The quick and dirty datastore for prototyping
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Failed to load latest commit information.


Cagette is a quick and dirty data store for prototyping. It allows to focus on designing your domain model, business rules, screens, user interactions, batch, reports, ... without bothering about how your data will eventually be persisted.

"Cagette saved my hackday" — Mr. Speaker

Setting up with sbt

Configure a new resolver:

resolvers += "Guillaume Bort" at ""

Add the library dependency:

libraryDependencies += "guillaume.bort" %% "cagette" % "0.2"

Using – as easy as 1,2,3

1. Create your domain class with an id field:

Let's try with a User class:

case class User(id: Long, email: String, groups: Seq[String])

2. Create a Cageot[DomainType,IdType]:

We usually name the Cageot with the same name as the domain class:

object User extends cagette.Cageot[User,Long]

3. Query the cageot using Scala functions

The Cageot type provides convenient higher order query functions you can use directly:


val allusers: Seq[User] = User.findAll()

findById(id: IdType)

val maybeUser: Option[User] = User.findById(88)

findOneBy(predicate: DomainType => Boolean)

val maybeUser: Option[User] = User.findOneBy( == "")

findBy(predicate: DomainType => Boolean)

val adminUsers: Seq[User] = User.findBy(_.groups.contains("admin"))

Setting up initial data set

While defining your Cageot you can define the initial data set it will contain:

object User extends cagette.Cageot[User,Long] {
	override def initialData = Seq(
		User(1, "", groups = Seq("admin", "user")),
		User(2, "", groups = Seq("user")),
		User(3, "", groups = Nil)


Updating data

save will insert a new instance

If you want to store a new instance into your Cageot, just use the save operation as: 
  User(88, "", groups = Seq("user")) 

but save will also update an existing instance

If the store already contains an instance with the same id, this instance will just be udpated:

User.findOneBy( == "").map { user => = (user.groups :+ "admin").distinct ))

then you can delete them later

You can either delete by using the id of the instances you want to remove as:


or batch delete using a predicate as:


How to avoid the evil id field

If you don't want to artificially specify an id field in your case class, you can still provide your own Identifier instance when creating your Cageot:

case class User(email: String, groups: Seq[String])
object User extends Cageot[User,String]()(Identifier( // Specify the identifier as `email`

Using with Play framework

You can directly use this for prototyping with Just add the resolver and library dependency to your Build.scala file, and start defining your model. The initialData set will be applied automatically at startup and each time your code change.


Check the source code and find secret features like pagination and autoincrement for domain class identifiers.