4.8 Executable Models
Executable models are built using graphic tools using such notations as finite-state machines,
UML
or
XML. There is a number of reasons for spending the time and money required to build such models:
??? Much of the system to be built will be constructed from existing components or using known
technologies. Some may be unknown and these need to be prototyped. The results will be small
demonstration fragments.
??? Some specifications are so large that they contain critical inconsistencies which are best exposed
by using a graphic notation such as
UML
.
UML
can now be used to generate executable code.
??? The feasibility of some models cannot be determined from simply reading the specification,
particularly where timing is concerned. An executable model can give early warning.
The models exist at three levels:
???
Prose
: the informal specification of the use case such as A.2.10 in Appendix A.
???
Diagram
: which can be any
UML
, state case, or other diagram supported by the modeling system.
???
Code
: which generates a pseudo user-interface. In the case of the
Informal specification
example,
the interface can be made to resemble an ATM as shown in Figure 4.6. Yes folks, you can get a
relatively realistic picture to play with.
FIGURE 4.4
UP phases and models
Phase Models
Use case Requirements
Inception
Analysis Analysis
Design Design Deploy-
ment
Elaboration
Implementation Implementation
Test Test
Construction
50
Manage Software Testing
FIGURE 4.
Pages:
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174