refactor imports

[redakcja.git] / apps / catalogue / models / document.py

diff --git a/apps/catalogue/models/document.py b/apps/catalogue/models/document.py
index 38c2110..9863e65 100755 (executable)
--- a/apps/catalogue/models/document.py
+++ b/apps/catalogue/models/document.py
@@ -10,11 +10,50 @@ from django.conf import settings
 from django.core.exceptions import ObjectDoesNotExist, MultipleObjectsReturned
 from django.db import models
 from django.template.loader import render_to_string
+from django.utils.encoding import force_unicode
 from django.utils.translation import ugettext_lazy as _
 from dvcs.models import Ref
 from organizations.models import Organization
 from catalogue.constants import STAGES
-from .tag import Tag
+from .tag import Tag, Category
+
+
+def metadata_from_text(text):
+    from lxml import etree
+    metadata = {}
+    text = text.replace(u'\ufeff', '')
+    # This is bad. The editor shouldn't spew unknown HTML entities.
+    text = text.replace(u' ', u'\u00a0')
+
+    try:
+        t = etree.fromstring(text)
+    except:
+        return {'title': '<<Resource invalid>>'}
+    header = t.find('.//header')
+    if header is None:
+        header = etree.fromstring(text).find('.//{http://nowoczesnapolska.org.pl/sst#}header')
+    metadata['title'] = getattr(header, 'text', ' ') or ' '
+    # print 'meta', d['title']
+
+    m = t.find('metadata')
+    if m is None:
+        m = t.find('{http://nowoczesnapolska.org.pl/sst#}metadata')
+    if m is not None:
+        c = m.find('{http://purl.org/dc/elements/1.1/}relation.coverimage.url')
+        if c is not None:
+            metadata['cover_url'] = c.text
+        for category in Category.objects.all():
+            for elem in m.findall('{http://purl.org/dc/elements/1.1/}' + category.dc_tag):
+                if elem.text is not None:
+                    if category.multiple:
+                        if category.dc_tag not in metadata:
+                            metadata[category.dc_tag] = []
+                        metadata[category.dc_tag].append(elem.text)
+                    else:
+                        if category.dc_tag in metadata:
+                            metadata['multiple_values'] = category.dc_tag
+                        metadata[category.dc_tag] = elem.text
+    return metadata
 
 
 class Document(Ref):
@@ -22,10 +61,12 @@ class Document(Ref):
 
     owner_user = models.ForeignKey(settings.AUTH_USER_MODEL, null=True)
     owner_organization = models.ForeignKey(Organization, null=True)
-    stage = models.CharField(_('stage'), max_length=128, blank=True, default=STAGES[0])
+    stage = models.CharField(_('stage'), max_length=128, blank=True, default=STAGES[0][0])
     assigned_to = models.ForeignKey(settings.AUTH_USER_MODEL, null=True, related_name='assignments')
     deleted = models.BooleanField(default=False)
     tags = models.ManyToManyField(Tag, blank=True)
+    # we need to know if it were ever published (for notifications)
+    published = models.BooleanField(default=False)
 
     # Where to cache searchable stuff from metadata?
     # Probably in some kind of search index.
@@ -33,43 +74,17 @@ class Document(Ref):
     class Meta:
         verbose_name = _('document')
         verbose_name_plural = _('documents')
+        ordering = ['-revision__created_at']
 
     def short_html(self):
         return render_to_string('catalogue/book_list/book.html', {'book': self})
 
     def meta(self):
-        from lxml import etree
-        metadata = {}
-
-        data = self.materialize()
-        data = data.replace(u'\ufeff', '')
-        # This is bad. The editor shouldn't spew unknown HTML entities.
-        data = data.replace(u'&nbsp;', u'\u00a0')
-
-        try:
-            t = etree.fromstring(data)
-        except:
-            return {'title': '<<Resource invalid>>'}
-        header = t.find('.//header')
-        if header is None:
-            header = etree.fromstring(data).find('.//{http://nowoczesnapolska.org.pl/sst#}header')
-        metadata['title'] = getattr(header, 'text', ' ') or ' '
-        # print 'meta', d['title']
-
-        m = t.find('metadata')
-        if m is None:
-            m = t.find('{http://nowoczesnapolska.org.pl/sst#}metadata')
-        if m is not None:
-            c = m.find('{http://purl.org/dc/elements/1.1/}relation.coverimage.url')
-            if c is not None:
-                metadata['cover_url'] = c.text
-            c = m.find('{http://purl.org/dc/elements/1.1/}audience')
-            if c is not None:
-                metadata['audience'] = c.text
-
-        return metadata
+        return metadata_from_text(self.materialize())
 
     def can_edit(self, user):
+        if user.is_superuser:
+            return True
         if self.owner_user:
             return self.owner_user == user
         else:
@@ -84,6 +99,9 @@ class Document(Ref):
             self.assigned_to = None
         self.save()
 
+    def stage_name(self):
+        return force_unicode(dict(STAGES)[self.stage]) if self.stage else None
+
     def get_plan(self):
         try:
             plan = self.plan_set.get(stage=self.stage)
@@ -94,3 +112,15 @@ class Document(Ref):
     def is_overdue(self):
         plan = self.get_plan()
         return plan is not None and plan.deadline and plan.deadline < date.today()
+
+    def commit(self, *args, **kwargs):
+        super(Document, self).commit(*args, **kwargs)
+        m = self.meta()
+        for category in Category.objects.all():
+            values = m.get(category.dc_tag)
+            if not category.multiple:
+                values = [values]
+            if not values:
+                values = []
+            tags = category.tag_set.filter(dc_value__in=values)
+            category.set_tags_for(self, tags)