TestWorks/Windows C/C++
Take a Tour of TestWorks
TestWorking Scribble

TestWorks, a fully integrated suite of software testing tools, will organize and automate your software testing process. TestWorks offers an end-to-end solution that covers all aspects of your process life-cycle: development, execution, management, reporting, organization, call-pair coverage, branch coverage analysis ... more!

TestWorks can be applied to the full spectrum of functions needed for producing quality software quicker, better, and less expensively.

This page shows you a sampling of screens taken from a TestWorks Application Note TestWorking Scribble on Windows.

About Scribble

Scribble is a C++ program that illustrates in progressive "chapters" the main features in the VC++ language. It combines keyboard input, mouse clicking and double-clicking, and drag-type drawing capabilities and, perhaps just as important, is a very strong test of VC++.

The version used as an example here, "chapter 8", consists of 1433 lines of C++ organized into 102 modules/functions located into 9 separate source files. Compiled without the debugging features its executable occupies about 40 KB; when fully instrumented by TCAT it occupies about 44 KB.

RECORDING TESTS WITH CAPBAK/MSW

Main Menu for CAPBAK/MSW. 1. Main Menu for CAPBAK/MSW. Each test you record with CAPBAK/MSW starts by using this simple control menu. It functions just like a VCR.

HotKey Popup Menu for CAPBAK/MSW. 2. HotKey Popup Menu for CAPBAK/MSW. You control how you make your recording using either function keys or the HotKey popup, shown here. Recordings that include extra synchronization beyond the Automatic Output Synchronization that is built into the product can enhance the reliability of playback in unusual circumstances, or can be used to synchronize playback on events such as the arrival of a particular ASCII text to a particular area of a window.

Example of Recording Scribbling. 3. Example Recording of Scribbling on "Scribble" With CAPBAK/MSW. Here is a TrueTime mode recording of a steady hand (?) drawing into Scribble's free-form area. Playback of this recording can play at exactly the same speed it was recorded: you re-enact exactly the same session you recorded. Or, you can play back your recording at continuously variable increased or decreased playback rates.

Extraction of Text from Window Area with OCR. 4. Extraction of Text from Window Area Using Built-In OCR Feature of CAPBAK/MSW. Here is an extraction from a graphical screen, where the OCR engine has identified the terms "To print the document" by taking the image from the screen as a pixel region and reconverting it back to ASCII. You can set up tests to PASS or FAIL based on the ASCII content of the images you have recorded.

Image Comparison with Built In Differencer. 5. Image Map Comparison of Two Saved Images with Built-In Image Differencer. You can use the built-in image comparison feature, which can be programmed with masks to help you disregard unimportant areas of the images, to decide if your PASSes or FAILs.

Here in the sample image are two image snaps of the clock at two different times. They are shown to be different -- the differential image is non-blank -- on the right hand side output image. Image compares when the two images differ by one or more pixels are deemed non-identical and the corresponding test FAILs.

ORGANIZING TESTS WITH SMARTS/MSW

SMARTS Main Window. 6. SMARTS Main GUI. You use SMARTS to catalog all of your tests into a test tree, specified by a user controlled ATS file (not shown here), that organizes each test into groups, of groups of groups, of groups of groups of groups, etc. You start up SMARTS with this simple GUI.

Input to ATS Creation Wizard. 7. Test Tree Creation Wizard. SMARTS comes with a powerful Automated Test Script (ATS) Creation Wizard. You organize the structure of your test suite in a simple tabulated "tree" structure like that shown here. This structure can be changed as your test suite evolves without having to re-create the entire test suite.

Output for SMARTS use from ATS Creation Wizard. 8. Result of Test Tree Creation Wizard: A Complete Test Tree. The ATS Creation Wizard wizard converts your simple tree into the complete Test Tree. In this example all of the individual test sequences and associated comparison commands are automatically inserted into the Test Tree.

SMARTS Test Tree Display with Source. 9. Test Tree with SMARTS and the Run-Tests Window. Here is part of a test tree -- the test tree can be as large as you like. Also shown is the source of the ATS file, which is expressed as a set of simple "C" functions. You don't have to be a "C" programmer to use SMARTS; there's a utility provided that helps you generate the initial test tree script from a simple outline format.

SMARTS Report Window. 10. SMARTS Report Window. Once a set of tests has been set up and run -- in the background or overnight -- you can look at any individual test result, the test results for a group of tests, a group of groups of tests, and so on. Here you see a very simple report showing that one particular test PASSed.

ANALYZING TEST COVERAGE WITH TCAT C/C++ FOR WINDOWS

You use TCAT on a test suite or on individual tests that you run to find out how good the tests are. What you didn't test is probably where your errors lie; you probably should focus testing on those untested areas if you want to improve coverage.

TCAT C/C++ Integrated with VC6+. 11. TCAT C/C++ Integrated with MS-Visual C++ v6.0 Main Window.
This image shows the complete VC 6.n window with six TCAT C/C++ action icons (see below) installed.


Visual Studio ToolBar Icons 12. Visual Studio ToolBar Icons.
Shown above are the six TCAT C/C++ toolbar icons plus a SMARTS/MSW and a CAPBAK/MSW toolbar icons that have been added to the Visual Studio VC++ environment during the TCAT C/C++ installation process.
These icons let programmer/testers perform these functions:

12a. Configure TCAT: Select among alternative modes of instrumentation.

12b. Build Instrumented Application: Instrument the application before compiling.

12c. Run Instrumented Application: Run the instrumented and compiled application.

12d. Analyze Coverage: Analyze the coverage achieved by one or more tests (see below). (C1 is the name used in TestWorks for branch coverage; TCAT measures both C1 = branch coverage and S1 = call-pair coverage).

12e. Show Digraph: Bring up the Digraph display for the selected object (see below).

12f. Show CallTree: Bring up the CallTree display for the selected object (see below).

12g. Run SMARTS Application: Organizes and executes a collection of tests which may use recorded scripts that are executed by CAPBAK from SMARTS control.

12h. Run CAPBAK Application: Captures and plays back keystrokes, mouse movements, object manipulations, and user-selected images and ASCII passages (screen-scraping).

Example of Scribble (Chapter 8) Execution. 13. Example of Scribble (Chapter 8) During Execution After Instrumentation Here is a sample of execution of the instrumented scribble2 program. Note that the recording includes a combination of both TrueTime and ObjectMode operation. CAPBAK/MSW always tries to make a recording, reverting to TrueTime mode when it cannot identify an object. This maximizes a testers productivity by assuring that tests recorded are going to play back without editing or other manual intervention.

Open Workspace Dialog Box. 14. Open Workspace Dialog Box
Here is the selection of the scribble2 project. The particular project is being selected for instrumentation.

Project Setting Dialog Box. 15. Project Setting Dialog Box
Here is the display showing the current C/C++ project selection. The particular project is being selected for instrumentation.

TCAT C/C++ is a powerful instrumenter and coverage analyzer that gives you three different views of your test data.

Source Code from Coverage Report. 16. Source Code Displayed From Coverage Report
This is the basic coverage report showing the connection between the cover values and the source code listing.

TCAT's CallTree Display. 17. TCAT's CallTree Display. The calltree display shows the connections between the various functions in your build. TCAT C/C++ generates this information when it processes your code prior to execution of the instrumented test object.


You can navigate your code by clicking on edges. Here the user has clicked on the connection from the CScribView function to the OnPrepareDC function, the latter of which is highlighted for you in the original source file from your project.

TCAT's Digraph Display. 18. TCAT's Digraph Display for a Function. TCAT also shows you the structural details of your individual functions. Here the digraph is for the CScribView function. You can tell which edges are associated with the parts of your code by clicking on the display; the corresponding part of the program shows up in color.

Viewing Source Code From Digraph. 19. Viewing Associated Source Code from Digraph
This is digraph -- an indication of control-flow structure -- with an edge selected to reflect to the source code, shown highlighted.

TCAT Coverage Report for Project. 20. TCAT Coverage Report for Project. Here you see the main window for TCAT C/C++ showing coverage values for the current project. Coverage is reported at the branch level (C1 Coverage) and at the call-pair level (S1 Coverage). The data in the picture is shown for the current test and for all prior tests (the data is kept in an Archive file that you can update when you want to accumulate coverage from test to test.

The display starts with filenames. Click on a filename and the display expands to the functions in that filename. Click on a function name and the display expands to show the segment and call-pair numbers for that function. Click on the segment number or click on the call-pair number and you're led directly to the line in your code that gave rise to that segment or callpair. You can pinpoint exactly where you missed a feature, where tests need to be beefed up, and where, during the next instrumentation cycle you may choose to de-instrument some files because you already have enough coverage.