Django 1.7, working version.
[wolnelektury.git] / apps / catalogue / models / book.py
index d3df67e..8f5f107 100644 (file)
@@ -2,54 +2,83 @@
 # This file is part of Wolnelektury, licensed under GNU Affero GPLv3 or later.
 # Copyright © Fundacja Nowoczesna Polska. See NOTICE for more information.
 #
+from collections import OrderedDict
 import re
 from django.conf import settings
-from django.core.cache import get_cache
+from django.core.cache import caches
 from django.db import models
 from django.db.models import permalink
 import django.dispatch
-from django.utils.datastructures import SortedDict
+from django.contrib.contenttypes.fields import GenericRelation
+from django.core.urlresolvers import reverse
 from django.utils.translation import ugettext_lazy as _
 import jsonfield
+from fnpdjango.storage import BofhFileSystemStorage
+from catalogue import constants
+from catalogue.fields import EbookField
 from catalogue.models import Tag, Fragment, BookMedia
-from catalogue.utils import create_zip, split_tags, truncate_html_words, book_upload_path
+from catalogue.utils import create_zip, split_tags, related_tag_name
+from catalogue import app_settings
 from catalogue import tasks
 from newtagging import managers
 
+bofh_storage = BofhFileSystemStorage()
 
-permanent_cache = get_cache('permanent')
+permanent_cache = caches['permanent']
+
+
+def _cover_upload_to(i, n):
+    return 'book/cover/%s.jpg' % i.slug
+
+def _cover_thumb_upload_to(i, n):
+    return 'book/cover_thumb/%s.jpg' % i.slug,
+
+def _ebook_upload_to(upload_path):
+    def _upload_to(i, n):
+        return upload_path % i.slug
+    return _upload_to
 
 
 class Book(models.Model):
     """Represents a book imported from WL-XML."""
     title         = models.CharField(_('title'), max_length=120)
     sort_key = models.CharField(_('sort key'), max_length=120, db_index=True, editable=False)
+    sort_key_author = models.CharField(_('sort key by author'), max_length=120, db_index=True, editable=False, default=u'')
     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)
+                    default=app_settings.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)
     parent_number = models.IntegerField(_('parent number'), default=0)
-    extra_info    = jsonfield.JSONField(_('extra information'), default='{}')
+    extra_info    = jsonfield.JSONField(_('extra information'), default={})
     gazeta_link   = models.CharField(blank=True, max_length=240)
     wiki_link     = models.CharField(blank=True, max_length=240)
     # files generated during publication
 
-    cover = models.FileField(_('cover'), upload_to=book_upload_path('png'),
-                null=True, blank=True)
-    ebook_formats = ['pdf', 'epub', 'mobi', 'fb2', 'txt']
+    cover = EbookField('cover', _('cover'),
+            null=True, blank=True,
+            upload_to=_cover_upload_to,
+            storage=bofh_storage, max_length=255)
+    # Cleaner version of cover for thumbs
+    cover_thumb = EbookField('cover_thumb', _('cover thumbnail'), 
+            null=True, blank=True,
+            upload_to=_cover_thumb_upload_to,
+            max_length=255)
+    ebook_formats = constants.EBOOK_FORMATS
     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)
+    tag_relations = GenericRelation(Tag.intermediary_table_model)
 
     html_built = django.dispatch.Signal()
     published = django.dispatch.Signal()
@@ -70,8 +99,9 @@ class Book(models.Model):
         from sortify import sortify
 
         self.sort_key = sortify(self.title)
+        self.title = unicode(self.title) # ???
 
-        ret = super(Book, self).save(force_insert, force_update)
+        ret = super(Book, self).save(force_insert, force_update, **kwargs)
 
         if reset_short_html:
             self.reset_short_html()
@@ -82,10 +112,21 @@ class Book(models.Model):
     def get_absolute_url(self):
         return ('catalogue.views.book_detail', [self.slug])
 
+    @staticmethod
+    @permalink
+    def create_url(slug):
+        return ('catalogue.views.book_detail', [slug])
+
     @property
     def name(self):
         return self.title
 
+    def language_code(self):
+        return constants.LANGUAGES_3TO2.get(self.language, self.language)
+
+    def language_name(self):
+        return dict(settings.LANGUAGES).get(self.language_code(), "")
+
     def book_tag_slug(self):
         return ('l-' + self.slug)[:120]
 
@@ -108,7 +149,7 @@ class Book(models.Model):
         if self.has_media(type_):
             if type_ in Book.formats:
                 return getattr(self, "%s_file" % type_)
-            else:                                             
+            else:
                 return self.media.filter(type=type_)
         else:
             return None
@@ -120,7 +161,7 @@ class Book(models.Model):
     def get_ogg(self):
         return self.get_media("ogg")
     def get_daisy(self):
-        return self.get_media("daisy")                       
+        return self.get_media("daisy")
 
     def reset_short_html(self):
         if self.id is None:
@@ -131,6 +172,14 @@ class Book(models.Model):
         for fragm in self.fragments.all().iterator():
             fragm.reset_short_html()
 
+        try:
+            author = self.tags.filter(category='author')[0].sort_key
+        except IndexError:
+            author = u''
+        type(self).objects.filter(pk=self.pk).update(sort_key_author=author)
+
+
+
     def has_description(self):
         return len(self.description) > 0
     has_description.short_description = _('description')
@@ -152,117 +201,33 @@ class Book(models.Model):
     has_daisy_file.short_description = 'DAISY'
     has_daisy_file.boolean = True
 
-    def wldocument(self, parse_dublincore=True):
+    def wldocument(self, parse_dublincore=True, inherit=True):
         from catalogue.import_utils import ORMDocProvider
         from librarian.parser import WLDocument
 
+        if inherit and self.parent:
+            meta_fallbacks = self.parent.cover_info()
+        else:
+            meta_fallbacks = None
+
         return WLDocument.from_file(self.xml_file.path,
                 provider=ORMDocProvider(self),
-                parse_dublincore=parse_dublincore)
-
-    def build_cover(self, book_info=None):
-        """(Re)builds the cover image."""
-        from StringIO import StringIO
-        from django.core.files.base import ContentFile
-        from librarian.cover import WLCover
-
-        if book_info is None:
-            book_info = self.wldocument().book_info
-
-        cover = WLCover(book_info).image()
-        imgstr = StringIO()
-        cover.save(imgstr, 'png')
-        self.cover.save(None, ContentFile(imgstr.getvalue()))
-
-    def build_html(self):
-        from django.core.files.base import ContentFile
-        from slughifi import slughifi
-        from sortify import sortify
-        from librarian import html
-
-        meta_tags = list(self.tags.filter(
-            category__in=('author', 'epoch', 'genre', 'kind')))
-        book_tag = self.book_tag()
-
-        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 = []
-            p = self.parent
-            while p:
-                ancestor_tags.append(p.book_tag())
-                p = p.parent
-
-            # Delete old fragments and create them from scratch
-            self.fragments.all().delete()
-            # Extract fragments
-            closed_fragments, open_fragments = html.extract_fragments(self.html_file.path)
-            for fragment in closed_fragments.values():
-                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 = sortify(theme_name.lower())
-                        tag.save()
-                    themes.append(tag)
-                if not themes:
-                    continue
-
-                text = fragment.to_string()
-                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.save()
-                new_fragment.tags = set(meta_tags + themes + [book_tag] + ancestor_tags)
-            self.save()
-            self.html_built.send(sender=self)
-            return True
-        return False
-
-    # Thin wrappers for builder tasks
-    def build_pdf(self, *args, **kwargs):
-        """(Re)builds PDF."""
-        return tasks.build_pdf.delay(self.pk, *args, **kwargs)
-    def build_epub(self, *args, **kwargs):
-        """(Re)builds EPUB."""
-        return tasks.build_epub.delay(self.pk, *args, **kwargs)
-    def build_mobi(self, *args, **kwargs):
-        """(Re)builds MOBI."""
-        return tasks.build_mobi.delay(self.pk, *args, **kwargs)
-    def build_fb2(self, *args, **kwargs):
-        """(Re)build FB2"""
-        return tasks.build_fb2.delay(self.pk, *args, **kwargs)
-    def build_txt(self, *args, **kwargs):
-        """(Re)builds TXT."""
-        return tasks.build_txt.delay(self.pk, *args, **kwargs)
+                parse_dublincore=parse_dublincore,
+                meta_fallbacks=meta_fallbacks)
 
     @staticmethod
     def zip_format(format_):
         def pretty_file_name(book):
             return "%s/%s.%s" % (
-                b.extra_info['author'],
-                b.slug,
+                book.extra_info['author'],
+                book.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.iterator()]
-        return create_zip(paths,
-                    getattr(settings, "ALL_%s_ZIP" % format_.upper()))
+        return create_zip(paths, app_settings.FORMAT_ZIPS[format_])
 
     def zip_audiobooks(self, format_):
         bm = BookMedia.objects.filter(book=self, type=format_)
@@ -270,13 +235,13 @@ class Book(models.Model):
         return create_zip(paths, "%s_%s" % (self.slug, format_))
 
     def search_index(self, book_info=None, index=None, index_tags=True, commit=True):
-        import search
         if index is None:
-            index = search.Index()
+            from search.index import Index
+            index = Index()
         try:
             index.index_book(self, book_info)
             if index_tags:
-                idx.index_tags()
+                index.index_tags()
             if commit:
                 index.index.commit()
         except Exception, e:
@@ -302,8 +267,11 @@ class Book(models.Model):
 
     @classmethod
     def from_text_and_meta(cls, raw_file, book_info, overwrite=False,
-            build_epub=True, build_txt=True, build_pdf=True, build_mobi=True, build_fb2=True,
-            search_index=True, search_index_tags=True):
+            dont_build=None, search_index=True,
+            search_index_tags=True):
+        if dont_build is None:
+            dont_build = set()
+        dont_build = set.union(set(dont_build), set(app_settings.DONT_BUILD))
 
         # check for parts before we do anything
         children = []
@@ -315,7 +283,6 @@ class Book(models.Model):
                     raise Book.DoesNotExist(_('Book "%s" does not exist.') %
                             part_url.slug)
 
-
         # Read book metadata
         book_slug = book_info.url.slug
         if re.search(r'[^a-z0-9-]', book_slug):
@@ -324,12 +291,17 @@ class Book(models.Model):
 
         if created:
             book_shelves = []
+            old_cover = None
         else:
             if not overwrite:
                 raise Book.AlreadyExists(_('Book %s already exists') % (
                         book_slug))
             # Save shelves for this book
             book_shelves = list(book.tags.filter(category='set'))
+            old_cover = book.cover_info()
+
+        # Save XML file
+        book.xml_file.save('%s.xml' % book.slug, raw_file, save=False)
 
         book.language = book_info.language
         book.title = book_info.title
@@ -344,46 +316,49 @@ class Book(models.Model):
 
         book.tags = set(meta_tags + book_shelves)
 
-        obsolete_children = set(b for b in book.children.all() if b not in children)
+        cover_changed = old_cover != book.cover_info()
+        obsolete_children = set(b for b in book.children.all()
+                                if b not in children)
+        notify_cover_changed = []
         for n, child_book in enumerate(children):
+            new_child = child_book.parent != book
             child_book.parent = book
             child_book.parent_number = n
             child_book.save()
+            if new_child or cover_changed:
+                notify_cover_changed.append(child_book)
         # Disown unfaithful children and let them cope on their own.
         for child in obsolete_children:
             child.parent = None
             child.parent_number = 0
             child.save()
             tasks.fix_tree_tags.delay(child)
-
-        # Save XML and HTML files
-        book.xml_file.save('%s.xml' % book.slug, raw_file, save=False)
-        book.build_cover(book_info)
-
-        # delete old fragments when overwriting
-        book.fragments.all().delete()
-
-        if book.build_html():
-            # No direct saves behind this point.
-            if not settings.NO_BUILD_TXT and build_txt:
-                book.build_txt()
-
-        if not settings.NO_BUILD_EPUB and build_epub:
-            book.build_epub()
-
-        if not settings.NO_BUILD_PDF and build_pdf:
-            book.build_pdf()
-
-        if not settings.NO_BUILD_MOBI and build_mobi:
-            book.build_mobi()
-
-        if not settings.NO_BUILD_FB2 and build_fb2:
-            book.build_fb2()
+            if old_cover:
+                notify_cover_changed.append(child)
+
+        # No saves beyond this point.
+
+        # Build cover.
+        if 'cover' not in dont_build:
+            book.cover.build_delay()
+            book.cover_thumb.build_delay()
+
+        # Build HTML and ebooks.
+        book.html_file.build_delay()
+        if not children:
+            for format_ in constants.EBOOK_FORMATS_WITHOUT_CHILDREN:
+                if format_ not in dont_build:
+                    getattr(book, '%s_file' % format_).build_delay()
+        for format_ in constants.EBOOK_FORMATS_WITH_CHILDREN:
+            if format_ not in dont_build:
+                getattr(book, '%s_file' % format_).build_delay()
 
         if not settings.NO_SEARCH_INDEX and search_index:
             tasks.index_book.delay(book.id, book_info=book_info, index_tags=search_index_tags)
 
-        tasks.fix_tree_tags.delay(book)
+        for child in notify_cover_changed:
+            child.parent_cover_changed()
+
         cls.published.send(sender=book)
         return book
 
@@ -403,7 +378,8 @@ class Book(models.Model):
             sub_parent_tags = parent_tags + [book.book_tag()]
             for frag in book.fragments.all():
                 affected_tags.update(frag.tags)
-                frag.tags = list(frag.tags.exclude(category='book')) + sub_parent_tags
+                frag.tags = list(frag.tags.exclude(category='book')
+                                    ) + sub_parent_tags
             for child in book.children.all():
                 affected_tags.update(fix_subtree(child, sub_parent_tags))
             return affected_tags
@@ -424,6 +400,41 @@ class Book(models.Model):
             book.reset_theme_counter()
             book = book.parent
 
+    def cover_info(self, inherit=True):
+        """Returns a dictionary to serve as fallback for BookInfo.
+
+        For now, the only thing inherited is the cover image.
+        """
+        need = False
+        info = {}
+        for field in ('cover_url', 'cover_by', 'cover_source'):
+            val = self.extra_info.get(field)
+            if val:
+                info[field] = val
+            else:
+                need = True
+        if inherit and need and self.parent is not None:
+            parent_info = self.parent.cover_info()
+            parent_info.update(info)
+            info = parent_info
+        return info
+
+    def parent_cover_changed(self):
+        """Called when parent book's cover image is changed."""
+        if not self.cover_info(inherit=False):
+            if 'cover' not in app_settings.DONT_BUILD:
+                self.cover.build_delay()
+                self.cover_thumb.build_delay()
+            for format_ in constants.EBOOK_FORMATS_WITH_COVERS:
+                if format_ not in app_settings.DONT_BUILD:
+                    getattr(self, '%s_file' % format_).build_delay()
+            for child in self.children.all():
+                child.parent_cover_changed()
+
+    def other_versions(self):
+        """Find other versions (i.e. in other languages) of the book."""
+        return type(self).objects.filter(common_slug=self.common_slug).exclude(pk=self.pk)
+
     def related_info(self):
         """Keeps info about related objects (tags, media) in cache field."""
         if self._related_info is not None:
@@ -435,8 +446,15 @@ class Book(models.Model):
                     'author', 'kind', 'genre', 'epoch'))
             tags = split_tags(tags)
             for category in tags:
-                rel['tags'][category] = [
-                        (t.name, t.slug) for t in tags[category]]
+                cat = []
+                for tag in tags[category]:
+                    tag_info = {'slug': tag.slug, 'name': tag.name}
+                    for lc, ln in settings.LANGUAGES:
+                        tag_name = getattr(tag, "name_%s" % lc)
+                        if tag_name:
+                            tag_info["name_%s" % lc] = tag_name
+                    cat.append(tag_info)
+                rel['tags'][category] = cat
 
             for media_format in BookMedia.formats:
                 rel['media'][media_format] = self.has_media(media_format)
@@ -519,19 +537,19 @@ class Book(models.Model):
 
     def pretty_title(self, html_links=False):
         book = self
-        names = list(book.tags.filter(category='author'))
-
-        books = []
-        while book:
-            books.append(book)
-            book = book.parent
-        names.extend(reversed(books))
+        rel_info = book.related_info()
+        names = [(related_tag_name(tag), Tag.create_url('author', tag['slug']))
+                    for tag in rel_info['tags'].get('author', ())]
+        if 'parents' in rel_info:
+            books = [(name, Book.create_url(slug))
+                        for name, slug in rel_info['parents']]
+            names.extend(reversed(books))
+        names.append((self.title, self.get_absolute_url()))
 
         if html_links:
-            names = ['<a href="%s">%s</a>' % (tag.get_absolute_url(), tag.name) for tag in names]
+            names = ['<a href="%s">%s</a>' % (tag[1], tag[0]) for tag in names]
         else:
-            names = [tag.name for tag in names]
-
+            names = [tag[0] for tag in names]
         return ', '.join(names)
 
     @classmethod
@@ -544,10 +562,11 @@ class Book(models.Model):
         """
         # get relevant books and their tags
         objects = cls.tagged.with_all(tags)
+        parents = objects.exclude(children=None).only('slug')
         # eliminate descendants
         l_tags = Tag.objects.filter(category='book',
-            slug__in=[book.book_tag_slug() for book in objects.iterator()])
-        descendants_keys = [book.pk for book in cls.tagged.with_any(l_tags).iterator()]
+            slug__in=[book.book_tag_slug() for book in parents.iterator()])
+        descendants_keys = [book.pk for book in cls.tagged.with_any(l_tags).only('pk').iterator()]
         if descendants_keys:
             objects = objects.exclude(pk__in=descendants_keys)
 
@@ -566,7 +585,7 @@ class Book(models.Model):
                 'title', 'parent', 'slug')
         if filter:
             books = books.filter(filter).distinct()
-            
+
             book_ids = set(b['pk'] for b in books.values("pk").iterator())
             for book in books.iterator():
                 parent = book.parent_id
@@ -578,11 +597,11 @@ class Book(models.Model):
                 books_by_parent.setdefault(book.parent_id, []).append(book)
 
         orphans = []
-        books_by_author = SortedDict()
+        books_by_author = OrderedDict()
         for tag in Tag.objects.filter(category='author').iterator():
             books_by_author[tag] = []
 
-        for book in books_by_parent.get(None,()):
+        for book in books_by_parent.get(None, ()):
             authors = list(book.tags.filter(category='author'))
             if authors:
                 for author in authors:
@@ -593,6 +612,7 @@ class Book(models.Model):
         return books_by_author, orphans, books_by_parent
 
     _audiences_pl = {
+        "SP": (1, u"szkoła podstawowa"),
         "SP1": (1, u"szkoła podstawowa"),
         "SP2": (1, u"szkoła podstawowa"),
         "P": (1, u"szkoła podstawowa"),
@@ -602,9 +622,17 @@ class Book(models.Model):
     }
     def audiences_pl(self):
         audiences = self.extra_info.get('audiences', [])
-        audiences = sorted(set([self._audiences_pl[a] for a in audiences]))
+        audiences = sorted(set([self._audiences_pl.get(a, (99, a)) for a in audiences]))
         return [a[1] for a in audiences]
 
+    def stage_note(self):
+        stage = self.extra_info.get('stage')
+        if stage and stage < '0.4':
+            return (_('This work needs modernisation'),
+                    reverse('infopage', args=['wymagajace-uwspolczesnienia']))
+        else:
+            return None, None
+
     def choose_fragment(self):
         tag = self.book_tag()
         fragments = Fragment.tagged.with_any([tag])
@@ -615,21 +643,18 @@ class Book(models.Model):
         else:
             return None
 
-
-def _has_factory(ftype):
-    has = lambda self: bool(getattr(self, "%s_file" % ftype))
-    has.short_description = ftype.upper()
-    has.__doc__ = None
-    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))
+for format_ in Book.formats:
+    field_name = "%s_file" % format_
+    # This weird globals() assignment makes Django migrations comfortable.
+    _upload_to = _ebook_upload_to('book/%s/%%s.%s' % (format_, format_))
+    _upload_to.__name__ = '_%s_upload_to' % format_
+    globals()[_upload_to.__name__] = _upload_to
+
+    EbookField(format_, _("%s file" % format_.upper()),
+        upload_to=_upload_to,
+        storage=bofh_storage,
+        max_length=255,
+        blank=True,
+        default=''
+    ).contribute_to_class(Book, field_name)