Unexplicable range error

I’m trying to implement this in a project but lately one of the passing tests which was using that library keeps on failing due to:

RangeError: Maximum call stack size exceeded                                                                                                                                                                                                      at MerkleTree.createHashes (node_modules/merkletreejs/index.js:41:15)                                                   at MerkleTree.createHashes (node_modules/merkletreejs/index.js:87:10)                                                   at MerkleTree.createHashes (node_modules/merkletreejs/index.js:87:10)                                                   at MerkleTree.createHashes (node_modules/merkletreejs/index.js:87:10)                                                   at MerkleTree.createHashes (node_modules/merkletreejs/index.js:87:10)                                                   at MerkleTree.createHashes (node_modules/merkletreejs/index.js:87:10)                                                   at MerkleTree.createHashes (node_modules/merkletreejs/index.js:87:10)                                                   at MerkleTree.createHashes (node_modules/merkletreejs/index.js:87:10)                                                   at MerkleTree.createHashes (node_modules/merkletreejs/index.js:87:10)                                                   at MerkleTree.createHashes (node_modules/merkletreejs/index.js:87:10)

Is it meant to act like that with a code like (using Block from here):

const block = new Block();
const SHA256 = require('crypto-js/sha256');
const BSHA3 = bufferify(require('crypto-js/sha3')); //see #5

const leaves = block.transactions.map(BSHA3);
console.log('leaves=', leaves); //[] as expected
block.merkleTree = new MerkleTree(leaves, SHA256); //Runs into the RangeError shown above

Or am I missing something?

EDIT: this doesn’t happen when the leaves point to a non-empty array but give the error whenever it uses an empty one (even before it got filled later on).

Would it be preferrable to only create and initialise the tree once the data passed to the leaves would already be there (say a bunch of transactions in a block) or was that edge case simply not taken into account?

[miguelmota]

@Berkmann18 ah sorry, you’re right. The leaves weren’t being converted into buffers. Pushed an update to convert the leaves into buffers so you don’t need to ‘bufferify’ the hash algo for the leaves

[miguelmota]

will take a look…