texml in requirements
[redakcja.git] / apps / catalogue / models / book.py
index db52128..f109474 100755 (executable)
@@ -3,46 +3,57 @@
 # This file is part of FNP-Redakcja, licensed under GNU Affero GPLv3 or later.
 # Copyright © Fundacja Nowoczesna Polska. See NOTICE for more information.
 #
-from django.db import models
+from django.contrib.sites.models import Site
+from django.db import models, transaction
 from django.template.loader import render_to_string
 from django.utils.translation import ugettext_lazy as _
-from slughifi import slughifi
-from catalogue.helpers import cached_in_field
-from catalogue.models import BookPublishRecord, ChunkPublishRecord
+from django.conf import settings
+from slugify import slugify
+
+
+import apiclient
+from catalogue.helpers import cached_in_field, GalleryMerger
+from catalogue.models import BookPublishRecord, ChunkPublishRecord, Project
 from catalogue.signals import post_publish
-from catalogue.tasks import refresh_instance
+from catalogue.tasks import refresh_instance, book_content_updated
 from catalogue.xml_tools import compile_text, split_xml
-
+from cover.models import Image
+import os
+import shutil
+import re
 
 class Book(models.Model):
     """ A document edited on the wiki """
 
     title = models.CharField(_('title'), max_length=255, db_index=True)
     slug = models.SlugField(_('slug'), max_length=128, unique=True, db_index=True)
+    public = models.BooleanField(_('public'), default=True, db_index=True)
     gallery = models.CharField(_('scan gallery name'), max_length=255, blank=True)
+    project = models.ForeignKey(Project, null=True, blank=True)
 
     #wl_slug = models.CharField(_('title'), max_length=255, null=True, db_index=True, editable=False)
-    parent = models.ForeignKey('self', null=True, blank=True, verbose_name=_('parent'), related_name="children")
-    parent_number = models.IntegerField(_('parent number'), null=True, blank=True, db_index=True)
+    parent = models.ForeignKey('self', null=True, blank=True, verbose_name=_('parent'), related_name="children", editable=False)
+    parent_number = models.IntegerField(_('parent number'), null=True, blank=True, db_index=True, editable=False)
 
     # Cache
     _short_html = models.TextField(null=True, blank=True, editable=False)
     _single = models.NullBooleanField(editable=False, db_index=True)
     _new_publishable = models.NullBooleanField(editable=False)
     _published = models.NullBooleanField(editable=False)
-
-    # Managers
-    objects = models.Manager()
+    _on_track = models.IntegerField(null=True, blank=True, db_index=True, editable=False)
+    dc_cover_image = models.ForeignKey(Image, blank=True, null=True,
+        db_index=True, on_delete=models.SET_NULL, editable=False)
+    dc_slug = models.CharField(max_length=128, null=True, blank=True,
+            editable=False, db_index=True)
 
     class NoTextError(BaseException):
         pass
 
     class Meta:
         app_label = 'catalogue'
-        ordering = ['parent_number', 'title']
+        ordering = ['title', 'slug']
         verbose_name = _('book')
         verbose_name_plural = _('books')
-        permissions = [('can_pubmark', 'Can mark for publishing')]
 
 
     # Representing
@@ -70,11 +81,20 @@ class Book(models.Model):
     def get_absolute_url(self):
         return ("catalogue_book", [self.slug])
 
+    def correct_about(self):
+        return "http://%s%s" % (
+            Site.objects.get_current().domain,
+            self.get_absolute_url()
+        )
 
     # Creating & manipulating
     # =======================
 
+    def accessible(self, request):
+        return self.public or request.user.is_authenticated()
+
     @classmethod
+    @transaction.atomic
     def create(cls, creator, text, *args, **kwargs):
         b = cls.objects.create(*args, **kwargs)
         b.chunk_set.all().update(creator=creator)
@@ -86,6 +106,7 @@ class Book(models.Model):
         return self.chunk_set.reverse()[0].split(*args, **kwargs)
 
     @classmethod
+    @transaction.atomic
     def import_xml_text(cls, text=u'', previous_book=None,
                 commit_args=None, **kwargs):
         """Imports a book from XML, splitting it into chunks as necessary."""
@@ -106,7 +127,7 @@ class Book(models.Model):
             if not title:
                 title = u'część %d' % (i + 1)
 
-            slug = slughifi(title)
+            slug = slugify(title)
 
             if i < book_len:
                 chunk = instance[i]
@@ -114,7 +135,7 @@ class Book(models.Model):
                 chunk.title = title[:255]
                 chunk.save()
             else:
-                chunk = instance.add(slug, title, adjust_slug=True)
+                chunk = instance.add(slug, title)
 
             chunk.commit(text, **commit_args)
 
@@ -126,14 +147,17 @@ class Book(models.Model):
         """
         slugs = set(c.slug for c in self)
         i = 1
-        new_slug = proposed
+        new_slug = proposed[:50]
         while new_slug in slugs:
-            new_slug = "%s_%d" % (proposed, i)
+            new_slug = "%s_%d" % (proposed[:45], i)
             i += 1
         return new_slug
 
+    @transaction.atomic
     def append(self, other, slugs=None, titles=None):
         """Add all chunks of another book to self."""
+        assert self != other
+
         number = self[len(self) - 1].number + 1
         len_other = len(other)
         single = len_other == 1
@@ -143,7 +167,7 @@ class Book(models.Model):
         if titles is not None:
             assert len(titles) == len_other
             if slugs is None:
-                slugs = [slughifi(t) for t in titles]
+                slugs = [slugify(t) for t in titles]
 
         for i, chunk in enumerate(other):
             # move chunk to new book
@@ -162,12 +186,11 @@ class Book(models.Model):
                     # just use the guessed title and original book slug
                     chunk.title = other_title_part
                     if other.slug.startswith(self.slug):
-                        chunk_slug = other.slug[len(self.slug):].lstrip('-_')
+                        chunk.slug = other.slug[len(self.slug):].lstrip('-_')
                     else:
-                        chunk_slug = other.slug
-                    chunk.slug = self.make_chunk_slug(chunk_slug)
+                        chunk.slug = other.slug
                 else:
-                    chunk.title = "%s, %s" % (other_title_part, chunk.title)
+                    chunk.title = ("%s, %s" % (other_title_part, chunk.title))[:255]
             else:
                 chunk.slug = slugs[i]
                 chunk.title = titles[i]
@@ -175,8 +198,49 @@ class Book(models.Model):
             chunk.slug = self.make_chunk_slug(chunk.slug)
             chunk.save()
             number += 1
+        assert not other.chunk_set.exists()
+
+        gm = GalleryMerger(self.gallery, other.gallery)
+        self.gallery = gm.merge()
+
+        # and move the gallery starts
+        if gm.was_merged:
+                for chunk in self[len(self) - len_other:]:
+                        old_start = chunk.gallery_start or 1
+                        chunk.gallery_start = old_start + gm.dest_size - gm.num_deleted
+                        chunk.save()
+
+        other.delete()
+
+
+    @transaction.atomic
+    def prepend_history(self, other):
+        """Prepend history from all the other book's chunks to own."""
+        assert self != other
+
+        for i in range(len(self), len(other)):
+            title = u"pusta część %d" % i
+            chunk = self.add(slugify(title), title)
+            chunk.commit('')
+
+        for i in range(len(other)):
+            self[i].prepend_history(other[0])
+
+        assert not other.chunk_set.exists()
         other.delete()
 
+    def split(self):
+        """Splits all the chunks into separate books."""
+        self.title
+        for chunk in self:
+            book = Book.objects.create(title=chunk.title, slug=chunk.slug,
+                    public=self.public, gallery=self.gallery)
+            book[0].delete()
+            chunk.book = book
+            chunk.number = 1
+            chunk.save()
+        assert not self.chunk_set.exists()
+        self.delete()
 
     # State & cache
     # =============
@@ -187,13 +251,34 @@ class Book(models.Model):
         except IndexError:
             return None
 
-    def publishable(self):
-        if not self.chunk_set.exists():
-            return False
-        for chunk in self:
-            if not chunk.publishable():
-                return False
-        return True
+    def assert_publishable(self):
+        assert self.chunk_set.exists(), _('No chunks in the book.')
+        try:
+            changes = self.get_current_changes(publishable=True)
+        except self.NoTextError:
+            raise AssertionError(_('Not all chunks have publishable revisions.'))
+
+        from librarian import NoDublinCore, ParseError, ValidationError
+
+        try:
+            bi = self.wldocument(changes=changes, strict=True).book_info
+        except ParseError, e:
+            raise AssertionError(_('Invalid XML') + ': ' + unicode(e))
+        except NoDublinCore:
+            raise AssertionError(_('No Dublin Core found.'))
+        except ValidationError, e:
+            raise AssertionError(_('Invalid Dublin Core') + ': ' + unicode(e))
+
+        valid_about = self.correct_about()
+        assert bi.about == valid_about, _("rdf:about is not") + " " + valid_about
+
+    def publishable_error(self):
+        try:
+            return self.assert_publishable()
+        except AssertionError, e:
+            return e
+        else:
+            return None
 
     def hidden(self):
         return self.slug.startswith('.')
@@ -221,6 +306,16 @@ class Book(models.Model):
         return self.publish_log.exists()
     published = cached_in_field('_published')(is_published)
 
+    def get_on_track(self):
+        if self.published:
+            return -1
+        stages = [ch.stage.ordering if ch.stage is not None else 0
+                    for ch in self]
+        if not len(stages):
+            return 0
+        return min(stages)
+    on_track = cached_in_field('_on_track')(get_on_track)
+
     def is_single(self):
         return len(self) == 1
     single = cached_in_field('_single')(is_single)
@@ -229,11 +324,47 @@ class Book(models.Model):
     def short_html(self):
         return render_to_string('catalogue/book_list/book.html', {'book': self})
 
+    def book_info(self, publishable=True):
+        try:
+            book_xml = self.materialize(publishable=publishable)
+        except self.NoTextError:
+            pass
+        else:
+            from librarian.dcparser import BookInfo
+            from librarian import NoDublinCore, ParseError, ValidationError
+            try:
+                return BookInfo.from_string(book_xml.encode('utf-8'))
+            except (self.NoTextError, ParseError, NoDublinCore, ValidationError):
+                return None
+
+    def refresh_dc_cache(self):
+        update = {
+            'dc_slug': None,
+            'dc_cover_image': None,
+        }
+
+        info = self.book_info()
+        if info is not None:
+            update['dc_slug'] = info.url.slug
+            if info.cover_source:
+                try:
+                    image = Image.objects.get(pk=int(info.cover_source.rstrip('/').rsplit('/', 1)[-1]))
+                except:
+                    pass
+                else:
+                    if info.cover_source == image.get_full_url():
+                        update['dc_cover_image'] = image
+        Book.objects.filter(pk=self.pk).update(**update)
+
     def touch(self):
+        # this should only really be done when text or publishable status changes
+        book_content_updated.delay(self)
+
         update = {
             "_new_publishable": self.is_new_publishable(),
             "_published": self.is_published(),
             "_single": self.is_single(),
+            "_on_track": self.get_on_track(),
             "_short_html": None,
         }
         Book.objects.filter(pk=self.pk).update(**update)
@@ -274,17 +405,26 @@ class Book(models.Model):
             changes = self.get_current_changes(publishable)
         return compile_text(change.materialize() for change in changes)
 
-    def publish(self, user):
+    def wldocument(self, publishable=True, changes=None, 
+            parse_dublincore=True, strict=False):
+        from catalogue.ebook_utils import RedakcjaDocProvider
+        from librarian.parser import WLDocument
+
+        return WLDocument.from_string(
+                self.materialize(publishable=publishable, changes=changes),
+                provider=RedakcjaDocProvider(publishable=publishable),
+                parse_dublincore=parse_dublincore,
+                strict=strict)
+
+    def publish(self, user, fake=False):
         """
             Publishes a book on behalf of a (local) user.
         """
-        raise NotImplementedError("Publishing not possible yet.")
-
-        from apiclient import api_call
-
+        self.assert_publishable()
         changes = self.get_current_changes(publishable=True)
         book_xml = self.materialize(changes=changes)
-        #api_call(user, "books", {"book_xml": book_xml})
+        if not fake:
+            apiclient.api_call(user, "books/", {"book_xml": book_xml})
         # record the publish
         br = BookPublishRecord.objects.create(book=self, user=user)
         for c in changes: