I have a ConfigLoader which I'm pretty sure works fine. I suspect I am not using application.conf file correctly or perhaps my directory containing the prd.env and dev.env is not in the correct place.
What I expect to happen:
I when I enter sbt run the ConfigLoader reads the application.conf file, sees there are variables within that file. Then checks if prd.env or dev.env depending on which environment it is in and then finally loads the variables. I want this so I can have a different database for dev and prd.
- I have not idea how it would find the Meta directory which contains the prd.env and dev.env (see image).
- I have no idea how it would be able to tell whether if the environment is in dev or prd.
- The goal is to now look towards deploying this app, which is why I need these environment variables to work.
I really appreciate all the help but please try and detail your answers because I am really stuck and short answers often assume I know more than I do. Thanks :)
Tech stack incase relevant:
- HTTP4S,
- CATS,
- Doobie,
- PostgreSQL
application.conf file below:
appone {
environment = ${ENV}
server-config {
url = ${?BASE_URL}
host = "0.0.0.0"
port = 8080
}
db-config {
driver = "org.postgresql.Driver"
url = ${?DATABASE_URL}
user = ${?APPONE_POSTGRES_USER}
password = ${?PASSWORD}
connection-threads = 4
}
}
Meta/dev.env file (I also have a Meta/prd.env but have shown it here due to contents)
ENV=dev
BASE_URL=http://localhost:8080/
DATABASE_URL=jdbc:postgresql://localhost:5400/bookswapdb
APPONE_POSTGRES_USER=su
PASSWORD=password
LoadConfig file below:
package com.fullstackryan.appone.config
import cats.ApplicativeError
import cats.implicits._
import pureconfig.error.ConfigReaderException
import pureconfig.{ConfigReader, ConfigSource, Derivation}
import shapeless.the
trait LoadConfig[F[_], TConfig] {
def load: F[TConfig]
}
object LoadConfig {
def load[F[_], TConfig](implicit loadConfig: LoadConfig[F, TConfig]): F[TConfig] =
the[LoadConfig[F, TConfig]].load
def apply[F[_], TConfig](
implicit reader: Derivation[ConfigReader[TConfig]], ae: ApplicativeError[F, Throwable]
): LoadConfig[F, TConfig] =
new LoadConfig[F, TConfig] {
def load: F[TConfig] = ApplicativeError[F, Throwable].fromEither {
ConfigSource.default
.at("appone")
.load[TConfig]
.leftMap(ConfigReaderException(_))
}
}
}
error
pureconfig.error.ConfigReaderException: Cannot convert configuration to a scala.runtime.Nothing$. Failures are:
at 'appone.db-config':
- (application.conf @ jar:file:/Users/ryanmcavoy/fullStackRyan/appone/target/bg-jobs/sbt_3cc4b1f5/job-11/target/419ddc2c/5befcb57/appone_2.13-0.0.1-SNAPSHOT.jar!/application.conf: 10) Key not found: 'url'.
- (application.conf @ jar:file:/Users/ryanmcavoy/fullStackRyan/appone/target/bg-jobs/sbt_3cc4b1f5/job-11/target/419ddc2c/5befcb57/appone_2.13-0.0.1-SNAPSHOT.jar!/application.conf: 10) Key not found: 'username'.
- (application.conf @ jar:file:/Users/ryanmcavoy/fullStackRyan/appone/target/bg-jobs/sbt_3cc4b1f5/job-11/target/419ddc2c/5befcb57/appone_2.13-0.0.1-SNAPSHOT.jar!/application.conf: 10) Key not found: 'pool-size'.
at com.fullstackryan.appone.config.LoadConfig$$anon$1.$anonfun$load$1(LoadConfig.scala:25)
at cats.syntax.EitherOps$.leftMap$extension(either.scala:172)
at com.fullstackryan.appone.config.LoadConfig$$anon$1.load(LoadConfig.scala:25)
at com.fullstackryan.appone.server.ApponeServer$.$anonfun$stream$1(ApponeServer.scala:32)
at com.fullstackryan.appone.server.ApponeServer$.$anonfun$stream$1$adapted(ApponeServer.scala:31)
at fs2.Stream$.$anonfun$flatMap$1(Stream.scala:1188)
at fs2.internal.FreeC$.go$2(Algebra.scala:609)
at fs2.internal.FreeC$.$anonfun$flatMapOutput$1(Algebra.scala:616)
at fs2.internal.FreeC$$anon$1.cont(Algebra.scala:53)
at fs2.internal.FreeC$ViewL$$anon$9$$anon$10.cont(Algebra.scala:242)
at fs2.internal.FreeC$ViewL$.mk(Algebra.scala:231)
at fs2.internal.FreeC$ViewL$.apply(Algebra.scala:220)
at fs2.internal.FreeC.viewL(Algebra.scala:106)
at fs2.internal.FreeC$.go$1(Algebra.scala:414)
at fs2.internal.FreeC$.$anonfun$compile$8(Algebra.scala:464)
at fs2.internal.FreeC$.$anonfun$compile$1(Algebra.scala:430)
addSbtPlugin("io.github.davidgregory084" % "sbt-tpolecat" % "0.1.14")
addSbtPlugin("io.spray" % "sbt-revolver" % "0.9.1")
addSbtPlugin("com.eed3si9n" % "sbt-assembly" % "0.14.10")
addSbtPlugin("au.com.onegeek" % "sbt-dotenv" % "2.1.204")
// deploy heroku
addSbtPlugin("com.typesafe.sbt" % "sbt-native-packager" % "1.7.6")
addSbtPlugin("com.heroku" % "sbt-heroku" % "2.1.0")

When running via
sbt run, the environment is set by the shell in which you're running sbt. Defining adev.envfile by itself does nothing.The mechanism for setting the environment in which sbt is running will vary depending on your shell.
For example if bash is your shell (this is worth trying in other Bourne-compatible shells), prefixing the environment variables with
export, e.g.:Then you incorporate the environment variables you've exported from
dev.envinto your bash environment withThose environment variables will then be set and thus incorporated into your config for the duration of your shell session (i.e. they'll continue across multiple
sbt runs until you exit the shell). If you changedev.env, they won't be available until yousource meta/dev.envagain.