queryset-refactor: Implemented a way to differentiate between filtering on a

single instance and filtering on multiple instances when spanning a
multi-valued relationship.


git-svn-id: http://code.djangoproject.com/svn/django/branches/queryset-refactor@7317 bcc190cf-cafb-0310-a4f2-bffc1f526a37

django/db/models/sql/query.py

@@ -809,7 +809,7 @@ class Query(object):
                     used, next, restricted)
 
     def add_filter(self, filter_expr, connector=AND, negate=False, trim=False,
-            merge_negated=False):
+            single_filter=False):
         """
         Add a single filter to the query. The 'filter_expr' is a pair:
         (filter_string, value). E.g. ('name__contains', 'fred')
@@ -818,9 +818,8 @@ class Query(object):
         automatically trim the final join group (used internally when
         constructing nested queries).
 
-        If 'merge_negated' is True, this negated filter will be merged with the
+        If 'single_filter' is True, we are processing a component of a
-        existing negated where node (if it exists). This is used when
+        multi-component filter (e.g. filter(Q1, Q2)).
-        constructing an exclude filter from combined subfilters.
         """
         arg, value = filter_expr
         parts = arg.split(LOOKUP_SEP)
@@ -849,7 +848,7 @@ class Query(object):
 
         try:
             field, target, opts, join_list, last = self.setup_joins(parts, opts,
-                    alias, (connector == AND), allow_many)
+                    alias, (connector == AND) and not single_filter, allow_many)
         except MultiJoin, e:
             self.split_exclude(filter_expr, LOOKUP_SEP.join(parts[:e.level]))
             return
@@ -917,7 +916,7 @@ class Query(object):
                 self.promote_alias(table)
 
         entry = (alias, col, field, lookup_type, value)
-        if merge_negated:
+        if negate and single_filter:
             # This case is when we're doing the Q2 filter in exclude(Q1, Q2).
             # It's different from exclude(Q1).exclude(Q2).
             for node in self.where.children:
@@ -957,7 +956,7 @@ class Query(object):
         else:
             subtree = False
         connector = AND
-        merge = False
+        internal = False
         for child in q_object.children:
             if isinstance(child, Node):
                 self.where.start_subtree(connector)
@@ -965,8 +964,8 @@ class Query(object):
                 self.where.end_subtree()
             else:
                 self.add_filter(child, connector, q_object.negated,
-                        merge_negated=merge)
+                        single_filter=internal)
-                merge = q_object.negated
+                internal = True
             connector = q_object.connector
         if subtree:
             self.where.end_subtree()

docs/db-api.txt

@@ -1559,7 +1559,7 @@ equivalent::
     model's primary key in queries.
 
 Lookups that span relationships
-~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+-------------------------------
 
 Django offers a powerful and intuitive way to "follow" relationships in
 lookups, taking care of the SQL ``JOIN``\s for you automatically, behind the
@@ -1582,8 +1582,66 @@ whose ``headline`` contains ``'Lennon'``::
 
     Blog.objects.filter(entry__headline__contains='Lennon')
 
+Spanning multi-valued relationships
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+**New in Django development version**
+
+.. note::
+    This is an experimental API and subject to change prior to
+    queryset-refactor being merged into trunk.
+
+When you are filtering an object based on a ``ManyToManyField`` or a reverse
+``ForeignKeyField``, there are two different sorts of filter you may be
+interested in. Consider the ``Blog``/``Entry`` relationship (``Blog`` to
+``Entry`` is a one-to-many relation). We might be interested in finding blogs
+that have an entry which has both *"Lennon"* in the headline and was published
+today. Or we might want to find blogs that have an entry with *"Lennon"* in
+the headline as well as an entry that was published today. Since there are
+multiple entries associated with a single ``Blog``, both of these queries are
+possible and make sense in some situations.
+
+The same type of situation arises with a ``ManyToManyField``. For example, if
+an ``Entry`` has a ``ManyToManyField`` called ``tags``, we might want to find
+entries linked to tags called *"music"* and *"bands"* or we might want an
+entry that contains a tag with a name of *"music"* and a status of *"public"*.
+
+To handle both of these situations, Django has a consistent way of processing
+``filter()`` and ``exclude()`` calls. Everything inside a single ``filter()``
+call is applied simultaneously to filter out items matching all those
+requirements. Successive ``filter()`` calls further restrict the set of
+objects, but for multi-valued relations, they apply to any object linked to
+the primary model, not necessarily those objects that were selected by an
+earlier ``filter()`` call.
+
+That may sound a bit confusing, so hopefully an example will clarify. To
+select all blogs that contains entries with *"Lennon"* in the headline and
+were published today, we would write::
+
+    Blog.objects.filter(entry__headline__contains='Lennon',
+            entry__pub_date=datetime.date.today())
+
+To select all blogs that contain an entry with *"Lennon"* in the headline
+**as well as** an entry that was published today, we would write::
+
+    Blog.objects.filter(entry__headline__contains='Lennon').filter(
+            entry__pub_date=datetime.date.today())
+
+In this second example, the first filter restricted the queryset to all those
+blogs linked to that particular type of entry. The second filter restricted
+the set of blogs *further* to those that are also linked to the second type of
+entry. The entries select by the second filter may or may not be the same as
+the entries in the first filter. We are filtering the ``Blog`` items with each
+filter statement, not the ``Entry`` items.
+
+All of this behaviour also applies to ``exclude()``: all the conditions in a
+single ``exclude()`` statement apply to a single instance (if those conditions
+are talking about the same multi-valued relation). Conditions in subsequent
+``filter()`` or ``exclude()`` calls that refer to the same relation may end up
+filtering on different linked objects.
+
 Escaping percent signs and underscores in LIKE statements
-~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+---------------------------------------------------------
 
 The field lookups that equate to ``LIKE`` SQL statements (``iexact``,
 ``contains``, ``icontains``, ``startswith``, ``istartswith``, ``endswith``

tests/regressiontests/queries/models.py

@@ -210,11 +210,24 @@ True
 
 >>> Item.objects.filter(Q(tags=t1)).order_by('name')
 [<Item: one>, <Item: two>]
->>> Item.objects.filter(Q(tags=t1) & Q(tags=t2))
-[<Item: one>]
 >>> Item.objects.filter(Q(tags=t1)).filter(Q(tags=t2))
 [<Item: one>]
 
+Each filter call is processed "at once" against a single table, so this is
+different from the previous example as it tries to find tags that are two
+things at once (rather than two tags).
+>>> Item.objects.filter(Q(tags=t1) & Q(tags=t2))
+[]
+
+>>> qs = Author.objects.filter(ranking__rank=2, ranking__id=rank1.id)
+>>> list(qs)
+[<Author: a2>]
+>>> qs.query.count_active_tables()
+2
+>>> qs = Author.objects.filter(ranking__rank=2).filter(ranking__id=rank1.id)
+>>> qs.query.count_active_tables()
+3
+
 Bug #4464
 >>> Item.objects.filter(tags=t1).filter(tags=t2)
 [<Item: one>]