1.People usually split repository for Terraform and Scala, this allows you to keep both in the same respository. I can create entire Infra and code needed from single repo running a single SBT/Gradle doAll task
2.It allows you to have central configs ie I can use HOCON style single config to create Infra and use in my spark jobs. Hence S3 bucket/EMR cluster name in Infra is always in sync. I can also pass in arguments based on environment rather than creating multiple TFvars
3.Scala as DSL layer I can mix and match Infra ie when running SBT I can decide if I want to create and run my app on AWS vs GCP vs Azure, can create or not create components directly, create modules in loop
Clone the project and run the following:
sbt publish local
Include in your project as "com.syedatifakhtar" %% "scalaterraform" %"0.1-SNAPSHOT"