Ticket #6422: ticket6422.diff

django/db/models/query.py

@@ -591 +591 @@
         obj.query.add_ordering(*field_names)
         return obj
 
-    def distinct(self, true_or_false=True):
+    def distinct(self, true_or_false=True, on_fields=None):
         """
         Returns a new QuerySet instance that will select only distinct results.
         """
         obj = self._clone()
-        obj.query.distinct = true_or_false
+        if on_fields:
+            if isinstance(on_fields, basestring):
+                on_fields = (on_fields,)
+            # Quote the field names
+            quote = connection.ops.quote_name
+            table_name = quote(self.model._meta.db_table)
+            obj.query.distinct = [table_name + "." + quote(field)
+                    for field in on_fields]
+        else:
+            obj.query.distinct = true_or_false
         return obj
 
     def extra(self, select=None, where=None, params=None, tables=None,

django/db/models/sql/query.py

@@ -408 +408 @@
 
         result = ['SELECT']
         if self.distinct:
-            result.append('DISTINCT')
+            if isinstance(self.distinct, list):
+                result.append('DISTINCT ON (%s)' % ', '.join(self.distinct))
+            else:
+                result.append('DISTINCT')
         result.append(', '.join(out_cols + self.ordering_aliases))
 
         result.append('FROM')
@@ -488 +491 @@
                 "Cannot combine queries on two different base models."
         assert self.can_filter(), \
                 "Cannot combine queries once a slice has been taken."
+        # TODO: if distinct lists do not match, this will fail
+        # whether this should be left as is or treated specially (e.g. by
+        # merging two lists into one) remains to be decided
         assert self.distinct == rhs.distinct, \
             "Cannot combine a unique query with a non-unique query."
 
@@ -912 +918 @@
             ordering = self.order_by or self.model._meta.ordering
         qn = self.quote_name_unless_alias
         qn2 = self.connection.ops.quote_name
+        # TODO: if self.distinct is a list, it needs to contain all the
+        # columns used in ordering. Automating this is rather tedious, so
+        # currently users should be simply warned that their queries will fail
+        # unless they provide the order_by paramters in distinct as well.
         distinct = self.distinct
         select_aliases = self._select_aliases
         result = []
@@ -2120 +2130 @@
                         "Cannot add count col with multiple cols in 'select': %r" % self.select
                 count = self.aggregates_module.Count(self.select[0])
         else:
+            # FIXME: the implications of when self.distinct is a list need to
+            # be thought through
             opts = self.model._meta
             if not self.select:
                 count = self.aggregates_module.Count((self.join((None, opts.db_table, None, None)), opts.pk.column),