In news, use a general feed of foundation's blog, not just wolne-lektury category.
[wolnelektury.git] / apps / catalogue / models.py
index 1f6210d..29106b1 100644 (file)
@@ -5,10 +5,9 @@
 from collections import namedtuple
 
 from django.db import models
-from django.db.models import permalink, Q
+from django.db.models import permalink
 import django.dispatch
 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.template.loader import render_to_string
@@ -16,22 +15,18 @@ 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.db.models.signals import post_save, m2m_changed, pre_delete
+from django.db.models.signals import post_save, 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 catalogue.utils import create_zip, split_tags
-from catalogue.tasks import touch_tag, index_book
-from shutil import copy
-from glob import glob
+from catalogue.fields import OverwritingFileField
+from catalogue.utils import create_zip, split_tags, truncate_html_words
+from catalogue import tasks
 import re
-from os import path
 
-import search
 
 # Those are hard-coded here so that makemessages sees them.
 TAG_CATEGORIES = (
@@ -58,6 +53,10 @@ class TagSubcategoryManager(models.Manager):
 
 
 class Tag(TagBase):
+    """A tag attachable to books and fragments (and possibly anything).
+    
+    Used to represent searchable metadata (authors, epochs, genres, kinds),
+    fragment themes (motifs) and some book hierarchy related kludges."""
     name = models.CharField(_('name'), max_length=50, db_index=True)
     slug = models.SlugField(_('slug'), max_length=120, db_index=True)
     sort_key = models.CharField(_('sort key'), max_length=120, db_index=True)
@@ -119,8 +118,8 @@ class Tag(TagBase):
             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)]
+                l_tags = Tag.objects.filter(slug__in=[book.book_tag_slug() for book in objects.iterator()])
+                descendants_keys = [book.pk for book in Book.tagged.with_any(l_tags).iterator()]
                 if descendants_keys:
                     objects = objects.exclude(pk__in=descendants_keys)
         return objects.count()
@@ -216,30 +215,8 @@ def book_upload_path(ext=None, maxlen=100):
     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):
+    """Represents media attached to a book."""
     FileFormat = namedtuple("FileFormat", "name ext")
     formats = SortedDict([
         ('mp3', FileFormat(name='MP3', ext='mp3')),
@@ -253,7 +230,7 @@ class BookMedia(models.Model):
     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)
-    extra_info  = JSONField(_('extra information'), default='{}', editable=False)
+    extra_info  = jsonfield.JSONField(_('extra information'), default='{}', editable=False)
     book = models.ForeignKey('Book', related_name='media')
     source_sha1 = models.CharField(null=True, blank=True, max_length=40, editable=False)
 
@@ -271,7 +248,7 @@ class BookMedia(models.Model):
 
         try:
             old = BookMedia.objects.get(pk=self.pk)
-        except BookMedia.DoesNotExist, e:
+        except BookMedia.DoesNotExist:
             old = None
         else:
             # if name changed, change the file name, too
@@ -285,9 +262,9 @@ class BookMedia(models.Model):
             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 = self.extra_info
         extra_info.update(self.read_meta())
-        self.set_extra_info_value(extra_info)
+        self.extra_info = extra_info
         self.source_sha1 = self.read_source_sha1(self.file.path, self.type)
         return super(BookMedia, self).save(*args, **kwargs)
 
@@ -350,6 +327,7 @@ class BookMedia(models.Model):
 
 
 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)
     slug = models.SlugField(_('slug'), max_length=120, db_index=True,
@@ -361,7 +339,7 @@ class Book(models.Model):
     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(_('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
@@ -425,18 +403,18 @@ class Book(models.Model):
             book_tag.save()
         return book_tag
 
-    def has_media(self, type):
-        if type in Book.formats:
-            return bool(getattr(self, "%s_file" % type))
+    def has_media(self, type_):
+        if type_ in Book.formats:
+            return bool(getattr(self, "%s_file" % type_))
         else:
-            return self.media.filter(type=type).exists()
+            return self.media.filter(type=type_).exists()
 
-    def get_media(self, type):
-        if self.has_media(type):
-            if type in Book.formats:
-                return getattr(self, "%s_file" % type)
+    def get_media(self, type_):
+        if self.has_media(type_):
+            if type_ in Book.formats:
+                return getattr(self, "%s_file" % type_)
             else:                                             
-                return self.media.filter(type=type)
+                return self.media.filter(type=type_)
         else:
             return None
 
@@ -455,7 +433,7 @@ class Book(models.Model):
 
         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():
+        for fragm in self.fragments.all().iterator():
             fragm.reset_short_html()
 
     def has_description(self):
@@ -501,71 +479,7 @@ class Book(models.Model):
         cover.save(imgstr, 'png')
         self.cover.save(None, ContentFile(imgstr.getvalue()))
 
-    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):
-        from django.core.files.base import ContentFile
-
-        text = self.wldocument().as_text()
-        self.txt_file.save('%s.txt' % self.slug, ContentFile(text.get_string()))
-
-
     def build_html(self):
-        from markupstring import MarkupString
         from django.core.files.base import ContentFile
         from slughifi import slughifi
         from librarian import html
@@ -609,9 +523,9 @@ class Book(models.Model):
                     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)
 
@@ -622,29 +536,42 @@ class Book(models.Model):
             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_txt(self, *args, **kwargs):
+        """(Re)builds TXT."""
+        return tasks.build_txt.delay(self.pk, *args, **kwargs)
+
     @staticmethod
     def zip_format(format_):
         def pretty_file_name(book):
             return "%s/%s.%s" % (
-                b.get_extra_info_value()['author'],
+                b.extra_info['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,
+                    for b in books.iterator()]
+        return create_zip(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()
+        return create_zip(paths, "%s_%s" % (self.slug, format_))
 
-    def search_index(self, book_info=None, reuse_index=False):
+    def search_index(self, book_info=None, reuse_index=False, index_tags=True):
+        import search
         if reuse_index:
             idx = search.ReusableIndex()
         else:
@@ -653,7 +580,8 @@ class Book(models.Model):
         idx.open()
         try:
             idx.index_book(self, book_info)
-            idx.index_tags()
+            if index_tags:
+                idx.index_tags()
         finally:
             idx.close()
 
@@ -676,9 +604,7 @@ 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,
-            search_index=True):
-        import re
-        from sortify import sortify
+            search_index=True, search_index_tags=True, search_index_reuse=False):
 
         # check for parts before we do anything
         children = []
@@ -686,7 +612,7 @@ class Book(models.Model):
             for part_url in book_info.parts:
                 try:
                     children.append(Book.objects.get(slug=part_url.slug))
-                except Book.DoesNotExist, e:
+                except Book.DoesNotExist:
                     raise Book.DoesNotExist(_('Book "%s" does not exist.') %
                             part_url.slug)
 
@@ -712,7 +638,7 @@ class Book(models.Model):
             book.common_slug = book_info.variant_of.slug
         else:
             book.common_slug = book.slug
-        book.set_extra_info_value(book_info.to_dict())
+        book.extra_info = book_info.to_dict()
         book.save()
 
         meta_tags = Tag.tags_from_info(book_info)
@@ -748,7 +674,8 @@ class Book(models.Model):
             book.build_mobi()
 
         if not settings.NO_SEARCH_INDEX and search_index:
-            index_book.delay(book.id, book_info)
+            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())
         descendants_tags = set()
@@ -758,12 +685,12 @@ class Book(models.Model):
             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():
+            for fragment in child_book.fragments.all().iterator():
                 fragment.tags = set(list(fragment.tags) + [book_tag])
             book_descendants += list(child_book.children.all())
 
         for tag in descendants_tags:
-            touch_tag.delay(tag)
+            tasks.touch_tag(tag)
 
         book.save()
 
@@ -780,18 +707,37 @@ class Book(models.Model):
             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.get_absolute_url()) for t in 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 related_themes(self):
+        theme_counter = self.theme_counter
+        book_themes = list(Tag.objects.filter(pk__in=theme_counter.keys()))
+        for tag in book_themes:
+            tag.count = theme_counter[tag.pk]
+        return book_themes
+
     def reset_tag_counter(self):
         if self.id is None:
             return
@@ -811,10 +757,10 @@ class Book(models.Model):
 
         if tags is None:
             tags = {}
-            for child in self.children.all().order_by():
+            for child in self.children.all().order_by().iterator():
                 for tag_pk, value in child.tag_counter.iteritems():
                     tags[tag_pk] = tags.get(tag_pk, 0) + value
-            for tag in self.tags.exclude(category__in=('book', 'theme', 'set')).order_by():
+            for tag in self.tags.exclude(category__in=('book', 'theme', 'set')).order_by().iterator():
                 tags[tag.pk] = 1
 
             if self.id:
@@ -840,8 +786,8 @@ class Book(models.Model):
 
         if tags is None:
             tags = {}
-            for fragment in Fragment.tagged.with_any([self.book_tag()]).order_by():
-                for tag in fragment.tags.filter(category='theme').order_by():
+            for fragment in Fragment.tagged.with_any([self.book_tag()]).order_by().iterator():
+                for tag in fragment.tags.filter(category='theme').order_by().iterator():
                     tags[tag.pk] = tags.get(tag.pk, 0) + 1
 
             if self.id:
@@ -867,7 +813,7 @@ class Book(models.Model):
 
     @classmethod
     def tagged_top_level(cls, tags):
-        """ Returns top-level books tagged with `tags'.
+        """ Returns top-level books tagged with `tags`.
 
         It only returns those books which don't have ancestors which are
         also tagged with those tags.
@@ -876,8 +822,9 @@ class Book(models.Model):
         # get relevant books and their tags
         objects = cls.tagged.with_all(tags)
         # eliminate descendants
-        l_tags = Tag.objects.filter(category='book', slug__in=[book.book_tag_slug() for book in objects])
-        descendants_keys = [book.pk for book in cls.tagged.with_any(l_tags)]
+        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()]
         if descendants_keys:
             objects = objects.exclude(pk__in=descendants_keys)
 
@@ -896,19 +843,20 @@ class Book(models.Model):
                 'title', 'parent', 'slug')
         if filter:
             books = books.filter(filter).distinct()
-            book_ids = set((book.pk for book in books))
-            for book in books:
+            
+            book_ids = set(b['pk'] for b in books.values("pk").iterator())
+            for book in books.iterator():
                 parent = book.parent_id
                 if parent not in book_ids:
                     parent = None
                 books_by_parent.setdefault(parent, []).append(book)
         else:
-            for book in books:
+            for book in books.iterator():
                 books_by_parent.setdefault(book.parent_id, []).append(book)
 
         orphans = []
         books_by_author = SortedDict()
-        for tag in Tag.objects.filter(category='author'):
+        for tag in Tag.objects.filter(category='author').iterator():
             books_by_author[tag] = []
 
         for book in books_by_parent.get(None,()):
@@ -930,14 +878,25 @@ class Book(models.Model):
         "LP": (3, u"liceum"),
     }
     def audiences_pl(self):
-        audiences = self.get_extra_info_value().get('audiences', [])
+        audiences = self.extra_info.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.short_description = ftype.upper()
+    has.__doc__ = None
     has.boolean = True
     has.__name__ = "has_%s_file" % ftype
     return has
@@ -954,6 +913,7 @@ for t in Book.formats:
 
 
 class Fragment(models.Model):
+    """Represents a themed fragment of a book."""
     text = models.TextField()
     short_text = models.TextField(editable=False)
     anchor = models.CharField(max_length=120)
@@ -979,6 +939,10 @@ class Fragment(models.Model):
         for lang, langname in settings.LANGUAGES:
             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())
@@ -1025,7 +989,7 @@ 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
     for tag in affected_tags:
-        touch_tag.delay(tag)
+        tasks.touch_tag(tag)
 
     # if book tags changed, reset book tag counter
     if isinstance(sender, Book) and \
@@ -1046,8 +1010,24 @@ def _pre_delete_handler(sender, instance, **kwargs):
         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)
+
+
+if not settings.NO_SEARCH_INDEX:
+    @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."""
+        import search
+        search.JVM.attachCurrentThread()
+        idx = search.Index()
+        idx.open(timeout=10000)  # 10 seconds timeout.
+        try:
+            idx.remove_book(instance)
+            idx.index_tags()
+        finally:
+            idx.close()