Download Lecture 1 for Chapter 9, Testing - Computer Science and Engineering

Using UML, Patterns, and Java
Object-Oriented Software Engineering
Chapter 11: Testing
These slides have been altered by Mohammad Ariful Hyder for CSE
321 using additional resources. (See references)
Outline





Terminology
Types of errors
Dealing with errors
Quality assurance vs Testing
Component Testing
 Unit testing
 Integration testing


Testing Strategy
Design Patterns & Testing
Bernd Bruegge & Allen H. Dutoit
Object-Oriented Software Engineering: Using UML, Patterns, and Java
2
Terminology

Reliability:
The measure of success with which the observed behavior of a
system confirms to some specification of its behavior. IEEE(The ability of a system or
component to perform its required functions under stated conditions for a specified period of time. It is often expressed
as a probability.)


Failure: Any deviation of the observed behavior from the
specified behavior
Error: The system is in a state such that further processing by
the system will lead to a failure.
 Stress or overload errors, Capacity or boundary errors, Timing errors, Throughput or
performance errors


Fault (Bug): The mechanical or algorithmic cause of an error.
Quality: IEEE(the degree to which a system, component, or process
meets customer or user needs or expectations.)
There are many different types of errors and different ways how
we can deal with them.
Bernd Bruegge & Allen H. Dutoit
Object-Oriented Software Engineering: Using UML, Patterns, and Java
3
Examples of Faults and Errors

Faults in the Interface
specification
 Mismatch between what the
client needs and what the
server offers
 Mismatch between
requirements and
implementation

Algorithmic Faults
 Missing initialization
 Branching errors (too soon,
too late)
 Missing test for nil
Bernd Bruegge & Allen H. Dutoit

Mechanical Faults (very
hard to find)
 Documentation does not
match actual conditions or
operating procedures

Errors




Stress or overload errors
Capacity or boundary errors
Timing errors
Throughput or performance
errors
Object-Oriented Software Engineering: Using UML, Patterns, and Java
4
Dealing with Errors

Verification:
 Assumes hypothetical environment that does not match real environment
 Proof might be buggy (omits important constraints; simply wrong)

Modular redundancy:
 Expensive

Declaring a bug to be a “feature”
 Bad practice

Patching
A patch (sometimes called a "fix") is a quick-repair job for a piece of programming. During a software product's beta test distribution or try-out
period and later after the product is formally released, problems (called bug) will almost invariably be found. A patch is the immediate solution that
is provided to users; it can sometimes be downloaded from the software maker's Web site. The patch is not necessarily the best solution for the
problem and the product developers often find a better solution to provide when they package the product for its next release. A patch is usually
developed and distributed as a replacement for or an insertion in compiled code (that is, in a binary file or object module). In larger operating
systems, a special program is provided to manage and keep track of the installation of patches.


Slows down performance
Testing (this lecture)
 Testing is never good enough
Bernd Bruegge & Allen H. Dutoit
Object-Oriented Software Engineering: Using UML, Patterns, and Java
5
Another View on How to Deal with Errors

Error prevention (before the system is released):
 Use good programming methodology to reduce complexity
 Use version control to prevent inconsistent system
 Apply verification to prevent algorithmic bugs

Error detection (while system is running):
 Testing: Create failures in a planned way
 Debugging: Start with an unplanned failuresMonitoring: Deliver
information about state. Find performance bugs

Error recovery (recover from failure once the system is released):
 Data base systems (atomic transactions)
 Modular redundancy
 Recovery blocks
Bernd Bruegge & Allen H. Dutoit
Object-Oriented Software Engineering: Using UML, Patterns, and Java
6
Some Observations

It is impossible to completely test any nontrivial module or any
system
 Theoretical limitations: Halting problem
 Practial limitations: Prohibitive in time and cost

Testing can only show the presence of bugs, not their absence
(Dijkstra).

Successful test cases:
 In unit and integration testing: A successful test case is one that
finds a previously undiscovered error.
 In acceptance testing: A successful test case is one that successfully
implements a requirement.
Bernd Bruegge & Allen H. Dutoit
Object-Oriented Software Engineering: Using UML, Patterns, and Java
7
Testing takes creativity

Testing often viewed as dirty work.

To develop an effective test, one must have:




Detailed understanding of the system
Knowledge of the testing techniques
Skill to apply these techniques in an effective and efficient manner
Testing is done best by independent testers
 We often develop a certain mental attitude that the program should
work in a certain way when in fact it does not.
 Programmers often stick to the data set that makes the program
work - "Don’t mess up my code!"
 A program often does not work when tried by somebody else. Don't
let this be the end-user.
Bernd Bruegge & Allen H. Dutoit
Object-Oriented Software Engineering: Using UML, Patterns, and Java
8
Testing Activities
Subsystem
Code
Subsystem
Code
Unit
Test
Unit
Test
Tested
Subsystem
Tested
Subsystem
Requirements
Analysis
Document
System
Design
Document
Integration
Test
Integrated
Subsystems
Functional
Test
User
Manual
Functioning
System
Tested Subsystem
Subsystem
Code
Unit
Test
Bernd Bruegge & Allen H. Dutoit
All tests by developer
Object-Oriented Software Engineering: Using UML, Patterns, and Java
9
Testing Activities continued
Client’s
Understanding
of Requirements
Global
Requirements
Validated
Functioning
System PerformanceSystem
Test
Accepted
System
Acceptance
Test
User
Environment
Installation
Test
Tests by client
Tests by developer
User’s understanding
Tests (?) by user
Bernd Bruegge & Allen H. Dutoit
Usable
System
Object-Oriented Software Engineering: Using UML, Patterns, and Java
System in
Use
10
Fault Handling Techniques
Fault Handling
Fault Avoidance
Design
Methodology
Verification
Fault Tolerance
Fault Detection
Atomic
Transactions
Reviews
Modular
Redundancy
Configuration
Management
Debugging
Testing
Unit
Testing
Bernd Bruegge & Allen H. Dutoit
Integration
Testing
System
Testing
Correctness
Debugging
Object-Oriented Software Engineering: Using UML, Patterns, and Java
Performance
Debugging
11
Quality Assurance encompasses Testing
Quality Assurance
Usability Testing
Scenario
Testing
Fault Avoidance
Verification
Prototype
Testing
Product
Testing
Fault Tolerance
Configuration
Management
Atomic
Transactions
Modular
Redundancy
Fault Detection
Reviews
Walkthrough
Inspection
Unit
Testing
Bernd Bruegge & Allen H. Dutoit
Debugging
Testing
Integration
Testing
System
Testing
Correctness
Debugging
Object-Oriented Software Engineering: Using UML, Patterns, and Java
Performance
Debugging
12
Types of Testing

Unit Testing:
 Perfomed on individual method or class
 Carried out by developers
 Goal: Confirm that method or class is correctly coded and carries
out the intended functionality
 Sometimes refers to testing on a subsystem

Integration Testing:
 Groups of subsystems (collection of classes) and eventually the
entire system
 Carried out by developers
 Goal: Test the interface among the subsystems
Bernd Bruegge & Allen H. Dutoit
Object-Oriented Software Engineering: Using UML, Patterns, and Java
13
System Testing

System Testing:
 The entire system
 Carried out by developers
 Goal: Determine if the system meets the requirements (functional
and global)

Acceptance Testing:
 Evaluates the system delivered by developers
 Carried out by the client. May involve executing typical
transactions on site on a trial basis
 Goal: Demonstrate that the system meets customer requirements
and is ready to use

Implementation (Coding) and testing go hand in hand
Bernd Bruegge & Allen H. Dutoit
Object-Oriented Software Engineering: Using UML, Patterns, and Java
14
Unit Testing

Informal:
 Incremental coding

Static Analysis:





Hand execution: Reading the source code (This is called desk checking)
Walkthrough (informal presentation to others)
Code Inspection (formal presentation to others)
Automated Tools checking for
 syntactic and semantic errors
 departure from coding standards
Dynamic Analysis:
 Black-box testing (Test the input/output behavior)
 White-box or Glass-box testing (Test the internal logic of the subsystem or
object)
 Data-structure based testing (Data types determine test cases)
Bernd Bruegge & Allen H. Dutoit
Object-Oriented Software Engineering: Using UML, Patterns, and Java
15
Black-box Testing

Focus: I/O behavior. If for any given input, we can predict the
output, then the module passes the test.
 Almost always impossible to generate all possible inputs
("test cases")

Goal: Reduce number of test cases by equivalence partitioning:
 Divide input conditions into equivalence classes
 Choose test cases for each equivalence class.

Example: If an object is supposed to accept a negative number,
testing one negative number is enough
Bernd Bruegge & Allen H. Dutoit
Object-Oriented Software Engineering: Using UML, Patterns, and Java
16
Black-box Testing (Continued)

Selection of equivalence classes (No rules, only guidelines):
 Input is valid across range of values. Select test cases from 3 equivalence
classes:



Below the range
Within the range
Above the range
 Input is valid if it is from a discrete set. Select test cases from 2 equivalence
classes:


Valid set member
Invalid set member
 Input is valid if it is the Boolean value true or false. Select


a true value
a false value
 Input is valid if it is an exact value of a particular data type. Select


the value
any other value of that type
Bernd Bruegge & Allen H. Dutoit
Object-Oriented Software Engineering: Using UML, Patterns, and Java
17
More on Equivalence Classes

Example
 The specifications for a DBMS state that the product must
handle any number of records between 1 and 16,383 (214–1)
 If the system can handle 34 records and 14,870 records,
then it probably will work fine for 8,252 records

Basic Idea
 If the system works for any one test case in the range
(1..16,383), then it will probably work for any other test
case in the range
 Range (1..16,383) constitutes an equivalence class
Bernd Bruegge & Allen H. Dutoit
Object-Oriented Software Engineering: Using UML, Patterns, and Java
18
Equivalence Classes and Boundary Value Analysis

Principle of Equivalence Classes
 Any one member of an equivalence class is as good a test case as any
other member of the equivalence class

Range Example: (1..16,383)
 Defines three different equivalence classes:




Equivalence Class 1: Fewer than 1 record
Equivalence Class 2: Between 1 and 16,383 records
Equivalence Class 3: More than 16,383 records
Principle of Boundary Value Analysis
 Select test cases on or just to one side of the boundary of
equivalence classes

This greatly increases the probability of detecting a fault
Bernd Bruegge & Allen H. Dutoit
Object-Oriented Software Engineering: Using UML, Patterns, and Java
19
More on Database Example

Test case 1:

Test case 2:
Test case 3:
Test case 4:
Test case 5:
Test case 6:
Test case 7:





Bernd Bruegge & Allen H. Dutoit
0 records
Member of equivalence class 1 and
adjacent to boundary value
1 record
Boundary value
2 records
Adjacent to boundary value
723 records
Member of equivalence class 2
16,382 records Adjacent to boundary value
16,383 records Boundary value
16,384 records Member of equivalence class 3 and
adjacent to boundary value
Object-Oriented Software Engineering: Using UML, Patterns, and Java
20
Equivalence Testing of Output Specifications

We also need to perform equivalence testing of the output
specifications

Example:
In 2003, the minimum Social Security (OASDI) deduction from any
one paycheck was $0.00, and the maximum was $5,394.00
 Test cases must include input data that should result in deductions
of exactly $0.00 and exactly $5,394.00
 Also, test data that might result in deductions of less than $0.00 or
more than $5,394.00
Bernd Bruegge & Allen H. Dutoit
Object-Oriented Software Engineering: Using UML, Patterns, and Java
21
Overall Strategy

Equivalence classes together with boundary value analysis to
test both input specifications and output specifications
 This approach generates a small set of test data with the potential
of uncovering a large number of faults
Bernd Bruegge & Allen H. Dutoit
Object-Oriented Software Engineering: Using UML, Patterns, and Java
22
White-box Testing

Focus: Thoroughness (Coverage). Every statement in the
component is executed at least once.

Four types of white-box testing




Statement Testing
Branch Testing
Path Testing
Loop Testing
Bernd Bruegge & Allen H. Dutoit
Object-Oriented Software Engineering: Using UML, Patterns, and Java
23
White-box Testing (Continued)

Statement Testing:
 Every statement should be tested at least once.

Branch Testing (Conditional Testing):
 Make sure that each possible outcome from a condition is tested
at least once
if ( i = TRUE) printf("YES\n"); else printf("NO\n");
Test cases: 1) i = TRUE; 2) i = FALSE

Path Testing:
 Make sure all paths in the program are executed.

This is infeasible. (Combinatorial explosion)
Bernd Bruegge & Allen H. Dutoit
Object-Oriented Software Engineering: Using UML, Patterns, and Java
24
White-box Testing Example
FindMean(float Mean, FILE ScoreFile)
{SumOfScores = 0.0; NumberOfScores = 0; Mean = 0;
Read(ScoreFile, Score); /*Read in and sum the scores*/
while (!EOF(ScoreFile) {
if ( Score > 0.0 ) {
SumOfScores = SumOfScores + Score;
NumberOfScores++;
}
Read(ScoreFile, Score);
}
/* Compute the mean and print the result */
if (NumberOfScores > 0 ) {
Mean = SumOfScores/NumberOfScores;
printf("The mean score is %f \n", Mean);
} else
printf("No scores found in file\n");
}
Bernd Bruegge & Allen H. Dutoit
Object-Oriented Software Engineering: Using UML, Patterns, and Java
25
White Box Testing: Path Testing

Draw a Flow Graph (see example next slide)

Use McCabe’s Complexity Measure
 Definition



Number of predicates (expressions returning a T/F value) + 1
#Edges – #Nodes + 2
Number of regions in flow graph
 Purpose:


To determine a basis (the test cases)
To limit the number of test cases
Bernd Bruegge & Allen H. Dutoit
Object-Oriented Software Engineering: Using UML, Patterns, and Java
26
White-box Testing Example: Determining the Paths
FindMean (FILE ScoreFile)
{ float SumOfScores = 0.0;
int NumberOfScores = 0;
1
float Mean=0.0; float Score;
Read(ScoreFile, Score);
2 while (!EOF(ScoreFile) {
3 if (Score > 0.0 ) {
SumOfScores = SumOfScores + Score;
4
NumberOfScores++;
}
5
Read(ScoreFile, Score);
6
}
/* Compute the mean and print the result */
7 if (NumberOfScores > 0) {
Mean = SumOfScores / NumberOfScores;
printf(“ The mean score is %f\n”, Mean);
} else
printf (“No scores found in file\n”);
9
}
Bernd Bruegge & Allen H. Dutoit
Object-Oriented Software Engineering: Using UML, Patterns, and Java
8
27
Constructing the Logic Flow Diagram
Start
1
F
2
T
3
T
F
5
4
6
7
T
F
9
8
Exit
Bernd Bruegge & Allen H. Dutoit
Object-Oriented Software Engineering: Using UML, Patterns, and Java
28
Determine a Basis: Use example on slides 27 and 28

Determine McCabe’s Complexity Measure
 The easiest formula to evaluate is to read the code:
Number of predicates (3) + 1 = 4

Check your answer by determining the number of regions:
Number of closed regions (3) + number of outside regions (always 1) = 4

Determine a basis (a linearly independent set of paths): Will contain 4 paths.
 Use the following technique:


Take the shortest path.
Add a single predicate to the next path.
 Basis:




Path 1: 1, 2, 7, 9, Exit (Could have the number 10) //at end of file after reading
Path 2: 1, 2, 3, 5, 6, 2, 7, 9, Exit
// reads a value that is negative or zero;
// then at end of file
Path 3: 1, 2, 3, 4, 5, 6, 2, 7, 8, Exit
// reads a single value that is positive, so
// completes mean
Path 4: 1, 2, 7, 8, Exit:
// infeasible path
Bernd Bruegge & Allen H. Dutoit
Object-Oriented Software Engineering: Using UML, Patterns, and Java
29
Finding the Test Cases
Start
1
a (Covered by any data)
2
b
(Positive score) d
c
4
(At end of file)
e
f
5
6
(NumberOfScores >i 0
8
k
Bernd Bruegge & Allen H. Dutoit
(File must contain at least one value)
3
7
(Negative score)
h (Reached through g if either f
or e is reached)
g
j (NumberOfScores < 0)
9
Exit
l
Object-Oriented Software Engineering: Using UML, Patterns, and Java
30
Another Example: Binary search flow graph
Bernd Bruegge & Allen H. Dutoit
Object-Oriented Software Engineering: Using UML, Patterns, and Java
31
Independent paths






1, 2, 3, 4, 5, 14
1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 14
1, 2, 3, 4, 5, 6, 7, 11, 12, 5, …
1, 2, 3, 4, 6, 7, 2, 11, 13, 5, …
Test cases should be derived so that all of these paths are
executed
A dynamic program analyser may be used to check that paths
have been executed
Bernd Bruegge & Allen H. Dutoit
Object-Oriented Software Engineering: Using UML, Patterns, and Java
32
Comparison of White & Black-box Testing

White-box Testing:

 Potentially infinite number of
paths have to be tested
 White-box testing often tests
what is done, instead of what
should be done
 Cannot detect missing use cases

Black-box Testing:
 Potential combinatorical
explosion of test cases (valid &
invalid data)
 Often not clear whether the
selected test cases uncover a
particular error
 Does not discover extraneous
use cases ("features")
Bernd Bruegge & Allen H. Dutoit


Both types of testing are needed
White-box testing and black box
testing are the extreme ends of a
testing continuum.
Any choice of test case lies in
between and depends on the
following:




Number of possible logical paths
Nature of input data
Amount of computation
Complexity of algorithms and
data structures
Object-Oriented Software Engineering: Using UML, Patterns, and Java
33
The 4 Testing Steps
1. Select what has to be
measured
3. Develop test cases
 Analysis: Completeness of
requirements
 Design: tested for cohesion
 Implementation: Code tests
2. Decide how the testing is
done




Code inspection
Proofs (Design by Contract)
Black-box, white box,
Select integration testing
strategy (big bang, bottom
up, top down, sandwich)
Bernd Bruegge & Allen H. Dutoit
 A test case is a set of test
data or situations that will be
used to exercise the unit
(code, module, system) being
tested or about the attribute
being measured
4. Create the test oracle
 An oracle contains the
predicted results for a set of
test cases
 The test oracle has to be
written down before the
actual testing takes place
Object-Oriented Software Engineering: Using UML, Patterns, and Java
34
Review: Design Principles

Cohesion: The interaction (or relatedness) between items
within a model.
 Goal: High
 A module should have one purpose so that it is more likely to be
reusable and maintainable.
 Ex. In a class, the data and methods should be highly related.

Coupling: The interaction among modules
 Goal: Low
 Limit the coupling between two modules to promote reuse.
 Ex. The number of modules called by the calling module should be
7±2
Bernd Bruegge & Allen H. Dutoit
Object-Oriented Software Engineering: Using UML, Patterns, and Java
35
Guidance for Test Case Selection

Use analysis knowledge about
functional requirements (blackbox testing):
 Use cases
 Expected input data
 Invalid input data


Use implementation knowledge
about algorithms:
 Examples:
 Force division by zero
 Use sequence of test cases for
interrupt handler
Use design knowledge about
system structure, algorithms, data
structures (white-box testing):
 Control structures

Test branches, loops, ...
 Data structures

Test records fields, arrays, ...
Bernd Bruegge & Allen H. Dutoit
Object-Oriented Software Engineering: Using UML, Patterns, and Java
36
Unit-testing Heuristics
1. Create unit tests as soon as object
design is completed:
 Black-box test: Test the use
cases & functional model
 White-box test: Test the
dynamic model
 Data-structure test: Test the
object model
5. Create a test harness
 Test drivers and test stubs are
needed for integration testing
6. Describe the test oracle
 Often the result of the first
successfully executed test
2. Develop the test cases
 Goal: Find the minimal number
of test cases to cover as many
paths as possible
7. Execute the test cases
 Don’t forget regression testing
 Re-execute test cases every time a
change is made.
3. Cross-check the test cases to eliminate
duplicates
 Don't waste your time!
8. Compare the results of the test with the test
oracle
 Automate as much as possible
4. Desk check your source code
 Reduces testing time
Bernd Bruegge & Allen H. Dutoit
Object-Oriented Software Engineering: Using UML, Patterns, and Java
37