epub creation on import, log produced files (Fixed #723)
[wolnelektury.git] / apps / catalogue / models.py
index 55a2b64..8c83faa 100644 (file)
@@ -13,11 +13,11 @@ from django.utils.translation import get_language
 from django.core.urlresolvers import reverse
 from datetime import datetime
 
-from newtagging.models import TagBase
+from newtagging.models import TagBase, tags_updated
 from newtagging import managers
 from catalogue.fields import JSONField
 
-from librarian import html, dcparser
+from librarian import dcparser, html, epub
 from mutagen import id3
 
 
@@ -43,7 +43,7 @@ class TagSubcategoryManager(models.Manager):
 
 class Tag(TagBase):
     name = models.CharField(_('name'), max_length=50, db_index=True)
-    slug = models.SlugField(_('slug'), max_length=120, unique=True, db_index=True)
+    slug = models.SlugField(_('slug'), max_length=120, db_index=True)
     sort_key = models.SlugField(_('sort key'), max_length=120, db_index=True)
     category = models.CharField(_('category'), max_length=50, blank=False, null=False,
         db_index=True, choices=TAG_CATEGORIES)
@@ -51,15 +51,26 @@ class Tag(TagBase):
     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'), default=0, blank=False, null=False)
+    book_count = models.IntegerField(_('book count'), blank=False, null=True)
     death = models.IntegerField(_(u'year of death'), blank=True, null=True)
     gazeta_link = models.CharField(blank=True, max_length=240)
     wiki_link = models.CharField(blank=True, max_length=240)
 
+    categories_rev = {
+        'autor': 'author',
+        'epoka': 'epoch',
+        'rodzaj': 'kind',
+        'gatunek': 'genre',
+        'motyw': 'theme',
+        'polka': 'set',
+    }
+    categories_dict = dict((item[::-1] for item in categories_rev.iteritems()))
+
     class Meta:
         ordering = ('sort_key',)
         verbose_name = _('tag')
         verbose_name_plural = _('tags')
+        unique_together = (("slug", "category"),)
 
     def __unicode__(self):
         return self.name
@@ -69,7 +80,7 @@ class Tag(TagBase):
 
     @permalink
     def get_absolute_url(self):
-        return ('catalogue.views.tagged_object_list', [self.slug])
+        return ('catalogue.views.tagged_object_list', [self.url_chunk])
 
     def has_description(self):
         return len(self.description) > 0
@@ -87,16 +98,67 @@ class Tag(TagBase):
         """ calculates the year of public domain entry for an author """
         return self.death + 71 if self.death is not None else None
 
+    def get_count(self):
+        """ returns global book count for book tags, fragment count for themes """
+
+        if self.book_count is None:
+            if self.category == 'book':
+                # never used
+                objects = Book.objects.none()
+            elif self.category == 'theme':
+                objects = Fragment.tagged.with_all((self,))
+            else:
+                objects = Book.tagged.with_all((self,)).order_by()
+                if self.category != 'set':
+                    # eliminate descendants
+                    l_tags = Tag.objects.filter(slug__in=[book.book_tag_slug() for book in objects])
+                    descendants_keys = [book.pk for book in Book.tagged.with_any(l_tags)]
+                    if descendants_keys:
+                        objects = objects.exclude(pk__in=descendants_keys)
+            self.book_count = objects.count()
+            self.save()
+        return self.book_count
+
     @staticmethod
     def get_tag_list(tags):
         if isinstance(tags, basestring):
-            tag_slugs = tags.split('/')
-            return [Tag.objects.get(slug=slug) for slug in tag_slugs]
+            real_tags = []
+            ambiguous_slugs = []
+            category = None
+            tags_splitted = tags.split('/')
+            for index, name in enumerate(tags_splitted):
+                if name in Tag.categories_rev:
+                    category = Tag.categories_rev[name]
+                else:
+                    if category:
+                        real_tags.append(Tag.objects.get(slug=name, category=category))
+                        category = None
+                    else:
+                        try:
+                            real_tags.append(Tag.objects.exclude(category='book').get(slug=name))
+                        except Tag.MultipleObjectsReturned, e:
+                            ambiguous_slugs.append(name)
+
+            if category:
+                # something strange left off
+                raise Tag.DoesNotExist()
+            if ambiguous_slugs:
+                # some tags should be qualified
+                e = Tag.MultipleObjectsReturned()
+                e.tags = real_tags
+                e.ambiguous_slugs = ambiguous_slugs
+                raise e
+            else:
+                return real_tags
         else:
             return TagBase.get_tag_list(tags)
 
+    @property
+    def url_chunk(self):
+        return '/'.join((Tag.categories_dict[self.category], self.slug))
+
 
-# TODO: why is this hard-coded ? 
+# TODO: why is this hard-coded ?
 def book_upload_path(ext):
     def get_dynamic_path(book, filename):
         return 'lektura/%s.%s' % (book.slug, ext)
@@ -130,9 +192,9 @@ class Book(models.Model):
     objects = models.Manager()
     tagged = managers.ModelTaggedItemManager(Tag)
     tags = managers.TagDescriptor(Tag)
-    
-    _tag_counter = JSONField(editable=False, default='')
-    _theme_counter = JSONField(editable=False, default='')
+
+    _tag_counter = JSONField(null=True, editable=False)
+    _theme_counter = JSONField(null=True, editable=False)
 
     class AlreadyExists(Exception):
         pass
@@ -148,8 +210,12 @@ class Book(models.Model):
     def save(self, force_insert=False, force_update=False, reset_short_html=True, refresh_mp3=True):
         if reset_short_html:
             # Reset _short_html during save
+            update = {}
             for key in filter(lambda x: x.startswith('_short_html'), self.__dict__):
+                update[key] = ''
                 self.__setattr__(key, '')
+            # Fragment.short_html relies on book's tags, so reset it here too
+            self.fragments.all().update(**update)
 
         book = super(Book, self).save(force_insert, force_update)
 
@@ -169,14 +235,16 @@ class Book(models.Model):
     @property
     def name(self):
         return self.title
-    
+
+    def book_tag_slug(self):
+        return ('l-' + self.slug)[:120]
+
     def book_tag(self):
-        slug = ('l-' + self.slug)[:120]
-        book_tag, created = Tag.objects.get_or_create(slug=slug)
+        slug = self.book_tag_slug()
+        book_tag, created = Tag.objects.get_or_create(slug=slug, category='book')
         if created:
             book_tag.name = self.title[:50]
             book_tag.sort_key = slug
-            book_tag.category = 'book'
             book_tag.save()
         return book_tag
 
@@ -252,7 +320,7 @@ class Book(models.Model):
         book_info = dcparser.parse(xml_file)
 
         if not isinstance(xml_file, File):
-            xml_file = File(xml_file)
+            xml_file = File(open(xml_file))
 
         try:
             return cls.from_text_and_meta(xml_file, book_info, overwrite)
@@ -264,6 +332,10 @@ 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
 
         # Read book metadata
         book_base, book_slug = book_info.url.rsplit('/', 1)
@@ -283,21 +355,25 @@ class Book(models.Model):
         book.save()
 
         book_tags = []
-        for category in ('kind', 'genre', 'author', 'epoch'):
-            tag_name = getattr(book_info, category)
-            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))
-            if created:
-                tag.name = tag_name
-                tag.sort_key = slughifi(tag_sort_key)
-                tag.category = category
-                tag.save()
-            book_tags.append(tag)
-
-        book.tags = book_tags
+        categories = (('kinds', 'kind'), ('genres', 'genre'), ('authors', 'author'), ('epochs', 'epoch'))
+        for field_name, category in categories:
+            try:
+                tag_names = getattr(book_info, field_name)
+            except:
+                tag_names = [getattr(book_info, category)]
+            for tag_name in tag_names:
+                tag_sort_key = tag_name
+                if category == 'author':
+                    tag_sort_key = tag_name.last_name
+                    tag_name = ' '.join(tag_name.first_names) + ' ' + tag_name.last_name
+                tag, created = Tag.objects.get_or_create(slug=slughifi(tag_name), category=category)
+                if created:
+                    tag.name = tag_name
+                    tag.sort_key = slughifi(tag_sort_key)
+                    tag.save()
+                book_tags.append(tag)
+
+        book.tags = book_tags + book_shelves
 
         book_tag = book.book_tag()
 
@@ -328,41 +404,50 @@ class Book(models.Model):
         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)
 
+            # Create EPUB
+            epub_file = StringIO()
+            epub.transform(book.xml_file, 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()
+
             # Extract fragments
             closed_fragments, open_fragments = html.extract_fragments(book.html_file.path)
-            book_themes = []
             for fragment in closed_fragments.values():
-                text = fragment.to_string()
-                short_text = ''
-                if (len(MarkupString(text)) > 240):
-                    short_text = unicode(MarkupString(text)[:160])
-                new_fragment, created = Fragment.objects.get_or_create(anchor=fragment.id, book=book,
-                    defaults={'text': text, 'short_text': short_text})
-
                 try:
                     theme_names = [s.strip() for s in fragment.themes.split(',')]
                 except AttributeError:
                     continue
                 themes = []
                 for theme_name in theme_names:
-                    tag, created = Tag.objects.get_or_create(slug=slughifi(theme_name))
+                    if not theme_name:
+                        continue
+                    tag, created = Tag.objects.get_or_create(slug=slughifi(theme_name), category='theme')
                     if created:
                         tag.name = theme_name
                         tag.sort_key = slughifi(theme_name)
-                        tag.category = 'theme'
                         tag.save()
                     themes.append(tag)
+                if not themes:
+                    continue
+
+                text = fragment.to_string()
+                short_text = ''
+                if (len(MarkupString(text)) > 240):
+                    short_text = unicode(MarkupString(text)[:160])
+                new_fragment, created = Fragment.objects.get_or_create(anchor=fragment.id, book=book,
+                    defaults={'text': text, 'short_text': short_text})
+
                 new_fragment.save()
-                new_fragment.tags = set(list(book.tags) + themes + [book_tag])
-                book_themes += themes
+                new_fragment.tags = set(book_tags + themes + [book_tag])
 
-            book_themes = set(book_themes)
-            book.tags = list(book.tags) + list(book_themes) + book_shelves
+        # refresh cache
+        book.tag_counter
+        book.theme_counter
 
         book.save()
         return book
-    
-    
+
+
     def refresh_tag_counter(self):
         tags = {}
         for child in self.children.all().order_by():
@@ -373,13 +458,18 @@ class Book(models.Model):
         self.set__tag_counter_value(tags)
         self.save(reset_short_html=False, refresh_mp3=False)
         return tags
-    
+
+    def reset_tag_counter(self):
+        self._tag_counter = None
+        self.save(reset_short_html=False, refresh_mp3=False)
+        if self.parent:
+            self.parent.reset_tag_counter()
+
     @property
     def tag_counter(self):
-        if self._tag_counter == '':
+        if self._tag_counter is None:
             return self.refresh_tag_counter()
         return dict((int(k), v) for k, v in self.get__tag_counter_value().iteritems())
-        #return self.get__tag_counter_value()
 
     def refresh_theme_counter(self):
         tags = {}
@@ -389,14 +479,19 @@ class Book(models.Model):
         self.set__theme_counter_value(tags)
         self.save(reset_short_html=False, refresh_mp3=False)
         return tags
-    
+
+    def reset_theme_counter(self):
+        self._theme_counter = None
+        self.save(reset_short_html=False, refresh_mp3=False)
+        if self.parent:
+            self.parent.reset_theme_counter()
+
     @property
     def theme_counter(self):
-        if self._theme_counter == '':
+        if self._theme_counter is None:
             return self.refresh_theme_counter()
         return dict((int(k), v) for k, v in self.get__theme_counter_value().iteritems())
-        return self.get__theme_counter_value()
-    
+
 
 
 class Fragment(models.Model):
@@ -424,11 +519,8 @@ class Fragment(models.Model):
         if short_html and len(short_html):
             return mark_safe(short_html)
         else:
-            book_authors = [mark_safe(u'<a href="%s">%s</a>' % (tag.get_absolute_url(), tag.name))
-                for tag in self.book.tags if tag.category == 'author']
-
             setattr(self, key, unicode(render_to_string('catalogue/fragment_short.html',
-                {'fragment': self, 'book': self.book, 'book_authors': book_authors})))
+                {'fragment': self})))
             self.save()
             return mark_safe(getattr(self, key))
 
@@ -461,3 +553,34 @@ class BookStub(models.Model):
         return self.title
 
 
+class FileRecord(models.Model):
+    slug = models.SlugField(_('slug'), max_length=120, db_index=True)
+    type = models.CharField(_('type'), max_length=20, db_index=True)
+    sha1 = models.CharField(_('sha-1 hash'), max_length=40)
+    time = models.DateTimeField(_('time'), auto_now_add=True)
+
+    class Meta:
+        ordering = ('-time','-slug', '-type')
+        verbose_name = _('file record')
+        verbose_name_plural = _('file records')
+
+    def __unicode__(self):
+        return "%s %s.%s" % (self.sha1,  self.slug, self.type)
+
+
+def _tags_updated_handler(sender, affected_tags, **kwargs):
+    # reset tag global counter
+    Tag.objects.filter(pk__in=[tag.pk for tag in affected_tags]).update(book_count=None)
+
+    # if book tags changed, reset book tag counter
+    if isinstance(sender, Book) and \
+                Tag.objects.filter(pk__in=(tag.pk for tag in affected_tags)).\
+                    exclude(category__in=('book', 'theme', 'set')).count():
+        sender.reset_tag_counter()
+    # if fragment theme changed, reset book theme counter
+    elif isinstance(sender, Fragment) and \
+                Tag.objects.filter(pk__in=(tag.pk for tag in affected_tags)).\
+                    filter(category='theme').count():
+        sender.book.reset_theme_counter()
+tags_updated.connect(_tags_updated_handler)
+