maXbox Starter 36 Software Testing

maXbox Starter 36
Start a Function Test
1.1 Measure Correctness and Stability
Today we step through the realm of testing and bug-finding and his visual
representation in metrics and audits.
The traditional approach to quality control and testing has been to apply it
after the development process completes. This approach is very limited as
uncovering defects or errors at a late stage can produce long delays with
high costs while the bugs are corrected, or can result in the publishing of a
low-quality software product.
Lets start with a first function and how to improve and test it:
function CountPos(const subtxt: string; Text: string): Integer;
begin
if (Length(subtxt)= 0) or (Length(Text)= 0) or (Pos(subtxt,Text)= 0) then
result:= 0
else
result:= (Length(Text)- Length(StringReplace(Text,subtxt,'',
[rfReplaceAll]))) div Length(subtxt);
end;
This function counts all sub-strings in a text or string. The point is that the
function uses another functions like Pos() and StringReplace(). Do we have
to test it also? No, but we need a reference value:
PrintF('CountPos: %d',[CountPos('max','this is max of maXbox a max numbermax')])
The result is 3, so we better extract our test-string in a constant:
Const TESTTEXT = 'this is max of maXbox a max numbermax';
PrintF('CountPos: %d',[CountPos('max',TESTTEXT)])

And with Assert we do automate that reference value (more on p. 5):
with procedure Assert(expr: Boolean; const msg: string); you set
an expression value which you expect to be true:
Assert(CountPos('max',TESTTEXT)=3, 'count pos assert')
 Find the position and count of all strings in another is the function to
test but is the function case sensitive or not?
Sometimes it is helpful to track your program's execution as your program
runs. Or maybe you want to see a variable's value without stopping
program execution at a breakpoint to test.
The OutputDebugString() function enables you to do exactly that. This
function is a convenient debugging tool that many programmers overlook,
primarily because of a general lack of discussion on the subject.
OutputDebugString(inttoStr(CountPos('max',TESTTEXT))+
'CountPos runs..');
That's all you have to do. Because Delphi, Lazarus or Free Pascal is
installed as the system debugger, any strings sent using OutputDebugString
will show up in the Event Log. You can place calls to OutputDebugString
anywhere you want in your code.
To view the value of a variable, you must format a string and send that
string to OutputDebugString, for example:
procedure GetOutputDebugString;
var
x : Integer;
S : string;
begin
{ Some code here...}
x:= CountPos('max',TESTTEXT);
S:= Format('X := %d', [X]);
OutputDebugString(PChar(S));
writeln(S)
end;
//OutputDebugString(PChar(Format('[%s][%s] %s',[aCaption, GetFormatDT(StartDT), aText])));
Also the Event Viewer is a Microsoft Management Console (MMC) snap-in
that enables you to browse and manage event logs. It is an indispensable
tool for monitoring the health of systems and troubleshooting issues when
they arise. Look at the last entry in the Event Log shown in the next
picture. That entry was generated using this code:
WriteToOSEventLog('mX3','maXbox','Details3',TESTTEXT);
2

When you use Event Viewer to troubleshoot or track a problem, you need
to locate events related to the problem, regardless of which event log they
appear in. Event Viewer enables you to filter for specific events across
multiple logs as of maXbox. We need values in
WriteToOSEventLog(const logName, logCaption,
logDetails: String; const logRawData: Str);
a logName, logCaption, logDetails and some RawData:
WriteToOSEventLog('mX3','maXbox','ID45:detail3',TESTTEXT);
The Application log (see picture) contains events logged by applications or
programs. For example, a database program might record a file error in
the application log. You as a developer decide which events to log.
To log a sound in a multimedia app for e.g. you have to set a song first:
playMP3(mp3path);
closeMP3;
DebugOutputString('closeMP3: '+ mp3path);
In a shortlink %windir%system32eventvwr.msc /s you start a device:
 Analytic or test events are published in high volume. They describe
program operation and indicate problems that can't be handled by user
tasks or intervention.
3

Lets go back to our function to test and build a test function too.
function CountPos(const subtxt: string; Text: string): Integer;
begin
if (Length(subtxt)= 0) or (Length(Text)= 0) or (Pos(subtxt,Text)= 0) then
result:= 0
else
result:= (Length(Text)- Length(StringReplace(Text,subtxt,'',
[rfReplaceAll]))) div Length(subtxt);
end;
Such a small function like above contains another 3 functions (yellow) and
before testing we should also check the understanding and the design of
the function. For example StringReplace() which replace one or more
substrings found within a string does have more parameters to test like
the [rfReplaceAll].
On the other side we can trust in a stable function of the runtime library
otherwise we loose time not to mention the frustration caused by a bug
that is difficult to find.
I asked a few lines above if the function to test is case-sensitive or not, do
you remember? Yes it is and we do have to consider that in our test-string
and test-function with Uppercase too.
-
function CountPosTest: boolean;
begin
result:= false;
if CountPos(Uppercase('max'),Uppercase(TESTTEXT))=4 then
result:= true;
end;
Each time we test (or change) our function we call the test-function with a
test-string and check the reference value which must be 4.
Normally you test a function by creating a test application or unit and if
something goes wrong or crashes you will run the test application under
the debugger and watching the values.
A debugger helps you find bugs or errors in your program at runtime. But
the debugging process isn't just for finding and fixing bugs - it's a develop-
ment and test tool as well.
Another great way to test a function is to compare the result with a
already well known standard function, for example CountStr():
Function CountStr(const ASearchFor, ASearchIn: string): Integer;
or check with a subroutine which both of them uses:
Function Unit SysUtils
StringReplace: Replace one or more substrings found within a string
4

function CountStr(const ASearchFor, ASearchIn: string): Integer;
var
Start: Integer;
begin
Result:= 0;
Start:= Pos(ASearchFor, ASearchIn);
while Start > 0 do begin
Inc(Result);
Start:= PosEx(ASearchFor, ASearchIn, Start+ 1);
end;
end;
Or you will look at the web and ask to a search machine: How do I find and
count the total amount of times that a particular word appears?1
Often, as programmers gain experience in building applications, they
develop a sixth sense for locating the cause of in-correctness or access
violations.
With asserts you can support that feeling what could crash. An assert
function is used to make sure that certain conditions which are assumed
to be true are never violated. Assert provides an opportunity to intercept
an unexpected condition and halt a program rather than allow execution
to continue under unanticipated conditions.
 If you encounter problems using any feature of Event Viewer, first
ensure that the Windows Event Log service is running. If you are working
with event subscriptions, ensure that the Windows Event Collector service
is running.
With this call in maXbox you can run the event log from the script:
ExecuteCommand('cmd','/c %windir%system32eventvwr.msc /s ')
Next we go to the topic of Unit Testing but not to generate units (this is
another story) but consuming as a test-function or protocol e.g. in maXbox
I do have some unit tests to check conditions before running a script!:
// test routines with assert or assert2 builds
SelftestPEM;
SelfTestCFundamentUtils;
SelfTestCFileUtils;
SelfTestCDateTime;
SelfTestCTimer;
SelfTestCRandom;
Writeln(' 6 Units Tested with success ')
ShowMessageBig(' 6 Units Tests with success!');
1
http://stackoverflow.com/questions/7283531/find-and-count-words-in-a-string-in-delphi
5

A unit is the smallest testable part of software. It usually has one or a few
inputs and usually a single output. In procedural programming a unit may
be an individual program, function, routine, procedure, etc. In object-
oriented programming, the smallest unit is a method, which may belong to
a base/ super class, abstract class or derived/ child class2
.
Lets say you have a program comprising of two units and the only test you
perform is system testing between Win and Linux:
SelfTestCFileUtils;
SelfTestCDateTime; as an extract in test-cases:
Assert2(FilePath('C', '..XY', 'AB', '') = 'AXYC', 'FilePath');
Assert2(PathCanonical('AB..CD........', '') = '', 'PathCanonical');
Assert2(UnixPathToWinPath('/c/d.f') = 'cd.f', 'UnixPathToWinPath');
Assert2(WinPathToUnixPath('cd.f') = '/c/d.f', 'WinPathToUnixPath');
A:= EncodeDateTime(2001, 09, 02, 12, 11, 10, 0);
Assert2(Month(A) = 9, 'EncodeDateTime');
S:= GMTTimeToRFC1123TimeA(A, True);
Assert2(S = '12:11:10 GMT','GMT');
S:= GMTDateTimeToRFC1123DateTimeA(A, True);
Assert2(S = 'Sun, 02 Sep 2001 12:11:10 GMT', 'GMTDateTimeToRFC1123DateTime');
Assert2(TickDeltaW($FFFFFFF6, 0) = 10,'TrickDelta');
Assert2(CPUClockFrequency > 0,'RTC Prepare');
Assert2(Length(RandomPassword(0, 0, True, True, True)) = 0,'random passwd');
Assert2(Length(RandomPassword(1, 1, True, True, True)) = 1,'random passwd');
CompareDateTime(DateOf(Now),DateOf(Now)+1)
 For me the important thing is to isolate the development environ-
ment from the test environment and make sure you are using a version
control system to keep track of your test scripts and constants.
V1.2.1 SelftestPEM;
V1.7.0 SelfTestCFundamentUtils;
V1.2.0 SelfTestCFileUtils;
V1.1.1 SelfTestCDateTime;
V1.2.5 SelfTestCTimer;
V0.9.2 SelfTestCRandom;
SingleCompareDelta = 1.0E-34;
DoubleCompareDelta = 1.0E-280;
{$IFDEF CLR}
ExtendedCompareDelta = DoubleCompareDelta;
{$ELSE}
ExtendedCompareDelta = 1.0E-4400;
{$ENDIF}
2
http://softwaretestingfundamentals.com/unit-testing/
6

// Default CompareDelta is set to SingleCompareDelta. This allows any type
// of floating-point value to be compared with any other.
// DefaultCompareDelta = SingleCompareDelta;
Const
Bytes1KB = 1024;
Bytes1MB = 1024 * Bytes1KB;
Bytes1GB = 1024 * Bytes1MB;
Bytes64KB = 64 * Bytes1KB;
Bytes64MB = 64 * Bytes1MB;
Bytes2GB = 2 * LongWord(Bytes1GB);
You see the screen with those test scripts notion and some calls with
external dependencies.
One of a question often arises is the following:
Why do I continue getting error messages even after I have written an
exception handler?
Answer: In its default settings, the Delphi IDE notifies you whenever an
exception occurs in your program.
What’s important to realize is that at that point, none of your program’s
exception-handling code has run yet. It’s all Delphi itself; its special status
as a debugger allows it to get first notification of any exception in your
program, even before your program knows about it.
Simply put, debugging when using exception handling can be a bit of a
pain. Each runtime an exception is raised, the debugger pauses program
execution at the except block just as if a breakpoint were set on that line.
If the except block is in your code, the execution point is displayed as it
would be if you had stopped at a breakpoint.
7

You can or must restart the program again by clicking the Run button, or
you can step through your code.
 You might not be able to continue execution of your program after an
exception is raised. Whether you can continue debugging depends on
what went wrong in your program.
If an exception passes through the tryblocks in the application code, the
application automatically calls the HandleException method. Unless the
exception object is EAbort, HandleException calls the OnException
handler, if one exists. Otherwise, it calls ShowException to display a
message box indicating an error occurred.
1.1.1 Performance Testing
A bit weird but also important is performance and tuning. Sometimes you
will need to call a function or message handler for a particular measure of
the time consuming:
with TStopWatch.Create() do begin
try
Start;
//TimeOutThisFunction() to test performance
Sleep(2); //2ms = 0.002 secs
Stop;
finally
Writeln('watchTime: ' +floatToStr(GetValueMSec/1000));
Writeln('watchTime: ' +GetTimeString);
Free; //TStopWatch Class
end;
Time has long been a major subject of philosophy, art, poetry and science.
Many fields use an operational definition in which the units of time are
defined. Scholars disagree on whether time itself can be measured or is
itself part of the measuring system3
.
Finally, if a high-resolution performance counter exists on the system, you
can use the QueryPerformanceFrequency Win API function to express the
frequency, in counts per second. The value of the count is processor
dependent!
QueryPerformanceFrequency returns the number of "ticks" per seconds -
it is the amount that the counter, QueryPerformanceCounter, will
increment in a second.
What about when you need to process millions of tree brunch leafs or
generate 10 quadrillions (10^16) of unique random numbers? In such
scenarios it is important that your code executes as quickly as possible.
3
A clock shows just the position of the sun – we name that time!
8

The calculation is almost self defining: cycles per second = cycles / sec s,
or, as reported, millions of cycles per sec s = cycles / microseconds. The
program sets priorities to high values before performing a calculation and
restores them after to increase the accuracy of the timing.
For example the voltage signals that an oscilloscope measures are
generically called waves, which is any repeating pattern (sound waves,
oceans waves, etc), and the signal displayed on the oscilloscope is called a
waveform. Those waveforms can be split up in a frequency spectrum too:
if QueryPerformanceFrequency(Frequency) then
QueryPerformanceCounter(Start1);
//Sleep(2000); //2000 millisecs = 2 secs
PerformanceDelayMS(2000000*10); //2000 micro secs = 2 milli secs
//2000*1000 = 2 seconds
QueryPerformanceCounter(Stop1);
Writeln('Delta: '+IntToStr(Stop1-Start1) + ' Freq: ' +
IntToStr(Frequency) + ' ' +
Format('%.6f', [(Stop1-Start1)/Frequency ])+ ' seconds')
 Just note that you have to check the boolean return value from
QueryPerformanceFrequency() which returns false if performance
counters are not implemented in your hardware.
PrintF('Delta: %d Freq: %d of %.6f micro secs!',[Stop1-Start1, Frequency,
(Stop1-Start1)/Frequency ])
By calling this function at the beginning and end of a section of code, an
application uses the counter as a high-resolution timer.
9

Time1:= Time;
PerformanceDelayMS(2000000*10); //2000 micro secs = 2 milli secs
PrintF('%d %s',[Trunc((Time-Time1)*24),
FormatDateTime('"h runtime:" nn:ss:zzz',Time-Time1)])
Important is to find any hot spots of an application and then measure time
or test the consuming cycles. Most of the time, you will be using the event
handler for notification that a particular event occurred. But this time is
user dependent, you can't accelerate a mouse move or something control
driven.
Take the OnMouseDown event, for example. If you handle the OnMouseDown
event, you are simply asking to be notified when the user clicks the
component with the mouse (remember that forms or dialogues are
components, too).
This picture below shows a profiler app in maXbox.
1.1.2 Code Review Integration
Code-coverage use.
So we have seen testing goes with automation to test. There are tools
which automate your tests driven by software metrics. With these tools at
hand, quality managers, developers and technical architects have a
complete workbench to rate application quality, perform quantitative
analysis and define corrective action. Those tools measure the amount of
test-cases and the metric is called code-coverage.
 The cost of fixing a defect detected during unit testing or code
metrics is lesser in comparison to that of defects detected at higher levels.
Unit Testing is the first level of testing and is performed and logged prior
to Integration Testing.
10

But most of the time we forget to check our code twice. To avoid this
problem, metric products provide support for the so called Continuous
Inspection paradigm by enabling real-time notifications when code quality
defects are introduced in an application.
This transparency ensures that internal quality assurance is an integral
part of the software development life-cycle, rather than an afterthought.
These features enable full support of a Continuous Inspection process and
produce three immediate benefits:
• They encourage developer adoption by enabling developers to
quickly identify code improvements and quick wins.
• They improve quality of code produced by increasing developer
knowledge and understanding of code quality issues.
• They reduce maintenance cost through early identification of
quality issues.
Every step in your development work-flow that can be automated should
be. You need this automation to be able to focus all your time on building
your product and providing value to your customers. Testing is one of
those steps where automation is key.
Test the script with F9 / F2 or press Compile. So far now we’ll open the
test-function example: 587_one_function_assert.txt
http://www.softwareschule.ch/examples/587_one_function_assert.txt
Hope you did already work with the Starter 24 on Clean Code topics:
https://bitbucket.org/max_kleiner/maxbox3/wiki/maXbox%20Tutorials
Conclusion Stepping through code is on e of the most basic
debugging operations, yet it still needs to be mentioned here. Sometimes
you fail to see the forest for the trees. (Just as sometimes authors of
programming books fail to include the obvious!) Reviewing the basics from
time to time can reveal something you were not previously aware of.
It is important to understand that when you catch an exception, code
execution continues after executing the except block. One of the
advantages to catching exceptions is that you can recover from the
exception and continue program execution.
Notice the Exit statement in a code snippet. In this case, you don't want
to continue code execution following the exception, so you exit the
procedure after handling the exception.
11

“If the brain were so simple we could understand it, we would be so simple
we couldn't.” Lyall Watson
Pic of script: 123_graphics_atom_save2.txt
Feedback @ max@kleiner.com
Literature: Kleiner et al., Patterns konkret, 2003, Software & Support
http://www.softwareschule.ch/download/codesign_2015.pdf
http://softwaretestingfundamentals.com/unit-testing/
http://www.sonarsource.com/products/features/continuous-inspection/
https://www.academia.edu/5129685/Software_Quality_with_SONAR
https://github.com/maxkleiner/maXbox3/releases
12

1.2 Appendix Assert Study with Assembler
PROCEDURE Assert(Cond: boolean; const Msg: string);
var
progSeg,
progOfs: word;
begin
asm
mov ax, [bp+04]
mov es, ax
mov ax, word ptr es:0
mov progSeg, ax
mov ax, [bp+02]
mov progOfs, ax
end;
if (Cond = FALSE) then begin
mmDebug.Lines.Add(Msg + ' at location ' +
IntToHex(progSeg, 4) +':' +IntToHex(progOfs, 4) );
ShowModal;
end;
end;
Check the script!
Script Example: maxbox3examples318_excel_export3
13

maXbox Starter 36 Software Testing

More Related Content

What's hot (20)

Similar to maXbox Starter 36 Software Testing (20)

More from Max Kleiner (20)

Recently uploaded (20)

maXbox Starter 36 Software Testing