search: don't index page id

As elasticsearch expects a long while we have an hash as string.
While at it remove references to _parent and _id as they are gone.

Author: xrmx

readthedocs/search/indexes.py

@@ -127,11 +127,8 @@ def bulk_index(self, data, index=None, chunk_size=500, parent=None,
             doc = {
                 '_index': index,
                 '_type': self._type,
-                '_id': source['id'],
                 '_source': source,
             }
-            if parent:
-                doc['_parent'] = parent
             if routing:
                 doc['_routing'] = routing
             docs.append(doc)
@@ -294,7 +291,7 @@ def get_mapping(self):
     def extract_document(self, data):
         doc = {}
 
-        attrs = ('id', 'project', 'title', 'headers', 'version', 'path',
+        attrs = ('project', 'title', 'headers', 'version', 'path',
                  'content', 'taxonomy', 'commit')
         for attr in attrs:
             doc[attr] = data.get(attr, '')