computing the resulting inventory from the merge changeset

[duchier at ps.uni-sb.de]

Aaron Bentley <aaron.bentley at utoronto.ca> writes:

>> Now Aaron's doubts are making sense (although his test cases have not
>> proven sufficient to expose the logic flaw).  I will now fix my
>> algorithm accordingly.  Sorry for being such a dimwit.
>
> Really?  What happened in the case I gave?

As I recall, not a thing - it worked just fine in all cases.  Devising
a failing case is made trickier by the fact that target entries are
sorted top-to-bottom according to the target tree, except for deleted
entries which remain sorted bottom-to-top according to the source
tree.

Cheers,

--Denys