This is a Java port of Segment's K-Sortable Globally Unique IDs. with 40 bit timestamp and epoch as 1 January 1970 UTC 00:00:00 and 29 bytes long
KSUID is for K-Sortable Unique IDentifier. It's a way to generate globally unique IDs similar to RFC 4122 UUIDs, but contain a time component so they can be "roughly" sorted by time of creation. The remainder of the KSUID is randomly generated bytes.
In summary:
- Roughly sortable by creation time;
- Can be stored as a string of 29 chars;
- Can be stored as an array of 21 bytes;
- String format is encoded to base-62 (0-9A-Za-z);
- String format is URL safe and has no hyphens.
For the story of how KSUIDs came about, see A brief history of the UUID.
To use this as a command-line program on Unix-like systems, run
wget https://repo1.maven.org/maven2/com/github/ksuid40/ksuid40/1.1.0/ksuid40-1.1.0-cli.jar
sudo mv ksuid40-1.1.0-cli.jar /usr/local/bin/ksuid40
sudo chmod +x /usr/local/bin/ksuid40
ksuid40
# prints 1HCpXwx2EK9oYluWbacgeCnFcLf
Then simply generate a ksuid40 string like this:
String ksuid40 = Ksuid40.newKsuid().toString();
System.out.println(ksuid40); // prints 1HCpXwx2EK9oYluWbacgeCnFcLf
For more complex use cases, create a 40
with a SecureRandom
and get a new Ksuid40
for use.
Note that Ksuid40Generator
is threadsafe and Ksuid40
is immutable (and therefore threadsafe).
// Construct a new KsuidGenerator object. Since it is threadsafe you only need one.
private static final Ksuid40Generator KSUID_GENERATOR = new Ksuid40Generator(new SecureRandom());
// Get a new Ksuid object.
final Ksuid ksuid40 = ksuid40Generator.newKsuid();
// The toString() method is the string representation of KSUID.
System.out.println("ksuid40:\n" + ksuid40 + "\n");
// The log string format shows some details on one line, suitable for logging.
System.out.println("ksuid40.toLogString():\n" + ksuid40.toLogString() + "\n");
// The inspect string format shows details.
System.out.println("ksuid40.toInspectString():\n" + ksuid40.toInspectString());
The output from the code block above is
ksuid40:
1HCpXwx2EK9oYluWbacgeCnFcLf
ksuid40.toLogString():
Ksuid[timestamp = 150215977, string = 1HCpXwx2EK9oYluWbacgeCnFcLf payload = [124, 76, 43, -110, 116, -6, \
-91, 45, 0, -125, -127, 109, 28, 24, 28, -17], ksuidBytes = [8, -12, 29, 41, 124, 76, 43, -110, 116, \
-6, -91, 45, 0, -125, -127, 109, 28, 24, 28, -17]]
ksuid40.toInspectString():
REPRESENTATION:
String: 1HCpXwx2EK9oYluWbacgeCnFcLf
Raw: 08F41D297C4C2B9274FAA52D0083816D1C181CEF
COMPONENTS:
Time: 2019-02-14 23:32:57 -0800 PST
Timestamp: 150215977
Payload: 7C4C2B9274FAA52D0083816D1C181CEF
A very rough performance profile for generating KSUIDs was run on a MacBook Pro with a 3.1 GHz Intel Core i7 and 16 GB 2133 MHz LPDDR3 RAM.
public static void main(final String[] args) {
final Ksuid40Generator generator = new Ksuid40Generator(new SecureRandom());
IntStream.range(0, 100).forEach(i -> generator.newKsuid()); // prime the random
IntStream.iterate(1000, operand -> operand * 10)
.limit(5)
.forEach(count -> {
final long start = System.nanoTime();
IntStream.range(0, count).forEach(i -> generator.newKsuid());
final long duration = TimeUnit.MILLISECONDS.convert(System.nanoTime() - start, TimeUnit.NANOSECONDS);
System.out.println(String.format("%,d in %,d ms. rate = %,d/ms", count, duration, count / duration));
});
}
The output from the code block above is
1,000 in 14 ms. rate = 71/ms
10,000 in 32 ms. rate = 312/ms
100,000 in 95 ms. rate = 1,052/ms
1,000,000 in 881 ms. rate = 1,135/ms
10,000,000 in 6,665 ms. rate = 1,500/ms
This library is Open Source software released under the MIT license.