(DocProvider):
-            """ used for joined EPUBs """
-
-            def __init__(self, book):
-                self.book = book
-
-            def by_slug(self, slug):
-                if slug == self.book.slug:
-                    return self.book.xml_file
-                else:
-                    return Book.objects.get(slug=slug).xml_file
 
         # Read book metadata
         book_base, book_slug = book_info.url.rsplit('/', 1)
@@ -417,10 +449,20 @@ class Book(models.Model):
         # Save XML and HTML files
         book.xml_file.save('%s.xml' % book.slug, raw_file, save=False)
 
+        # delete old fragments when overwriting
+        book.fragments.all().delete()
+
         html_file = NamedTemporaryFile()
         if html.transform(book.xml_file.path, html_file, parse_dublincore=False):
             book.html_file.save('%s.html' % book.slug, File(html_file), save=False)
 
+            # get ancestor l-tags for adding to new fragments
+            ancestor_tags = []
+            p = book.parent
+            while p:
+                ancestor_tags.append(p.book_tag())
+                p = p.parent
+
             # Extract fragments
             closed_fragments, open_fragments = html.extract_fragments(book.html_file.path)
             for fragment in closed_fragments.values():
@@ -449,25 +491,18 @@ class Book(models.Model):
                     defaults={'text': text, 'short_text': short_text})
 
                 new_fragment.save()
-                new_fragment.tags = set(book_tags + themes + [book_tag])
+                new_fragment.tags = set(book_tags + themes + [book_tag] + ancestor_tags)
 
-        # Create EPUB
-        epub_file = StringIO()
-        try:
-            epub.transform(BookImportDocProvider(book), book.slug, epub_file)
-            book.epub_file.save('%s.epub' % book.slug, ContentFile(epub_file.getvalue()), save=False)
-            FileRecord(slug=book.slug, type='epub', sha1=sha1(epub_file.getvalue()).hexdigest()).save()
-        except NoDublinCore:
-            pass
+        if not book.parent:
+            book.build_epub(remove_descendants=False)
 
-        delete_epubs = book.has_epub_file()
         book_descendants = list(book.children.all())
         # add l-tag to descendants and their fragments
         # delete unnecessary EPUB files
         while len(book_descendants) > 0:
             child_book = book_descendants.pop(0)
             child_book.tags = list(child_book.tags) + [book_tag]
-            if delete_epubs:
+            if child_book.has_epub_file():
                 child_book.epub_file.delete()
             child_book.save()
             for fragment in child_book.fragments.all():
@@ -475,8 +510,8 @@ class Book(models.Model):
             book_descendants += list(child_book.children.all())
 
         # refresh cache
-        book.tag_counter
-        book.theme_counter
+        book.reset_tag_counter()
+        book.reset_theme_counter()
 
         book.save()
         return book