nunit.core.interfaces

The TestRunner Interface allows client code, such as the NUnit console and gui runners, to load and run tests. This is the lowest level interface generally supported for running tests and is implemented by the RemoteTestRunner class in the NUnit core as well as by other classes running on the client side. The Load method is used to load a suite of tests from one or more assemblies, returning a tree of TestNodes to the caller. The CountTestCases family of methods returns the number of test cases in the loaded suite, either in its entirety or by using a filter to count a subset of tests. The Run family of methods performs a test run synchronously, returning a TestResult or TestResult[] to the caller. If provided, an EventListener interface will be notified of significant events in the running of the tests. A filter may be used to run a subset of the tests. BeginRun and EndRun provide a simplified form of the asynchronous invocation pattern used in many places within the .NET framework. Because the current implementation allows only one run to be in process at a time, an IAsyncResult is not used at this time. Methods to cancel a run and to wait for a run to complete are also provided. The result of the last run may be obtained by querying the TestResult property.

Load the assemblies in a test package

The test package to be loaded True if the tests were loaded successfully, otherwise false

Unload all tests previously loaded

Count Test Cases using a filter

The filter to apply The number of test cases found

Run selected tests and return a test result. The test is run synchronously, and the listener interface is notified as it progresses. Arguments control whether trace and log output are passed back to the caller.

Interface to receive EventListener notifications. The filter to apply when running the tests If true, trace output is returned as a TestOutput event The threshold for log output to be returned as a TestOutput event

Start a run of selected tests. The tests are run aynchronously and the listener interface is notified as it progresses. Arguments control whether trace and log output are passed back to the caller.

Wait for an asynchronous run to complete and return the result.

A TestResult for the entire run

Cancel the test run that is in progress. For a synchronous run, a client wanting to call this must create a separate run thread.

Wait for the test run in progress to complete. For a synchronous run, a client wanting to call this must create a separate run thread. In particular, a gui client calling this method is likely to hang, since events will not be able to invoke methods on the gui thread.

TestRunners are identified by an ID. So long as there is only one test runner or a single chain of test runners, the default id of 0 may be used. However, any client that creates multiple runners must ensure that each one has a unique ID in order to locate and run specific tests.

Running indicates whether a test is in progress. To retrieve the results from an asynchronous test run, wait till Running is false.

Returns information about loaded assemblies

The loaded test, converted to a tree of TestNodes so they can be serialized and marshalled to a remote client.

Result of the last test run.

TestName encapsulates all info needed to identify and locate a test that has been loaded by a runner. It consists of a three components: the simple name of the test, an int id that is unique to a given tree of tests and an int runner id that identifies the particular runner that holds the test instance.

ID that uniquely identifies the test

The simple name of the test, without qualification

The fully qualified name of the test

Parse a string representation of a TestName, returning a TestName.

The string to parse A TestName

Compares two TestNames for equality

the other TestID True if the two TestIDs are equal

Calculates a hashcode for this TestID

The hash code.

Override ToString() to display the UniqueName

Override the == operator

Override the != operator

Returns a duplicate of this TestName

Gets or sets the TestID that uniquely identifies this test

Gets the ID for the runner that created the test from the TestID, or returns -1 if the TestID is null.

Gets or sets the simple name of the test

Gets or sets the full (qualified) name of the test

Get the string representation of this test name, incorporating all the components of the name.

The ITestCaseBuilder interface is exposed by a class that knows how to build a test case from certain methods.

Examine the method and determine if it is suitable for this builder to use in building a TestCase. Note that returning false will cause the method to be ignored in loading the tests. If it is desired to load the method but label it as non-runnable, ignored, etc., then this method must return true.

The test method to examine True is the builder can use this method

Build a TestCase from the provided MethodInfo.

The method to be used as a test case A TestCase or null

ITestCaseBuilder2 extends ITestCaseBuilder with methods that include the suite for which the test case is being built. Test case builders not needing the suite can continue to implement ITestCaseBuilder.

Examine the method and determine if it is suitable for this builder to use in building a TestCase to be included in the suite being populated. Note that returning false will cause the method to be ignored in loading the tests. If it is desired to load the method but label it as non-runnable, ignored, etc., then this method must return true.

The test method to examine The suite being populated True is the builder can use this method

Build a TestCase from the provided MethodInfo for inclusion in the suite being constructed.

The method to be used as a test case The test suite being populated, or null A TestCase or null

The Addin class holds information about an addin.

Construct an Addin for a type.

The type to be used

Return true if two Addins have teh same type name

The other addin to be compared

Return a hash code for this addin

The name of the Addin

Brief description of what the Addin does

The type or types of extension provided, using one or more members of the ExtensionType enumeration.

The AssemblyQualifiedName of the type that implements the addin.

The status of the addin

Any message that clarifies the status of the Addin, such as an error message or an explanation of why the addin is disabled.

TestInfo holds common info about a test. It represents only a single test or a suite and contains no references to other tests. Since it is informational only, it can easily be passed around using .Net remoting. TestInfo is used directly in all EventListener events and in TestResults. It contains an ID, which can be used by a runner to locate the actual test. TestInfo also serves as the base class for TestNode, which adds hierarchical information and is used in client code to maintain a visible image of the structure of the tests.

Common interface supported by all representations of a test. Only includes informational fields. The Run method is specifically excluded to allow for data-only representations of a test.

Count the test cases that pass a filter. The result should match those that would execute when passing the same filter to Run.

The filter to apply The count of test cases

Gets the completely specified name of the test encapsulated in a TestName object.

Gets a string representing the type of test, e.g.: "Test Case"

Indicates whether the test can be run using the RunState enum.

Reason for not running the test, if applicable

Count of the test cases ( 1 if this is a test case )

Categories available for this test

Return the description field.

Return additional properties of the test

True if this is a suite

Gets the parent test of this test

For a test suite, the child tests or suites Null if this is not a test suite

TestName that identifies this test

Reason for not running the test

Number of test cases in this test or suite

True if this is a suite

The test description

A list of all the categories assigned to a test

A dictionary of properties, used to add information to tests without requiring the class to change.

Construct from an ITest

Test from which a TestNode is to be constructed

Construct as a parent to multiple tests.

The name to use for the new test An array of child tests

Construct given a test name

The TestName for the new test

Counts the test cases that would be run if this test were executed using the provided filter.

The filter to apply A count of test cases

Gets the completely specified name of the test encapsulated in a TestName object.

Gets a string representing the kind of test this object represents for display purposes.

The test description

Gets the RunState for this test

The reason for ignoring a test

Count of test cases in this test.

Gets the parent test of this test

Gets a list of the categories applied to this test

Gets a list of any child tests

True if this is a suite, false if a test case

Gets the Properties dictionary for this test

Test Class.

TestName that identifies this test

Indicates whether the test should be executed

Test suite containing this test, or null

A dictionary of properties, used to add information to tests without requiring the class to change.

Get the current apartment state of the test

Constructs a test given its name

The name of the test

Constructs a test given the path through the test hierarchy to its parent and a name.

The parent tests full name The name of the test

Constructs a test given a TestName object

The TestName for this test

Sets the runner id of a test and optionally its children

The runner id to be used True if all children should get the same id

Gets a count of test cases that would be run using the specified filter.

Runs the test under a particular filter, sending notifications to a listener.

An event listener to receive notifications A filter used in running the test

Compares this test to another test for sorting purposes

The other test Value of -1, 0 or +1 depending on whether the current test is less than, equal to or greater than the other test

Return true if the test requires a thread

Get the desired apartment state for running the test

Gets a boolean value indicating whether this test should run on it's own thread.

Gets the TestName of the test

Gets a string representing the kind of test that this object represents, for use in display.

Whether or not the test should be run

Reason for not running the test, if applicable

Gets a count of test cases represented by or contained under this test.

Gets a list of categories associated with this test.

Gets a description associated with this test.

Gets the property dictionary for this test

Indicates whether this test is a suite

Gets the parent test of this test

Gets the parent as a Test object. Used by the core to set the parent.

Gets this test's child tests

Gets the Type of the fixture used in running this test

Gets or sets a fixture object for running this test

The PropertyNames struct lists common property names, which are accessed by reflection in the NUnit core. This provides a modicum of type safety as opposed to using the strings directly.

Exception Type expected from a test

Exception Type expected from a test (pre-2.5)

FullName of the Exception Type expected from a test

FullName of the Exception Type expected from a test (pre-2.5)

ExpectedException Message

ExpectedException MatchType

Expected return result from test

Indicates whether there is an expected return result

Description of the test

Alternate test name

Arguments for the test

Indicates test case is ignored

Indicates test case is explicit

The reason a test case is ignored

Properties of the test

Categories of the test

Name of a category

Reason for not running a test

Flag indicating excluded test should be marked as Ignored

Name of an addin that must be present to run a test

The ResultState enum indicates the result of running a test

The result is inconclusive

The test was not runnable.

The test has been skipped.

The test has been ignored.

The test succeeded

The test failed

The test encountered an unexpected exception

The test was cancelled by the user

The FailureSite enum indicates the stage of a test in which an error or failure occured.

Failure in the test itself

Failure in the SetUp method

Failure in the TearDown method

Failure of a parent test

Failure of a child test

The IService interface is implemented by all Services.

Initialize the Service

Do any cleanup needed before terminating the service

DecoratorPriority wraps constants that may be used to represent the relative priority of TestDecorators. Decorators with a lower priority are applied first so that higher priority decorators wrap them. NOTE: This feature is subject to change.

The default priority, equivalent to Normal

Priority for Decorators that must apply first

Normal Decorator priority

Priority for Decorators that must apply last

The ITestDecorator interface is exposed by a class that knows how to enhance the functionality of a test case or suite by decorating it.

Examine the a Test and either return it as is, modify it or return a different TestCase.

The Test to be decorated The MethodInfo used to construct the test The resulting Test

Enumeration expressing the level of log messages to be captured by NUnit and sent to the runner. These happen to match the standard levels used by log4net, but will be translated for any other loggers we support.

No logging

Severe error beyond which continuation is not possible

Error that may allow continuation

A warning message

An informational message

Messages used for debugging

All of the preceding plus more detailled messages if supported

CategoryFilter is able to select or exclude tests based on their categories.

Interface to be implemented by filters applied to tests. The filter applies when running the test, after it has been loaded, since this is the only time an ITest exists.

Determine if a particular test passes the filter criteria. Pass may examine the parents and/or descendants of a test, depending on the semantics of the particular filter

The test to which the filter is applied True if the test passes the filter, otherwise false

Determine whether the test itself matches the filter criteria, without examining either parents or descendants.

The test to which the filter is applied True if the filter matches the any parent of the test

Indicates whether this is the EmptyFilter

Unique Empty filter.

Determine if a particular test passes the filter criteria. The default implementation checks the test itself, its parents and any descendants. Derived classes may override this method or any of the Match methods to change the behavior of the filter.

The test to which the filter is applied True if the test passes the filter, otherwise false

Determine whether the test itself matches the filter criteria, without examining either parents or descendants.

The test to which the filter is applied True if the filter matches the any parent of the test

Determine whether any ancestor of the test mateches the filter criteria

The test to which the filter is applied True if the filter matches the an ancestor of the test

Determine whether any descendant of the test matches the filter criteria.

The test to be matched True if at least one descendant matches the filter criteria

Indicates whether this is the EmptyFilter

Nested class provides an empty filter - one that always returns true when called, unless the test is marked explicit.

Construct an empty CategoryFilter

Construct a CategoryFilter using a single category name

A category name

Construct a CategoryFilter using an array of category names

An array of category names

Add a category name to the filter

A category name

Check whether the filter matches a test

The test to be matched

Return the string representation of a category filter

Gets the list of categories from this filter

ParameterSet encapsulates method arguments and other selected parameters needed for constructing a parameterized test case.

The ITestCaseData interface is implemented by a class that is able to return complete testcases for use by a parameterized test method. NOTE: This interface is used in both the framework and the core, even though that results in two different types. However, sharing the source code guarantees that the various implementations will be compatible and that the core is able to reflect successfully over the framework implementations of ITestCaseData.

Gets the argument list to be provided to the test

Gets the expected result

Indicates whether a result has been specified. This is necessary because the result may be null, so it's value cannot be checked.

Gets the expected exception Type

Gets the FullName of the expected exception

Gets the name to be used for the test

Gets the description of the test

Gets a value indicating whether this is ignored.

true if ignored; otherwise, false.

Gets a value indicating whether this is explicit.

true if explicit; otherwise, false.

Gets the ignore reason.

The ignore reason.

A dictionary of properties, used to add information to tests without requiring the class to change.

Construct a non-runnable ParameterSet, specifying the provider excetpion that made it invalid.

Construct an empty parameter set, which defaults to being Runnable.

Constructs a ParameterSet from another object, accessing properties by reflection. The object must expose at least an Arguments property in order for the test to be runnable.

The RunState for this set of parameters.

Holds any exception thrown by the parameter provider

The arguments to be used in running the test, which must match the method signature.

The original arguments supplied by the user, used for display purposes.

The Type of any exception that is expected.

The FullName of any exception that is expected

The Message of any exception that is expected

Gets or sets the type of match to be performed on the expected message

The expected result of the test, which must match the method return type.

Returns true if an expected result has been specified for this parameter set.

A description to be applied to this test case

A name to be used for this test case in lieu of the standard generated name containing the argument list.

Gets or sets a value indicating whether this is ignored.

true if ignored; otherwise, false.

Gets or sets a value indicating whether this is explicit.

true if explicit; otherwise, false.

Gets or sets the ignore reason.

The ignore reason.

Gets a list of categories associated with this test.

Gets the property dictionary for this test

NUnitAddinAttribute is used to mark all add-ins. The marked class must implement the IAddin interface.

The name of this addin

A description for the addin

The type of extension provided

Default Constructor

Summary description for TestFramework.

The name of the framework

The file name of the assembly that defines the framwork

Constructs a TestFramwork object given its name and assembly name.

The ISuiteBuilder interface is exposed by a class that knows how to build a suite from one or more Types.

Examine the type and determine if it is suitable for this builder to use in building a TestSuite. Note that returning false will cause the type to be ignored in loading the tests. If it is desired to load the suite but label it as non-runnable, ignored, etc., then this method must return true.

The type of the fixture to be used True if the type can be used to build a TestSuite

Build a TestSuite from type provided.

The type of the fixture to be used A TestSuite

The IExtensionHost interface is implemented by each of NUnit's Extension hosts. Currently, there is only one host, which resides in the test domain.

Return an extension point by name, if present

The name of the extension point The extension point, if found, otherwise null

Get a list of the ExtensionPoints provided by this host.

Get an interface to the framework registry

Gets the ExtensionTypes supported by this host

An enum indicating the ExtensionTypes supported

The TestResult class represents the result of a test and is used to communicate results across AppDomains.

Indicates the result of the test

Indicates the location of a failure

The elapsed time for executing this test

The test that this result pertains to

The stacktrace at the point of failure

Message giving the reason for failure

List of child results

Number of asserts executed by this test

Construct a test result given a TestInfo

The test to be used

Construct a TestResult given an ITest

Construct a TestResult given a TestName

A TestName

Mark the test as succeeding

Mark the test as succeeding and set a message

Mark the test as ignored.

The reason the test was not run

Mark the test as ignored.

The ignore exception that was thrown

Mark the test as ignored.

The reason the test was not run Stack trace giving the location of the command

Mark the test as skipped.

The reason the test was not run

Mark the test a not runnable with a reason

The reason the test is invalid

Mark the test as not runnable due to a builder exception

The exception thrown by the builder or an addin

Set the result of the test

The ResultState to use in the result The reason the test was not run Stack trace giving the location of the command The location of the failure, if any

Set the result of the test

The ResultState to use in the result The reason the test was not run Stack trace giving the location of the command

Set the result of the test.

The ResultState to use in the result The exception that caused this result The site at which an error or failure occured

Mark the test as a failure due to an assertion having failed.

Message to display Stack trace giving the location of the failure

Mark the test as a failure due to an assertion having failed.

Message to display Stack trace giving the location of the failure The site of the failure

Marks the result as an error due to an exception thrown by the test.

The exception that was caught

Marks the result as an error due to an exception thrown from the indicated FailureSite.

The exception that was caught The site from which it was thrown

Add a child result

The child result to be added

Gets the ResultState of the test result, which indicates the success or failure of the test.

Gets the stage of the test in which a failure or error occured.

Indicates whether the test executed

Gets the name of the test result

Gets the full name of the test result

Gets the test associated with this result

Indicates whether the test ran successfully

Indicates whether the test failed

Indicates whether the test had an error (as opposed to a failure)

Gets a description associated with the test

Gets the elapsed time for running the test

Gets the message associated with a test failure or with not running the test

Gets any stacktrace associated with an error or failure.

Gets or sets the count of asserts executed when running the test.

Return true if this result has any child results

Gets a list of the child results of this TestResult

Represents a single point of extension for NUnit. Some extension points may accept only a single extension, while others may accept more than one at the same time.

Install an extension at this extension point. If the extension object does not meet the requirements for this extension point, an exception is thrown.

The extension to install

Removes an extension from this extension point. If the extension object is not present, the method returns without error.

Get the name of this extension point

Get the host that provides this extension point

Represents a single point of extension for NUnit. Some extension points may accept only a single extension, while others may accept more than one at the same time. This interface enhances IExtensionPoint by allowing specification of a priority order for applying addins.

Install an extension at this extension point specifying an integer priority value for the extension.If the extension object does not meet the requirements for this extension point, or if the extension point does not support the requested priority level, an exception is thrown.

The extension to install The priority level for this extension

TestID encapsulates a unique identifier for tests. As currently implemented, this is an integer and is unique within the AppDomain. TestID is one component of a TestName. We use this object, rather than a raw int, for two reasons: (1) to hide the implementation so it may be changed later if necessary and (2) so that the id may be null in a "weak" TestName.

The int key that distinguishes this test from all others created by the same runner.

Static value to seed ids. It's started at 1000 so any uninitialized ids will stand out.

Construct a new TestID

Construct a TestID with a given value. Used in parsing test names and in order to construct an artificial test node for aggregating multiple test runners.

Parse a TestID from it's string representation

Override of Equals method to allow comparison of TestIDs

Override of GetHashCode for TestIDs

Override ToString() to display the int id

Operator == override

Operator != override

Clone this TestID

An identical TestID

Summary description for NameFilter.

Construct an empty NameFilter

Construct a NameFilter for a single TestName

Add a TestName to a NameFilter

Check if a test matches the filter

The test to match True if it matches, false if not

The IDataPointProvider interface is used by extensions that provide data for a single test parameter.

Determine whether any data is available for a parameter.

A ParameterInfo representing one argument to a parameterized test True if any data is available, otherwise false.

Return an IEnumerable providing data for use with the supplied parameter.

A ParameterInfo representing one argument to a parameterized test An IEnumerable providing the required data

The IDataPointProvider2 interface extends IDataPointProvider by making the test fixture for which the test is being built available for use.

Determine whether any data is available for a parameter.

A ParameterInfo representing one argument to a parameterized test The test suite for which the test is being built True if any data is available, otherwise false.

Return an IEnumerable providing data for use with the supplied parameter.

A ParameterInfo representing one argument to a parameterized test The test suite for which the test is being built An IEnumerable providing the required data

SimpleName filter selects tests based on their name

Construct an empty SimpleNameFilter

Construct a SimpleNameFilter for a single name

The name the filter will recognize.

Construct a SimpleNameFilter for an array of names

The names the filter will recognize.

Add a name to a SimpleNameFilter

The name to be added.

Add an array of names to a SimpleNameFilter

The name to be added.

Check whether the filter matches a test

The test to be matched True if it matches, otherwise false

The IFrameworkRegistry allows extensions to register new frameworks or emulations of other frameworks.

The name of the framework The name of the assembly that the tests reference

The TestOutput class holds a unit of output from a test to either stdOut or stdErr

Construct with text and an ouput destination type

Text to be output Destination of output

Return string representation of the object for debugging

Get the text

Get the output type

Enum representing the output destination It uses combinable flags so that a given output control can accept multiple types of output. Normally, each individual output uses a single flag value.

Send output to stdOut

Send output to stdErr

Send output to Trace

Send output to Log

TestNode represents a single test or suite in the test hierarchy. TestNode holds common info needed about a test and represents a single node - either a test or a suite - in the hierarchy of tests. TestNode extends TestInfo, which holds all the information with the exception of the list of child classes. When constructed from a Test, TestNodes are always fully populated with child TestNodes. Like TestInfo, TestNode is purely a data class, and is not able to execute tests.

For a test suite, the child tests or suites Null if this is not a test suite

Construct from an ITest

Test from which a TestNode is to be constructed

Construct a TestNode given a TestName and an array of child tests.

The TestName of the new test An array of tests to be added as children of the new test

Gets the parent test of the current test

Array of child tests, null if this is a test case.

The RunState enum indicates whether a test can be executed. When used on a TestResult it may also indicate whether the test has been executed. See individual values for restrictions on use.

The test is not runnable.

The test is runnable. This value would normally not appear on a TestResult, since it would change to Executed.

The test can only be run explicitly. Would normally not appear on a TestResult, since it would change to Executed or Skipped.

The test has been skipped. This value may appear on a Test when certain attributes are used to skip the test.

The test has been ignored. May appear on a Test, when the IgnoreAttribute is used. Appears on a TestResult in that case or if the test is dynamically ignored.

Represents the manner in which test assemblies are distributed across processes.

Use the default setting, depending on the runner and the nature of the tests to be loaded.

Run tests directly in the NUnit process

Run tests in a single separate process

Run tests in a separate process per assembly

Represents the manner in which test assemblies use AppDomains to provide isolation

Use the default setting, depending on the runner and the nature of the tests to be loaded.

Don't create a test domain - run in the primary AppDomain

Run tests in a single separate test domain

Run tests in a separate domain per assembly

TestPackage holds information about a set of tests to be loaded by a TestRunner. It may represent a single assembly or a set of assemblies. It supports selection of a single test fixture for loading.

Construct a package, specifying the name of the package. If the package name is an assembly file type (dll or exe) then the resulting package represents a single assembly. Otherwise it is a container for multiple assemblies.

The name of the package

Construct a package, specifying the name to be used and a list of assemblies.

The package name, used to name the top-level test node The list of assemblies comprising the package

Return the value of a setting or a default.