uuid62

Go implementation of short ID generator by using UUIDv1, V4, V6, V7 and base62 inspired by npm uuid62 package.

It always return strings whose length is 21 or 22. It is shorter than string represention of UUID (36). And the results includes only alphabet and numeric characters (no symbols).

package main

import (
	"fmt"
	"github.com/shibukawa/uuid62/v2"
)

func main() {
	// it just returns string
	idString, err := uuid62.V1()
	if err != nil {
		fmt.Println(idString)
	}

	// it also provides API decode/encode between UUID
	uuid, err := Decode(idString)
	if err != nil {
		fmt.Println(idString)
	}
	idStringAgain := Encode(uuid) // idStringAgain == idString
	fmt.Println(idStringAgain)
}

API

Generator Functions

• func uuid62.V1() (string, error)
• func uuid62.V4() (string, error)
• func uuid62.V6() (string, error)
• func uuid62.V7() (string, error)

They generates UUID string in base62 format.

Helper Functions

• func uuid62.Decode(uuidString string) (uuid.UUID, error)

Convert string to github.com/gofrs/uuid's UUID object.

• func uuid62.Encode(uuid.UUID uuidObj) (string, error)

Convert github.com/gofrs/uuid's UUID object to string.

• func uuid62.Timestamp(uuidString string) (time.Time, error)

Get timestamp from uuid string. It accepts only V1 and V6.

Change History

v1 -> v2

uuid62.V7Nano and uuid62.V7Micro and uuid62.V7Milli and removed and introduced uuid62.V7 because the latest uuid draft removes precision option from V7.

Change base62 encode character order. Now resulting strings have also same order with original UUID. So the resulting string is also k-sortable.

License

Apache 2