47degrees / memeid   0.8.0

!

memeid is a JVM library for generating RFC-compliant Universal Unique Identifiers (UUIDs).

A universally unique identifier (UUID) is a 128-bit number used to identify information in computer systems.

When generated according to the standard methods, UUIDs are for practical purposes unique. Their uniqueness does not depend on a central registration authority or coordination between the parties generating them, unlike most other numbering schemes. -- Wikipedia article on UUIDs

• Rationale
• Install
  • Java
    • Using maven
    • Using gradle
  • Scala
• Usage
  • Scala
    • UUID construction
      • Time-based (v1)
      • Random (v4)
      • Namespaced (v3, v5)
      • Semi-sequential, random (SQUUID)
    • Java interoperability
    • Literal syntax
    • Integrations
      • Doobie
      • Circe
      • Http4s
        • Path parameters
        • Query parameters
      • Tapir
      • FUUID
      • Cats & Cats-effect
        • Typeclasses
        • Constructors
      • Scalacheck
• Benchmarking
• References

The UUID type that ships with the JVM java.util.UUID has a number of problems, namely:

• A bug in the comparison function that will never be fixed
• Only provides UUID generation for random (V4) and non-namespaced pseudo-V3 UUIDs

This library aims to solve the aforementioned issues and provide an RFC-compliant UUID type with coherent comparison, a rich API to get its different fields and constructors for the UUID variants.

<dependency>
    <groupId>com.47deg</groupId>
    <artifactId>memeid</artifactId>
    <version>0.8.0</version>
</dependency>

compile group: 'com.47deg', name: 'memeid', version: '0.8.0'

Add this to your build.sbt file:

libraryDependencies += "com.47deg" %% "memeid4s" % "0.8.0"

The time-based (V1) variant of UUIDs is the fastest to generate. It uses a monotonic clock and node information to generate UUIDs.

import memeid4s.UUID

UUID.V1.next

The cryptographically random variant, equivalent to java.util.UUID/randomUUID.

UUID.V4.random

Namespaced UUIDs are generated from a UUID (namespace) and a hashed value (name), V3 uses MD5 and V5 uses SHA1 hash.

val namespace = UUID.V1.next

We can now create UUIDs with the namespace and an arbitrary value as the name. It automatically works with Strings and UUIDs:

UUID.V3(namespace, "my-secret-code")

If you want to hash a custom type, you must provide an implicit memeid4s.digest.Digestible instance.

import memeid4s.digest.Digestible

case class User(firstName: String, lastName: String)

implicit val digestibleUser: Digestible[User] =
  (u: User) => u.firstName.getBytes ++ u.lastName.getBytes

The implicit instance is used to convert your type into a byte array for hashing:

UUID.V3(namespace, User("Federico", "García Lorca"))

SQUUIDs are a non-standard variaton of V4 UUIDs that are semi-sequential. They incorporate a time-component in their 32 most significant bits to generate UUIDs that don't fragment DB indexes.

UUID.V4.squuid

memeid provides conversion method between UUID and java.util.UUID through:

val j = java.util.UUID.fromString("a5fa7934-501c-46eb-9ea7-16de3086e6d8")

val u = memeid.UUID.fromString("8b4d1529-5fd0-4a91-8f4f-ceee10d1c060")

UUID.fromUUID(j)
// res5: UUID = a5fa7934-501c-46eb-9ea7-16de3086e6d8

u.asJava
// res6: java.util.UUID = 8b4d1529-5fd0-4a91-8f4f-ceee10d1c060

memeid provides literal syntax with compile-time verification for UUIDs with the uuid interpolator. To use it, add this to your build.sbt:

libraryDependencies += "com.47deg" %% "memeid4s-literal" % "0.8.0"

We can now create UUIDs with literal syntax by importing memeid.literal._

import memeid4s.literal._

uuid"cb096727-6a82-4abd-bc79-fc92be8c5d88"
// res7: UUID = cb096727-6a82-4abd-bc79-fc92be8c5d88

Invalid UUID literals will fail at compile time:

uuid"not-a-uuid"
// error: invalid UUID: not-a-uuid
// uuid"not-a-uuid"
// ^^^^^^^^^^^^^^^^

memeid provides several modules which integrate with popular third-party libraries. If you see something missing don't hesitate to open an issue or send a patch.

The Doobie integration allows you to use the UUID type mapped to your database's UUID type.

libraryDependencies += "com.47deg" %% "memeid4s-doobie" % "0.8.0"

To have the UUID mappings available in scope you can import memeid.doobie.implicits.

import memeid4s.doobie.implicits._

def select(uuid: UUID): Query0[UUID] =
  sql"""SELECT id from test where id = $uuid""".query[UUID]

def insert(uuid: UUID): Update0 =
  sql"""insert into test (id) values ($uuid)""".update

val example = uuid"58d61328-1b08-1171-1ee7-1283ed639e77"

{
  for {
    _ <- insert(example).run.transact(transactor)
    u <- select(example).unique.transact(transactor)
  } yield u
}.unsafeRunSync()
// res10: UUID = 58d61328-1b08-1171-1ee7-1283ed639e77

libraryDependencies += "com.47deg" %% "memeid4s-circe" % "0.8.0"

You can import memeid.circe.implicits to have the Encoder and Decoder instances for UUID in scope.

import io.circe.Json

import io.circe.Encoder
import io.circe.Decoder

import memeid4s.circe.implicits._

val uuid = uuid"58d61328-1b08-1171-1ee7-1283ed639e77"

val json = Json.fromString(uuid.toString)

Encoder[UUID].apply(uuid)
// res11: Json = JString(value = "58d61328-1b08-1171-1ee7-1283ed639e77")
Decoder[UUID].decodeJson(json)
// res12: Decoder.Result[UUID] = Right(
//   value = 58d61328-1b08-1171-1ee7-1283ed639e77
// )

libraryDependencies += "com.47deg" %% "memeid4s-http4s" % "0.8.0"

Using UUID companion object we can extract UUIDs from path parameters in URLs:

import cats.effect._

import org.http4s._
import org.http4s.dsl.io._

HttpRoutes.of[IO] { case GET -> Root / "user" / UUID(uuid) =>
  Ok(s"Hello, $uuid!")
}

The http4s integrations provides implicit instances for QueryParamDecoder[UUID] and QueryParamEncoder[UUID], which you can use to derive matchers for query parameters or send UUID in request query parameters.

import cats.effect._

import org.http4s._
import org.http4s.dsl.io._

import memeid4s.http4s.implicits._

object UUIDParamDecoder extends QueryParamDecoderMatcher[UUID]("uuid")

HttpRoutes.of[IO] { case GET -> Root / "user" :? UUIDParamDecoder(uuid) =>
  Ok(s"Hello, $uuid!")
}

libraryDependencies += "com.47deg" %% "memeid4s-tapir" % "0.8.0"

The Tapir integration provides implicit instances for Codec[UUID] and Schema[UUID], which allow using UUID as type for query/path params or headers in endpoints. As well as enriching documentation when a UUID field is used.

import memeid4s.tapir.implicits._

import sttp.tapir._

endpoint.get.in("hello" / path[UUID])

libraryDependencies += "com.47deg" %% "memeid4s-fuuid" % "0.8.0"

The FUUID integration provides both semi (via extension methods) and auto conversions between memeid's UUID type and FUUID.

import memeid4s.UUID

import io.chrisdavenport.fuuid.FUUID
import memeid4s.fuuid.syntax._

val fuuid: FUUID = UUID.V4.random.toFUUID

val uuid: UUID = fuuid.toUUID

import memeid4s.UUID

import io.chrisdavenport.fuuid.FUUID
import memeid4s.fuuid.auto._

def usingFUUID(fuuid: FUUID) = fuuid

def usingUUID(uuid: UUID) = uuid

val uuid: UUID = UUID.V4.random

val fuuid: FUUID = FUUID.fromUUID(java.util.UUID.randomUUID)

usingFUUID(uuid)
usingUUID(fuuid)

libraryDependencies += "com.47deg" %% "memeid4s-cats" % "0.8.0"

The cats integration provides typeclass implementation for UUID, as well as effectful constructors for UUIDs for integration with programs that use cats-effect.

import cats._

import memeid4s.cats.implicits._
import cats.effect.IO

Order[UUID]
Hash[UUID]
Eq[UUID]
Show[UUID]

UUID.random[IO]

val namespace = UUID.V4.random

UUID.v3[IO, String](namespace, "my-secret-code")
UUID.v5[IO, String](namespace, "my-secret-code")

libraryDependencies += "com.47deg" %% "memeid4s-scalacheck" % "0.8.0"

The scalacheck integration provides Arbitrary instances for the UUID, as well as for the different version classes.

import org.scalacheck.Arbitrary.arbitrary

import memeid4s.scalacheck.arbitrary.instances._

arbitrary[UUID]
arbitrary[UUID.V1]
arbitrary[UUID.V2]
arbitrary[UUID.V3]
arbitrary[UUID.V4]
arbitrary[UUID.V5]

sbt-jmh is used for executing the benchmarking tests.

There are 2 kind of benchmarking:

• runAvgtime: Measures the average time it takes for the benchmark method to execute (a single execution). Generates the master.avgtime.csv file in the bench folder.
• runThroughput: Measures the number of operations per second, meaning the number of times per second your benchmark method could be executed. Generates the master.throughput.csv file in the bench folder.

• RFC 4122 - A Universally Unique Identifier
• JVM UUID type
• clj-uuid Clojure implementation