International Business Machines Corporation

James Clive Stewart

Ian Gerald Craggs

Ian Derek Heritage

Malcolm A Couldwell

Michael Anthony Ricketts

Thomas Jack Geraint Bailey

This invention provides a method, computer program and system for updating of a model of a system under test for use in model-based testing. In one embodiment, the system includes a rules engine, a comparator and a message dictionary. The rules engine runs an action against the system under test and runs the action against a model under test. The comparator compares an output from the system under test with a corresponding output from the model under test, wherein the output from the model under test is selected from one of a plurality of output message records. The dictionary engine updates the output message record by changing the message record if the output of the model under test is different from the output of the system under test. If the output message record is empty, the dictionary engine adds an output message record based on the system under test.

Model-based testing of a graphical user interface

This invention provides a method, computer program and system for updating a model of a system under test for use in model-based testing. In one embodiment, the system includes a rules engine, a comparator and a message dictionary. The rules engine runs an action against the system under test and runs the action against a model under test. The comparator compares an output from the system under test with a corresponding output from the model under test, wherein the output from the model under test is selected from one of a plurality of output message records. The dictionary engine updates the output message record by changing the message record if the output of the model under test is different from the output of the system under test. If the output message record is empty, the dictionary engine adds an output message record based on the system under test.

Method and apparatus for model-based testing of a graphical user interface

A method and system are provided for testing a graphical user (GUI) (). The method includes applying a test step () to a GUI under test () and finding a control in the GUI under test () to which the test step is to be applied. A model () is built dynamically of representations of the contents of the GUI under test () as a new control is found. The method includes finding a collection of dependent controls and representing the collection of controls in the GUI model () as a generalised pattern of the controls. The method applies a test step to the model () simultaneously with the GUI under test () and compares the results of the test step from the GUI under test () and from the model (). If a test step is applied to the model and it is determined that a representation of the control to which the test step is to be applied is not provided in the model (), a building rule () is executed to add a representation of the control to the model ().

Method and system for graphical user interface testing

A method and system are provided for testing a graphical user interface (GUI). The method includes providing a pattern of a collection of generalized interface object types, searching an interface or a model of an interface for instances of collections of interface objects matching the pattern, and identifying a control provided by the interface represented by the collection of interface object types and its location in the interface hierarchy. A pattern includes a first generalized object type and a second generalized object type with a specified relative position in the interface to the first object type.

A method and apparatus for concurrency testing within a model-based testing environment is provided. One implementation involves concurrency testing within a model-based software system testing environment, by receiving concurrent test service requests; duplicating a model representing software under test, based on the number of requests; sending the concurrent requests to each of the duplicated models sequentially but in different orders; storing the outcomes of each of the models; comparing the outcomes from the models to the outcome from the software system under test; and if all comparisons fail, then indicating test failure.

Growing community of inventors

Evesham, United Kingdom

James Clive Stewart