Scala 3

When defining a field, the Scala compiler creates up to four accessors for it: a getter, a setter, and if the field is annotated with @BeanProperty, a bean getter and a bean setter.

For instance in the following class definition

class C(@myAnnot @BeanProperty var c: Int)

there are six entities which can carry the annotation @myAnnot: the constructor parameter, the generated field and the four accessors.

By default, annotations on (val-, var- or plain) constructor parameters end up on the parameter, not on any other entity. Annotations on fields by default only end up on the field.

The meta-annotations in package scala.annotation.meta are used to control where annotations on fields and class parameters are copied. This is done by annotating either the annotation type or the annotation class with one or several of the meta-annotations in this package.

Annotating the annotation type

The target meta-annotations can be put on the annotation type when instantiating the annotation. In the following example, the annotation @Id will be added only to the bean getter getX.

import javax.persistence.Id
class A {
 @(Id @beanGetter) @BeanProperty val x = 0
}

In order to annotate the field as well, the meta-annotation @field would need to be added.

The syntax can be improved using a type alias:

object ScalaJPA {
 type Id = javax.persistence.Id @beanGetter
}
import ScalaJPA.Id
class A {
 @Id @BeanProperty val x = 0
}

Annotating the annotation class

For annotations defined in Scala, a default target can be specified in the annotation class itself, for example

@getter
class myAnnotation extends Annotation

This only changes the default target for the annotation myAnnotation. When instantiating the annotation, the target can still be specified as described in the last section.

This package object contains primitives for concurrent and parallel programming.

Guide

A more detailed guide to Futures and Promises, including discussion and examples can be found at https://docs.scala-lang.org/overviews/core/futures.html.

Common Imports

When working with Futures, you will often find that importing the whole concurrent package is convenient:

import scala.concurrent._

When using things like Futures, it is often required to have an implicit ExecutionContext in scope. The general advice for these implicits are as follows.

If the code in question is a class or method definition, and no ExecutionContext is available, request one from the caller by adding an implicit parameter list:

def myMethod(myParam: MyType)(implicit ec: ExecutionContext) = …
//Or
class MyClass(myParam: MyType)(implicit ec: ExecutionContext) { … }

This allows the caller of the method, or creator of the instance of the class, to decide which ExecutionContext should be used.

For typical REPL usage and experimentation, importing the global ExecutionContext is often desired.

import scala.concurrent.ExcutionContext.Implicits.global

Specifying Durations

Operations often require a duration to be specified. A duration DSL is available to make defining these easier:

import scala.concurrent.duration._
val d: Duration = 10.seconds

Using Futures For Non-blocking Computation

Basic use of futures is easy with the factory method on Future, which executes a provided function asynchronously, handing you back a future result of that function without blocking the current thread. In order to create the Future you will need either an implicit or explicit ExecutionContext to be provided:

import scala.concurrent._
import ExecutionContext.Implicits.global  // implicit execution context

val firstZebra: Future[Int] = Future {
 val words = Files.readAllLines("/etc/dictionaries-common/words").asScala
 words.indexOfSlice("zebra")
}

Avoid Blocking

Although blocking is possible in order to await results (with a mandatory timeout duration):

import scala.concurrent.duration._
Await.result(firstZebra, 10.seconds)

and although this is sometimes necessary to do, in particular for testing purposes, blocking in general is discouraged when working with Futures and concurrency in order to avoid potential deadlocks and improve performance. Instead, use callbacks or combinators to remain in the future domain:

val animalRange: Future[Int] = for {
 aardvark <- firstAardvark
 zebra <- firstZebra
} yield zebra - aardvark

animalRange.onSuccess {
 case x if x > 500000 => println("It's a long way from Aardvark to Zebra")
}

The jdk package contains utilities to interact with JDK classes.

This packages offers a number of converters, that are able to wrap or copy types from the scala library to equivalent types in the JDK class library and vice versa:

- CollectionConverters, converting collections like scala.collection.Seq, scala.collection.Map, scala.collection.Set, scala.collection.mutable.Buffer, scala.collection.Iterator and scala.collection.Iterable to their JDK counterparts - OptionConverters, converting between Option and java.util.Optional and primitive variations - StreamConverters, to create JDK Streams from scala collections - DurationConverters, for conversions between scala scala.concurrent.duration.FiniteDuration and java.time.Duration - FunctionConverters, from scala Functions to java java.util.function.Function, java.util.function.UnaryOperator, java.util.function.Consumer and java.util.function.Predicate, as well as primitive variations and Bi-variations.

By convention, converters that wrap an object to provide a different interface to the same underlying data structure use .asScala and .asJava extension methods, whereas converters that copy the underlying data structure use .toScala and .toJava.

In the javaapi package, the same converters can be found with a java-friendly interface that don't rely on implicit enrichments.

Additionally, this package offers Accumulators, capable of efficiently traversing JDK Streams.

The package object scala.math contains methods for performing basic numeric operations such as elementary exponential, logarithmic, root and trigonometric functions.

All methods forward to java.lang.Math unless otherwise noted.

This package handles the execution of external processes. The contents of this package can be divided in three groups, according to their responsibilities:

• Indicating what to run and how to run it.

• Handling a process input and output.

• Running the process.

For simple uses, the only group that matters is the first one. Running an external command can be as simple as "ls".!, or as complex as building a pipeline of commands such as this:

import scala.sys.process._
"ls" #| "grep .scala" #&& Seq("sh", "-c", "scalac *.scala") #|| "echo nothing found" lazyLines

We describe below the general concepts and architecture of the package, and then take a closer look at each of the categories mentioned above.

Concepts and Architecture

The underlying basis for the whole package is Java's Process and ProcessBuilder classes. While there's no need to use these Java classes, they impose boundaries on what is possible. One cannot, for instance, retrieve a process id for whatever is executing.

When executing an external process, one can provide a command's name, arguments to it, the directory in which it will be executed and what environment variables will be set. For each executing process, one can feed its standard input through a java.io.OutputStream, and read from its standard output and standard error through a pair of java.io.InputStream. One can wait until a process finishes execution and then retrieve its return value, or one can kill an executing process. Everything else must be built on those features.

This package provides a DSL for running and chaining such processes, mimicking Unix shells ability to pipe output from one process to the input of another, or control the execution of further processes based on the return status of the previous one.

In addition to this DSL, this package also provides a few ways of controlling input and output of these processes, going from simple and easy to use to complex and flexible.

When processes are composed, a new ProcessBuilder is created which, when run, will execute the ProcessBuilder instances it is composed of according to the manner of the composition. If piping one process to another, they'll be executed simultaneously, and each will be passed a ProcessIO that will copy the output of one to the input of the other.

What to Run and How

The central component of the process execution DSL is the scala.sys.process.ProcessBuilder trait. It is ProcessBuilder that implements the process execution DSL, that creates the scala.sys.process.Process that will handle the execution, and return the results of such execution to the caller. We can see that DSL in the introductory example: #|, #&& and #!! are methods on ProcessBuilder used to create a new ProcessBuilder through composition.

One creates a ProcessBuilder either through factories on the scala.sys.process.Process's companion object, or through implicit conversions available in this package object itself. Implicitly, each process is created either out of a String, with arguments separated by spaces -- no escaping of spaces is possible -- or out of a scala.collection.Seq, where the first element represents the command name, and the remaining elements are arguments to it. In this latter case, arguments may contain spaces.

To further control what how the process will be run, such as specifying the directory in which it will be run, see the factories on scala.sys.process.Process's companion object.

Once the desired ProcessBuilder is available, it can be executed in different ways, depending on how one desires to control its I/O, and what kind of result one wishes for:

• Return status of the process (! methods)

• Output of the process as a String (!! methods)

• Continuous output of the process as a LazyList[String] (lazyLines methods)

• The Process representing it (run methods)

Some simple examples of these methods:

import scala.sys.process._

// This uses ! to get the exit code
def fileExists(name: String) = Seq("test", "-f", name).! == 0

// This uses !! to get the whole result as a string
val dirContents = "ls".!!

// This "fire-and-forgets" the method, which can be lazily read through
// a LazyList[String]
def sourceFilesAt(baseDir: String): LazyList[String] = {
 val cmd = Seq("find", baseDir, "-name", "*.scala", "-type", "f")
 cmd.lazyLines
}

We'll see more details about controlling I/O of the process in the next section.

Handling Input and Output

In the underlying Java model, once a Process has been started, one can get java.io.InputStream and java.io.OutputStream representing its output and input respectively. That is, what one writes to an OutputStream is turned into input to the process, and the output of a process can be read from an InputStream -- of which there are two, one representing normal output, and the other representing error output.

This model creates a difficulty, which is that the code responsible for actually running the external processes is the one that has to take decisions about how to handle its I/O.

This package presents an alternative model: the I/O of a running process is controlled by a scala.sys.process.ProcessIO object, which can be passed _to_ the code that runs the external process. A ProcessIO will have direct access to the java streams associated with the process I/O. It must, however, close these streams afterwards.

Simpler abstractions are available, however. The components of this package that handle I/O are:

• scala.sys.process.ProcessIO: provides the low level abstraction.

• scala.sys.process.ProcessLogger: provides a higher level abstraction for output, and can be created through its companion object.

• scala.sys.process.BasicIO: a library of helper methods for the creation of ProcessIO.

• This package object itself, with a few implicit conversions.

Some examples of I/O handling:

import scala.sys.process._

// An overly complex way of computing size of a compressed file
def gzFileSize(name: String) = {
 val cat = Seq("zcat", name)
 var count = 0
 def byteCounter(input: java.io.InputStream) = {
   while(input.read() != -1) count += 1
   input.close()
 }
 val p = cat run new ProcessIO(_.close(), byteCounter, _.close())
 p.exitValue()
 count
}

// This "fire-and-forgets" the method, which can be lazily read through
// a LazyList[String], and accumulates all errors on a StringBuffer
def sourceFilesAt(baseDir: String): (LazyList[String], StringBuffer) = {
 val buffer = new StringBuffer()
 val cmd = Seq("find", baseDir, "-name", "*.scala", "-type", "f")
 val lazyLines = cmd lazyLines_! ProcessLogger(buffer append _)
 (lazyLines, buffer)
}

Instances of the java classes java.io.File and java.net.URL can both be used directly as input to other processes, and java.io.File can be used as output as well. One can even pipe one to the other directly without any intervening process, though that's not a design goal or recommended usage. For example, the following code will copy a web page to a file:

import java.io.File
import java.net.URL
import scala.sys.process._
new URL("https://www.scala-lang.org/") #> new File("scala-lang.html") !

More information about the other ways of controlling I/O can be found in the Scaladoc for the associated objects, traits and classes.

Running the Process

Paradoxically, this is the simplest component of all, and the one least likely to be interacted with. It consists solely of scala.sys.process.Process, and it provides only two methods:

• exitValue(): blocks until the process exit, and then returns the exit value. This is what happens when one uses the ! method of ProcessBuilder.

• destroy(): this will kill the external process and close the streams associated with it.