Pragmatic Design Quality Assessment - (Tutorial at ICSE 2008)

Pragmatic Design Quality Assessment

Tudor Gîrba
University of Bern, Switzerland

Michele Lanza
University of Lugano, Switzerland

Radu Marinescu
Politehnica University of Timisoara, Romania

Software is complex.

29% Succeeded

18% Failed

53% Challenged

The Standish Group, 2004

How large is your project?

1’000’000 lines of code


* 2 = 2’000’000 seconds


* 2 = 2’000’000 seconds
/ 3600 = 560 hours


* 2 = 2’000’000 seconds
/ 3600 = 560 hours
/ 8 = 70 days


* 2 = 2’000’000 seconds
/ 3600 = 560 hours
/ 8 = 70 days
/ 20 = 3 months

But, code is for the computer.
Why would we ever read it?

}
}
{
{

}
}
{
{

g
rin
ee
gin
en
d
ar
rw
fo

fo
rw
ar
d
en
gin
ee
rin
g
{ {
{ { { {
} { {
}
} actual development } }
} { } } }

What is the current state?

fo
rw
ar
What should we do?

d
en
Where to start?

gin
ee
How to proceed?

rin
g
{ {
{ { { {
} { {
}
} { } } }

fo
rw
g
rin

ar
ee

d
gin

en
gin
en

ee
se

rin
erv

g
re

{ {
{ { { {
} { {
}
} { } } }

Reverse engineering is analyzing a subject system to:
identify components and their relationships, and
create more abstract representations.

Chikofky & Cross, 90

{ {
{ {
}
}
}
} { }

A large system contains lots of details.

ity?
its qual
ju dge
How to { {
{ {
}
}
}
} { }

A large system contains lots of details.

http://moose.unibe.ch

http://loose.upt.ro/incode

1
Software in
2
Software in
numbers pictures

3
Software in
4
Software in
time tools

Software in numbers

1

Youcannot control
what you cannot measure.

Tom de Marco

Metrics are functions that assign numbers to
products, processes and resources.

Software metrics are measurements which
relate to software systems, processes or
related documents.

Metrics compress system traits into numbers.

Examples of size metrics

NOM - number of methods
NOA - number of attributes
LOC - number of lines of code
NOS - number of statements
NOC - number of children

Lorenz, Kidd, 1994
Chidamber, Kemerer, 1994

McCabe cyclomatic complexity (CYCLO) counts
the number of independent paths through the code of a
function.

McCabe, 1977

 it reveals the minimum number of tests to write

 interpretation can’t directly lead to improvement action

Weighted Method Count (WMC) sums up the
complexity of class’ methods (measured by the metric
of your choice; usually CYCLO).


 it is conﬁgurable, thus adaptable to our precise needs

 interpretation can’t directly lead to improvement action

Depth of Inheritance Tree (DIT) is the (maximum)
depth level of a class in a class hierarchy.


 inheritance is measured

 only the potential and not the real impact is quantiﬁed

Coupling between objects (CBO) shows the number
of classes from which methods or attributes are used.


 it takes into account real dependencies not just declared ones

 no differentiation of types and/or intensity of coupling

Tight Class Cohesion (TCC) counts the relative
number of method-pairs that access attributes of the
class in common.

Bieman, Kang, 1995

TCC = 2 / 10 = 0.2

 interpretation can lead to improvement action

 ratio values allow comparison between systems

Metrics Assess and Improve Quality!

Metrics Assess and Improve Quality!

a lly ?
Re

Problem 1: metrics granularity

?
capture symptoms, not causes of problems

in isolation,
they don’t lead to improvement solutions

Problem 1: metrics granularity

?
capture symptoms, not causes of problems

in isolation,
they don’t lead to improvement solutions

Problem 2: implicit mapping
we don’t reason in terms of metrics,
but in terms of design principles

2 big obstacles in using metrics:

Thresholds make metrics hard to interpret

Granularity make metrics hard to use in isolation

Can metrics help me
in what I really care for? :)

fo
rw
How do I understand code?

ar
d
en
gin
ee
rin
g
{ {
{ { { {
} { {
}
} { } } }

fo
rw

ar
d
How do I improve code?

en
gin
ee
rin
g
{ {
{ { { {
} { {
}
} { } } }

fo
rw

ar
d

en
gin
How do I improve myself?

ee
rin
g
{ {
{ { { {
} { {
}
} { } } }

etr ics!
w ith m
o do
i ng t
tn oth
I wan

fo
rw

ar
d

en
gin

ee
rin
g
{ {
{ { { {
} { {
}
} { } } }

How to get an initial understanding of a system?

Metric Value
LOC 35175
NOM 3618
NOC 384
CYCLO 5579
NOP 19
CALLS 15128
FANOUT 8590
AHH 0.12
ANDC 0.31

Metric Value
LOC 35175
NOM 3618
NOC 384
CYCLO 5579
NOP 19
CALLS 15128
FANOUT 8590
ha t?
ww
AHH
An d no 0.12
ANDC 0.31

hierarchies?

coupling?

The Overview Pyramid provides a metrics
overview. Lanza, Marinescu 2006

Inheritance
ANDC 0.31
AHH 0.12
20.21 NOP 19
9.42 NOC 384
9.72 NOM 3618 NOM 418
0.15 LOC 35175 15128 CALLS 0.56
CYCLO 5579 8590 FANOUT

Size Communication


ANDC 0.31
AHH 0.12
20.21 NOP 19
9.42 NOC 384
9.72 NOM 3618 NOM 418
0.15 LOC 35175 15128 CALLS 0.56

Size


ANDC 0.31
AHH 0.12
20.21 NOP 19
9.42 NOC 384
9.72 NOM 3618 NOM 418
0.15 LOC 35175 15128 CALLS 0.56

Communication


Inheritance
ANDC 0.31
AHH 0.12
20.21 NOP 19
9.42 NOC 384
9.72 NOM 3618 NOM 418
0.15 LOC 35175 15128 CALLS 0.56


ANDC 0.31
AHH 0.12
20.21 NOP 19
9.42 NOC 384
9.72 NOM 3618 NOM 418
0.15 LOC 35175 15128 CALLS 0.56

Java C++
LOW AVG HIGH LOW AVG HIGH

CYCLO/LOC 0.16 0.20 0.24 0.20 0.25 0.30

LOC/NOM 7 10 13 5 10 16

NOM/NOC 4 7 10 4 9 15

...


ANDC 0.31
AHH 0.12
20.21 NOP 19
9.42 NOC 384
9.72 NOM 3618 NOM 418
0.15 LOC 35175 15128 CALLS 0.56

close to high close to average close to low


close to high close to average close to low

etr ics!
w ith m
o do
i ng t
oth
nt n
I wa

fo
rw

ar
d

en
gin

ee
rin
g
{ {
{ { { {
} { {
}
} { } } }

Quality is more than 0 bugs.

Breaking design principles, rules and best practices
deteriorates the code;
it leads to design problems.

Imagine changing just a small design fragment

Imagine changing just a small design fragment

and33%
of all classes
would require changes

expensive
Design problems
are frequent
unavoidable

expensive
Design problems
are frequent
unavoidable

th em?
ate
limin
nd e
ete ct a
to d
How

God Classes tend to centralize the intelligence of the
system, to do everything and to use data from small
data-classes.
Riel, 1996

God Classes tend
to centralize the intelligence of the system,
to do everything and
to use data from small data-classes.

God Classes
centralize the intelligence of the system,
do everything and
use data from small data-classes.

God Classes
are complex,
are not cohesive,
access external data.

God Classes
are complex,
are not cohesive,
access external data.

sing
uer ies u
to q s
s in ator
etric per
se m ical o
Co mpo log

Detection Strategies are metric-based queries to
detect design ﬂaws. Lanza, Marinescu 2006

Rule 1

METRIC 1 > Threshold 1

AND Quality problem

Rule 2

METRIC 2 < Threshold 2

Shotgun
Surgery has

uses is

has (partial) Feature Data
Envy uses Class

is partially
God has
Intensive Class
Coupling Brain has
has
Method
Extensive Brain has Significant
Coupling Class Duplication
has
is
is
has
Refused
is Tradition
Parent
Breaker
Bequest
has (subclass)

Futile
Hierarchy Lanza, Marinescu 2006

Identity Collaboration Classification
Disharmonies Disharmonies Disharmonies

A God Class centralizes too much intelligence in
the system. Lanza, Marinescu 2006

Class uses directly more than a
few attributes of other classes

ATFD > FEW

Functional complexity of the
class is very high
AND GodClass
WMC ! VERY HIGH

Class cohesion is low

TCC < ONE THIRD

An Envious Method is more interested in data
from a handful of classes. Lanza, Marinescu 2006

Method uses directly more than
a few attributes of other classes

ATFD > FEW

Method uses far more attributes
of other classes than its own
AND Feature Envy
LAA < ONE THIRD

The used quot;foreignquot; attributes
belong to very few other classes

FDP ! FEW

Data Classes are dumb data holders.
Lanza, Marinescu 2006

Interface of class reveals data
rather than offering services

WOC < ONE THIRD
AND Data Class

Class reveals many attributes and is
not complex

Data Classes are dumb data holders.
Lanza, Marinescu 2006

More than a few public
data

NOAP + NOAM > FEW
AND
Complexity of class is not
high

WMC < HIGH
Class reveals many
OR attributes and is not
Class has many public complex
data

NOAP + NOAM > MANY
AND
Complexity of class is not
very high

WMC < VERY HIGH

Follow a clear and repeatable process

Follow a clear and repeatable process

mb ers!
so f nu
in term
qu ality
ab out
re ason
D on’t

QA is part of the the Development Process


Can we understand the beauty of a painting
by measuring its frame or counting its colors?

Software in pictures

2

1854,
London,
cholera
epidemic

1812, Napoleon’s Campaign in Russia

Numbers..

ANDC 0.31
AHH 0.12
20.21 NOP 19
9.42 NOC 384
9.72 NOM 3618 NOM 418
0.15 LOC 35175 15128 CALLS 0.56

Visualization compresses the system into pictures.

A picture is worth

a thousand words... anonymous

...depends on the picture
Lanza

U ML
han
mo re t
tio n is
liza
isua
wa re v
Soft

We are

visual
beings ...

... and we’re
good at
spotting
patterns

Gestalt principles

proximity similarity

enclosure connectivity

More Gestalt principles

closure continuity

We do not see with our eyes, but with our brain.

Our brain works like a computer, with 3 types of memory
Iconic memory, the visual sensory register
Short-term memory, the working memory
Long-term memory

Sensation Perception
(Physical Process) (Cognitive Process)

Stimulus Sensory Organ Perceptual Organ

Brain
Iconic Memory - Short-term Memory - Long-term Memory

Iconic Short-term
memory memory

Iconic Short-term
memory memory

< 1 second
very fast
automatic
subconscious
preattentive

Iconic Short-term
memory memory

< 1 second couple of seconds
very fast 3-9 chunks
automatic
subconscious
preattentive

Categorizing Preattentive Attributes

Category Form Color Spatial Motion Motion

Orientation Hue 2D position Flicker

Line length Intensity Direction

Line width

Size
Attribute
Shape

Curvature

Added marks

Enclosure

Attributes of Form

Orientation Line Length Line Width Size

Shape Curvature Added Marks Enclosure

Attributes of Form

Line Length Line Width Size


Attributes of Form

Line Width Size


Attributes of Form

Size


Attributes of Form


Attributes of Form

Curvature Added Marks Enclosure

Attributes of Form

Added Marks Enclosure

Attributes of Form

Enclosure

Exemplifying Preattentive Processing


8789364082376403128764532984732984732094873290845
389274-0329874-32874-23198475098340983409832409832
049823-0984903281453209481-0839393947896587436598


8789364082376403128764532984732984732094873290845
389274-0329874-32874-23198475098340983409832409832
049823-0984903281453209481-0839393947896587436598

8789364082376403128764532984732984732094873290845
389274-0329874-32874-23198475098340983409832409832
049823-0984903281453209481-0839393947896587436598

70%
of all external
inputs come
through the eyes

Software visualization is
the use of the crafts of typography, graphic design, animation,
and cinematography with modern human-computer interaction
and computer graphics technology to facilitate both the human
understanding and effective use of computer software.
Price, Becker, Small

A picture is worth

a thousand words.

era lly :)
ok it lit
L to
UM

Polymetric Views show up to 5 metrics.
Lanza, 2003

Width metric

Height metric

Position metrics

Color
metric

A simple & powerful concept

LOC

NOS
parameters

parameters lines

System Complexity shows class hierarchies.
Lanza, Ducasse, 2003

attributes

methods lines

Class Blueprint shows class internals.

Initialize Interface Internal Accessor Attribute

invocation and access direction

Class Blueprint has a rich vocabulary.

internal access
Access
external
Attribute
access Invocation

Regular Constant

invocations
Overriding Delegating

lines Method
Extending Setter

Abstract Getter

Class Blueprint reveals patterns.

twin classes schizophrenic class

Distribution Map shows properties over
structure. Ducasse etal, 2006

31 parts, 394 elements and 9 properties

Softwarenaut explores the package structure.
Lungu etal, 2006

Code City shows where your code lives.
Wettel, Lanza, 2007

classes are buildings grouped in quarters of packages

{
{
} }
}
{

{
{
}

}
re
v er
se
en
gin
ee
rin
g

reverse engineering

fo
actual development rw
ar
d
en
gin
ee
rin
g
{
{
}
}
{
{
}
}

{ { { { { { { { { {
{ { { { { { { { { {
} } } } }
}
} }
} }
} }
} }
}
} { } } { } } { } } { } } { }

The history of a large system contains even more details.

Most often time is put on the horizontal
and a property on the vertical axis.

Lehman etal, 2001

Spectographs show change activity.
Wu etal, 2004

commit

time

Evolution Matrix shows changes in classes.

Idle class

Pulsar class

Supernova class

White dwarf class

Evolution Matrix shows changes in classes.

What changed? When did it change? ...

2 4 3 5 7

2 2 3 4 9

2 2 1 2 3

2 2 2 2 2

1 5 3 4 4

Evolution of
Number of Methods LENOM(C) = ∑ |NOMi(C)-NOMi-1(C)| 2i-n

LENOM(C) = 4 + 2 + 1 + 0 = 7

1 5 3 4 4

Latest Evolution of
Number of Methods LENOM(C) = ∑ |NOMi(C)-NOMi-1(C)| 2i-n

Earliest Evolution of
Number of Methods EENOM(C) = ∑ |NOMi(C)-NOMi-1(C)| 22-i

-3 -2 -1 0
LENOM(C) = 42 + 22 + 12 + 02 = 1.5

1 5 3 4 4

EENOM(C) = 4 20 + 2 2-1 + 1 2-2 + 0 2-3 = 5.25

ENOM LENOM EENOM

2 4 3 5 7 7 3.5 3.25

2 2 3 4 9 7 5.75 1.37

2 2 1 2 3 3 1 2

2 2 2 2 2 0 0 0

1 5 3 4 4 7 1.25 5.25

ENOM LENOM EENOM

balanced changer 7 3.5 3.25

late changer 7 5.75 1.37

3 1 2

dead stable 0 0 0

early changer 7 1.25 5.25

ENOM LENOM EENOM

balanced changer 7 3.5 3.25

late changer 7 5.75 1.37

ed.
measur3 1 2
be
ry can
H isto
dead stable 0 0 0

early changer 7 1.25 5.25

History can be measured in many ways.

Evolution Number of Methods
Stability Number of Lines of Code
Historical Max of Cyclomatic Complexity
Growth Trend Number of Modules
... ...

The recently changed parts are likely to change in the
near future.

Common wisdom

The recently changed parts are likely to change in the
near future.

ally?
Common wisdom
re
re they
A

past future

present

past future

YesterdayWeatherHit(present):

past:=histories.topLENOM(start, present)

future:=histories.topEENOM(present, end)

past.intersectWith(future).notEmpty()

present

past future

YesterdayWeatherHit(present):

past:=histories.topLENOM(start, present)

future:=histories.topEENOM(present, end)

past.intersectWith(future).notEmpty()

present
prediction hit

Yesterday’s Weather shows the localization of changed in
time. Girba etal, 2004

hit hit hit

YW = 3 / 8 = 37%

hit hit hit hit hit hit hit

YW = 7 / 8 = 87%

the system.


ATFD > FEW

Functional complexity of the
class is very high
AND GodClass
WMC ! VERY HIGH


TCC < ONE THIRD

the system.


ATFD > FEW

tab le?
f it is s
wh
Functional complexity of the at i
ut,
class is very high
B AND GodClass
WMC ! VERY HIGH


TCC < ONE THIRD

History-based Detection Strategies take evolution
into account. Ratiu etal, 2004

God Class
in the last version

isGodClass(last)

AND Harmless God Class

Stable throughout
the history

Stability > 90%

What happens with inheritance?

A A A A A

B C B C B C B B

D D D E

ver .1 ver. 2 ver. 3 ver. 4 ver. 5

A is persistent, B is stable, C was removed, E is newborn ...

Hierarchy Evolution encapsulates time.
Girba etal, 2005
A
changed
methods

changed
age
lines
C B

Removed

Removed
D E

A is persistent, B is stable, C was removed, E is newborn ...

Hierarchy Evolution reveals patterns.
Girba etal, 2005

Co-change analysis recovers hidden dependencies.
Time is the lines.

Gall etal, 2003

Evolution Radar shows co-change relationships.
D’Ambros, Lanza 2006

one package and its co-change relationships

Software is developed by people.

Alphabetical order is no order.

Ownership Map reveals development patterns.
Girba etal, 2006

Who copied from whom?

(john 23.06.03) public boolean stillValid (ToDoItem I, Designer dsgr) {
(bill 09.01.05) if (!isActive()) {
(bill 09.01.05) return false
(bill 09.01.05) }
(steve 16.02.05) List offs = i.getOffenders();
(john 23.06.03) Object dm = offs.firstElement();
(steve 16.02.05) ListSet newOffs = computeOffenders(dm);
(john 23.06.03) boolean res = offs.equals(newOffs);
(john 23.06.03) return res;

(george 13.02.05) public boolean stillValid (ToDoItem I, Designer dsgr) {
(bill 11.13.05) }
(george 13.02.05) Object dm = offs.firstElement();
(george 13.02.05) boolean res = offs.equals(newOffs);
(george 13.02.05) return res;

What is useless?

(john 23.06.03) public boolean stillValid (ToDoItem I, Designer dsgr) {
(bill 09.01.05) }
(john 23.06.03) Object dm = offs.firstElement();
(john 23.06.03) boolean res = offs.equals(newOffs);
(john 23.06.03) return res;

(george 13.02.05) public boolean stillValid (ToDoItem I, Designer dsgr) {
(bill 11.13.05) }
(george 13.02.05) Object dm = offs.firstElement();
(george 13.02.05) boolean res = offs.equals(newOffs);
(george 13.02.05) return res;

When did changes happen?

23.06.03 public boolean stillValid (ToDoItem I, Designer dsgr) {
09.01.05 if (!isActive()) {
09.01.05 return false
09.01.05 }
16.02.05 List offs = i.getOffenders();
23.06.03 Object dm = offs.firstElement();
16.02.05 ListSet newOffs = computeOffenders(dm);
23.06.03 boolean res = offs.equals(newOffs);
23.06.03 return res;

13.02.05 public boolean stillValid (ToDoItem I, Designer dsgr) {
11.13.05 if (!isActive()) {
11.13.05 return false
11.13.05 }
16.02.05 List offs = i.getOffenders();
13.02.05 Object dm = offs.firstElement();
16.02.05 ListSet newOffs = computeOffenders(dm);
13.02.05 boolean res = offs.equals(newOffs);
13.02.05 return res;

Clone Evolution shows how developers copy.
Balint etal, 2006

Software in tools

4

http://moose.unibe.ch


http://www.inf.unisi.ch/phd/wettel/codecity.html

Tudor Gîrba, Michele Lanza, Radu Marinescu

http://creativecommons.org/licenses/by/3.0/

Pragmatic Design Quality Assessment - (Tutorial at ICSE 2008)

Recommandé

Recommandé

Contenu connexe

Similaire à Pragmatic Design Quality Assessment - (Tutorial at ICSE 2008)

Similaire à Pragmatic Design Quality Assessment - (Tutorial at ICSE 2008) (18)

Plus de Tudor Girba

Plus de Tudor Girba (20)

Dernier

Dernier (20)

Pragmatic Design Quality Assessment - (Tutorial at ICSE 2008)