There's still a LOT of work going on and continuing questions on asn 7...

1) I managed to tell a couple of people they had 100% even though their
root.count did not agree with mine.  Since that is somewhat of a special
case anyway don't worry about it and I won't consider it in any of the scores.

2) Also, don't worry if your nodes_used is one more than I showed.  I didn't
mention that I had only counted the allocated nodes and not the initial
root node that was already there.  Either way is OK.

3) internal counts
Many of you are only increasing the counts for the END nodes.  I guess that
will be OK although my intention had been that the counts everywhere would
indicate the number of times that each node had been passed through during
the insertion process.  That way we could get the word count statistics from
the leaves of the tree during treewalk() but also have other stats for
internal tree nodes that would correspond to the frequency of word prefixes.

Bottom line -- just get the main approach as close to right as you can by 6PM