This repository was forked from Pigi, which was published under the MIT license.

TODO Finish README, add function definitions & use guide

Sparse Merkle Trees

Original paper describing how to implemenet a sparse merkle tree efficiently.

A sparse merkle tree is a merkle tree of intractable size, i.e. a tree which is so large that it would be impossible to actually store all of its leaf nodes.

The essential idea which makes a sparse merkle tree possible is the pre-calculation of a default node for each level in the tree: if the default value of a leaf node is D1 = 0, then the default value for the second level is D2 = h(h(0) ++ h(0)), the default value for the third level is D3 = h(D1 ++ D1), etc. This allows you to efficiently calculate the root hash without actually having to hash every node in the tree.

Sparse merkle trees make it possible to have a merkle tree with a leaf node for every possible output value a hash algorithm can produce. It also makes it very simple to produce append-only merkle trees where a node for every possible index in some integer range is initialized with a default value.

TypeScript

This repo has a typescript implementation of a sparse merkle tree which can be found in src/merkle-tree.ts.

SparseMerkleTree uses memdown for its database.

Modifications

The modifications to the code from Pigi were primarily to minimize dependencies on non-essential libraries, since the Pigi code was part of a large monorepo.

TODO

• Add checkpoint functionality to the tree
  • Can use some code from merkle-patricia-tree.
• Add bitfield for missing default hashes in proof function.

Contracts

This repo has a Solidity implementation of a sparse merkle tree, including verification and storage/production.

Modifications

Addition of verifyAndUpate function This is a pure function which takes a standard merkle proof (rootHash, leafValue, leafIndex, siblings) and an additional parameter newValue. It calculates the merkle root from the provided leaf, index and siblings, and calculates a second root newRootHash through the same process but using newValue instead of leafValue. It returns a tuple of (valid, newRootHash) where valid is a boolean stating whether the calculated root hash (using leafValue) matches the rootHash parameter that was provided for comparison.

TODO

• Add bitfield for missing default hashes in proof verification function.