Merge p3yk branch with the trunk up to revision 45595. This breaks a fair
number of tests, all because of the codecs/_multibytecodecs issue described
here (it's not a Py3K issue, just something Py3K discovers):
http://mail.python.org/pipermail/python-dev/2006-April/064051.html

Hye-Shik Chang promised to look for a fix, so no need to fix it here. The
tests that are expected to break are:

test_codecencodings_cn
test_codecencodings_hk
test_codecencodings_jp
test_codecencodings_kr
test_codecencodings_tw
test_codecs
test_multibytecodec

This merge fixes an actual test failure (test_weakref) in this branch,
though, so I believe merging is the right thing to do anyway.
diff --git a/Lib/test/test_random.py b/Lib/test/test_random.py
index 9c2e0d0..bba4c7c 100644
--- a/Lib/test/test_random.py
+++ b/Lib/test/test_random.py
@@ -93,10 +93,29 @@
         self.gen.sample(set(range(20)), 2)
         self.gen.sample(range(20), 2)
         self.gen.sample(xrange(20), 2)
-        self.gen.sample(dict.fromkeys('abcdefghijklmnopqrst'), 2)
         self.gen.sample(str('abcdefghijklmnopqrst'), 2)
         self.gen.sample(tuple('abcdefghijklmnopqrst'), 2)
 
+    def test_sample_on_dicts(self):
+        self.gen.sample(dict.fromkeys('abcdefghijklmnopqrst'), 2)
+
+        # SF bug #1460340 -- random.sample can raise KeyError
+        a = dict.fromkeys(range(10)+range(10,100,2)+range(100,110))
+        self.gen.sample(a, 3)
+
+        # A followup to bug #1460340:  sampling from a dict could return
+        # a subset of its keys or of its values, depending on the size of
+        # the subset requested.
+        N = 30
+        d = dict((i, complex(i, i)) for i in xrange(N))
+        for k in xrange(N+1):
+            samp = self.gen.sample(d, k)
+            # Verify that we got ints back (keys); the values are complex.
+            for x in samp:
+                self.assert_(type(x) is int)
+        samp.sort()
+        self.assertEqual(samp, range(N))
+
     def test_gauss(self):
         # Ensure that the seed() method initializes all the hidden state.  In
         # particular, through 2.2.1 it failed to reset a piece of state used