Making a Client

To start with S3, we want to make an instance of S3 to represent all the interactions we can make with S3.

We will want cats-effect for this example will use cats.effect.IO.

We'll also be using the Blaze client from http4s, so we'll need the builder for that.

Following that, we need all of the common classes in Fawn, and all of the S3 classes.

import cats.effect._
import org.http4s.client.blaze.BlazeClientBuilder
import com.meltwater.fawn.common._
import com.meltwater.fawn.s3._

import scala.concurrent.ExecutionContext

In order to create our client, we will need some extra settings, including AWS credentials. These would normally be brought in via a config file or command line parser, but for this example we will instantiate them here.

val credentials = AWSCredentials("KEYID", "SECRET")
// credentials: AWSCredentials = AWSCredentials(
//   accessKeyId = "KEYID",
//   secretAccessKey = "SECRET"
// )
val region      = AWSRegion.`eu-west-1`
// region: AWSRegion = AWSRegion(value = "eu-west-1")
val accountId   = 123456L
// accountId: Long = 123456L

We can then make a Resource for our http4s client and then map it into an S3.

import org.typelevel.log4cats.SelfAwareStructuredLogger
import org.typelevel.log4cats.slf4j.Slf4jLogger

implicit def unsafeLogger[F[_]: Sync]: SelfAwareStructuredLogger[F] = Slf4jLogger.getLogger[F]

val s3Resource: Resource[IO, S3[IO]] =
  BlazeClientBuilder[IO](ExecutionContext.global).resource.map { client =>
    S3[IO](client, credentials, region)
  }

You would then use it or tie it in with your program's other Resources:

s3Resource.use { s3 =>
  s3.createBucket("hello-world-bucket")
}