A programming methodology based on data abstraction |
| |
Authors: | John F. Isner |
| |
Affiliation: | (1) National Geodetic Survey, National Ocean Survey, NOAA, 6001 Executive Boulevard, 20852 Rockville, MD |
| |
Abstract: | The view of data offered by traditional programming languages and data management facilities is not sufficiently abstract. A recent development in computer science is the concept of abstract data type, or data abstraction. A data abstraction is a class of objects whose behavior is completely defined by a representation-independent specification. Data abstraction creates a natural separation of concerns, making it an ideal basis for a system design methodology. The methodology requires a formal specification for each data abstraction identified. Parnas'state machine specification technique is illustrated for a level net data abstraction. It a language supporting abstraction at the compiler level is not available, data abstractions may be simulated in any of the popular high-level languages. Finally, an object oriented approach to data base design is suggested. Presented at International Symposium on Management of Geodetic Data, Copenhagen, August 24–26, 1981. |
| |
Keywords: | |
本文献已被 SpringerLink 等数据库收录! |
|