ParallelTestExecution
Trait that causes that the tests of any suite it is mixed into to be run in parallel if a Distributor
is passed to runTests
.
ScalaTest's normal approach for running suites of tests in parallel is to run different suites in parallel, but the tests of any one suite sequentially. This approach should provide sufficient distribution of the work load in most cases, but some suites may encapsulate multiple long-running tests. Such suites may dominate the execution time of the run. If so, mixing in this trait into just those suites will allow their long-running tests to run in parallel with each other, thereby helping to reduce the total time required to run an entire run.
To make it easier for users to write tests that run in parallel, this trait runs each test in its own instance of the class. Running each test in its own instance enables tests to use the same instance vars
and mutable objects referenced from instance variables without needing to synchronize. Although ScalaTest provides functional approaches to factoring out common test code that can help avoid such issues, running each test in its own instance is an insurance policy that makes running tests in parallel easier and less error prone.
For the details on how ParallelTestExecution
works, see the documentation for methods run
, runTests
, and runTest
, which this trait overrides.
Note: This trait's implementation of runTest
is final
, to ensure that behavior related to individual tests are executed by the same thread that executes the actual test. This means, for example, that you won't be allowed to write ...with ParallelTestExecution with BeforeAndAfter
. Instead, you'd need to put ParallelTestExecution
last, as in: with BeforeAndAfter with ParallelTestExecution
. For more details, see the documentation for the runTest
method.
Attributes
- Graph
-
- Supertypes
- Self type
Members list
Value members
Abstract methods
Construct a new instance of this Suite
.
Construct a new instance of this Suite
.
This trait's implementation of runTests
invokes this method to create a new instance of this Suite
for each test. This trait's implementation of this method uses reflection to call this.getClass.newInstance
. This approach will succeed only if this Suite
's class has a public, no-arg constructor. In most cases this is likely to be true, because to be instantiated by ScalaTest's Runner
a Suite
needs a public, no-arg constructor. However, this will not be true of any Suite
defined as an inner class of another class or trait, because every constructor of an inner class type takes a reference to the enclosing instance. In such cases, and in cases where a Suite
class is explicitly defined without a public, no-arg constructor, you will need to override this method to construct a new instance of the Suite
in some other way.
Here's an example of how you could override newInstance
to construct a new instance of an inner class:
import org.scalatest.Suite class Outer { class InnerSuite extends Suite with ParallelTestExecution { def testOne() {} def testTwo() {} override def newInstance = new InnerSuite } }
Attributes
- Definition Classes
Concrete methods
Modifies the behavior of super.run
to facilitate parallel test execution.
Modifies the behavior of super.run
to facilitate parallel test execution.
This trait's implementation of this method only changes the supertrait implementation if both testName
and args.distributedTestSorter
are defined. If either testName
or args.distributedTestSorter
is empty, it simply invokes super.run
, passing along the same testName
and args
object.
If both testName
and args.distributedTestSorter
are defined, however, this trait's implementation of this method will create a "test-specific reporter" whose apply
method will invoke the apply
method of the DistributedTestSorter
, which takes a test name as well as the event. It will then invoke super.run
passing along the same testName
and an Args
object that is the same except with the original reporter replaced by the test-specific reporter.
Value parameters
- args
-
the
Args
for this run - testName
-
an optional name of one test to execute. If
None
, all relevant tests should be executed. I.e.,None
acts like a wildcard that means execute all relevant tests in thisSuite
.
Attributes
- Returns
-
a
Status
object that indicates when all tests and nested suites started by this method have completed, and whether or not a failure occurred.
Modifies the behavior of super.runTest
to facilitate parallel test execution.
Modifies the behavior of super.runTest
to facilitate parallel test execution.
This trait's implementation of this method only changes the supertrait implementation if args.distributor
is defined. If args.distributor
is empty, it simply invokes super.runTests
, passing along the same testName
and args
object.
If args.distributor
is defined, then it uses the args.runTestInNewInstance
flag to decide what to do. If runTestInNewInstance
is true
, this is the general instance responsible for running all tests, so it first notifies args.distributedTestSorter
(if defined) that it is distributing this test by invoking distributingTest
on it, passing in the testName
. Then it wraps a new instance of this class, obtained by invoking newInstance
in a suite whose run method will ensure that only the test whose name was passed to this method as testName
is executed. Finally, this trait's implementation of this method submits this wrapper suite to the distributor.
If runTestInNewInstance
is false
, this is the test-specific (distributed) instance, so this trait's implementation of this method simply invokes super.runTest
, passing along the same testName
and args
object, delegating responsibility for actually running the test to the super implementation. After super.runTest
returns (or completes abruptly by throwing an exception), it notifies args.distributedTestSorter
(if defined) that it has completed running the test by invoking completedTest
on it, passing in the testName
.
Note: this trait's implementation of this method is final
to ensure that any other desired runTest
behavior is executed by the same thread that executes the test. For example, if you were to mix in BeforeAndAfter
after ParallelTestExecution
, the before
and after
code would be executed by the general instance on the main test thread, rather than by the test-specific instance on the distributed thread. Marking this method final
ensures that traits like BeforeAndAfter
can only be "super" to ParallelTestExecution
and, therefore, that its before
and after
code will be run by the same distributed thread that runs the test itself.
Value parameters
- args
-
the
Args
for this run - testName
-
the name of one test to execute.
Attributes
- Returns
-
a
Status
object that indicates when the test started by this method has completed, and whether or not it failed .
Modifies the behavior of super.runTests
to facilitate parallel test execution.
Modifies the behavior of super.runTests
to facilitate parallel test execution.
This trait's implementation of this method always invokes super.runTests
to delegate to OneInstancePerTest
's implementation, but it may pass in a modified args
object. If args.runTestInNewInstance
is false
and args.distributor
is defined, this trait's implementation of this method will wrap the passed args.reporter
in a new Reporter
that can sort events fired by parallel tests back into sequential order, with a timeout. It will pass this new reporter to super.runTests
(in args.reporter
) as well as a defined DistributedTestSorter
(in args.distributedTestSorter) that can be used to communicate with the sorting reporter. Otherwise, if args.runTestInNewInstance
is true
or args.distributor
is empty, this trait's implementation of this method simply calls super.runTests
, passing along the same testName
and args
.
Value parameters
- args
-
the
Args
for this run - testName
-
an optional name of one test to run. If
None
, all relevant tests should be run. I.e.,None
acts like a wildcard that means run all relevant tests in thisSuite
.
Attributes
- Returns
-
a
Status
object that indicates when all tests started by this method have completed, and whether or not a failure occurred.
A maximum amount of time to wait for out-of-order events generated by running the tests of this Suite
in parallel while sorting the events back into a more user-friendly, sequential order.
A maximum amount of time to wait for out-of-order events generated by running the tests of this Suite
in parallel while sorting the events back into a more user-friendly, sequential order.
The default implementation of this method returns the value specified via -T
to Runner, or 2 seconds, if no -T
was supplied.
Attributes
- Returns
-
a maximum amount of time to wait for events while resorting them into sequential order
Inherited methods
The total number of tests that are expected to run when this Suite
's run
method is invoked.
The total number of tests that are expected to run when this Suite
's run
method is invoked.
Value parameters
- filter
-
a
Filter
with which to filter tests to count based on their tags
Attributes
- Inherited from:
- SuiteMixin
An immutable IndexedSeq
of this SuiteMixin
object's nested Suite
s. If this SuiteMixin
contains no nested Suite
s, this method returns an empty IndexedSeq
.
An immutable IndexedSeq
of this SuiteMixin
object's nested Suite
s. If this SuiteMixin
contains no nested Suite
s, this method returns an empty IndexedSeq
.
Attributes
- Inherited from:
- SuiteMixin
The fully qualified name of the class that can be used to rerun this suite.
The fully qualified name of the class that can be used to rerun this suite.
Attributes
- Inherited from:
- SuiteMixin
Runs zero to many of this suite's nested suites.
Runs zero to many of this suite's nested suites.
Value parameters
- args
-
the
Args
for this run
Attributes
- Returns
-
a
Status
object that indicates when all nested suites started by this method have completed, and whether or not a failure occurred. - Throws
-
NullArgumentException
if
args
isnull
. - Inherited from:
- SuiteMixin
A string ID for this Suite
that is intended to be unique among all suites reported during a run.
A string ID for this Suite
that is intended to be unique among all suites reported during a run.
The suite ID is intended to be unique, because ScalaTest does not enforce that it is unique. If it is not unique, then you may not be able to uniquely identify a particular test of a particular suite. This ability is used, for example, to dynamically tag tests as having failed in the previous run when rerunning only failed tests.
Attributes
- Returns
-
this
Suite
object's ID. - Inherited from:
- SuiteMixin
A user-friendly suite name for this Suite
.
A user-friendly suite name for this Suite
.
This trait's implementation of this method returns the simple name of this object's class. This trait's implementation of runNestedSuites
calls this method to obtain a name for Report
s to pass to the suiteStarting
, suiteCompleted
, and suiteAborted
methods of the Reporter
.
Attributes
- Returns
-
this
Suite
object's suite name. - Inherited from:
- SuiteMixin
A Map
whose keys are String
names of tagged tests and whose associated values are the Set
of tag names for the test. If a test has no associated tags, its name does not appear as a key in the returned Map
. If this Suite
contains no tests with tags, this method returns an empty Map
.
A Map
whose keys are String
names of tagged tests and whose associated values are the Set
of tag names for the test. If a test has no associated tags, its name does not appear as a key in the returned Map
. If this Suite
contains no tests with tags, this method returns an empty Map
.
Subclasses may override this method to define and/or discover tags in a custom manner, but overriding method implementations should never return an empty Set
as a value. If a test has no tags, its name should not appear as a key in the returned Map
.
Attributes
- Inherited from:
- SuiteMixin
Provides a TestData
instance for the passed test name, given the passed config map.
Provides a TestData
instance for the passed test name, given the passed config map.
This method is used to obtain a TestData
instance to pass to withFixture(NoArgTest)
and withFixture(OneArgTest)
and the beforeEach
and afterEach
methods of trait BeforeAndAfterEach
.
Value parameters
- testName
-
the name of the test for which to return a
TestData
instance - theConfigMap
-
the config map to include in the returned
TestData
Attributes
- Returns
-
a
TestData
instance for the specified test, which includes the specified config map - Inherited from:
- SuiteMixin
A Set
of test names. If this Suite
contains no tests, this method returns an empty Set
.
A Set
of test names. If this Suite
contains no tests, this method returns an empty Set
.
Although subclass and subtrait implementations of this method may return a Set
whose iterator produces String
test names in a well-defined order, the contract of this method does not required a defined order. Subclasses are free to implement this method and return test names in either a defined or undefined order.
Attributes
- Inherited from:
- SuiteMixin