brisbane:CHKMap.iteritems() tweaks

[John Arbash Meinel]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Ian Clatworthy wrote:
> John Arbash Meinel wrote:
>> I think I've gotten the fixes I wanted for CHKMap.iteritems(). It is
>> available at:
>>
>> http://bzr.arbash-meinel.com/branches/bzr/brisbane/iter_changes_fixes
> 
> Looks good on the whole.
> 
>>                  for length, length_filter in length_filters:
>> -                    if prefix[:length] in length_filter:
>> -                        if type(node) == tuple:
>> -                            keys[node] = prefix
>> -                        else:
>> -                            yield node
>> -                        break
> 
> So the previous code broke out of the inner loop on the first match while
> we now continue going so we can collect all matches. I suspect it will
> make very little difference in practice because I'm guessing 90%+ of non-None
> search keys only contain a single tuple anyway?

So for IE.children() yes. We have maybe 30 children, and that quickly
gets spread out across all of the 255-way nodes.

Other places in our code seem to only pass in a single key anyway.

For parent_id,basename access, from what I could see, we only really
look for one directory at a time. So again, 1 entry in the length_filter.

Note that before my change to _iter_nodes, we would have 32 keys at this
point.

> 
>> +    def test__iter_nodes_splits_key_filter(self):
>> +        node = InternalNode('')
>> +        child = LeafNode()
>> +        child.set_maximum_size(100)
>> +        child.map(None, ("foo",), "bar")
>> +        node.add_node("f", child)
>> +        child = LeafNode()
>> +        child.set_maximum_size(100)
>> +        child.map(None, ("bar",), "baz")
>> +        node.add_node("b", child)
> 
> This and the previous test have the same setup code so you
> might want to pull it out into a common method.
> 
>> +        key_filter = (('foo',), ('bar',), ('cat',))
>> +        for child, node_key_filter in node._iter_nodes(None,
>> +                                                       key_filter=key_filter):
>> +            # each child could only match one key filter, so make sure it was
>> +            # properly filtered
>> +            self.assertEqual(1, len(node_key_filter))
> 
> I don't understand why ('cat',) gives a node_key_length of 1.
> Please explain that in some comments.
> 
> Ian C.

('cat',) doesn't give *any* nodes. The point of adding it is to ensure
that it doesn't accidentally end up in one of the other children. (Test
that X hits, but also test that Y *doesn't* hit.)

I'm happy to add some more comments.

John
=:->

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (Cygwin)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iEYEARECAAYFAknI/AgACgkQJdeBCYSNAAO3yQCeOqd+GAFB3CGojJQNQf/YeGj9
pCMAnirK/v8tkFAsHUuOhYpVaigGChCf
=Rn7u
-----END PGP SIGNATURE-----