SCRU128: Sortable, Clock and Random number-based Unique identifier

SCRU128 ID is yet another attempt to supersede UUID for the users who need decentralized, globally unique time-ordered identifiers. SCRU128 is inspired by ULID and KSUID and has the following features:

• 128-bit unsigned integer type
• Sortable by generation time (as integer and as text)
• 25-digit case-insensitive textual representation (Base36)
• 48-bit millisecond Unix timestamp that ensures useful life until year 10889
• Up to 281 trillion time-ordered but unpredictable unique IDs per millisecond
• 80-bit three-layer randomness for global uniqueness

// generate a new identifier object
let x = scru128::new();
println!("{}", x); // e.g., "036z951mhjikzik2gsl81gr7l"
println!("{}", x.to_u128()); // as a 128-bit unsigned integer

// generate a textual representation directly
println!("{}", scru128::new_string()); // e.g., "036z951mhzx67t63mq9xe6q0j"

See SCRU128 Specification for details.

Crate features

Default features:

• std configures Scru128Generator with the system clock. Without std, this crate provides basic SCRU128 primitives available under no_std environments.
• rand enables a blanket implementation for rand::RngCore to use rand and any other conforming random number generators with Scru128Generator.
• default_rng (implies std and rand) provides the default random number generator for Scru128Generator and enables the Scru128Generator::new() constructor.
• global_gen (implies default_rng) provides the process-wide default SCRU128 generator and enables the new() and new_string() functions.

Optional features:

• serde enables serialization/deserialization of Scru128Id via serde.

License

Licensed under the Apache License, Version 2.0.

See also

• docs.rs/scru128