Software Testing (introduction)

advertisement
Software Analysis
and Testing
Paolo Tonella
Alessandro Marchetto
Cu D. Nguyen
Mariano Ceccato
Software Engineering (SE) unit, Fondazione Bruno Kessler (FBK)
Team
Paolo Tonella
Cu D. Nguyen
Mariano Ceccato
Alessandro Marchetto
Software Engineering research unit
http://se.fbk.eu
at Fondazione Bruno Kessler
http://www.fbk.eu
2
FBK
FBK (more than 350 researchers) is a research organization of the
Autonomous Province of Trento that promotes research in the areas of
science, technology and humanities.
FBK objectives are to: (i) conduct research that obtains recognition at an
international level; (ii) carry out applied research of strategic importance to
the province; (iii) publicize scientific results and promote economic
development; and (iv) encourage innovation throughout the province.
3
FBK – CIT
Three main areas of Information Technology:
- Engineering
- Content
- Interaction
FBK is organized into Research Units.
- Research units are research
groups that are above critical mass
but still manageable.
- Research Units include senior
and young researchers, postdocs,
PhD students, project managers,
system architects and programmers.
- Research Units are strongly
encouraged towards collaborations
and projects that can exploit
synergies
among
difference
competences.
4
Web site
Material for the course:
- Course objectives and program
- Course agenda
- Notes and slides of the lectures
- Project deliverables
- Exam rules and dates
http://selab.fbk.eu/swat
5
Lecture schedule
Wed.
Fri.
16.30 -18.30
14.30 -16.30
room A107
room A107
Please, come to the Lab lectures with your laptop or
ask us for having a laptop.
… Is any laptop needed?
6
Exam
The exam consists of an oral discussion of the project
carried out during the laboratory.


During the course we will work on a project.
The exam will consist of some questions on the project implementation
plus one question on the theory behind the project.
 Project deliverables include both implementation artifacts (Eclipse
project, test cases, etc.) and documentation artifacts (reports). Both are
mandatory for admission to the exam.
 The next lecture will be devoted to the presentation of the course
project.
Project must be delivered one week before
the exam date!
7
Objective

This course aims at teaching students how to analyze and test a
software system, when it is evolved to accommodate a set of
change requirements (e.g., adding new functionalities, bug fixing,
adaptation or restructuring, etc.), by executing a software project
which involves:



definition of acceptance tests for the change requirements;
definition of tests for the user interface;
definition of unit tests for the modules implemented to realize the
change requirements;
 application of adequacy testing criteria to the implemented modules;
 regression testing w.r.t. the preserved functionalities of the evolved
software
8
Program
Testing:

Background and Context:

Software maintenance and
evolution
 Code analysis
 Software testing
Acceptance Test





GUI testing
Unit test
Structural test (path and data flow)
Mutation test
Automated path generation and path
testing
 Regression test
 Test case prioritization
Laboratory:






program analysis and understanding
user interface test
acceptance test creation and execution
unit test creation and execution
coverage and mutation test
regression test
Questions:
- Which testing to apply?
- How to define test cases?
- How to automate test cases execution?
- How to decide when to stop testing?
Tools:
- Fitnesse
- Junit
- MuJava
- MuClipse
- Clover
- Jumble
- Emma
-…
9
Agenda
Intro & course project: Sep 14, 16, 21, 23
Acceptance testing: Sep 28, 30; Oct 5
GUI testing: Oct 7, 12
Unit testing: Oct 14, 19
Analysis and testing theory: Oct 21, 26, 28; Nov 2, 30
Debugging: Nov 4, 9, 11, 16
Coverage and mutation testing: Nov 18, 23, 25
Regression testing: Dec 2, 7
Advanced topic: Dec 14, 16
10
Software Maintenance
and
Software Testing
11
Software maintenance
Software maintenance has been defined as “the modification of a
software product after delivery to correct faults, to improve
performance or other attributes, or to adapt the product to a changed
environment.” (ANSI/IEEE, 1983)
Maintenance activities can be classified into four categories:

Perfective maintenance
 Adaptive maintenance
 Corrective maintenance
 Preventive maintenance
12
Software maintenance
Maintenance activities are difficult:





Continuing change.
Increasing complexity.
Fundamental law of program evolution.
Conservation of organizational stability (invariant work rate).
Conservation of familiarity (perceived complexity).
13
Software maintenance steps
Change
Request
Program
Understanding
Change location
identification
Change
implementation
Testing
Ripple effects
14
Legacy system
• They were implemented years ago ( 1970)
• Their technology became obsolete (obsolete languages, language styles,
hardware, …)
• They have been maintained for a long time ( 30 years)
• Their structure is deteriorated and does not facilitate understanding
• Their documentation (if it exists) has become obsolete
• Original authors are not available
Attention!
• They contain business rules not recorded elsewhere
• They can not be easily replaced (important!)
• They represent a large investment
15
Legacy dilemma
What should we do with legacy code?
•
•
to build the new system from scratch.
trying to understand the legacy code and to reconstitute it in a new form.
16
Reverse Engineering
Reverse engineering is the process of taking something (a device, an
electrical component, a car, a software, …) apart and analyzing its working
in details, usually with the intention to construct a new device or program
that does the same thing.
 Forward engineering is the traditional process of moving from high-level
abstractions to the physical implementation of a system.
Requirements

Design
Implementation
Reverse engineering is “the inverse” of Forward engineering
Requirements
Design
“Abstract Code Representation”
Implementation
Code
17
Re-engineering
Re-engineering is the examination (reverse engineering) of a system to
reconstitute it (forward engineering) in a new form.
This process may include modifications with respect to new requirements not
met by the original system (Semantics cannot be preserved).
The re-engineering process takes many forms, depending on its
objectives.
Sample objectives are:
- code migration/porting (ex. C to C++)
- reengineering code for reusing it
…
18
Horse-shoe model of reengineering
19
Restructuring
Restructuring is the transformation from one representation to another at the
same relative abstraction level - while preserving the system external
behavior (functionality and semantics).
Examples:
Code level: - from an unstructured (“spaghetti”) form to a
structured form (“goto-less”)
- conversion of set of “if-statements” into a
“case structure”.
Design level: to introduce design patterns (e.g., Model-View-Controller
architecture).
20
Program analysis
Program analysis is the (automated) inspection of a program to infer some
properties. In some cases, properties can be inferred without running the
program (static analysis). In other cases, properties can be inferred only
running the program (dynamic analysis).
Examples are:
- Type analysis (type inference)
- Dead code analysis
- Clone analysis
- Pointer Analysis
…
21
Software Testing
Software testing is a key activity, both in software development and in
software maintenance. Its objectives are opposed to that of development:
instead of making the system work, it aims at breaking it, thus revealing the
presence of defects. This is the main reason why development and testing
teams should be separate.
Code analyses are extremely helpful to testing. Specifically, structural testing
(aka white-box testing), as opposed to functional testing (aka black-box
testing), assumes the possibility to access the internal structure of the
program, analyze it and derive testing criteria from such an analysis. The
outcome supports:
• Test case production/automatic generation.
• Definition of stopping criteria.
22
Testing
•
One of the practical methods commonly used to detect the
presence of errors (failures) in a computer program is to
test it for a set of inputs. Testing detects errors; only
exhaustive testing, usually infeasible, can prove
correctness (absence of errors).
I1, I2, I3,
…, In, …
Our program
The output
is correct?
Expected results
=?
Obtained results
“Inputs”
23
Examples of test case


Test Input Description:
1. Login to <Abc page> as administrator
2. Go to Reports page
3. Click on the ‘Schedule reports' button
4. Add reports
5. Update
Expected Results:
The report schedule should get added to the report
schedule Table
Test case for sort:
• Test data: <12 -29 32 >
• Expected output: -29 12 32
24
Terminology …

Failure: it is an observable incorrect behavior or state of a given
system. In this case, the system displays a behavior that is
contrary to its specifications/requirements. Thus, a failure is tied
(only) to system executions/behaviors and it occurs at runtime
when some part of the system enters an unexpected state.

Fault: (commonly named “bug/defect”) it is a defect in a system.
A failure may be caused by the presence of one or more faults on
a given system. However, the presence of a fault in a system may
or may not lead to a failure, e.g., a system may contain a fault in
its code but on a fragment of code that is never exercised so this
kind of fault do not lead to a software failure.

Error: it is the developer mistake that produce a fault. Often, it
has been caused by human activities such as the typing errors.
25
LOC
Code
1
program double ();
2
var x,y: integer;
3
begin
4
read(x);
5
y := x * x;
6
write(y)
7
end
Example …
Failure: x = 3 means y =9 Failure!
• This is a failure of the system since the correct output would be 6
Fault: The fault that causes the failure is in line 5. The * operator is used
instead of +.
Error: The error that conduces to this fault may be:
• a typing error (the developer has written * instead of +)
• a conceptual error (e.g., the developer doesn't know how to double a number)
26
Terminology …

Test Case: input sequence and associated expected output


Testing: testing is the process of executing a program with
the intent of finding errors




Test Suite: a set of test cases for a system
Testing cannot guarantee the absence of faults,
Strategies for defining test suites,
Formal methods (e.g., model checking) can be used to statically
verify software properties, this is not testing.
Debugging: finding and fixing faults in the code
27
Sources for test cases definition …
• The requirements to the program (its specification)
• An informal description
• A set of scenarios (use cases)
• A set of sequence diagrams
• A state machine
• The system itself (the code or the execution of the application)
• A set of selection criteria
• Heuristics (e.g., guidelines for testing)
• Experience (of the tester)
28
Testing: three main questions …

At which level conducting the testing?




System
Unit
Integration
How to choose inputs?

Considering the program as a black box


using the specifications/use cases/requirements
Considering the program as a white box


A randomly
selected set of
inputs is often
not adequate
…
using the structure
How to identify the expected output?

Test oracles
29
Test phases





Acceptance Testing – this checks if the end user functionalities
are actually delivered. It is often a contractual prerequisite for the
user to accept and pay for the software.
Unit testing – this is testing of a single function, procedure, class.
It is usually done by the developer, not by a separate testing
team.
Integration testing – this checks that units tested in isolation
work properly when put together. It often requires drivers and
stubs to simulate the missing components while integrating the
system.
System testing – here the goal is to ensure that the whole
system works properly.
Regression Testing – this checks that the system preserves its
functionality after maintenance and/or evolution.
30
Acceptance vs. Unit Testing




Acceptance Tests are specified by the customer and
analyst to test that the overall system is functioning as
required (Did developers build the right system?).
Acceptance tests typically test the entire system, or
some large chunk of it.
Unit Tests are tests written by the developers to test a
functionality as they write it.
Unit tests typically test each unit of a system in
isolation.
31
“At different points in the process”
Iterative Software development
Write
acceptance
tests
Write
and
execute
unit tests
Execute
acceptance
tests
increment
+ system
“Written before”
Prioritized
functionalities
“Executed after the development”
32
Jemmy/Abbot/JFCUnit/…
Testing tools
FIT/Fitnesse (High level)
Cactus
GUI
Perfomance and
Load Testing
JMeter/JUnitPerf
Business
Logic
HttpUnit/Canoo/Selenium
Junit (Low level)
Web UI
Persistence
Layer
Junit/SQLUnit/XMLUnit
33
Badly designed systems makes testing
difficult




We have a thick GUI
that has program logic.
The interfaces between
the modules are not
clearly defined.
Testing
of
specific
functions cannot be
isolated.
Testing has to be done
through the GUI
Testing is difficult.
“Badly designed system”
GUI-test drivers
34
Well architected applications makes testing
simple
Design for testability



The GUI does not contain
any program logic other
than
dealing
with
presentation.
The interfaces between the
modules are well defined.
This give us testing
advantages.
Unit
and
System acceptance testing
are simpler and they can
be automated
“Well architected application”
35
How good these test cases are?
Adequacy = level of confidence of a test suite applied to the
system under test
Several criteria:

coverage of the requirements (what if 100% covered wrt the
implemented features)
 coverage of the code (what if 100% covered wrt the code)
 fault detection (what if no fault found)
36
Coverage Testing
Coverage measures describe the degree to which a
program has been tested
Many type of coverage measures





statements
branches
paths
methods, classes
requirement specifications, etc.
Example with two input data:
(1) X=7 => a=8 and b=6
Statements [1, 2, 3, 4, 5, 6, 8]
(2) X=22 => a=23 and b=21
Statements [1, 2, 3, 4, 6, 7, 8]
1
2
3
4
5
6
7
8
scanf("%d", &x);
a = x + 1;
b = x - 1;
if (a < 10)
x++;
if (b > 20)
x--;
printf("%d\n", x);
37
Mutation Testing

Mutant: a copy of the original
program with a small change
(seeded fault)

Mutant killed: if its
behaviors/outputs differ from
those of the original program

Mutant score: number of killed
mutants
38
Regression testing
Selective re-testing of a system or component to verify that modifications
have not caused unintended effects
- Ripple effects
Can be conducted at each of the test levels: unit, integration, system
Modifications
Version X
T1
T2
T3
Changed
Version X+1
• find affected testcases (red)
• change affected testcases (red)
• execute them
• define new test cases, if
necessary
Tn
…
T1
T2
T3
Tn
…
Testcases
to
modify
39
Test case prioritization
The faults revealed by a test case are unknown until the
test case is executed and its output is evaluated
against the oracle.
The order in which test cases are executed affects:


The rate of fault detection: good orderings reveal faults
earlier than bad ones;
The rate of code coverage: good orderings meet the required
coverage level earlier than bad ones.
The information can be used in regression testing to prioritize
the re-executions of the test cases, thus early finding faults.
40
Conclusions







Software maintenance implies several activities (reverse
engineering, analysis, etc.)
One of the most relevant and hard activities is software testing
Different types of testing exist: acceptance, unit, etc.
The motivation for acceptance testing is demonstrating working
functionalities from the end-user perspective.
The motivation for unit testing is finding faults in the unit the
developer is working on.
Several issues impact testing: test case definition, test case
execution, adequacy criteria, etc.
Badly designed systems make testing difficult.
41
Additional references
- V.R. Basili and R.W. Selby. Comparing the Effectiveness of Software Testing
Strategies. IEEE Transactions on Software Engineering. 1987
- Jim Heumann. Generating Test Cases From Use Cases. Online IBM journal. 2001
- Peter Zielczynski. Traceability from Use Cases to Test Cases. online IBM journal
2006
- R.C.Martin and G.Melnik. Tests and Requirements, Requirements and Tests: A
Möbius Strip. IEEE Software 2008
- Y.K.Malaiya, M.N.Li, J.M.Bieman, R.Karcich. Software reliability growth with test
coverage. IEEE Transactions on Reliability
- Q. Yang, J.Jenny Li and D.M. WEISS. A Survey of Coverage-Based Testing Tools.
The Computer Journal 2007
- Ben H. Smith and Laurie Williams. Should Software Testers Use Mutation Analysis
to Augment a Test Set. Journal of Software Systems 2009
-Florentin Ipate, Raluca Lefticaru. State-based Testing is Functional Testing! TAICPART 2007
-Jeff Offutt, Shaoying Liu, Aynur Abdurazik and Paul Ammann. Generating test data
from state-based specifications. Jounal on Software Testing, Verification and
Reliability. 2003
42
Download