ulid

Implementation of ULID - Universally Unique Lexicographically Sortable Identifier

https://github.com/ad-si/ulid

LTS Haskell 23.14:0.3.3.0
Stackage Nightly 2025-03-11:0.3.3.0
Latest on Hackage:0.3.3.0

See all snapshots ulid appears in

BSD-3-Clause licensed by Steve Kollmansberger, Adrian Sieber
Maintained by [email protected]
This version can be pinned in stack with:ulid-0.3.3.0@sha256:5fb9adeb269e84e0c2882a0ed7e17b7abbe91213c46a03a15f69227f1fb30f3a,2807

Module documentation for 0.3.3.0

Used by 1 package in nightly-2025-02-10(full list with versions):

ULID Implementation in Haskell

Lexicographically sortable, 128-bit identifier with 48-bit timestamp and 80 random bits. Canonically encoded as a 26 character string, as opposed to the 36 character UUID.

Original implementation and spec: github.com/alizain/ulid

 01an4z07by   79ka1307sr9x4mv3

|----------| |----------------|
 Timestamp       Randomness
  48 bits         80 bits

Universally Unique Lexicographically Sortable Identifier

UUID can be suboptimal for many uses-cases because:

  • It isn’t the most character efficient way of encoding 128 bits of randomness
  • UUID v1/v2 is impractical in many environments, as it requires access to a unique, stable MAC address
  • UUID v3/v5 requires a unique seed and produces randomly distributed IDs, which can cause fragmentation in many data structures
  • UUID v4 provides no other information than randomness, which can cause fragmentation in many data structures

Instead, herein is proposed ULID:

  • 128-bit compatibility with UUID
  • 1.21e+24 unique ULIDs per millisecond
  • Lexicographically sortable
  • Canonically encoded as a 26 character string, as opposed to the 36 character UUID
  • Uses Douglas Crockford’s base 32 for better efficiency and readability (5 bits per character)
  • Case insensitive
  • No special characters (URL safe)

Known Issues

  • No monotonicity guarantees (official spec)
  • Lexicographically sorted based on the random component if timestamps are the same. This causes the sort order to be non-deterministic for ULIDs with the same timestamp, but is necessary to avoid incorrect Map and Set behavior.

Usage

A simple usage example:

module Main where

import Data.ULID

main :: IO ()
main = do
  -- Derive a ULID using the current time and default random number generator
  ulid1 <- getULID
  print ulid1

  -- Derive a ULID using a specified time and default random number generator
  ulid2 <- getULIDTime 1469918176.385 -- POSIX Time, millisecond precision
  print ulid2

As per the spec, it is also possible to use a cryptographically-secure random number generator to contribute the randomness. However, the programmer must manage the generator on their own.

Example:

module Main where

import Data.ULID

import qualified Crypto.Random       as CR
import qualified Data.ULID.Random    as UR
import qualified Data.ULID.TimeStamp as TS

main :: IO ()
main = do
  g <- (CR.newGenIO :: IO CR.SystemRandom)

  -- Generate timestamp from current time
  t <- TS.getULIDTimeStamp

  let ulid3 = case UR.mkCryptoULIDRandom g of
        Left err        -> error $ show err
        -- Use g2, …, to continue generating secure ULIDs
        Right (rnd, g2) -> ULID t rnd

  print ulid3

Test Suite

stack test

Performance

stack bench
Running 1 benchmarks...
Benchmark ulid-bench: RUNNING...
217,868 op/s generate
Benchmark ulid-bench: FINISH