Rev 3768: Activate test for common node skipping. - 50 times performance improvement. in http://people.ubuntu.com/~robertc/baz2.0/repository

[Robert Collins]

At http://people.ubuntu.com/~robertc/baz2.0/repository

------------------------------------------------------------
revno: 3768
revision-id: robertc at robertcollins.net-20081113032056-djjm0lykam7okvit
parent: robertc at robertcollins.net-20081113022855-8fb2mhizoehchwf1
committer: Robert Collins <robertc at robertcollins.net>
branch nick: repository
timestamp: Thu 2008-11-13 14:20:56 +1100
message:
  Activate test for common node skipping. - 50 times performance improvement.
modified:
  bzrlib/chk_map.py              chk_map.py-20081001014447-ue6kkuhofvdecvxa-1
  bzrlib/tests/test_chk_map.py   test_chk_map.py-20081001014447-ue6kkuhofvdecvxa-2
=== modified file 'bzrlib/chk_map.py'
--- a/bzrlib/chk_map.py	2008-11-13 02:28:55 +0000
+++ b/bzrlib/chk_map.py	2008-11-13 03:20:56 +0000
@@ -190,7 +190,9 @@
                     return True
             return False
 
+        loop_counter = 0
         while self_pending or basis_pending:
+            loop_counter += 1
             if not self_pending:
                 # self is exhausted: output remainder of basis
                 for prefix, node, path in basis_pending:
@@ -266,8 +268,15 @@
                             yield (tuple(self_details[0].split('\x00')),
                                 basis_details[1], self_details[1])
                         continue
-                    # At least one side wasn't a string, we need to expand it
-                    # before we can continue
+                    # At least one side wasn't a string.
+                    if (self._node_key(self_pending[0][1]) ==
+                        self._node_key(basis_pending[0][1])):
+                        # Identical pointers, skip (and don't bother adding to
+                        # excluded, it won't turn up again.
+                        heapq.heappop(self_pending)
+                        heapq.heappop(basis_pending)
+                        continue
+                    # Now we need to expand this node before we can continue
                     if read_self:
                         prefix, node, path = heapq.heappop(self_pending)
                         if check_excluded(path):
@@ -278,6 +287,7 @@
                         if check_excluded(path):
                             continue
                         process_node(prefix, node, path, basis, basis_pending)
+        # print loop_counter
 
     def iteritems(self, key_filter=None):
         """Iterate over the entire CHKMap's contents."""

=== modified file 'bzrlib/tests/test_chk_map.py'
--- a/bzrlib/tests/test_chk_map.py	2008-11-13 02:28:55 +0000
+++ b/bzrlib/tests/test_chk_map.py	2008-11-13 03:20:56 +0000
@@ -205,16 +205,16 @@
         basis = self._get_map(basis_dict, maximum_size=10)
         target = self._get_map(target_dict, maximum_size=10,
             chk_bytes=basis._store)
+        basis_get = basis._store.get_record_stream
+        def get_record_stream(keys, order, fulltext):
+            if ('sha1:1adf7c0d1b9140ab5f33bb64c6275fa78b1580b7',) in keys:
+                self.fail("'aaa' pointer was followed %r" % keys)
+            return basis_get(keys, order, fulltext)
+        basis._store.get_record_stream = get_record_stream
         result = sorted(list(target.iter_changes(basis)))
         for change in result:
             if change[0] == ('aaa',):
                 self.fail("Found unexpected change: %s" % change)
-        return
-        # check both target and basis did not load the aaa pointer
-        self.assertIsInstance(target._root_node._items['aa']._items['aaa'],
-            tuple)
-        self.assertIsInstance(basis._root_node._items['a']._items['aaa'],
-            tuple)
 
     def test_iter_changes_unchanged_keys_in_multi_key_leafs_ignored(self):
         # Within a leaf there are no hash's to exclude keys, make sure multi