bloomf

This package implements a bloom filter for a general type of usage. It uses FNV and a simple trick to compute the k hashes required.

features

• Initializing Bloom Filter requires only the size of the filter and the number of hashing functions.
• Usage of Uint8Array TypedArray to ensure minimal memory footprint.
• Usage of bit operation to operate on our bit set implies better performance (to be test thouroughtly).
• FNV hash and simple linear hash used as hashing functions.
• Possibility to insert different element types (Number, String) by converting them into strings.

Usage

 
const BloomFilter = require('bloomf');
 
const filterSize = 10;
const kHashes = 3;
 
const bl = new BloomFilter(filterSize, kHashes);
 
bl.insert(3);
bl.insert("bloblo");
 
bl.test(3) // returns true
bl.test("bloblo") // returns true
 

Improvements

• Use better hashing functions than the linear hashing function
• offer multiple constructors
• Check how feasable it is to transform any object into bytes in JavaScript