hot-fns

Lightweight, zero-dependency utility functions optimized for hot paths.

npm install hot-fns

Philosophy

This library contains only utilities that are genuinely faster and more memory efficient than lodash:

• No intermediate arrays - Uses for...in instead of Object.keys() for object iteration
• Indexed for loops - Direct arr[i] access instead of iterator abstractions
• Thin abstractions - Single iteratee helper vs lodash's layered internal wrappers

If a utility isn't measurably faster or more memory efficient, it's not here. Use lodash for everything else.

Requirements

• Node.js >= 20
• ESM or CommonJS

Quick Start

import { groupByHot, partitionHot, mapValuesHot, evolveHot } from 'hot-fns';

// Single-pass grouping (returns Map by default)
const byStatus = groupByHot(orders, 'status');

// Or get a plain object instead
const byStatusObj = groupByHot(orders, 'status', true);

// Single-pass partition (2x faster than dual filter)
const [active, inactive] = partitionHot(users, u => u.isActive);

// Transform values without intermediate Object.keys() array
const doubled = mapValuesHot(prices, v => v * 2);

// Ramda-style evolve for nested transformations
const updated = evolveHot(
  { name: s => s.toUpperCase(), data: { count: n => n + 1 } },
  { name: 'foo', data: { count: 5 }, other: true },
);
// => { name: 'FOO', data: { count: 6 }, other: true }

Imports

// Main entry - all utilities
import { groupByHot, partitionHot } from 'hot-fns';

// Subpath imports - array utilities only
import { partitionHot, intersectionHot } from 'hot-fns/array';

// Subpath imports - object utilities only
import { groupByHot, mapKeysHot } from 'hot-fns/object';

API

Array

| Function | Description | | ----------------------------------------- | --------------------------------------------------- | | intersectionHot(...arrays) | Common elements (smallest-array-first optimization) | | intersectionByHot(arr1, arr2, iteratee) | Intersection with custom key | | partitionHot(arr, predicate) | Split by predicate (single-pass) | | uniqueByKeyHot(arr, iteratee) | Dedupe by key (Map-based O(n)) |

Object

| Function | Description | | ----------------------------------- | ------------------------------------------ | | countByHot(arr, iteratee) | Count occurrences by key | | evolveHot(spec, obj) | Ramda-style recursive transformations | | groupByHot(arr, iteratee, asObj) | Group by key (Map default, Object if true) | | indexByHot(arr, iteratee) | Create lookup table | | mapKeysHot(obj, mapper) | Transform object keys (no Object.keys()) | | mapObjectHot(obj, keyFn, valueFn) | Transform both keys and values | | mapValuesHot(obj, mapper) | Transform object values (no Object.keys()) | | omitHot(obj, keys) | Exclude keys (Set-based O(1) lookup) | | omitByHot(obj, predicate) | Exclude by predicate | | pickHot(obj, keys) | Select keys | | pickByHot(obj, predicate) | Select by predicate |

Iteratee Shorthand

Most functions accept an iteratee that can be:

• A function: x => x.user.id
• A property path: 'user.id'

groupByHot(users, 'role'); // string shorthand
groupByHot(users, u => u.role); // function
uniqueByKeyHot(items, 'nested.id'); // deep path

Publishing

pnpm version patch|minor|major
pnpm build
npm publish

License

GPLv3