Loading...
Please wait, while we are loading the content...
Similar Documents
From OOA to C++: The Missing Link
| Content Provider | Semantic Scholar |
|---|---|
| Author | Mellor, Stephen J. |
| Copyright Year | 1994 |
| Abstract | In 1912, one Charles Dawson discovered a skull fragment that was accepted as evidence of the long-sought missing link between apes and humans. In 1953, Piltdown Man as the fossil was known, was shown to be a forgery, a combination of a 600-year-old human skull and a modern orangutan. Sadly, this tale is redolent of today's methods for making the transition from object-oriented analysis to code: two related, but fundamentally disjoint parts are stuck together to make what seems to be a rational progression from one step to another. As a solution to this problem, this paper will show how to build a detailed and specific mapping from an object-oriented analysis into C++ code. The code will be built using a revolutionary technique called Recursive Design, a part of the Shlaer-Mellor Method. The approach is based on the construction of an architecture that defines the rules for the organization of data, control and algorithm, and then a translation, via the architecture, from the application directly to C++. One of the many advantages claimed for object-oriented analysis and design is that the transition between the analysis and design is significantly smoother than for other methods. There are two main reasons for this claim. First, in most object-oriented methods, the notation for analysis and design activities is the same. A single notation helps alleviate the discontinuity between the two activities. On the other hand, in structured analysis and design there are two notations: the data flow diagram for analysis and the structure chart for design. There is then a need to say the same thing twice, once in analysis notation and once in design notation. The second reason for the claim is that the structure of an object-oriented design and implementation can be the same as for the analysis. The argument runs as follows. The object-oriented analyst abstracts objects on the basis of the real-world subject matter under analysis. The resulting objects tightly connect data and function to act as a unit. As in the real world, each of the objects is distinct from the other objects and loosely coupled. The designer then turns the objects in the analysis directly into the objects in the design. Any change in the real world leads to a corresponding change in the analysis, which, in turn, leads to a corresponding change in the design. The size and cost of making the changes are as low as can be, because the analysis models represent the real world exactly, |
| File Format | PDF HTM / HTML |
| Alternate Webpage(s) | http://ooatool.com/docs/Missing94.pdf |
| Language | English |
| Access Restriction | Open |
| Content Type | Text |
| Resource Type | Article |
