hyperdyperid

Fastest random ID and random string generation for Node.js

• str3_36 — generates random 3 character Base36 alphanumeric string.
• str5_36 — generates random 5 character Base36 alphanumeric string.
• str6_36 — generates random 6 character Base36 alphanumeric string.
• str10_36 — generates random 10 character alphanumeric string.
• str(len [, alphabet]) — generates any length random string using an alphabet. Use like str(10, '0123456789').
• hyperid — generates "hyperid" like ID.
• xorshift32 — iterates through pseudo random numbers using "xorshift32" algorithm. This used to be the algorithm behind Math.random() in V8, then it switched to "xorshift64".
• randomU32(min, max) — generate a random 32-bit integer.

Usage

Install

yarn add hyperdyperid

Generate base36 10-character string:

const {str10_36} = require('hyperdyperid/lib/str10_36');

str10_36();
// lxifpmidcr
// cn41vpmtxo
// dj2u5ghp3r
// 6yro5d0g9l

Performance

Ran on Mac Mini M1.

Random ID generation:

node -v
v16.14.2
node benchmarks/main.js 
hyperdyperid xorshift32() x 241,258,490 ops/sec ±1.97% (78 runs sampled)
hyperdyperid str3_36() x 203,361,942 ops/sec ±1.57% (82 runs sampled)
hyperdyperid str5_36() x 124,134,559 ops/sec ±1.74% (85 runs sampled)
hyperdyperid str6_36() x 125,011,053 ops/sec ±2.48% (70 runs sampled)
hyperdyperid str10_36() x 19,445,199 ops/sec ±1.92% (84 runs sampled)
hyperdyperid str(5) x 5,940,801 ops/sec ±1.70% (82 runs sampled)
hyperdyperid str(10) x 2,798,758 ops/sec ±1.45% (65 runs sampled)
hyperdyperid hyperid() x 26,317,209 ops/sec ±2.61% (76 runs sampled)
hyperid x 17,018,022 ops/sec ±1.66% (84 runs sampled)
Math.random() x 72,294,752 ops/sec ±1.65% (83 runs sampled)
Math.random().toString() x 4,503,932 ops/sec ±2.68% (80 runs sampled)
nanoid x 470,470 ops/sec ±3.17% (77 runs sampled)
shortid x 41,870 ops/sec ±6.21% (67 runs sampled)
Fastest is hyperdyperid xorshift32()

Random integer generation:

node -v
v16.14.2
node benchmarks/randomU32.js 
randomU32(min, max) x 248,233,310 ops/sec ±1.91% (73 runs sampled)
Math.random() * (max - min + 1) + min x 35,034,863 ops/sec ±1.75% (81 runs sampled)
Fastest is randomU32(min, max)

License

MIT © Vadim Dalecky.