I have a Scala trait for persistence and transaction management (which I will blog about in more detail later). The trait looks like:<\/p>\n
trait DomainManager {\r\n def get[E](id: Long)(implicit m: ClassManifest[E]): Option[E]\r\n\r\n def find[E](namedQuery: String, params: Map[String, Any] = null): Option[E]\r\n\r\n \/\/...more methods\r\n\r\n def withTransaction[R](f: (TransactionStatus) => R,\r\n readOnly: Boolean = false,\r\n propagationBehavior: PropagationBehavior = PropagationRequired): R\r\n}<\/pre>\nLet’s take a look at withTransaction()<\/code> specifically. It is called like:<\/p>\nval result = domainManager.withTransaction { txStatus =>\r\n \/\/ Access database here\r\n}<\/pre>\nIf your application is written in Scala or Java it is sometimes handy to have certain pieces of it written in Groovy, to be able to easily change a class and reload it without restarting the application. Your Groovy code will be able to call any of your Java classes, but what about your Scala classes? For example, what if we want to call withTransaction()<\/code> on DomainManager<\/code>? How do we deal with the f<\/code> parameter? And what about the default parameter values?<\/p>\nGroovy and Scala both have the concept of functions as first class objects. In Groovy they are called closures and they are implemented by the class groovy.lang.Closure<\/code><\/a>. In Scala they are called functions and they are implemented by the traits scala.Function0, scala.Function1<\/a>, ... scala.FunctionN<\/code>, where N is the number of parameters to the function.<\/p>\n