Update proof logic to use merged proofs

[rianhughes]

Currently we handle left and right proofs separately but it seems that we need to merge the proofs. Implement the logic to merge and de-merge the proofs

You should implement two functions.

func mergeProofPaths(leftPath,rightPath []ProofNode) (mergedPath []ProofNode,error)
leftPath and rightPath are lists of ProofNodes, which define paths from the root node to some leaf
mergedPath represents the merged list of leftPath and rightPath, with no duplicates.

func splitProofPath(mergedPath []ProofNode) (leftPath,rightPath []ProofNode,error)
This function should do the reverse of mergeProofPaths, ie take a mergedPath, and split it into leftPath and rightPath.

It might be useful to look at this, and an example trie here. Note that ProofNodes are either Binary or Edge.