Software Engineering (phần 10) pptx

Now Figure 1 1 6 the doors close, and the elevator starts : Returning to the state diagram of Figure 12.6, consider what happens when the.. Send message to Floor Bu/on to turn on button

: i : :

'

In contrast, the representation of a UM L state diagram is somew hat Iess formal ' l

The three aspects of a state machine (state,event,and predicate)are distributed overthe !UML diagram For example, the state Go in'o W oii Sto'e in Figure 12.6 is entered j

if t he present stat e is El evotor Even, toop and the predicate ë el evator st opped, j 1

no r equest s pendi ngl i s t rue ( UML predi cat es or ' ' guar ds' ' appear n bracket s as j 1 I l

shown in Fi gure 12 6) When the state Go i ne W ci , Sioie has been entered action

l

Cl ose el evai or doors cher timeout is to be canied out.current versions of O OA i !'

di agram accordingly is no probl em However, when the object-ori ented paradigm

matures, itis likely that more formal versions willbe developed and the correspondingdynamic m odels w ill be som ewhat closer to finite state machines r

To see the equivalence of the state diagram of Figure 1 2.6 and the STDS of Figures j

1 l14 through 1 l 16, consider various scenarios For exam ple consider the hrst part '

the elevator goes i nto state S(U, 3), that is, i t stops at fl oor 3, about to go up (Because l ithe sim plifying assum ption has been m ade of only one elevator, the argument e in j

i Figure l l I6 is suppressed here).From Figure 1 l 5, when the elevator arrives, the ; i.i

floor button is turned off Now (Figure 1 1 6) the doors close, and the elevator starts :

Returning to the state diagram of Figure 12.6, consider what happens when the

elevator nears floor 3 Because the elevator is in motion, the next state entered is :

' Dei ermine I f S'op Reques/d The requests are checked and, because User A has ' i

!'

requested t he el evator to st op there, the next st at e is Stop oi Fl oor The elevator stops j t

at oor 3 and t he door s open The el evat or but t on f or oor 3 has not been pressed so 1 ' 1

User A enters and presses the elevator button for floor 7 Therefore, the next !

has st opped and t wo r equest s ar e pendi ng, so st at e Cl ose El evoi or Doors next and , ! 1

. t the doors close after a ti meout.The floor button at floor 3 w as pressed by U ser A , t

Protess Nexl Requesl is next and the elevator starts to move toward floor 4 The j

relevant aspects of the corresponding diagrams clearly are equivalent w ith respect to 'this scenario', you may w ish to consider other scenarios as well y !

I

From the preceding discussion. it should com e as no surprise to learn that Figure :

I12

.6 was constructed from the scenarios M ore precisely, the specific events of the !scenarios w ere generalized For example, consider the first event of the scenario of f '

Fi gure l 2 2, User A presses t he Up fl oor butt on at fl oor 3 This specihc event i s generali zed to an arbit rary butt on being pushed Then there are two possi bi lit ies : j

l @' j

;

t 3V% < H A p T E R IQ * Obiee-ori en#ed Anol ysis Phose

i

! Either the button already is Iit (in which case nothing happens) or the button is unlit

1

i case of an al ready Ii t butt on is model ed by the do-nothing loop with guard (but t on

@ pushed, buf t on l i f) in the top l eft-hand corner of Fi gure 12 6 The other case, an unli t

E button, i s model ed by the arrow labeled wi th t he guard Ibuf t on pushed, buft on unl i

' leading to state Protess Request From event 2 of the scenario it is clear that the action

E Turn on butt on i s needed in this state.Furthermore, the purpose of the user's action

'

of pressing an arbitrary but ton is to request an el evator, so aetion Updot e request s

also m ust be carried out in the state Protess Request

Now consi der event 3 of the scenario, An el evat or arri ves at fl oor 3 Thi s

!

) was generalized to the concept of an arbitrary elevator m oving between Qoors The

moti on of the el evat or i s modeled by the guard Iel evat or moving i n di rect i on d,

oor fs next) and the stat e Delermi ne i f Slop Requesled But there again are two

l possibilities.Ei ther there is a request to stop at floorf or no such request ln the former

;

i case con- espondi ng t o guar d ( no r equest o st op at oor , he el evat or si mpl y mus t

onl inue Moving one more floor i n direction d In t he latter case (correspondi ng t o

t 7 ! g uard (user I ncs requested stop ct fl oor f!) from t he scenari o of Figure l 2 2 it is dear

j ! t ha t t s ne c es s r y t St op el evat or f om e ve nt 3) , an d t hen Open door s an d s or f

timer (from events 5 and 6) Also, similar to the case of the Protess Requesi state,

it becomes apparent that i t i s necessary to Updat e request s in state Slop G Fl oor.

!

. :1 Finally, generalizing event 4 of the scenario leads to the realization that the elevator '

i

i: button has to be turned off if it is Iit.This is modeled by stat e El evolor Bu/on 0: ,

r (the st ate on t he bottom l eft of Figure l 2 6), together wit h the two guards above t ll

'

Doors', generalizing event 10 yields the state Process Next Requesi However, the(

:

: need for the state Go inlo W oif Stote and guard (no request s pendi ng, doors cl osedl

is deduced by generalizing an eventof a different scenario, one in which the user exit

The state diagrams for the other classes are relatively straightforward and t

here-: fore left as an exercise (Problem l2.)

g as suggested in Section l 2.4.2, is to use CRC cards

1 :k

'

@ j:

t

.

: i'

7 O pen elevator doors and storttimer i

s CRC card highli ghts t wo majorprobl ems wit h the hrst terati on of t he object- ; '

i t ed anal ysi s Fi rst consi der es ponsi bi l y 1 Tur n on el evat or bueon Thi s com- ! l

responsible for turning themsel ves on or off Also, from the viewpoi nt of informa- : l

t on hi di ng ( Sect i on 7 6) ,he El evot or Conl rol l er s houl d not have t he knowl edge of l

c han ges ar e ne ede d t br es pons i bi l i es hr ou gh n Fi gur e 1 2 7 The s i cor r ec - t

ti ons are refl ected in Figure l 2 8 the second i teration of the CRC card for the El evoe r

i e

,

v A second problem is that a class has been overlooked Consider the responsibility I

jr ' The attributes of a class sometim es are termed state variables The reason for this ; !

1 - '' t erminol ogy i s that, in most object-oriented implement ations, the state of the product

i d termined by the val ues of the attri butes of the various component objects The I e s e

: ! i r

!

: rl'j

1 aao t u A p T : R lx obi eu-ori en'ed Anol ysis phose

! 1 send message to Elevo'or Bu/on to turn on button

1 2 send message to Elevofor Bu/on to turn off button

3 Send message to Floor Bu/on to turn on button

4 Send message to Floor Bu/on to turn off bufton

5 Send message to Elevolor to move up one floor

4 6 send message to Elevador to move down one floor

1 ylgvee 1Q.@ second iteration ofCRC card for the class

i is not surprisi ng t hat the concept of t he state plays an important rol e i n t he

object-oriented paradigm This concept can be used to help determ ine whether a componentshould be m odeled as a class lf the component in question possesses a state that

i will be changed during execution of the im plem entation, then it probably should be

i m odeled as a class.Cl earl y, the doors of the elevator possess a state (open or cl osed),

-i oriented paradigm al lows the st ate to be hidden wit hin an object and hence protected

1 from unauthorized change.lf there is an El evoi or Doors object, the only way that the

f j

' '

' I Doors object Serious acci dents can be caused by opening or closing the doors of an

I I el evat or at t he wr ong t me' , see t he Jus tn Case You Want ed t o Know Box on page

j

1 1 38 l.Therefore, for certain types of products, safety considerations should be added '

' 1 , to the other advantages of objects Iist ed i n chapters 7 and 8.

; 12.7 need to be changed analogously to responsibilities l through 6 That is,m essages

.Responsibil ity 7 is Open el evat or doors and st art ti mer.

l '

J us' I N G s: You W ANTZP To Kwow q j

. Some years ago I was on the 10th floor of a building, Perhaps if that elevator control system had been

:. w aiIstartting ied tm patientlo step forw ary for d- only no elan elevator.The doors opened,evator was there. develappropriate oped using the object-oriented paradiopening of the doors on the 10th gm,floor the in-m ight l!

.

l What aved me was he ot al bl acknes s aw as was ha ve been avoi ded

about to st ep into the el evat or shaft and l i nsti ncti vel y j j !

'

j I

-.( p

'

1 i

1 -Thi

s must be s pl i i nt o t wo s eparat e r es ponsi bi l i es A message i ndeed must be s ent ë i

t o t he El evuù or Door s t o open.However. t he timer is part of the El evcior Conl rol l er,

l The second it erati on of the CRC card for t he El evoior Control l er cl ass (Figure 12.8) q r , hows that t his separation of responsi bili ti es has been achieved satisfactoril y ( k 1 s

In additi on to the two major problems highl ighted by the CRC card of Figure , I 12.7 responsibil ities Check requesis and Updat e request s of El evol or Coniroll er

requests are dehned simply to be of type requestType; a data structure for requests , : .

: i

The corrected class diagram is shown in Figure 12.9 Having modified the class r !

'

diagram the use-case and state diagram s m ust be reexam ined to see if they,too need i ' '

! IIi

! 1 !

!!

1

1 2 The floor button informs the elevator controller that the floor button has been pushed.

1 3 Tie elevator controller sends a message to the Up floor button to turn itselfon

! h Ievator controller sends a series of messages to the elevator to move itself up .

! to floor 3 The elevatorcontains User B, who has entered the elevator ctfloor 1 and

P

5 The elevator controller sends a messoge to the Up floor button to turn itselfoff

E 6 The elevator controller sends a message to the elevator doors to open themselves 'i

( Z The elevatorcontrol starts the timer

User A enters the elevator

:

8 User A presses elevotor button for floor 7

9 The elevator button informs the elevator controller that the elevator button has been

. 10 The elevatorcontroller sends a messcge to the elevator button for floor 7 to turn itself on

i 1 1 The elevatorcontroller sends c message to the elevator doors to close themselves oher o

' 1 4 The el evot or conf r ol l er ends a mess age t o t e el evat or door s o open t hems el ves

: j

cllow UserA to exit from the elevator

15 The elevotor controller starts the timer '(

' User A exits from ihe elevator

Flgvee 1Q.1@ Second iteration of u normal scenario

furtherrelinement The use-case diagram clearly stillis adequate However,the actions

in the state diagram of Figure l2.6 must be m odified to reiect the responsibilities of

1 i teration) Also, the set of st at e diagrams must be extended to include the additi onal (

k 1 class The scenarios need to be to updated to reiectthese changes', Figure 12.10 shows

ê

!: been made and checked (including the modified CRC cards), it still may be necessary

# ' 5 duri ng t he object- or i ent ed desi gn phase t o r et ur n t o t he object - ori ent ed anal ysi s and 2 l

to that of A ppendix E O nce approved, the specilication document is handed to the

j design team , together with the various U M L diagram s

' ! -

1 l

i l

l2.e AIR G OURMET CASE STUDY: O BJK T-O RIENTID A NAW SIS a@a 1,

!

! k

1Q.y QA SK Toots FoR TH: @ BJE<T-@ RIKNTEP ' j l

I

l1B

earing in mind the role pl ayed by diagrams i n object-oriented analysi s, it is not I surpri sing that a number of CASE tools have been devel oped t o support object- ( I

1oriented analysis ln its basic form , such a tool essentially is a drawing tool that : t

k es eas y t per f or m e ch of he mo del i ng s eps Mor e i mp or t nt t s a mpl er ' 1 j ma

to modify a diagram constructed with a drawing tool than to attem pt to change a 2

!hand-draw n figure Thuss a CASE tool of this type supports the graphical aspects of ;

object-oriented analysi s In additi on, some t ool s of this type not only draw all the I I ë

1 levant diagram s but CRC cards as well A strength of these tools is that a change to 7

:l

; the underlying m odel is reoected autom atically in aIl the affected diagram s', after all, )

the various diagram s are m erely different views of the underlying m odel '

'

, a considerable portion ot the rest of the object-oriented li fe cycle as wel l Nowadays 1 1

Is support UML (Rumbaugh, Jacobson and Booch, 19991 : i 1 virtually al1 of these too

j I r

? : Object-ori ent ed analysi s consi sts of three steps: use-case modeling, class model-

ing, and dynam ic modeling These steps are canied out iteratively, starting w ith the :

first step, use-case modeling The main menu of the Air Gourmet rapid prototype

( Appendi x C or D) yi el ds a l i st of use cas es These al s o coul d have been deduced j

C wi th considerably more effort di rectly from the requirements of Section 1 0 14 j 1 ' T

The use case for m aking a reservation is shown in Figure l2 l Normal and E '.I)exception scenarios are needed for this use-case diagram One approach is to con- I '

! I

struct separate normaland exception scenarios', this was done for the elevator problem (

l

( Fi gures l 2 2 and l 2 3) For he Ai r Gourmet probl em, we use a sl i ght l y di f erent or- 4 j

m alism , an extended scenario that retlects possible alternatives.A n extended scenario '

Figure l2.2 was constructed from the prototype (Appendix C or D) Choosing 'Ent er a Res er vat i on f rom t he mai n menu (r epr esent ed by event l of Fi gur e 1 2 2) 1 t i

and 9), special meals information (events 6 and 7), and passenger information (events )

2 and 3) It was quickly realized that the rapid prototype solicits information in the ' '

wrong order, and that the paym ent for the reservation had been om itted Rearranging k ,

1I 'the requests and adding the payment portion yielded events 1 through l 1 of Figure j: 12 l2 The last three events of the scenario w ere added to com plete the normal part ' l

2 Tlae Phone Operaàor requesfs àlne name and address of àlne Passenger.

1 3 The Possenger provides fhe requested information

1 zt The Phone Operatorasks the Fassenger the date ofthe flightand the flight number

j ! 7 The P os s enger ct es whot n d of peci al meal , f any , i s des i ed.

'' 8 The Phone Operoforasks the Fossenger for his or her seating preference.

: 9 The Fassenger provides his or her seofing preference

; ' 1 0 The Fhone Operctor requests payment information

, I

1 1 The Possenger provides o volîd credif cord number

;

! 1 2 The Fhone Operator commits the reservation to the Air Gourmet database.

1 1 3 The Air Gourmet datobase issues the reservation ID to the Phone Operator.

t 1z1' The Fhone Operator gives the reservation ID to the Passenger and informs the

C The Fassenger has an unusuol meol requestthatAir Gourmet cannot fulf

D There are problems with the creditcard number provided by the Fassenger,(

! Flgvze 12.12 Extended scenario corresponding to Figure 1.1 1 1

j corresponding extended scenario is shown in Figure l2 14 Event 3 of this scenalio

was obtai ned from the scon post card porti on of the rapi d prot otype Events 1 and

2 then were added to com plete the scenario As with the previous scenarios the four

.

1 possible alternatives were deduced while working on the normal events

7/

!

1

return/scan postcard

2 : l

; 1' j j

l

1

l

t o t he Passenger who recei ved a speci al meal

2 The Fassenger receives the postcard, rates the meal, and mail the postcard back to Air j )

3 The Air Gourmet Staff Member receives the postcard and updates the appropriote flight 1

r ecor d i t Ai r Gour met dat cbas e t o r ecor d t e Fas s en ger ' r espons e ! 1I

'

k l P

i IA

IB

' C The F as s enger mar ks he post c ar d er r oneous l y; or exompl e, by pr ovi di ng an ans wer ; : I

outside the ronge ofvalid values ' !i

D A Fassenger who did not requesta special meal receives the postcard and mails it bock i 1

:: ! ï

Flgvre 1Q.1* Extended scenario corresponding to Figure 12 3 ' i

l

:

The other use cases (create and shade a special meals list, transport special j '

meals, view é'low sodium '' report view E'percentage statistics'' report, view çtm eals I ' l )

! '

not loaded''report, and view ûçpoor quality'' report) are sim ilar as are their respective I I i

find their attributes,and determine their interrelati onshi ps and interactions This was l j

performed using the three-stage noun extraction method of Section l2.4, rather than p, k

scenarios or CRC cards Stage 1 is to define the product as briefly and concisely as l

'

A computeri zed system is needed t o provi de i ntorma

I :

ln stage 2 an inform al strategy is draw n up, preferably in a single paragraph l

II àO

!R

eport s are to be generat ed to document the eff icacy of the speci al meals program ' 1

. The reports concern m eals Ioaded on flights, llights boarded by passengers nam es ' !

d ddresses of passengers, meal qualitys and Iow-sodium meal quality ! l

1 ln stage 3 the nouns in this paragraph are identified, yielding

Reports are to be generated to document t he effi cacy of t he special meals pr ogr om.

i

i ' rhe reporf s concern percent oges of meal s l oaded on fl i ght s and boording of f ght s

,meal qualis and Iow-sodium

; by passengers, names and a resses o p

!

'

, The nouns are report , effi cacy, program, percentage meal fl i ght boarding,

passenger, name, oddress, and qual i ty Effi cacy, progrom, percent oge, boording, C

' and qual i yare abstractnouns and therefore unlikely to beclasses Nameandaddress ;

clearly are attributes of passenger The presence or absence of a special meal on a

i ven flight is a central issue W hat i s l ess cl ear is whether the meal itsel f or the

lirst iteration This left two nouns and, thercfore, two candidate classes: Report and

j Possenger The resulting first iteration ofthe class diagram is shown in Figure l2.15

i This hgure reflects the two candidate classes, plus four subclasses. one for each of the

i

. j and ending date Each subclass com prises the lields specific to that report

! There are a number of problem s with this class diagram '

, onboard the flight for which they ordered a special m eal, the data have to be é

j2

. Each report has to access multi pl e fli ght records, and each ii ght has multi ple 4 j

Section 10.14 specifies four num bered reports,as reflected in Figure 12.15 How- 1

3.

!2

ever, the f irst numbered report actually consi sts of three separat e report s There- 1

,the class diagram should reqect six (and not four) subclasses of Report j

$ These events are rectiied in the class diagram of Figure l2.16 There are in- '

t ' deed now six subclasses of Report Furtherm ore there is a one-to-m any relationship I E.) (Section 1 1 5) between Repod and Fl ighl Record, rel lecti ng that each report con- t I

, t ains information from many qights Fi nall y the open diamond next to Fl igl / Retord , l i

!

l a one-to-m any relationship.That is, each llight record comprises many passenger , ',

ds These records could have been modeled as components of the Fl i ghl Reeord : !

) I nst ead, f or si mpl i ci t y and cl ar i y, t he Pcssenger cl ass i s t r eat ed as a separat e en- E j

candidate class As previously pointed out, it is easy to add classes but m uch harder ( ;

t wo candi dat e cl ass es or he i ni t al t erat i on ' j

lt i s less easy to justify t he decision to i ncorporate only four repol t subclasses. ' !

I IAfter all.the rapi d prototype (Appendix C or Appendix D) cont ains six reports On ' l

(

. be better to note the ease with which the second iteration was produced, once it was ) t

he passengers on those Pights ln other words, iterati on is intri nsic t o the object- i ' 1

)

oriented paradigm in general and object-oriented anal ysis in part icular lteration does l I

discarding an artifact w hen a fault is detected, but rather modifying that : ''

J not mea n

m ifact stepwise to correct the fault II 'i

The thi rd st ep in object-oriented anal ysis is dynami c modeli ng A1l t he scenari os ( !

flected in the state diagram of Figure l2 7 for the com plete product Strictly lare re

i

speaking, a state di agram appertains to a class not the product as a whol e, but because j i

l asses of the Air Gourmet product do not move from state to state, a di agram ' ' ( j the c

The object-oriented analysis appears compl ete lt t herefore should be revi ewed ! q

.Neither the client nor the development team should be w illing to accept the set

. of UM L diagram s as an adequate specification document A fter all, as pointed out in r: Section 1 l the specihcation docum ent is a contract between client and developers ' '

Accordingly, after the OOA is hnished, a m ore conventional speciEcation docum ent yhas to be drawn up Because that docum ent would be so sim ilar to m uch of the m aterial

!

.) aea t u A p 4 R 42 @ O bleu -o rien'ed Anolysis Phose

meal Iooded : Boolean oddress 2 : string

!

.

'

onboord : Boolean city : strins

E perc meal quoliy : int state/province : string

f l

' Ioaded as specifed,

'

k j

q

2 l

! ' I

A2.* A1 R GOURMET CM E STUDY: OBJECT-O RIENTED ANAWSIS 3@@ l

i (

iMake Reservation Order Special Meal

1

Store reservation in Air Create Iistor caterers : 1. !

(ake-offtime - 1 hrS check-in time < take-off time - 10 min) ;

i i

E1 Passenger Check-in

-) )

c

l

j ) '

Scan postcard

:data complete

- - j :

ê l 2

1 , (.

1

Get st ar t end dat es @ j p

j '

i .

I (i 1.t

l ) :

1 Just as w hen the structured paradigm was used for the specihcation of the Air

1 Gourmet product (Sect ion 1 1 5), at thi s poi nt the software project management pl an

l

i is drawn up Appendix F contai ns a software project management plan for develop- E

i Object-ori ent ed analysis is a specihc approach to the specifi cation phase,so the

gen-1 l challenges of the specihcation phase described in Section 1 l

g ;' Recal l that, as described i n Secti on l 6, the transiti on from object-oriented anal

-' ysis to object-oriented design is far smoother than the transition in the classi cal

l paradigm from the specification phase to the design phase In the classical paradigm,

i an initial task of the design phase is to decompose the product into modules In

con-' trast, the classes, the i t modul es' ' of the object -oriented design phase are extracted

h

during the object-orient ed analysis phase, ready for refinement during t he object t

-r oriented design phase The presence of classes from early in the OOA phase means

that the tem ptation to carry the OOA too far can be extremely strong

5

For example, eonsider the issue of allocation of m ethods to elasses One task

of the classical specihcation phase is to determ ine the data and actions of the target

r product How ever, allocation of the various actions to specihc modules should be

delayed until the design phase, because as pointed out in Section 1 l 16, we first have

' i At each step of the O OA process it is a good idea to m inim ize the information

' èi that would have to be reorganized during iteration.Therefore, allocation of methods

; to classes should wait until the design phase, no matter how tem pting it m ay be to go

just a li ttle further during the object-ori ented anal ysis phase.

ar e di scus sed i n Sect i on 12 7 The chapt erconcl udes wi t h t he ob ject - ori ent ed anal ys i i J '

of the Air Gourmet case study (Section 12.8) and a discussion of the challenges of 1,

t he object -oriented analysis phase (Secti on l 2 9) I 1 j

!

j 'p5 : I

1 1

ly books descri bi ng different versions of object-ori ented analysis include îcoad l

Ea r

and Yourdon, l 99 la; Rumbaugh et a1 , 199 l ; Shl aer and M ellor, l 992 , and Booch, I I !

19941 As mentioned in the chapter, these techni ques (and others not li sted here) are i 1

I ib

Objectory ( Jacobson, Christ erson Jonsson and Overgaard, l 9921 The Unif ied Soft- y ,

:

cobson, Booch, and Rumbaugh, 1 9991 Catalysis is another important object-ori ented : ,

ROOM is an object-oriented methodology for real -time software gsel ic, ) )

I

l 1 i Gull ekson

,and W ard 19951 Further i nformation on real -ti me object-oriented tech- I I j !

( '

and Rumbaugh, Jacobson, and Booch, 19991 The October 1999 issue of Communi- ! jcati ons ofthe ACM contai ns a broad variet y of papers on the use of UML 1

'

The noun-extraction technique used i n this chapter to extract candidate classes i s l

, : j formalized in t lurist o

'

I p

r

-good source of information on CRC cards i t

ber of articles on object-oriented analysis Several eomparisons of object-ori ented ' !

: l isons of object-ori ent ed and structured analysis t echni ques appears i n gFi chman and . r

j

: 1

i f scenarios to assist wi th object-oriented analysi s.

i I

, ; object-oriented anal ysi s?

f 1 2 4 Use object-ori ent ed analysis to specify the software for controlli ng the ATM of

Prob-E lem 8.9 There is no need to consider the details of the constituent hardware

compo-nents such as the card reader, printer, and cash dispenser lnstead, sim ply assum e that

be introduced wi thout adversel y atfecti ng t he project?

1 Q.6 W hat is the earli est point i n the object-oriented paradigm at which cl asses can be

m eaningfully introduced?

i

! 1Q.7' ls it possible to represent the dynamic m odel using a form alism other than the statel

! diagram described in this chapter? Explain your answ er

1 Q.8 W hy do you think that the attributes of the classes but not the methods are determined

during object-ori ent ed anal ysis?

i 1 2.@ (Term Project) Use object-ori ent ed analysis to specify the Broadlands Area Chi ldren' s i

Hospit al product descri bed in Appendix A.

I )

! j

, ,1

l '- I

I - h 1

j j

wi th the object-ori ent ed analysis of Section l 2.8 ' 1

j

! ! )

; !

I

11

Technologyjbr Real-lime Svstems, Prentice Hall, Upper Saddle River, NJs l996 I I y

Beck and Cunni ngham, 19891 K BECK AND W CUNNINGHAM, G A Labor atory for Teaching k 1

, ( j ,

(Bellinzona, Fugini and Pernici, 19951 R BELLINZONA, M G FIJGINI, AND B PERNICI, ! I 1 '

' 'Reusing Specifications in 0 0 Applications, ,,IEEE Sojtware 12 (M arch 1995) ; 1 '

pp.656-75.

' h 4 I i cat i ons, 2nd ed , l ( Booch, 1 9941 G Booci l Ob ject - or i nt e dAnal ysi s and Desi gn w? / pp

(Booch, Rumbaugh, and Jacobson, 19991 G Boocu, J RUMBAUGH, AND 1 JACOBSON, The !

UM L Users G uide, A ddison W esley, Reading, M As 1999 8

' (Coad an d Yourdon, 199 1al P COAD ANt) E YOURDON, Object-oriented Analyxiy 2nd ed., l t ''

,j Y

ourdon Press, Englewood Cliffs, NJ, 199 1 ( ! t '(Coleman, Hayes, and Bear, 19921 D COLEMAN, F HAYES, AND S BEAR, ttlntroducing i j

'

:Jn Software Engineering 18 (January 1992), pp 9- I8 '

(Coleman et al., 19941 D COLEMAN, P ARNOI-D, S BODOFF, C DOLLIN, l1 GILCHRIST, (t *

F HAYES, AND P JEREMAES, Object-oriented Development The Fusion M ethod, ! '

$ j

!de Champeaux and Faure, l9921 D DE CHAMPEAIJX ANo P FAURE, ''A Comparative Study ' L

(Embley Jackson, and W oodfield, 19951 D W EMBI.EY, R B JAclksoN ANI) 1

S N W OODFIEI-D, 00 Systems Analysis: Is lt or Isn t It? IEEE s'q/rktlre 12 (July : j .

Fowlers 1997b1 M FOWLER WITH K SCOTT, UM L Distilled, Addison-Wesley, Reading, M A, 1

g

-lareland Gery, 19971 D HAREL AND E GERY, ''Executable Object Modeling with ' 1

Statecharts

(acobson, Booch, and Rumbaugh, l999) J RUMBAIJGH, G Boocll, AND 1 JACOBSON, The

Unfi ed Soft ware Devel opment #mc

' !

l '

e n > * e < $ l I !

l (

!i

: j)j '

i .

,m any have been used only by their authors Som e I

design strategies particularly those developed by academics have a lirm theoretical basis Others includingmanv draw n uo bv academ ics, are more praem atic in nature', thev were put forward because their authors E ,

; found that they worked well in practice M ost design techniques are m anual, but automation increasingly is ' ,

' becoming an im portant aspect of design, if only to assistn the management of documentation ! 1

1 1

' Notwit hstandi ng t hi s pl ethora of desi gn techniques, there is a certain underlying pattern A major t heme j j ,

j 'j

of this book is that two essential aspects of a product are its actions and the data on which the actions operate I

'

I jfirst

A weakness of action-oriented design techniques is that they concentrate on the actions', the data are of : 1

I 1 only secondary im portance Data-oriented design techniques sim ilarly emphasize the data, to the detrim ent of I

j ' the actions.The solution is to use object-ori ented techniques, which gi ve equal wei ght to actions and data ln ! 'rthis chapter, acti on- and data-ori ent ed desi gn are described first, then object-ori ented desi gn i s described Just 1 q I

as an object i ncorporat es both acti on and data, so object-ori ent ed design combi nes features of acti on-oriented I '

and dat a- or i ent ed desi gn Ther efor e, a basi c unders t andi ng of act i on- and dat a- ori ent ed des i gn i s needed t o I L 1 '

( get a full understanding of object-oriented design ( 1

:?Before specilic design techniques are exam ined, some general rem arks m ust be m ade regarding the !. ,

1

i ''design phase

; I

J -

-:

I '

The software design phase consists of three activities: architectural design, detail- :

ed design, and design testing The input to the design process is the specihcation I '

;

I '

h j ' '

I t I

:

i descripti on of what the product is to do

.The output is the design document, a

docu-1

ment, a description of how the product is to achieve this

t During architectural design (also known as general design,Iogical design, or

j high-level design) a modular decomposition of the product is developed That i

! the specifications are analyzed carefully, and a m odule structure that has the desired' functionality is produced The output from this activity is a list of the modules and a

deseription of how they are to be interconnected From the viewpoint of abstraction,during architectural design, the existence of certain modules is assumed', the design

during the object -oriented analysis phase (Chapt er 12) This is because the hrst step

in OOA is to determ ine the classes Because a class is a type of module, some of themodular decom position has been perform ed during the analysis phase

The next activity is detailed design, also known as modular design, physical

s two-stage process is typical of abstraction as described in Chapter 7 First,

) exist Then each m odule is designed in turn, w ithout regard to its being a com ponent

E j

. j lt was stated previously thatthe design phase has three activities and thatthe third

i activity is testing The word activity was used, rather than stage or step, to emphasize

j that testing is an integral part of design, just as it i s an int egral part of t he enti re

software development and m aintenance process Testing is not something performedonly after the architectural design and detailed design have been com pleted

A variety of different design techniques now are described, hrst action-oriented

techni ques, then data-oriented techniques, and final ly object-ori ent ed techni ques.

: j

mod-1 ules with high cohesion and low coupling.W e now describe two practi cal techniques

' j

for achi eving t hi s design objecti ve, data f low analysis (Secti on 13 3) and

transac-. (

i the specihcations can be represented by a data flow diagram, and because (at least

;: t in theory) every product can be represented by a DFD, data flow analysis is uni-

) i processi ng products (Transaction analysi s, described in Secti on 13 4, is a good way

f of decom posing transaction processing products into modules.) i

'

j

1

' j:i1

1 i

I I

Data flow analysis (DFA) is a design technique for achieving modules with high

h i on I t can be used i n conjunct i on wi t h most s peci hcat i on t echni ques Her e 1

i (Sect i on l l 3) The ë I DFA ipr es ent ed i n con junct i on wi t h st r uct ured syst ems anal ys

' :input tothe technique is adataqow diagram.A key pointis that,oncethe D FD has been

com pleted, the softw are designer has precise and complete infbrm ation regarding the 1 ;

s :input to and output from the product.

Consider the flow of data in the product represented by the DFD of Figure 13.1 j .

point in the flow of data at which the output can be identified as such rather than as '

Using the points ot

posed into t hree modules: the i nput modul e transform modul e, and output modul e !

i: Now each m odule is taken in turn, its points of highest abstraction found, and the !

d l es decomposed agai n Thi s pr ocedur e iconti nued s epwi s e unt ieach modul e 1

1 In tai-rness it s ou h ld be pointed out that m inor m oditications might have to be 1; l ,

: made to the decom position to achieve the lowestpossible coupling Data flow analysis r' is a way of achieving high cohesion The aim of com posite-structured design is high

cohesion butalso low coupling.To achieve the latter, isometimes is necessaryto m ake !

1

gure laa Dof a f l ow di agram showi ng fl ow of dat o and act i ons of product j ,

j , '