fix buggy migration
[wolnelektury.git] / apps / catalogue / models.py
index 37ea4e8..5b33dcc 100644 (file)
@@ -8,16 +8,19 @@ 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.defaultfilters import slugify
 from django.utils.safestring import mark_safe
 from django.utils.translation import get_language
 from django.core.urlresolvers import reverse
 from datetime import datetime
 
-from newtagging.models import TagBase
+from django.conf import settings
+
+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, NoDublinCore
 from mutagen import id3
 
 
@@ -31,6 +34,12 @@ TAG_CATEGORIES = (
     ('book', _('book')),
 )
 
+MEDIA_FORMATS = (
+    ('odt', _('ODT file')),
+    ('mp3', _('MP3 file')),
+    ('ogg', _('OGG file')),
+    ('daisy', _('DAISY file')), 
+)
 
 class TagSubcategoryManager(models.Manager):
     def __init__(self, subcategory):
@@ -44,14 +53,14 @@ class TagSubcategoryManager(models.Manager):
 class Tag(TagBase):
     name = models.CharField(_('name'), max_length=50, 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)
+    sort_key = models.CharField(_('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)
     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'), 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)
@@ -98,6 +107,27 @@ 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):
@@ -138,39 +168,56 @@ class Tag(TagBase):
 
 
 # TODO: why is this hard-coded ?
-def book_upload_path(ext):
-    def get_dynamic_path(book, filename):
-        return 'lektura/%s.%s' % (book.slug, ext)
+def book_upload_path(ext=None):
+    def get_dynamic_path(media, filename, ext=ext):
+        # how to put related book's slug here?
+        if not ext:
+            ext = media.type
+        if not media.name:
+            name = slugify(filename.split(".")[0])
+        else:
+            name = slugify(media.name)
+        return 'lektura/%s.%s' % (name, ext)
     return get_dynamic_path
 
 
-class Book(models.Model):
-    title = models.CharField(_('title'), max_length=120)
-    slug = models.SlugField(_('slug'), max_length=120, unique=True, db_index=True)
-    description = models.TextField(_('description'), blank=True)
-    created_at = models.DateTimeField(_('creation date'), auto_now=True)
-    _short_html = models.TextField(_('short HTML'), editable=False)
-    parent_number = models.IntegerField(_('parent number'), default=0)
-    extra_info = JSONField(_('extra information'))
-    gazeta_link = models.CharField(blank=True, max_length=240)
-    wiki_link = models.CharField(blank=True, max_length=240)
-
+class BookMedia(models.Model):
+    type        = models.CharField(_('type'), choices=MEDIA_FORMATS, max_length="100")
+    name        = models.CharField(_('name'), max_length="100", blank=True)
+    file        = models.FileField(_('file'), upload_to=book_upload_path(), blank=True)    
+    uploaded_at = models.DateTimeField(_('creation date'), auto_now_add=True, editable=False)
 
-    # Formats
-    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)
-    odt_file = models.FileField(_('ODT file'), upload_to=book_upload_path('odt'), blank=True)
-    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)
+    def __unicode__(self):
+        return "%s (%s)" % (self.name, self.file.name.split("/")[-1])
 
-    parent = models.ForeignKey('self', blank=True, null=True, related_name='children')
+    class Meta:
+        ordering            = ('type', 'name')
+        verbose_name        = _('book media')
+        verbose_name_plural = _('book media')
 
-    objects = models.Manager()
-    tagged = managers.ModelTaggedItemManager(Tag)
-    tags = managers.TagDescriptor(Tag)
+class Book(models.Model):
+    title         = models.CharField(_('title'), max_length=120)
+    slug          = models.SlugField(_('slug'), max_length=120, unique=True, db_index=True)
+    description   = models.TextField(_('description'), blank=True)
+    created_at    = models.DateTimeField(_('creation date'), auto_now_add=True)
+    _short_html   = models.TextField(_('short HTML'), editable=False)
+    parent_number = models.IntegerField(_('parent number'), default=0)
+    extra_info    = JSONField(_('extra information'))
+    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)        
+    # other files
+    medias        = models.ManyToManyField(BookMedia, blank=True)
+    
+    parent        = models.ForeignKey('self', blank=True, null=True, related_name='children')
+    objects  = models.Manager()
+    tagged   = managers.ModelTaggedItemManager(Tag)
+    tags     = managers.TagDescriptor(Tag)
 
     _tag_counter = JSONField(null=True, editable=False)
     _theme_counter = JSONField(null=True, editable=False)
@@ -186,7 +233,7 @@ class Book(models.Model):
     def __unicode__(self):
         return self.title
 
-    def save(self, force_insert=False, force_update=False, reset_short_html=True, refresh_mp3=True):
+    def save(self, force_insert=False, force_update=False, reset_short_html=True, refresh_mp3=True, **kwargs):
         if reset_short_html:
             # Reset _short_html during save
             update = {}
@@ -198,13 +245,13 @@ class Book(models.Model):
 
         book = super(Book, self).save(force_insert, force_update)
 
-        if refresh_mp3 and self.mp3_file:
-            print self.mp3_file, self.mp3_file.path
+        if refresh_mp3 and self.has_media('mp3'):
+            file = self.get_media('mp3')[0]
+            #print file, file.path
             extra_info = self.get_extra_info_value()
             extra_info.update(self.get_mp3_info())
             self.set_extra_info_value(extra_info)
             book = super(Book, self).save(force_insert, force_update)
-
         return book
 
     @permalink
@@ -215,15 +262,76 @@ class Book(models.Model):
     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]
+        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.sort_key = self.title.lower()
             book_tag.save()
         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                          
+        else:
+            if self.medias.filter(book=self, type=type).count() > 0:
+                return True
+            else:
+                return False
+
+    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.html_file                
+            elif type == "txt":
+                return self.txt_file
+            elif type == "pdf":
+                return self.pdf_file
+            else:                                             
+                return self.medias.filter(book=self, type=type)
+        else:
+            return None
+
+    def get_mp3(self):
+        return self.get_media("mp3")
+    def get_odt(self):
+        return self.get_media("odt")
+    def get_ogg(self):
+        return self.get_media("ogg")
+    def get_daisy(self):
+        return self.get_media("daisy")                       
+
     def short_html(self):
         key = '_short_html_%s' % get_language()
         short_html = getattr(self, key)
@@ -235,20 +343,20 @@ class Book(models.Model):
             tags = [mark_safe(u'<a href="%s">%s</a>' % (tag.get_absolute_url(), tag.name)) for tag in tags]
 
             formats = []
-            if self.html_file:
+            # 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.pdf_file:
-                formats.append(u'<a href="%s">PDF</a>' % self.pdf_file.url)
-            if self.epub_file:
-                formats.append(u'<a href="%s">EPUB</a>' % self.epub_file.url)
-            if self.odt_file:
-                formats.append(u'<a href="%s">ODT</a>' % self.odt_file.url)
-            if self.txt_file:
-                formats.append(u'<a href="%s">TXT</a>' % self.txt_file.url)
-            if self.mp3_file:
-                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.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("odt"):
+                formats.append(u'<a href="%s">ODT</a>' % self.get_media('odt').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.medias.order_by('type'):
+                formats.append(u'<a href="%s">%s</a>' % m.type, m.file.url)
 
             formats = [mark_safe(format) for format in formats]
 
@@ -258,9 +366,21 @@ class Book(models.Model):
             return mark_safe(getattr(self, key))
 
 
+    @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 get_mp3_info(self):
         """Retrieves artist and director names from audio ID3 tags."""
-        audio = id3.ID3(self.mp3_file.path)
+        audio = id3.ID3(self.get_media('mp3')[0].file.path)
         artist_name = ', '.join(', '.join(tag.text) for tag in audio.getall('TPE1'))
         director_name = ', '.join(', '.join(tag.text) for tag in audio.getall('TPE3'))
         return {'artist_name': artist_name, 'director_name': director_name}
@@ -270,6 +390,7 @@ class Book(models.Model):
     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'
@@ -280,16 +401,77 @@ class Book(models.Model):
     has_epub_file.short_description = 'EPUB'
     has_epub_file.boolean = True
 
-    def has_odt_file(self):
-        return bool(self.odt_file)
-    has_odt_file.short_description = 'ODT'
-    has_odt_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'
+    has_mp3_file.boolean = True
+
+    def has_ogg_file(self):
+        return bool(self.has_media("ogg"))
+    has_ogg_file.short_description = 'OGG'
+    has_ogg_file.boolean = True
+    
+    def has_daisy_file(self):
+        return bool(self.has_media("daisy"))
+    has_daisy_file.short_description = 'DAISY'
+    has_daisy_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(refresh_mp3=False)
+            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(refresh_mp3=False)
+            book_descendants += list(child_book.children.all())
+
+
     @classmethod
     def from_xml_file(cls, xml_file, overwrite=False):
         # use librarian to parse meta-data
@@ -308,6 +490,18 @@ class Book(models.Model):
         from tempfile import NamedTemporaryFile
         from slughifi import slughifi
         from markupstring import MarkupString
+        from django.core.files.storage import default_storage
+
+        # 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))
+                except Book.DoesNotExist, e:
+                    raise Book.DoesNotExist(_('Book with slug = "%s" does not exist.') % slug)
+
 
         # Read book metadata
         book_base, book_slug = book_info.url.rsplit('/', 1)
@@ -341,70 +535,83 @@ class Book(models.Model):
                 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.sort_key = tag_sort_key.lower()
                     tag.save()
                 book_tags.append(tag)
 
-        book.tags = book_tags
+        book.tags = book_tags + book_shelves
 
         book_tag = book.book_tag()
 
-        if hasattr(book_info, 'parts'):
-            for n, part_url in enumerate(book_info.parts):
-                base, slug = part_url.rsplit('/', 1)
-                try:
-                    child_book = Book.objects.get(slug=slug)
-                    child_book.parent = book
-                    child_book.parent_number = n
-                    child_book.save()
-                except Book.DoesNotExist, e:
-                    raise Book.DoesNotExist(_('Book with slug = "%s" does not exist.') % slug)
-
-        book_descendants = list(book.children.all())
-        while len(book_descendants) > 0:
-            child_book = book_descendants.pop(0)
-            child_book.tags = list(child_book.tags) + [book_tag]
+        for n, child_book in enumerate(children):
+            child_book.parent = book
+            child_book.parent_number = n
             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())
 
         # 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)
-            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:
+                    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.sort_key = theme_name.lower()
                         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] + ancestor_tags)
+
+        if not settings.NO_BUILD_EPUB:
+            book.root_ancestor().build_epub()
+
+        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]
+            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())
 
-            book_themes = set(book_themes)
-            book.tags = list(book.tags) + list(book_themes) + book_shelves
+        # refresh cache
+        book.reset_tag_counter()
+        book.reset_theme_counter()
 
         book.save()
         return book
@@ -421,6 +628,12 @@ class Book(models.Model):
         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 is None:
@@ -436,14 +649,18 @@ class Book(models.Model):
         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 is None:
             return self.refresh_theme_counter()
         return dict((int(k), v) for k, v in self.get__theme_counter_value().iteritems())
 
-
-
 class Fragment(models.Model):
     text = models.TextField()
     short_text = models.TextField(editable=False)
@@ -503,3 +720,34 @@ 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)
+