AsyncFeatureSpec

Enables testing of asynchronous code without blocking, using a style consistent with traditional AnyFeatureSpec tests.

Given a Future returned by the code you are testing, you need not block until the Future completes before performing assertions against its value. You can instead map those assertions onto the Future and return the resulting Future[Assertion] to ScalaTest. The test will complete asynchronously, when the Future[Assertion] completes.

Although not required, AsyncFeatureSpec is often used together with GivenWhenThen to express acceptance requirements in more detail. Here's an example AsyncFeatureSpec:

package org.scalatest.examples.asyncfeaturespec

import org.scalatest._
import scala.concurrent.Future
import scala.concurrent.ExecutionContext

// Defining actor messages
case object IsOn
case object PressPowerButton

class TVSetActor { // Simulating an actor
  private var on: Boolean = false
  def !(msg: PressPowerButton.type): Unit =
    synchronized {
      on = !on
    }
  def ?(msg: IsOn.type)(implicit c: ExecutionContext): Future[Boolean] =
    Future {
      synchronized { on }
    }
}

class TVSetActorSpec extends featurespec.AsyncFeatureSpec with GivenWhenThen {

  implicit override def executionContext =
    scala.concurrent.ExecutionContext.Implicits.global

  info("As a TV set owner")
  info("I want to be able to turn the TV on and off")
  info("So I can watch TV when I want")
  info("And save energy when I'm not watching TV")

  Feature("TV power button") {
    Scenario("User presses power button when TV is off") {

      Given("a TV set that is switched off")
      val tvSetActor = new TVSetActor

      When("the power button is pressed")
      tvSetActor ! PressPowerButton

      Then("the TV should switch on")
      val futureBoolean = tvSetActor ? IsOn
      futureBoolean map { isOn => assert(isOn) }
    }

    Scenario("User presses power button when TV is on") {

      Given("a TV set that is switched on")
      val tvSetActor = new TVSetActor
      tvSetActor ! PressPowerButton

      When("the power button is pressed")
      tvSetActor ! PressPowerButton

      Then("the TV should switch off")
      val futureBoolean = tvSetActor ? IsOn
      futureBoolean map { isOn => assert(!isOn) }
    }
  }
}

Note: for more information on the calls to Given, When, and Then, see the documentation for trait GivenWhenThen and the Informers section below.

An AsyncFeatureSpec contains feature clauses and scenarios. You define a feature clause with feature, and a scenario with scenario. Both feature and scenario are methods, defined in AsyncFeatureSpec, which will be invoked by the primary constructor of TVSetActorSpec. A feature clause describes a feature of the subject (class or other entity) you are specifying and testing. In the previous example, the subject under specification and test is a TV set. The feature being specified and tested is the behavior of a TV set when its power button is pressed. With each scenario you provide a string (the spec text) that specifies the behavior of the subject for one scenario in which the feature may be used, and a block of code that tests that behavior. You place the spec text between the parentheses, followed by the test code between curly braces. The test code will be wrapped up as a function passed as a by-name parameter to scenario, which will register the test for later execution. The result type of the by-name in an AsyncFeatureSpec must be Future[Assertion].

Starting with version 3.0.0, ScalaTest assertions and matchers have result type Assertion. The result type of the first test in the example above, therefore, is Future[Assertion]. When an AsyncFeatureSpec is constructed, any test that results in Assertion will be implicitly converted to Future[Assertion] and registered. The implicit conversion is from Assertion to Future[Assertion] only, so you must end synchronous tests in some ScalaTest assertion or matcher expression. If a test would not otherwise end in type Assertion, you can place succeed at the end of the test. succeed, a field in trait Assertions, returns the Succeeded singleton:

scala> succeed
res2: org.scalatest.Assertion = Succeeded

Thus placing succeed at the end of a test body will satisfy the type checker.

An AsyncFeatureSpec's lifecycle has two phases: the registration phase and the ready phase. It starts in registration phase and enters ready phase the first time run is called on it. It then remains in ready phase for the remainder of its lifetime.

Scenarios can only be registered with the scenario method while the AsyncFeatureSpec is in its registration phase. Any attempt to register a scenario after the AsyncFeatureSpec has entered its ready phase, i.e., after run has been invoked on the AsyncFeatureSpec, will be met with a thrown TestRegistrationClosedException. The recommended style of using AsyncFeatureSpec is to register tests during object construction as is done in all the examples shown here. If you keep to the recommended style, you should never see a TestRegistrationClosedException.

Each scenario represents one test. The name of the test is the spec text passed to the scenario method. The feature name does not appear as part of the test name. In a AsyncFeatureSpec, therefore, you must take care to ensure that each test has a unique name (in other words, that each scenario has unique spec text).

When you run a AsyncFeatureSpec, it will send Formatters in the events it sends to the Reporter. ScalaTest's built-in reporters will report these events in such a way that the output is easy to read as an informal specification of the subject being tested. For example, were you to run TVSetSpec from within the Scala interpreter:

scala> org.scalatest.run(new TVSetActorSpec)

You would see:

TVSetActorSpec:
As a TV set owner
I want to be able to turn the TV on and off
So I can watch TV when I want
And save energy when I'm not watching TV
Feature: TV power button
  Scenario: User presses power button when TV is off
    Given a TV set that is switched off
    When the power button is pressed
    Then the TV should switch on
  Scenario: User presses power button when TV is on
    Given a TV set that is switched on
    When the power button is pressed
    Then the TV should switch off

Or, to run just the “Feature: TV power button Scenario: User presses power button when TV is on” method, you could pass that test's name, or any unique substring of the name, such as "TV is on". Here's an example:

scala> org.scalatest.run(new TVSetActorSpec, "TV is on")
TVSetActorSpec:
As a TV set owner
I want to be able to turn the TV on and off
So I can watch TV when I want
And save energy when I'm not watching TV
Feature: TV power button
  Scenario: User presses power button when TV is on
    Given a TV set that is switched on
    When the power button is pressed
    Then the TV should switch off

Asynchronous execution model

AsyncFeatureSpec extends AsyncTestSuite, which provides an implicit scala.concurrent.ExecutionContext named executionContext. This execution context is used by AsyncFeatureSpec to transform the Future[Assertion]s returned by each test into the FutureOutcome returned by the test function passed to withFixture. This ExecutionContext is also intended to be used in the tests, including when you map assertions onto futures.

On both the JVM and Scala.js, the default execution context provided by ScalaTest's asynchronous testing styles confines execution to a single thread per test. On JavaScript, where single-threaded execution is the only possibility, the default execution context is scala.scalajs.concurrent.JSExecutionContext.Implicits.queue. On the JVM, the default execution context is a serial execution context provided by ScalaTest itself.

When ScalaTest's serial execution context is called upon to execute a task, that task is recorded in a queue for later execution. For example, one task that will be placed in this queue is the task that transforms the Future[Assertion] returned by an asynchronous test body to the FutureOutcome returned from the test function. Other tasks that will be queued are any transformations of, or callbacks registered on, Futures that occur in your test body, including any assertions you map onto Futures. Once the test body returns, the thread that executed the test body will execute the tasks in that queue one after another, in the order they were enqueued.

ScalaTest provides its serial execution context as the default on the JVM for three reasons. First, most often running both tests and suites in parallel does not give a significant performance boost compared to just running suites in parallel. Thus parallel execution of Future transformations within individual tests is not generally needed for performance reasons.

Second, if multiple threads are operating in the same suite concurrently, you'll need to make sure access to any mutable fixture objects by multiple threads is synchronized. Although access to mutable state along the same linear chain of Future transformations need not be synchronized, this does not hold true for callbacks, and in general it is easy to make a mistake. Simply put: synchronizing access to shared mutable state is difficult and error prone. Because ScalaTest's default execution context on the JVM confines execution of Future transformations and call backs to a single thread, you need not (by default) worry about synchronizing access to mutable state in your asynchronous-style tests.

Third, asynchronous-style tests need not be complete when the test body returns, because the test body returns a Future[Assertion]. This Future[Assertion] will often represent a test that has not yet completed. As a result, when using a more traditional execution context backed by a thread-pool, you could potentially start many more tests executing concurrently than there are threads in the thread pool. The more concurrently execute tests you have competing for threads from the same limited thread pool, the more likely it will be that tests will intermitently fail due to timeouts.

Using ScalaTest's serial execution context on the JVM will ensure the same thread that produced the Future[Assertion] returned from a test body is also used to execute any tasks given to the execution context while executing the test body—and that thread will not be allowed to do anything else until the test completes. If the serial execution context's task queue ever becomes empty while the Future[Assertion] returned by that test's body has not yet completed, the thread will block until another task for that test is enqueued. Although it may seem counter-intuitive, this blocking behavior means the total number of tests allowed to run concurrently will be limited to the total number of threads executing suites. This fact means you can tune the thread pool such that maximum performance is reached while avoiding (or at least, reducing the likelihood of) tests that fail due to timeouts because of thread competition.

This thread confinement strategy does mean, however, that when you are using the default execution context on the JVM, you must be sure to never block in the test body waiting for a task to be completed by the execution context. If you block, your test will never complete. This kind of problem will be obvious, because the test will consistently hang every time you run it. (If a test is hanging, and you're not sure which one it is, enable slowpoke notifications.) If you really do want to block in your tests, you may wish to just use a traditional AnyFeatureSpec with ScalaFutures instead. Alternatively, you could override the executionContext and use a traditional ExecutionContext backed by a thread pool. This will enable you to block in an asynchronous-style test on the JVM, but you'll need to worry about synchronizing access to shared mutable state.

To use a different execution context, just override executionContext. For example, if you prefer to use the runNow execution context on Scala.js instead of the default queue, you would write:

// on Scala.js
implicit override def executionContext =
    org.scalatest.concurrent.TestExecutionContext.runNow

If you prefer on the JVM to use the global execution context, which is backed by a thread pool, instead of ScalaTest's default serial execution contex, which confines execution to a single thread, you would write:

// on the JVM (and also compiles on Scala.js, giving
// you the queue execution context)
implicit override def executionContext =
    scala.concurrent.ExecutionContext.Implicits.global

Serial and parallel test execution

By default (unless you mix in ParallelTestExecution), tests in an AsyncFeatureSpec will be executed one after another, i.e., serially. This is true whether those tests return Assertion or Future[Assertion], no matter what threads are involved. This default behavior allows you to re-use a shared fixture, such as an external database that needs to be cleaned after each test, in multiple tests in async-style suites. This is implemented by registering each test, other than the first test, to run as a continuation after the previous test completes.

If you want the tests of an AsyncFeatureSpec to be executed in parallel, you must mix in ParallelTestExecution and enable parallel execution of tests in your build. You enable parallel execution in Runner with the -P command line flag. In the ScalaTest Maven Plugin, set parallel to true. In sbt, parallel execution is the default, but to be explicit you can write:

parallelExecution in Test := true // the default in sbt

On the JVM, if both ParallelTestExecution is mixed in and parallel execution is enabled in the build, tests in an async-style suite will be started in parallel, using threads from the Distributor, and allowed to complete in parallel, using threads from the executionContext. If you are using ScalaTest's serial execution context, the JVM default, asynchronous tests will run in parallel very much like traditional (such as AnyFeatureSpec) tests run in parallel: 1) Because ParallelTestExecution extends OneInstancePerTest, each test will run in its own instance of the test class, you need not worry about synchronizing access to mutable instance state shared by different tests in the same suite. 2) Because the serial execution context will confine the execution of each test to the single thread that executes the test body, you need not worry about synchronizing access to shared mutable state accessed by transformations and callbacks of Futures inside the test.

If ParallelTestExecution is mixed in but parallel execution of suites is not enabled, asynchronous tests on the JVM will be started sequentially, by the single thread that invoked run, but without waiting for one test to complete before the next test is started. As a result, asynchronous tests will be allowed to complete in parallel, using threads from the executionContext. If you are using the serial execution context, however, you'll see the same behavior you see when parallel execution is disabled and a traditional suite that mixes in ParallelTestExecution is executed: the tests will run sequentially. If you use an execution context backed by a thread-pool, such as global, however, even though tests will be started sequentially by one thread, they will be allowed to run concurrently using threads from the execution context's thread pool.

The latter behavior is essentially what you'll see on Scala.js when you execute a suite that mixes in ParallelTestExecution. Because only one thread exists when running under JavaScript, you can't "enable parallel execution of suites." However, it may still be useful to run tests in parallel on Scala.js, because tests can invoke API calls that are truly asynchronous by calling into external APIs that take advantage of non-JavaScript threads. Thus on Scala.js, ParallelTestExecution allows asynchronous tests to run in parallel, even though they must be started sequentially. This may give you better performance when you are using API calls in your Scala.js tests that are truly asynchronous.

Futures and expected exceptions

If you need to test for expected exceptions in the context of futures, you can use the recoverToSucceededIf and recoverToExceptionIf methods of trait RecoverMethods. Because this trait is mixed into supertrait AsyncTestSuite, both of these methods are available by default in an AsyncFeatureSpec.

If you just want to ensure that a future fails with a particular exception type, and do not need to inspect the exception further, use recoverToSucceededIf:

recoverToSucceededIf[IllegalStateException] { // Result type: Future[Assertion]
  emptyStackActor ? Peek
}

The recoverToSucceededIf method performs a job similar to assertThrows, except in the context of a future. It transforms a Future of any type into a Future[Assertion] that succeeds only if the original future fails with the specified exception. Here's an example in the REPL:

scala> import org.scalatest.RecoverMethods._
import org.scalatest.RecoverMethods._

scala> import scala.concurrent.Future
import scala.concurrent.Future

scala> import scala.concurrent.ExecutionContext.Implicits.global
import scala.concurrent.ExecutionContext.Implicits.global

scala> recoverToSucceededIf[IllegalStateException] {
     |   Future { throw new IllegalStateException }
     | }
res0: scala.concurrent.Future[org.scalatest.Assertion] = ...

scala> res0.value
res1: Option[scala.util.Try[org.scalatest.Assertion]] = Some(Success(Succeeded))

Otherwise it fails with an error message similar to those given by assertThrows:

scala> recoverToSucceededIf[IllegalStateException] {
     |   Future { throw new RuntimeException }
     | }
res2: scala.concurrent.Future[org.scalatest.Assertion] = ...

scala> res2.value
res3: Option[scala.util.Try[org.scalatest.Assertion]] =
    Some(Failure(org.scalatest.exceptions.TestFailedException: Expected exception
      java.lang.IllegalStateException to be thrown, but java.lang.RuntimeException
      was thrown))

scala> recoverToSucceededIf[IllegalStateException] {
     |   Future { 42 }
     | }
res4: scala.concurrent.Future[org.scalatest.Assertion] = ...

scala> res4.value
res5: Option[scala.util.Try[org.scalatest.Assertion]] =
    Some(Failure(org.scalatest.exceptions.TestFailedException: Expected exception
      java.lang.IllegalStateException to be thrown, but no exception was thrown))

The recoverToExceptionIf method differs from the recoverToSucceededIf in its behavior when the assertion succeeds: recoverToSucceededIf yields a Future[Assertion], whereas recoverToExceptionIf yields a Future[T], where T is the expected exception type.

recoverToExceptionIf[IllegalStateException] { // Result type: Future[IllegalStateException]
  emptyStackActor ? Peek
}

In other words, recoverToExpectionIf is to intercept as recovertToSucceededIf is to assertThrows. The first one allows you to perform further assertions on the expected exception. The second one gives you a result type that will satisfy the type checker at the end of the test body. Here's an example showing recoverToExceptionIf in the REPL:

scala> val futureEx =
     |   recoverToExceptionIf[IllegalStateException] {
     |     Future { throw new IllegalStateException("hello") }
     |   }
futureEx: scala.concurrent.Future[IllegalStateException] = ...

scala> futureEx.value
res6: Option[scala.util.Try[IllegalStateException]] =
    Some(Success(java.lang.IllegalStateException: hello))

scala> futureEx map { ex => assert(ex.getMessage == "world") }
res7: scala.concurrent.Future[org.scalatest.Assertion] = ...

scala> res7.value
res8: Option[scala.util.Try[org.scalatest.Assertion]] =
    Some(Failure(org.scalatest.exceptions.TestFailedException: "[hello]" did not equal "[world]"))

Ignored tests

To support the common use case of temporarily disabling a test, with the good intention of resurrecting the test at a later time, AsyncFeatureSpec provides registration methods that start with ignore instead of scenario. Here's an example:

package org.scalatest.examples.asyncfeaturespec.ignore

import org.scalatest.featurespec.AsyncFeatureSpec
import scala.concurrent.Future

class AddSpec extends AsyncFeatureSpec {

  def addSoon(addends: Int*): Future[Int] = Future { addends.sum }
  def addNow(addends: Int*): Int = addends.sum

  Feature("The add methods") {

    ignore("addSoon will eventually compute a sum of passed Ints") {
      val futureSum: Future[Int] = addSoon(1, 2)
      // You can map assertions onto a Future, then return
      // the resulting Future[Assertion] to ScalaTest:
      futureSum map { sum => assert(sum == 3) }
    }

    Scenario("addNow will immediately compute a sum of passed Ints") {
      val sum: Int = addNow(1, 2)
      // You can also write synchronous tests. The body
      // must have result type Assertion:
      assert(sum == 3)
    }
  }
}

If you run class AddSpec with:

scala> org.scalatest.run(new AddSpec)

It will run only the second test and report that the first test was ignored:

AddSpec:
Feature: The add methods
- Scenario: addSoon will eventually compute a sum of passed Ints !!! IGNORED !!!
- Scenario: addNow will immediately compute a sum of passed Ints

If you wish to temporarily ignore an entire suite of tests, you can (on the JVM, not Scala.js) annotate the test class with @Ignore, like this:

package org.scalatest.examples.asyncfeaturespec.ignoreall

import org.scalatest.featurespec.AsyncFeatureSpec
import scala.concurrent.Future
import org.scalatest.Ignore

@Ignore
class AddSpec extends AsyncFeatureSpec {

  def addSoon(addends: Int*): Future[Int] = Future { addends.sum }
  def addNow(addends: Int*): Int = addends.sum

  Feature("The add methods") {

    Scenario("addSoon will eventually compute a sum of passed Ints") {
      val futureSum: Future[Int] = addSoon(1, 2)
      // You can map assertions onto a Future, then return
      // the resulting Future[Assertion] to ScalaTest:
      futureSum map { sum => assert(sum == 3) }
    }

    Scenario("addNow will immediately compute a sum of passed Ints") {
      val sum: Int = addNow(1, 2)
      // You can also write synchronous tests. The body
      // must have result type Assertion:
      assert(sum == 3)
    }
  }
}

When you mark a test class with a tag annotation, ScalaTest will mark each test defined in that class with that tag. Thus, marking the AddSpec in the above example with the @Ignore tag annotation means that both tests in the class will be ignored. If you run the above AddSpec in the Scala interpreter, you'll see:

AddSpec:
Feature: The add methods
- Scenario: addSoon will eventually compute a sum of passed Ints !!! IGNORED !!!
- Scenario: addNow will immediately compute a sum of passed Ints !!! IGNORED !!!

Note that marking a test class as ignored won't prevent it from being discovered by ScalaTest. Ignored classes will be discovered and run, and all their tests will be reported as ignored. This is intended to keep the ignored class visible, to encourage the developers to eventually fix and “un-ignore” it. If you want to prevent a class from being discovered at all (on the JVM, not Scala.js), use the DoNotDiscover annotation instead.

If you want to ignore all tests of a suite on Scala.js, where annotations can't be inspected at runtime, you'll need to change it to ignore at each test site. To make a suite non-discoverable on Scala.js, ensure it does not declare a public no-arg constructor. You can either declare a public constructor that takes one or more arguments, or make the no-arg constructor non-public. Because this technique will also make the suite non-discoverable on the JVM, it is a good approach for suites you want to run (but not be discoverable) on both Scala.js and the JVM.

Informers

One of the parameters to AsyncFeatureSpec's run method is a Reporter, which will collect and report information about the running suite of tests. Information about suites and tests that were run, whether tests succeeded or failed, and tests that were ignored will be passed to the Reporter as the suite runs. Most often the default reporting done by AsyncFeatureSpec's methods will be sufficient, but occasionally you may wish to provide custom information to the Reporter from a test. For this purpose, an Informer that will forward information to the current Reporter is provided via the info parameterless method. You can pass the extra information to the Informer via its apply method. The Informer will then pass the information to the Reporter via an InfoProvided event.

One use case for the Informer is to pass more information about a scenario to the reporter. For example, the GivenWhenThen trait provides methods that use the implicit info provided by AsyncFeatureSpec to pass such information to the reporter. You can see this in action in the initial example of this trait's documentation.

Documenters

AsyncFeatureSpec also provides a markup method that returns a Documenter, which allows you to send to the Reporter text formatted in Markdown syntax. You can pass the extra information to the Documenter via its apply method. The Documenter will then pass the information to the Reporter via an MarkupProvided event.

Here's an example FlatSpec that uses markup:

package org.scalatest.examples.asyncfeaturespec.markup

import collection.mutable
import org.scalatest._

class SetSpec extends featurespec.AsyncFeatureSpec with GivenWhenThen {

  markup { """

Mutable Set
-----------

A set is a collection that contains no duplicate elements.

To implement a concrete mutable set, you need to provide implementations
of the following methods:

    def contains(elem: A): Boolean
    def iterator: Iterator[A]
    def += (elem: A): this.type
    def -= (elem: A): this.type

If you wish that methods like `take`,
`drop`, `filter` return the same kind of set,
you should also override:

     def empty: This

It is also good idea to override methods `foreach` and
`size` for efficiency.

  """ }

  Feature("An element can be added to an empty mutable Set") {
    Scenario("When an element is added to an empty mutable Set") {
      Given("an empty mutable Set")
      val set = mutable.Set.empty[String]

      When("an element is added")
      set += "clarity"

      Then("the Set should have size 1")
      assert(set.size === 1)

      And("the Set should contain the added element")
      assert(set.contains("clarity"))

      markup("This test finished with a **bold** statement!")
      succeed
    }
  }
}

Although all of ScalaTest's built-in reporters will display the markup text in some form, the HTML reporter will format the markup information into HTML. Thus, the main purpose of markup is to add nicely formatted text to HTML reports. Here's what the above SetSpec would look like in the HTML reporter:

Notifiers and alerters

ScalaTest records text passed to info and markup during tests, and sends the recorded text in the recordedEvents field of test completion events like TestSucceeded and TestFailed. This allows string reporters (like the standard out reporter) to show info and markup text after the test name in a color determined by the outcome of the test. For example, if the test fails, string reporters will show the info and markup text in red. If a test succeeds, string reporters will show the info and markup text in green. While this approach helps the readability of reports, it means that you can't use info to get status updates from long running tests.

To get immediate (i.e., non-recorded) notifications from tests, you can use note (a Notifier) and alert (an Alerter). Here's an example showing the differences:

package org.scalatest.examples.asyncfeaturespec.note

import collection.mutable
import org.scalatest._

class SetSpec extends featurespec.AsyncFeatureSpec {

  Feature("An element can be added to an empty mutable Set") {
    Scenario("When an element is added to an empty mutable Set") {

      info("info is recorded")
      markup("markup is *also* recorded")
      note("notes are sent immediately")
      alert("alerts are also sent immediately")

      val set = mutable.Set.empty[String]
      set += "clarity"
      assert(set.size === 1)
      assert(set.contains("clarity"))
    }
  }
}

Because note and alert information is sent immediately, it will appear before the test name in string reporters, and its color will be unrelated to the ultimate outcome of the test: note text will always appear in green, alert text will always appear in yellow. Here's an example:

scala> org.scalatest.run(new SetSpec)
SetSpec:
Feature: An element can be added to an empty mutable Set
  + notes are sent immediately
  + alerts are also sent immediately
  Scenario: When an element is added to an empty mutable Set
    info is recorded
  + markup is *also* recorded

Another example is slowpoke notifications. If you find a test is taking a long time to complete, but you're not sure which test, you can enable slowpoke notifications. ScalaTest will use an Alerter to fire an event whenever a test has been running longer than a specified amount of time.

In summary, use info and markup for text that should form part of the specification output. Use note and alert to send status notifications. (Because the HTML reporter is intended to produce a readable, printable specification, info and markup text will appear in the HTML report, but note and alert text will not.)

Pending tests

A pending test is one that has been given a name but is not yet implemented. The purpose of pending tests is to facilitate a style of testing in which documentation of behavior is sketched out before tests are written to verify that behavior (and often, before the behavior of the system being tested is itself implemented). Such sketches form a kind of specification of what tests and functionality to implement later.

To support this style of testing, a test can be given a name that specifies one bit of behavior required by the system being tested. At the end of the test, it can call method pending, which will cause it to complete abruptly with TestPendingException.

Because tests in ScalaTest can be designated as pending with TestPendingException, both the test name and any information sent to the reporter when running the test can appear in the report of a test run. (In other words, the code of a pending test is executed just like any other test.) However, because the test completes abruptly with TestPendingException, the test will be reported as pending, to indicate the actual test, and possibly the functionality, has not yet been implemented. Here's an example:

package org.scalatest.examples.asyncfeaturespec.pending

import org.scalatest.featurespec.AsyncFeatureSpec
import scala.concurrent.Future

class AddSpec extends AsyncFeatureSpec {

  def addSoon(addends: Int*): Future[Int] = Future { addends.sum }
  def addNow(addends: Int*): Int = addends.sum

  Feature("The add methods") {

    Scenario("addSoon will eventually compute a sum of passed Ints") (pending)

    Scenario("addNow will immediately compute a sum of passed Ints") {
      val sum: Int = addNow(1, 2)
      // You can also write synchronous tests. The body
      // must have result type Assertion:
      assert(sum == 3)
    }
  }
}

(Note: "(pending)" is the body of the test. Thus the test contains just one statement, an invocation of the pending method, which throws TestPendingException.) If you run this version of AddSpec with:

scala> org.scalatest.run(new AddSpec)

It will run both tests, but report that first test is pending. You'll see:

AddSpec:
Feature: The add methods
- Scenario: addSoon will eventually compute a sum of passed Ints (pending)
- Scenario: addNow will immediately compute a sum of passed Ints

One difference between an ignored test and a pending one is that an ignored test is intended to be used during significant refactorings of the code under test, when tests break and you don't want to spend the time to fix all of them immediately. You can mark some of those broken tests as ignored temporarily, so that you can focus the red bar on just failing tests you actually want to fix immediately. Later you can go back and fix the ignored tests. In other words, by ignoring some failing tests temporarily, you can more easily notice failed tests that you actually want to fix. By contrast, a pending test is intended to be used before a test and/or the code under test is written. Pending indicates you've decided to write a test for a bit of behavior, but either you haven't written the test yet, or have only written part of it, or perhaps you've written the test but don't want to implement the behavior it tests until after you've implemented a different bit of behavior you realized you need first. Thus ignored tests are designed to facilitate refactoring of existing code whereas pending tests are designed to facilitate the creation of new code.

One other difference between ignored and pending tests is that ignored tests are implemented as a test tag that is excluded by default. Thus an ignored test is never executed. By contrast, a pending test is implemented as a test that throws TestPendingException (which is what calling the pending method does). Thus the body of pending tests are executed up until they throw TestPendingException.

Tagging tests

An AsyncFeatureSpec's tests may be classified into groups by tagging them with string names. As with any suite, when executing an AsyncFeatureSpec, groups of tests can optionally be included and/or excluded. To tag an AsyncFeatureSpec's tests, you pass objects that extend class org.scalatest.Tag to methods that register tests. Class Tag takes one parameter, a string name. If you have created tag annotation interfaces as described in the Tag documentation, then you will probably want to use tag names on your test functions that match. To do so, simply pass the fully qualified names of the tag interfaces to the Tag constructor. For example, if you've defined a tag annotation interface with fully qualified name, com.mycompany.tags.DbTest, then you could create a matching tag for AsyncFeatureSpecs like this:

package org.scalatest.examples.asyncfeaturespec.tagging

import org.scalatest.Tag

object DbTest extends Tag("com.mycompany.tags.DbTest")

Given these definitions, you could place AsyncFeatureSpec tests into groups with tags like this:

import org.scalatest.featurespec.AsyncFeatureSpec
import org.scalatest.tagobjects.Slow
import scala.concurrent.Future

class AddSpec extends AsyncFeatureSpec {

  def addSoon(addends: Int*): Future[Int] = Future { addends.sum }
  def addNow(addends: Int*): Int = addends.sum

  Feature("The add methods") {

    Scenario("addSoon will eventually compute a sum of passed Ints",
         Slow) {

      val futureSum: Future[Int] = addSoon(1, 2)
      // You can map assertions onto a Future, then return
      // the resulting Future[Assertion] to ScalaTest:
      futureSum map { sum => assert(sum == 3) }
    }

    Scenario("addNow will immediately compute a sum of passed Ints",
        Slow, DbTest) {

      val sum: Int = addNow(1, 2)
      // You can also write synchronous tests. The body
      // must have result type Assertion:
      assert(sum == 3)
    }
  }
}

This code marks both tests with the org.scalatest.tags.Slow tag, and the second test with the com.mycompany.tags.DbTest tag.

The run method takes a Filter, whose constructor takes an optional Set[String] called tagsToInclude and a Set[String] called tagsToExclude. If tagsToInclude is None, all tests will be run except those those belonging to tags listed in the tagsToExclude Set. If tagsToInclude is defined, only tests belonging to tags mentioned in the tagsToInclude set, and not mentioned in tagsToExclude, will be run.

It is recommended, though not required, that you create a corresponding tag annotation when you create a Tag object. A tag annotation (on the JVM, not Scala.js) allows you to tag all the tests of an AsyncFeatureSpec in one stroke by annotating the class. For more information and examples, see the documentation for class Tag. On Scala.js, to tag all tests of a suite, you'll need to tag each test individually at the test site.

Shared fixtures

A test fixture is composed of the objects and other artifacts (files, sockets, database connections, etc.) tests use to do their work. When multiple tests need to work with the same fixtures, it is important to try and avoid duplicating the fixture code across those tests. The more code duplication you have in your tests, the greater drag the tests will have on refactoring the actual production code.

ScalaTest recommends three techniques to eliminate such code duplication in async styles:

• Refactor using Scala
• Override withFixture
• Mix in a before-and-after trait

Each technique is geared towards helping you reduce code duplication without introducing instance vars, shared mutable objects, or other dependencies between tests. Eliminating shared mutable state across tests will make your test code easier to reason about and eliminate the need to synchronize access to shared mutable state on the JVM.

The following sections describe these techniques, including explaining the recommended usage for each. But first, here's a table summarizing the options:

Calling get-fixture methods

If you need to create the same mutable fixture objects in multiple tests, and don't need to clean them up after using them, the simplest approach is to write one or more get-fixture methods. A get-fixture method returns a new instance of a needed fixture object (or a holder object containing multiple fixture objects) each time it is called. You can call a get-fixture method at the beginning of each test that needs the fixture, storing the returned object or objects in local variables. Here's an example:

package org.scalatest.examples.asyncfeaturespec.getfixture

import org.scalatest.featurespec.AsyncFeatureSpec
import scala.concurrent.Future

class ExampleSpec extends AsyncFeatureSpec {

  def fixture: Future[String] = Future { "ScalaTest is designed to " }

  Feature("Simplicity") {
    Scenario("User needs to read test code written by others") {
      val future = fixture
      val result = future map { s => s + "encourage clear code!" }
      result map { s =>
        assert(s == "ScalaTest is designed to encourage clear code!")
      }
    }

    Scenario("User needs to understand what the tests are doing") {
      val future = fixture
      val result = future map { s => s + "be easy to reason about!" }
      result map { s =>
        assert(s == "ScalaTest is designed to be easy to reason about!")
      }
    }
  }
}

If you need to configure fixture objects differently in different tests, you can pass configuration into the get-fixture method. For example, you could pass in an initial value for a fixture object as a parameter to the get-fixture method.

Overriding withFixture(NoArgAsyncTest)

Although the get-fixture method approach takes care of setting up a fixture at the beginning of each test, it doesn't address the problem of cleaning up a fixture at the end of the test. If you just need to perform a side-effect at the beginning or end of a test, and don't need to actually pass any fixture objects into the test, you can override withFixture(NoArgAsyncTest), a method defined in trait AsyncTestSuite, a supertrait of AsyncFeatureSpec.

Trait AsyncFeatureSpec's runTest method passes a no-arg async test function to withFixture(NoArgAsyncTest). It is withFixture's responsibility to invoke that test function. The default implementation of withFixture simply invokes the function and returns the result, like this:

// Default implementation in trait AsyncTestSuite
protected def withFixture(test: NoArgAsyncTest): FutureOutcome = {
  test()
}

You can, therefore, override withFixture to perform setup before invoking the test function, and/or perform cleanup after the test completes. The recommended way to ensure cleanup is performed after a test completes is to use the complete-lastly syntax, defined in supertrait CompleteLastly. The complete-lastly syntax will ensure that cleanup will occur whether future-producing code completes abruptly by throwing an exception, or returns normally yielding a future. In the latter case, complete-lastly will register the cleanup code to execute asynchronously when the future completes.

The withFixture method is designed to be stacked, and to enable this, you should always call the super implementation of withFixture, and let it invoke the test function rather than invoking the test function directly. In other words, instead of writing “test()”, you should write “super.withFixture(test)”, like this:

// Your implementation
override def withFixture(test: NoArgTest) = {

  // Perform setup here

  complete {
    super.withFixture(test) // Invoke the test function
  } lastly {
    // Perform cleanup here
  }
}

If you have no cleanup to perform, you can write withFixture like this instead:

// Your implementation
override def withFixture(test: NoArgTest) = {

  // Perform setup here

  super.withFixture(test) // Invoke the test function
}

If you want to perform an action only for certain outcomes, you'll need to register code performing that action as a callback on the Future using one of Future's registration methods: onComplete, onSuccess, or onFailure. Note that if a test fails, that will be treated as a scala.util.Success(org.scalatest.Failed). So if you want to perform an action if a test fails, for example, you'd register the callback using onSuccess.

Here's an example in which withFixture(NoArgAsyncTest) is used to take a snapshot of the working directory if a test fails, and send that information to the standard output stream:

package org.scalatest.examples.asyncfeaturespec.noargasynctest

import java.io.File
import org.scalatest._
import scala.concurrent.Future

class ExampleSpec extends featurespec.AsyncFeatureSpec {

  override def withFixture(test: NoArgAsyncTest) = {

    super.withFixture(test) onFailedThen { _ =>
      val currDir = new File(".")
      val fileNames = currDir.list()
      info("Dir snapshot: " + fileNames.mkString(", "))
    }
  }

  def addSoon(addends: Int*): Future[Int] = Future { addends.sum }

  Feature("addSoon") {
    Scenario("succeed case") {
      addSoon(1, 1) map { sum => assert(sum == 2) }
    }

    Scenario("fail case") {
      addSoon(1, 1) map { sum => assert(sum == 3) }
    }
  }
}

Running this version of ExampleSpec in the interpreter in a directory with two files, hello.txt and world.txt would give the following output:

scala> org.scalatest.run(new ExampleSpec)
ExampleSpec:
Feature: addSoon
- Scenario: succeed case
- Scenario: fail case *** FAILED ***
  2 did not equal 3 (:33)

Note that the NoArgAsyncTest passed to withFixture, in addition to an apply method that executes the test, also includes the test name and the config map passed to runTest. Thus you can also use the test name and configuration objects in your withFixture implementation.

Lastly, if you want to transform the outcome in some way in withFixture, you'll need to use either the map or transform methods of Future, like this:

// Your implementation
override def withFixture(test: NoArgAsyncTest) = {

  // Perform setup here

  val futureOutcome = super.withFixture(test) // Invoke the test function

  futureOutcome change { outcome =>
    // transform the outcome into a new outcome here
  }
}

Note that a NoArgAsyncTest's apply method will return a scala.util.Failure only if the test completes abruptly with a "test-fatal" exception (such as OutOfMemoryError) that should cause the suite to abort rather than the test to fail. Thus usually you would use map to transform future outcomes, not transform, so that such test-fatal exceptions pass through unchanged. The suite will abort asynchronously with any exception returned from NoArgAsyncTest's apply method in a scala.util.Failure.

Calling loan-fixture methods

If you need to both pass a fixture object into a test and perform cleanup at the end of the test, you'll need to use the loan pattern. If different tests need different fixtures that require cleanup, you can implement the loan pattern directly by writing loan-fixture methods. A loan-fixture method takes a function whose body forms part or all of a test's code. It creates a fixture, passes it to the test code by invoking the function, then cleans up the fixture after the function returns.

The following example shows three tests that use two fixtures, a database and a file. Both require cleanup after, so each is provided via a loan-fixture method. (In this example, the database is simulated with a StringBuffer.)

package org.scalatest.examples.asyncfeaturespec.loanfixture

import java.util.concurrent.ConcurrentHashMap

import scala.concurrent.Future
import scala.concurrent.ExecutionContext

object DbServer { // Simulating a database server
  type Db = StringBuffer
  private final val databases = new ConcurrentHashMap[String, Db]
  def createDb(name: String): Db = {
    val db = new StringBuffer // java.lang.StringBuffer is thread-safe
    databases.put(name, db)
    db
  }
  def removeDb(name: String): Unit = {
    databases.remove(name)
  }
}

// Defining actor messages
sealed abstract class StringOp
case object Clear extends StringOp
case class Append(value: String) extends StringOp
case object GetValue

class StringActor { // Simulating an actor
  private final val sb = new StringBuilder
  def !(op: StringOp): Unit =
    synchronized {
      op match {
        case Append(value) => sb.append(value)
        case Clear => sb.clear()
      }
    }
  def ?(get: GetValue.type)(implicit c: ExecutionContext): Future[String] =
    Future {
      synchronized { sb.toString }
    }
}

import org.scalatest._
import DbServer._
import java.util.UUID.randomUUID

class ExampleSpec extends featurespec.AsyncFeatureSpec {

  def withDatabase(testCode: Future[Db] => Future[Assertion]) = {
    val dbName = randomUUID.toString // generate a unique db name
    val futureDb = Future { createDb(dbName) } // create the fixture
    complete {
      val futurePopulatedDb =
        futureDb map { db =>
          db.append("ScalaTest is designed to ") // perform setup
        }
      testCode(futurePopulatedDb) // "loan" the fixture to the test code
    } lastly {
      removeDb(dbName) // ensure the fixture will be cleaned up
    }
  }

  def withActor(testCode: StringActor => Future[Assertion]) = {
    val actor = new StringActor
    complete {
      actor ! Append("ScalaTest is designed to ") // set up the fixture
      testCode(actor) // "loan" the fixture to the test code
    } lastly {
      actor ! Clear // ensure the fixture will be cleaned up
    }
  }

  Feature("Simplicity") {
    // This test needs the actor fixture
    Scenario("User needs to read test code written by others") {
      withActor { actor =>
        actor ! Append("encourage clear code!")
        val futureString = actor ? GetValue
        futureString map { s =>
          assert(s === "ScalaTest is designed to encourage clear code!")
        }
      }
    }
    // This test needs the database fixture
    Scenario("User needs to understand what the tests are doing") {
      withDatabase { futureDb =>
        futureDb map { db =>
          db.append("be easy to reason about!")
          assert(db.toString === "ScalaTest is designed to be easy to reason about!")
        }
      }
    }
    // This test needs both the actor and the database
    Scenario("User needs to write tests") {
      withDatabase { futureDb =>
        withActor { actor => // loan-fixture methods compose
          actor ! Append("be easy to remember how to write!")
          val futureString = actor ? GetValue
          val futurePair: Future[(Db, String)] =
            futureDb zip futureString
          futurePair map { case (db, s) =>
            db.append("be easy to learn!")
            assert(db.toString === "ScalaTest is designed to be easy to learn!")
            assert(s === "ScalaTest is designed to be easy to remember how to write!")
          }
        }
      }
    }
  }
}

As demonstrated by the last test, loan-fixture methods compose. Not only do loan-fixture methods allow you to give each test the fixture it needs, they allow you to give a test multiple fixtures and clean everything up afterwards.

Also demonstrated in this example is the technique of giving each test its own "fixture sandbox" to play in. When your fixtures involve external side-effects, like creating databases, it is a good idea to give each database a unique name as is done in this example. This keeps tests completely isolated, allowing you to run them in parallel if desired.

Overriding withFixture(OneArgTest)

If all or most tests need the same fixture, you can avoid some of the boilerplate of the loan-fixture method approach by using a FixtureAsyncTestSuite and overriding withFixture(OneArgAsyncTest). Each test in a FixtureAsyncTestSuite takes a fixture as a parameter, allowing you to pass the fixture into the test. You must indicate the type of the fixture parameter by specifying FixtureParam, and implement a withFixture method that takes a OneArgAsyncTest. This withFixture method is responsible for invoking the one-arg async test function, so you can perform fixture set up before invoking and passing the fixture into the test function, and ensure clean up is performed after the test completes.

To enable the stacking of traits that define withFixture(NoArgAsyncTest), it is a good idea to let withFixture(NoArgAsyncTest) invoke the test function instead of invoking the test function directly. To do so, you'll need to convert the OneArgAsyncTest to a NoArgAsyncTest. You can do that by passing the fixture object to the toNoArgAsyncTest method of OneArgAsyncTest. In other words, instead of writing “test(theFixture)”, you'd delegate responsibility for invoking the test function to the withFixture(NoArgAsyncTest) method of the same instance by writing:

withFixture(test.toNoArgAsyncTest(theFixture))

Here's a complete example:

package org.scalatest.examples.asyncfeaturespec.oneargasynctest

import org.scalatest._
import scala.concurrent.Future
import scala.concurrent.ExecutionContext

// Defining actor messages
sealed abstract class StringOp
case object Clear extends StringOp
case class Append(value: String) extends StringOp
case object GetValue

class StringActor { // Simulating an actor
  private final val sb = new StringBuilder
  def !(op: StringOp): Unit =
    synchronized {
      op match {
        case Append(value) => sb.append(value)
        case Clear => sb.clear()
      }
    }
  def ?(get: GetValue.type)(implicit c: ExecutionContext): Future[String] =
    Future {
      synchronized { sb.toString }
    }
}

class ExampleSpec extends featurespec.FixtureAsyncFeatureSpec {

  type FixtureParam = StringActor

  def withFixture(test: OneArgAsyncTest): FutureOutcome = {

    val actor = new StringActor
    complete {
      actor ! Append("ScalaTest is designed to ") // set up the fixture
      withFixture(test.toNoArgAsyncTest(actor))
    } lastly {
      actor ! Clear // ensure the fixture will be cleaned up
    }
  }

  Feature("Simplicity") {
    Scenario("User needs to read test code written by others") { actor =>
      actor ! Append("encourage clear code!")
      val futureString = actor ? GetValue
      futureString map { s =>
        assert(s === "ScalaTest is designed to encourage clear code!")
      }
    }

    Scenario("User needs to understand what the tests are doing") { actor =>
      actor ! Append("be easy to reason about!")
      val futureString = actor ? GetValue
      futureString map { s =>
        assert(s === "ScalaTest is designed to be easy to reason about!")
      }
    }
  }
}

In this example, the tests required one fixture object, a StringActor. If your tests need multiple fixture objects, you can simply define the FixtureParam type to be a tuple containing the objects or, alternatively, a case class containing the objects. For more information on the withFixture(OneArgAsyncTest) technique, see the documentation for FixtureAsyncFeatureSpec.

Mixing in BeforeAndAfter

In all the shared fixture examples shown so far, the activities of creating, setting up, and cleaning up the fixture objects have been performed during the test. This means that if an exception occurs during any of these activities, it will be reported as a test failure. Sometimes, however, you may want setup to happen before the test starts, and cleanup after the test has completed, so that if an exception occurs during setup or cleanup, the entire suite aborts and no more tests are attempted. The simplest way to accomplish this in ScalaTest is to mix in trait BeforeAndAfter. With this trait you can denote a bit of code to run before each test with before and/or after each test each test with after, like this:

package org.scalatest.examples.asyncfeaturespec.beforeandafter

import org.scalatest.featurespec.AsyncFeatureSpec
import org.scalatest.BeforeAndAfter
import scala.concurrent.Future
import scala.concurrent.ExecutionContext

// Defining actor messages
sealed abstract class StringOp
case object Clear extends StringOp
case class Append(value: String) extends StringOp
case object GetValue

class StringActor { // Simulating an actor
  private final val sb = new StringBuilder
  def !(op: StringOp): Unit =
    synchronized {
      op match {
        case Append(value) => sb.append(value)
        case Clear => sb.clear()
      }
    }
  def ?(get: GetValue.type)(implicit c: ExecutionContext): Future[String] =
    Future {
      synchronized { sb.toString }
    }
}

class ExampleSpec extends AsyncFeatureSpec with BeforeAndAfter {

  final val actor = new StringActor

  before {
    actor ! Append("ScalaTest is designed to ") // set up the fixture
  }

  after {
    actor ! Clear // clean up the fixture
  }

  Feature("Simplicity") {
    Scenario("User needs to read test code written by others") {
      actor ! Append("encourage clear code!")
      val futureString = actor ? GetValue
      futureString map { s =>
        assert(s == "ScalaTest is designed to encourage clear code!")
      }
    }

    Scenario("User needs to understand what the tests are doing") {
      actor ! Append("be easy to reason about!")
      val futureString = actor ? GetValue
      futureString map { s =>
        assert(s == "ScalaTest is designed to be easy to reason about!")
      }
    }
  }
}

Note that the only way before and after code can communicate with test code is via some side-effecting mechanism, commonly by reassigning instance vars or by changing the state of mutable objects held from instance vals (as in this example). If using instance vars or mutable objects held from instance vals you wouldn't be able to run tests in parallel in the same instance of the test class (on the JVM, not Scala.js) unless you synchronized access to the shared, mutable state.

Note that on the JVM, if you override ScalaTest's default serial execution context, you will likely need to worry about synchronizing access to shared mutable fixture state, because the execution context may assign different threads to process different Future transformations. Although access to mutable state along the same linear chain of Future transformations need not be synchronized, it can be difficult to spot cases where these constraints are violated. The best approach is to use only immutable objects when transforming Futures. When that's not practical, involve only thread-safe mutable objects, as is done in the above example. On Scala.js, by contrast, you need not worry about thread synchronization, because in effect only one thread exists.

Although BeforeAndAfter provides a minimal-boilerplate way to execute code before and after tests, it isn't designed to enable stackable traits, because the order of execution would be non-obvious. If you want to factor out before and after code that is common to multiple test suites, you should use trait BeforeAndAfterEach instead, as shown later in the next section, composing fixtures by stacking traits.

Composing fixtures by stacking traits

In larger projects, teams often end up with several different fixtures that test classes need in different combinations, and possibly initialized (and cleaned up) in different orders. A good way to accomplish this in ScalaTest is to factor the individual fixtures into traits that can be composed using the stackable trait pattern. This can be done, for example, by placing withFixture methods in several traits, each of which call super.withFixture. Here's an example in which the StringBuilderActor and StringBufferActor fixtures used in the previous examples have been factored out into two stackable fixture traits named Builder and Buffer:

package org.scalatest.examples.asyncfeaturespec.composingwithasyncfixture

import org.scalatest._
import org.scalatest.SuiteMixin
import collection.mutable.ListBuffer
import scala.concurrent.Future
import scala.concurrent.ExecutionContext

// Defining actor messages
sealed abstract class StringOp
case object Clear extends StringOp
case class Append(value: String) extends StringOp
case object GetValue

class StringBuilderActor { // Simulating an actor
  private final val sb = new StringBuilder
  def !(op: StringOp): Unit =
    synchronized {
      op match {
        case Append(value) => sb.append(value)
        case Clear => sb.clear()
      }
    }
  def ?(get: GetValue.type)(implicit c: ExecutionContext): Future[String] =
    Future {
      synchronized { sb.toString }
    }
}

class StringBufferActor {
  private final val buf = ListBuffer.empty[String]
  def !(op: StringOp): Unit =
    synchronized {
      op match {
        case Append(value) => buf += value
        case Clear => buf.clear()
      }
    }
  def ?(get: GetValue.type)(implicit c: ExecutionContext): Future[List[String]] =
    Future {
      synchronized { buf.toList }
    }
}

trait Builder extends AsyncTestSuiteMixin { this: AsyncTestSuite =>

  final val builderActor = new StringBuilderActor

  abstract override def withFixture(test: NoArgAsyncTest) = {
    builderActor ! Append("ScalaTest is designed to ")
    complete {
      super.withFixture(test) // To be stackable, must call super.withFixture
    } lastly {
      builderActor ! Clear
    }
  }
}

trait Buffer extends AsyncTestSuiteMixin { this: AsyncTestSuite =>

  final val bufferActor = new StringBufferActor

  abstract override def withFixture(test: NoArgAsyncTest) = {
    complete {
      super.withFixture(test) // To be stackable, must call super.withFixture
    } lastly {
      bufferActor ! Clear
    }
  }
}

class ExampleSpec extends AsyncFeatureSpec with Builder with Buffer {

  Feature("Simplicity") {
    Scenario("User needs to read test code written by others") {
      builderActor ! Append("encourage clear code!")
      val futureString = builderActor ? GetValue
      val futureList = bufferActor ? GetValue
      val futurePair: Future[(String, List[String])] = futureString zip futureList
      futurePair map { case (str, lst) =>
        assert(str == "ScalaTest is designed to encourage clear code!")
        assert(lst.isEmpty)
        bufferActor ! Append("sweet")
        succeed
      }
    }

    Scenario("User needs to understand what the tests are doing") {
      builderActor ! Append("be easy to reason about!")
      val futureString = builderActor ? GetValue
      val futureList = bufferActor ? GetValue
      val futurePair: Future[(String, List[String])] = futureString zip futureList
      futurePair map { case (str, lst) =>
        assert(str == "ScalaTest is designed to be easy to reason about!")
        assert(lst.isEmpty)
        bufferActor ! Append("awesome")
        succeed
      }
    }
  }
}

By mixing in both the Builder and Buffer traits, ExampleSpec gets both fixtures, which will be initialized before each test and cleaned up after. The order the traits are mixed together determines the order of execution. In this case, Builder is “super” to Buffer. If you wanted Buffer to be “super” to Builder, you need only switch the order you mix them together, like this:

class Example2Spec extends AsyncFeatureSpec with Buffer with Builder

If you only need one fixture you mix in only that trait:

class Example3Spec extends AsyncFeatureSpec with Builder

Another way to create stackable fixture traits is by extending the BeforeAndAfterEach and/or BeforeAndAfterAll traits. BeforeAndAfterEach has a beforeEach method that will be run before each test (like JUnit's setUp), and an afterEach method that will be run after (like JUnit's tearDown). Similarly, BeforeAndAfterAll has a beforeAll method that will be run before all tests, and an afterAll method that will be run after all tests. Here's what the previously shown example would look like if it were rewritten to use the BeforeAndAfterEach methods instead of withFixture:

package org.scalatest.examples.asyncfeaturespec.composingbeforeandaftereach

import org.scalatest._
import org.scalatest.BeforeAndAfterEach
import collection.mutable.ListBuffer
import scala.concurrent.Future
import scala.concurrent.ExecutionContext

// Defining actor messages
sealed abstract class StringOp
case object Clear extends StringOp
case class Append(value: String) extends StringOp
case object GetValue

class StringBuilderActor { // Simulating an actor
  private final val sb = new StringBuilder
  def !(op: StringOp): Unit =
    synchronized {
      op match {
        case Append(value) => sb.append(value)
        case Clear => sb.clear()
      }
    }
  def ?(get: GetValue.type)(implicit c: ExecutionContext): Future[String] =
    Future {
      synchronized { sb.toString }
    }
}

class StringBufferActor {
  private final val buf = ListBuffer.empty[String]
  def !(op: StringOp): Unit =
    synchronized {
      op match {
        case Append(value) => buf += value
        case Clear => buf.clear()
      }
    }
  def ?(get: GetValue.type)(implicit c: ExecutionContext): Future[List[String]] =
    Future {
      synchronized { buf.toList }
    }
}

trait Builder extends BeforeAndAfterEach { this: Suite =>

  final val builderActor = new StringBuilderActor

  override def beforeEach() {
    builderActor ! Append("ScalaTest is designed to ")
    super.beforeEach() // To be stackable, must call super.beforeEach
  }

  override def afterEach() {
    try super.afterEach() // To be stackable, must call super.afterEach
    finally builderActor ! Clear
  }
}

trait Buffer extends BeforeAndAfterEach { this: Suite =>

  final val bufferActor = new StringBufferActor

  override def afterEach() {
    try super.afterEach() // To be stackable, must call super.afterEach
    finally bufferActor ! Clear
  }
}

class ExampleSpec extends featurespec.AsyncFeatureSpec with Builder with Buffer {

  Feature("Simplicity") {

    Scenario("User needs to read test code written by others") {
      builderActor ! Append("encourage clear code!")
      val futureString = builderActor ? GetValue
      val futureList = bufferActor ? GetValue
      val futurePair: Future[(String, List[String])] = futureString zip futureList
      futurePair map { case (str, lst) =>
        assert(str == "ScalaTest is designed to encourage clear code!")
        assert(lst.isEmpty)
        bufferActor ! Append("sweet")
        succeed
      }
    }

    Scenario("User needs to understand what the tests are doing") {
      builderActor ! Append("be easy to reason about!")
      val futureString = builderActor ? GetValue
      val futureList = bufferActor ? GetValue
      val futurePair: Future[(String, List[String])] = futureString zip futureList
      futurePair map { case (str, lst) =>
        assert(str == "ScalaTest is designed to be easy to reason about!")
        assert(lst.isEmpty)
        bufferActor ! Append("awesome")
        succeed
      }
    }
  }
}

To get the same ordering as withFixture, place your super.beforeEach call at the end of each beforeEach method, and the super.afterEach call at the beginning of each afterEach method, as shown in the previous example. It is a good idea to invoke super.afterEach in a try block and perform cleanup in a finally clause, as shown in the previous example, because this ensures the cleanup code is performed even if super.afterEach throws an exception.

The difference between stacking traits that extend BeforeAndAfterEach versus traits that implement withFixture is that setup and cleanup code happens before and after the test in BeforeAndAfterEach, but at the beginning and end of the test in withFixture. Thus if a withFixture method completes abruptly with an exception, it is considered a failed test. By contrast, if any of the beforeEach or afterEach methods of BeforeAndAfterEach complete abruptly, it is considered an aborted suite, which will result in a SuiteAborted event.

Shared tests

Sometimes you may want to run the same test code on different fixture objects. In other words, you may want to write tests that are "shared" by different fixture objects. To accomplish this in an AsyncFeatureSpec, you first place shared tests in behavior functions. These behavior functions will be invoked during the construction phase of any AsyncFeatureSpec that uses them, so that the tests they contain will be registered as tests in that AsyncFeatureSpec. For example, given this StackActor class:

package org.scalatest.examples.asyncfeaturespec.sharedtests

import scala.collection.mutable.ListBuffer
import scala.concurrent.Future
import scala.concurrent.ExecutionContext

// Stack operations
case class Push[T](value: T)
sealed abstract class StackOp
case object Pop extends StackOp
case object Peek extends StackOp
case object Size extends StackOp

// Stack info
case class StackInfo[T](top: Option[T], size: Int, max: Int) {
  require(size >= 0, "size was less than zero")
  require(max >= size, "max was less than size")
  val isFull: Boolean = size == max
  val isEmpty: Boolean = size == 0
}

class StackActor[T](Max: Int, name: String) {

  private final val buf = new ListBuffer[T]

  def !(push: Push[T]): Unit =
    synchronized {
      if (buf.size != Max)
        buf.prepend(push.value)
      else
        throw new IllegalStateException("can't push onto a full stack")
    }

  def ?(op: StackOp)(implicit c: ExecutionContext): Future[StackInfo[T]] =
    synchronized {
      op match {
        case Pop =>
          Future {
            if (buf.size != 0)
              StackInfo(Some(buf.remove(0)), buf.size, Max)
            else
              throw new IllegalStateException("can't pop an empty stack")
          }
        case Peek =>
          Future {
            if (buf.size != 0)
              StackInfo(Some(buf(0)), buf.size, Max)
            else
              throw new IllegalStateException("can't peek an empty stack")
          }
        case Size =>
          Future { StackInfo(None, buf.size, Max) }
      }
    }

  override def toString: String = name
}

You may want to test the stack represented by the StackActor class in different states: empty, full, with one item, with one item less than capacity, etc. You may find you have several tests that make sense any time the stack is non-empty. Thus you'd ideally want to run those same tests for three stack fixture objects: a full stack, a stack with a one item, and a stack with one item less than capacity. With shared tests, you can factor these tests out into a behavior function, into which you pass the stack fixture to use when running the tests. So in your AsyncFeatureSpec for StackActor, you'd invoke the behavior function three times, passing in each of the three stack fixtures so that the shared tests are run for all three fixtures.

You can define a behavior function that encapsulates these shared tests inside the AsyncFeatureSpec that uses them. If they are shared between different AsyncFeatureSpecs, however, you could also define them in a separate trait that is mixed into each AsyncFeatureSpec that uses them. For example, here the nonEmptyStackActor behavior function (in this case, a behavior method) is defined in a trait along with another method containing shared tests for non-full stacks:

import org.scalatest.featurespec.AsyncFeatureSpec

trait AsyncFeatureSpecStackBehaviors { this: AsyncFeatureSpec =>

  def nonEmptyStackActor(createNonEmptyStackActor: => StackActor[Int],
        lastItemAdded: Int, name: String): Unit = {

    Scenario("Size is fired at non-empty stack actor: " + name) {
      val stackActor = createNonEmptyStackActor
      val futureStackInfo = stackActor ? Size
      futureStackInfo map { stackInfo =>
        assert(!stackInfo.isEmpty)
      }
    }

    Scenario("Peek is fired at non-empty stack actor: " + name) {
      val stackActor = createNonEmptyStackActor
      val futurePair: Future[(StackInfo[Int], StackInfo[Int])] =
        for {
          beforePeek <- stackActor ? Size
          afterPeek <- stackActor ? Peek
        } yield (beforePeek, afterPeek)
      futurePair map { case (beforePeek, afterPeek) =>
        assert(afterPeek.top == Some(lastItemAdded))
        assert(afterPeek.size == beforePeek.size)
      }
    }

    Scenario("Pop is fired at non-empty stack actor: " + name) {
      val stackActor = createNonEmptyStackActor
      val futurePair: Future[(StackInfo[Int], StackInfo[Int])] =
        for {
          beforePop <- stackActor ? Size
          afterPop <- stackActor ? Pop
        } yield (beforePop, afterPop)
      futurePair map { case (beforePop, afterPop) =>
        assert(afterPop.top == Some(lastItemAdded))
        assert(afterPop.size == beforePop.size - 1)
      }
    }
  }

  def nonFullStackActor(createNonFullStackActor: => StackActor[Int], name: String): Unit = {

    Scenario("Size is fired at non-full stack actor: " + name) {
      val stackActor = createNonFullStackActor
      val futureStackInfo = stackActor ? Size
      futureStackInfo map { stackInfo =>
        assert(!stackInfo.isFull)
      }
    }

    Scenario("Push is fired at non-full stack actor: " + name) {
      val stackActor = createNonFullStackActor
      val futurePair: Future[(StackInfo[Int], StackInfo[Int])] =
        for {
          beforePush <- stackActor ? Size
          afterPush <- { stackActor ! Push(7); stackActor ? Peek }
        } yield (beforePush, afterPush)
      futurePair map { case (beforePush, afterPush) =>
        assert(afterPush.size == beforePush.size + 1)
        assert(afterPush.top == Some(7))
      }
    }
  }
}

Given these behavior functions, you could invoke them directly, but AsyncFeatureSpec offers a DSL for the purpose, which looks like this:

ScenariosFor(nonEmptyStackActor(almostEmptyStackActor, LastValuePushed, almostEmptyStackActorName))
ScenariosFor(nonFullStackActor(almostEmptyStackActor, almostEmptyStackActorName))

Here's an example:

class StackSpec extends AsyncFeatureSpec with AsyncFeatureSpecStackBehaviors {

  val Max = 10
  val LastValuePushed = Max - 1

  // Stack fixture creation methods
  val emptyStackActorName = "empty stack actor"
  def emptyStackActor = new StackActor[Int](Max, emptyStackActorName )

  val fullStackActorName = "full stack actor"
  def fullStackActor = {
    val stackActor = new StackActor[Int](Max, fullStackActorName )
    for (i <- 0 until Max)
      stackActor ! Push(i)
    stackActor
  }

  val almostEmptyStackActorName = "almost empty stack actor"
  def almostEmptyStackActor = {
    val stackActor = new StackActor[Int](Max, almostEmptyStackActorName )
    stackActor ! Push(LastValuePushed)
    stackActor
  }

  val almostFullStackActorName = "almost full stack actor"
  def almostFullStackActor = {
    val stackActor = new StackActor[Int](Max, almostFullStackActorName)
    for (i <- 1 to LastValuePushed)
      stackActor ! Push(i)
    stackActor
  }

  Feature("A Stack is pushed and popped") {

    Scenario("Size is fired at empty stack actor") {
      val stackActor = emptyStackActor
      val futureStackInfo = stackActor ? Size
      futureStackInfo map { stackInfo =>
        assert(stackInfo.isEmpty)
      }
    }

    Scenario("Peek is fired at empty stack actor") {
      recoverToSucceededIf[IllegalStateException] {
        emptyStackActor ? Peek
      }
    }

    Scenario("Pop is fired at empty stack actor") {
      recoverToSucceededIf[IllegalStateException] {
        emptyStackActor ? Pop
      }
    }

    ScenariosFor(nonEmptyStackActor(almostEmptyStackActor, LastValuePushed, almostEmptyStackActorName))
    ScenariosFor(nonFullStackActor(almostEmptyStackActor, almostEmptyStackActorName))

    ScenariosFor(nonEmptyStackActor(almostFullStackActor, LastValuePushed, almostFullStackActorName))
    ScenariosFor(nonFullStackActor(almostFullStackActor, almostFullStackActorName))

    Scenario("full is invoked on a full stack") {
      val stackActor = fullStackActor
      val futureStackInfo = stackActor ? Size
      futureStackInfo map { stackInfo =>
        assert(stackInfo.isFull)
      }
    }

    ScenariosFor(nonEmptyStackActor(fullStackActor, LastValuePushed, fullStackActorName))

    Scenario("push is invoked on a full stack") {
      val stackActor = fullStackActor
      assertThrows[IllegalStateException] {
        stackActor ! Push(10)
      }
    }
  }
}

If you load these classes into the Scala interpreter (with scalatest's JAR file on the class path), and execute it, you'll see:

scala> org.scalatest.run(new StackSpec)
StackSpec:
Feature: A Stack actor
- Scenario: Size is fired at empty stack actor
- Scenario: Peek is fired at empty stack actor
- Scenario: Pop is fired at empty stack actor
- Scenario: Size is fired at non-empty stack actor: almost empty stack actor
- Scenario: Peek is fired at non-empty stack actor: almost empty stack actor
- Scenario: Pop is fired at non-empty stack actor: almost empty stack actor
- Scenario: Size is fired at non-full stack actor: almost empty stack actor
- Scenario: Push is fired at non-full stack actor: almost empty stack actor
- Scenario: Size is fired at non-empty stack actor: almost full stack actor
- Scenario: Peek is fired at non-empty stack actor: almost full stack actor
- Scenario: Pop is fired at non-empty stack actor: almost full stack actor
- Scenario: Size is fired at non-full stack actor: almost full stack actor
- Scenario: Push is fired at non-full stack actor: almost full stack actor
- Scenario: Size is fired at full stack actor
- Scenario: Size is fired at non-empty stack actor: full stack actor
- Scenario: Peek is fired at non-empty stack actor: full stack actor
- Scenario: Pop is fired at non-empty stack actor: full stack actor
- Scenario: Push is fired at full stack actor

One thing to keep in mind when using shared tests is that in ScalaTest, each test in a suite must have a unique name. If you register the same tests repeatedly in the same suite, one problem you may encounter is an exception at runtime complaining that multiple tests are being registered with the same test name. Although in an AsyncFeatureSpec, the feature clause is a nesting construct analogous to AsyncFunSpec's describe clause, you many sometimes need to do a bit of extra work to ensure that the test names are unique. If a duplicate test name problem shows up in an AsyncFeatureSpec, you'll need to pass in a prefix or suffix string to add to each test name. You can call toString on the shared fixture object, or pass this string the same way you pass any other data needed by the shared tests. This is the approach taken by the previous AsyncFeatureSpecStackBehaviors example.

Given this AsyncFeatureSpecStackBehaviors trait, calling it with the almostEmptyStackActor fixture, like this:

ScenariosFor(nonEmptyStackActor(almostEmptyStackActor, LastValuePushed, almostEmptyStackActorName))

yields test names:

• Size is fired at non-empty stack actor: almost empty stack actor
• Peek is fired at non-empty stack actor: almost empty stack actor
• Pop is fired at non-empty stack actor: almost empty stack actor

Whereas calling it with the almostFullStackActor fixture, like this:

ScenariosFor(nonEmptyStack(almostFullStackActor, lastValuePushed, almostFullStackActorName))

yields different test names:

• Size is fired at non-empty stack actor: almost full stack actor
• Peek is fired at non-empty stack actor: almost full stack actor
• Pop is fired at non-empty stack actor: almost full stack actor