Object Oriented Testing
Sharif University of Technology
Object Oriented Testing
l Find the greatest possible number of errors with a
manageable amount of effort applied over a realistic time
l The nature of OO programs changes both testing strategy
and testing tactics.
l Do we need less efforts for testing because of greater reuse of
l Answer is no! Binder argues that more testing is needed to
obtain high reliability in OO systems, since each reuse is a
new context of usage.
Testing Object-Oriented Applications:
Why is it Different?
l No sequential procedural executions
l No functional decomposition
l No structure charts to design integration testing
l Iterative O-O development and its impact on testing and
l To cover the strategies and tools associated with
object oriented testing
Analysis and Design Testing
l To discuss test plans and execution for projects
l When should testing begin?
l Analysis and Design:
Testing begins by evaluating the OOA and OOD models
How do we test OOA models (requirements and use cases)?
How do we test OOD models (class and sequence diagrams)?
Structured walk-throughs, prototypes
Formal reviews of correctness, completeness and consistency
How does OO make testing different from procedural programming?
Concept of a ‘unit’ broadens due to class encapsulation
Integration focuses on classes and their execution across a ‘thread’
or in the context of a use case scenario
Validation may still use conventional black box methods
Criteria for Completion of Testing
l When are we done testing? (Are we there yet?)
l How to answer this question is still a research question
1. One view: testing is never done… the burden simply shifts
from the developer to the customer
2. Or: testing is done when you run out of time or money
3. Or use a statistical model:
Assume that errors decay logarithmically with testing time
Measure the number of errors in a unit period
Fit these measurements to a logarithmic curve
Can then say: “with our experimentally valid statistical model we
have done sufficient testing to say that with 95% confidence the
probability of 1000 CPU hours of failure free operation is at least
l Issues to address for a successful software testing strategy:
Specify product requirements long before testing commences
For example: portability, maintainability, usability
Do so in a manner that is unambiguous and quantifiable
Understand the users of the software, with use cases
Develop a testing plan that emphasizes “rapid cycle testing”
Get quick feedback from a series of small incremental tests
Build robust software that is designed to test itself
Use assertions, exception handling and automated testing tools (Junit).
Conduct formal technical reviews to assess test strategy & test cases
“Who watches the watchers?”
Testing OOA and OOD Models (1)
l The review of OO analysis and design models is especially useful because the
same semantic constructs (e.g., classes, attributes, operations, messages) appear
at the analysis, design, and code level.
l Therefore, a problem in the definition of class attributes that is uncovered
during analysis will circumvent side effects that might occur if the problem
were not discovered until design or code (or even the next iteration of analysis).
l By fixing the number of attributes of a class during the first iteration of OOA,
the following problems may be avoided:
Creation of unnecessary subclasses.
Incorrect class relationships.
Improper behavior of the system or its classes.
l If the error is not uncovered during analysis and propagated further more
efforts needed during design or coding stages.
Testing OOA and OOD Models (2)
l Analysis and design models cannot be tested in the conventional sense, because
they cannot be executed.
l Formal technical review can be used to examine the correctness and consistency
of both analysis and design models.
Syntax: Each model is reviewed to ensure that proper modeling conventions
have been maintained.
Semantic: Must be judged based on the model’s conformance to the real world
problem domain by domain experts.
May be judged by considering the relationship among entities in the model.
Each class and its connections to other classes should be examined.
The Class-responsibility-collaboration model and object-relationship diagram
can be used.
l Early informal models are tested informally
l The criteria should be interpretive in the context of
iterative incremental approach
Model Testing Approaches
l Testing by comparison
compares each model to its predecessor or to previous
forms of the model
l Testing by inspection
uses checklists to make sure that the model meets certain
l Testing by verification
follows certain steps to assure completeness and
consistency of one part of the model with another
Examples of Analysis and Design
Models to be Tested
l CRC cards
l Class specifications
l Use cases
l State-Transition Models
State transition diagrams for classes, clusters, and
l Object network
Message sequence between methods in classes
l Transaction-Flow Models
Testing the Class Model
1. Revisit the Use Cases, CRC cards and UML class model.
Check that all collaborations are properly represented. Inspect
the description of each CRC index card to determine if a
delegated responsibility is part of the collaborator’s definition
Example: in a point of sale system.
A read credit card responsibility of a credit sale class is accomplished if
satisfied by a credit card collaborator
2. Invert connections to ensure that each collaborator asked for a
service is receiving requests from a reasonable source
Example: a credit card being asked for a purchase amount
l Have you tested your analysis and design?
If not, who will do it?
3. These steps are applied iteratively to each class and through
each evolution of the OOA model. 13
Testing OO Code
Class/Unit tests Integration
The Structured Testing Pyramid
Requirement Definition Systems Testing
Preliminary Design Integration Testing
Detailed Design Unit Testing
The OO Testing Pyramid
Use Case Analysis Use Case Testing
Cluster (Integration) Testing
Class Hierarchy Design
Test OOD Class Testing
Method Design Method Testing
l What is a unit?
A single, cohesive function?
A function whose code fits on one page?
The amount of code that can be written in 4 to 40 hours?
Code that is assigned to one person?
l We can no longer test a single operation in isolation but
rather as part of a class.
l In object-oriented programs, a unit is a method within a
l Smallest testable unit is the encapsulated class
Class Testing Process
How to test?
engineer test cases Why a loop?
Methods for Generating Test Cases
For Unit Testing
l Statement coverage
l Graph based
l All unit testing methods are also applicable to
testing methods within a class.
Class Test Case Design
Berard [BER93,94] proposes the following approach:
1. Identify each test case uniquely
- Associate test case explicitly with the class and/or method to be tested
2. State the purpose of the test
3. Each test case should contain:
a. list of specified states for the object that is to be tested
b. A list of messages and operations that will be exercised as a consequence of
c. A list of exceptions that may occur as the object is tested
d. A list of external conditions for setup (i.e., changes in the environment
external to the software that must exist in order to properly conduct the
e. Supplementary information that will aid in understanding or implementing
Automated unit testing tools facilitate these requirements
Challenges of Class Testing
l OO class is the target for test case design.
Difficult to obtain a snapshot of a class without building extra methods
which display the classes’ state
l Inheritance and polymorphism:
Each new context of use (subclass) requires re-testing because a method
may be implemented differently (polymorphism).
Other unaltered methods within the subclass may use the redefined method
and need to be tested
l How is class testing different from conventional testing?
- Conventional testing focuses on input-process-output, whereas class testing
focuses on each method, then designing sequences of methods to exercise
states of a class
l White box tests:
Basis path, condition, data flow and loop tests can all apply to individual
methods, but don’t test interactions between methods
Notes on OO Testing 1.
l Applicability of conventional test case design:
Although white-box testing is useful, the efforts might be redirected to
tests at a class level.
Use cases can provide useful input in the design of black-box and state-
l Fault-based testing: (Tests have high likelihood of uncovering faults)
Its planning begins with the analysis model.
If real faults in an OO system are perceived to be implausible then this
approach is really no better than any random testing technique.
If the analysis and design models can provide insight into what is likely to
go wrong, then fault-based testing can find significant numbers of
errors with relatively low efforts.
Notes on OO Testing 2.
l Impact of OO programming on Testing:
When an operation is invoked, it may be hard to tell exactly what code
It can be hard to determine the exact type or class of a parameter.
OO operations are smaller, more time needed for integration. So
integration faults become more plausible.
In 3 large scale studies: 50% of methods are less than 2 C++ statements or 4
Smalltalk statements (Law of Demeter connection)
l Scenario-based test design:
Concentrates on what the user does, not what the product does.
Capturing the tasks via use-cases that the user has to perform.
l Testing surface structure and deep structure
Surface structure: externally observable structure (end-user).
Deep structure: internal technical details of an OO program.
OOT Methods: Class Level 1
l Random testing
identify operations applicable to a class
define constraints on their use
identify a minimum test sequence
l an operation sequence that defines the minimum life history of the class
generate a variety of random (but valid) test sequences
l exercise other (more complex) class instance life histories
Operations: open, setup, deposit, withdraw, balance, summarize, creditlimit, close.
1. Open – setup – deposit – withdraw – close
2. Open – setup – deposit –[deposit | withdraw | balance | summarize] * –
withdraw – close. Generate random test sequences using this template
OOT Methods: Class Level 2
l Partition Testing
reduces the number of test cases required to test a class in much the same way as
equivalence partitioning for conventional software
l categorize and test operations based on their ability to change the state of a
class (e.g.: deposit, withdraw)
l categorize and test operations based on the attributes that they use ( e.g.:
l categorize and test operations based on the generic function each performs.
(e.g.: (Init OP: open , setup) (Comp. OP: deposit, withdraw) (queries: balance,
summarize, creditlimit) (Termination OP: close))
What Methods to Test Within Classes
l New methods: defined in the class under test and
not inherited or overloaded by methods in a
superclass: Complete testing
l Inherited methods: defined in a superclass of the
class under test: Retest only if the methods
interacts with new or redefined method.
l Redefined methods: defined in a superclass of but
redefined in the class under test : complete Retest
reusing tests from the superclass.
Class Testing Techniques
In addition to testing methods within a class
(either glass box or black box), the following three
techniques can be used to perform functional
testing for the class:
l A state-transition model describes the different
states and transitions of a class in the context of its
position in the inheritance hierarchy.
(OMT dynamic model)
l The state of an object is the combination of all the
attribute values and objects that the object
l An object may transition from a state to state as a
result of an event, which may yield an action.
Prospect Receive application Member
l Create test cases corresponding to each transition
path that represent a full object life cycle
l Make sure each transition is exercised at least
 Integration Testing
l OO does not have a hierarchical control structure so
conventional top-down and bottom-up integration tests
have little meaning
l Integrating operations one at a time into a class is often impossible
because of the direct and indirect interactions of the components that
make up the class.
l Integration applied three different incremental strategies:
Thread-based testing: integrates classes required to respond to one
input or event
Use-based testing: integrates classes required by one use case
Cluster testing: integrates classes required to demonstrate one
- What integration testing
strategies will you use? 31
Cluster (Integration) Testing:
Why is it Different?
l Event-Driven execution
l No functional decomposition
l The sequence of execution is not known
l Integration testing must be driven by how object
will behave dynamically
l Object composition introduces a new dimension of
Random Integration Testing
l Multiple Class Random Testing
1. For each client class, use the list of class methods to
generate a series of random test sequences.
Methods will send messages to other server classes.
2. For each message that is generated, determine the
collaborating class and the corresponding method in the
3. For each method in the server object (that has been
invoked by messages sent from the client object),
determine the messages that it transmits
4. For each of the messages, determine the next level of
methods that are invoked and incorporate these into the
Cluster (Integration) Testing
l A cluster is a collection of classes (possibly from
different systems) cooperating with each other via
l A cluster specification should include methods from
each class that will be accessed
l Cluster testing focuses on the interaction among the
instances of the classes in the cluster
l It assumes that each class has been tested individually
l Cluster testing is considered a second level of
Methods for Forming Clusters
l Function-based clustering
Based on requirements and use cases
Difficult to perform if requirements were not available
during the design phase
l Subject-based clustering
Based on subject areas that need to test separately
l Project Schedule-based clustering
l Contract-based clustering
Types of Errors Found During
l Messaging errors:
Failure to meet a requirement, i.e., no method to send or
receive a message
Incompatible method and message in sender and receiver
Incorrect event timing between object actions
Incorrect instantiation or destruction of objects
Types of Errors Found During
Integration Testing (cont.)
l User interface errors:
A given sequence of user actions does not have the
expected effect on the component.
The timing of events received from the user results in
incorrect functioning of the component
Techniques for Object-Oriented
l Message Quiescence
l Event Quiescence
l A Method/Message path (MM-Path) is a sequence
of method executions linked by messages.
l An MM-Path starts with a method and ends when
it reaches a method that does not issue a message
of its own, i.e., reaches a message Quiescence.
l An input port event followed by a set of MM-
Paths, and terminated by an output event is called
Atomic System Function (ASF).
INPUT PORT EVENT
A OUTPUT PORT EVENT
Class 1 A Class 3
ASF INPUT PORT ASF OUTPUT PORT
2 Class 2
ATM PIN Entry
l Customer enters card (event)
l Screen requesting PIN entry is displayed
l An interleaved sequence of digit key toughs with audible
and visual feedback
l Possibility of cancellation by customer
l System disposition (valid PIN or card retained)
ASF ends here ASF Starts here
Customer inserts card
Message is displayed
 Validation Testing
l Are we building the right product?
l Validation succeeds when software functions in a manner
that can be reasonably expected by the customer.
l Focus on user-visible actions and user-recognizable outputs
l Details of class connections disappear at this level
Use-case scenarios from the software requirements spec
Black-box testing to create a deficiency list
Acceptance tests through alpha (at developer’s site) and beta (at
customer’s site) testing with actual customers
l How will you validate your software product?
 System Testing
l Software may be part of a larger system. This often leads
to “finger pointing” by other system dev teams
l Finger pointing defence:
1. Design error-handling paths that test external information
2. Conduct a series of tests that simulate bad data
3. Record the results of tests to use as evidence
l Types of System Testing:
Recovery testing: how well and quickly does the system recover
Security testing: verify that protection mechanisms built into the
system will protect from unauthorized access (hackers,
disgruntled employees, fraudsters)
Stress testing: place abnormal load on the system
Performance testing: investigate the run-time performance within
the context of an integrated system 46
l All rules and methods of traditional systems
testing are also applicable applicable to object-
l A thread can be “re-defined” as a sequence of
method executions linked by messages in the object
l Use cases provide a more precise mechanism for
writing test cases.
Use Case # 1 of the Telephone Banking System:
Task: Making a balance transfer
User Class: Current customer
Frequency: one per week per customer
User Action System Response
User dials the number System plays greeting and ask for
User enters invalid account number System informs user and ask for
account number again
User enter a valid account number System asks for PIN#
l Junit at Junit.org
l CPPUnit on SourceForge.net,
l NUnit for .Net
l Rational Robot
OO Testing Methodology
l Jorgensen and Erickson propose 5 levels:
A Method Unit testing
Message Quiescence Integration
Event Quiescence Integration
Thread testing System
Thread interaction System
Stages of Test [RUP]
l Developer Testing
l Unit Test
l Integration Test
l System Test
l Acceptance Test
Test Planning [RUP]
l What iteration you are you in?
l What stage of test (unit test, integration test,
system test) you are performing?
l Types of test (functional, stress, volume,
performance, usability, distribution, and so on).
l Evaluation criteria used (code-based test coverage,
requirements-based test coverage, number of
defects, mean time between failure, and so on.)
l Testing techniques used (manual and automated)
Completion Criteria [RUP]
Manual/Automated Testing [RUP]