Abstract
Technical software systems contain extensive and complex computations that are frequently implemented in an optimized and unstructured way. Computations are, therefore, hard to comprehend from source code. If no other documentation exists, it is a tedious endeavor to understand which input data impact on a particular computation and how a program does achieves a particular result. We apply symbolic execution to automatically extract computations from source code. Symbolic execution makes it possible to identify input and output data, the actual computation as well as constraints of a particular computation, independently of encountered optimizations and unstructured program elements. The proposed technique may be used to improve maintenance and reengineering activities concerning legacy code in scientific and engineering domains.
Original language | English |
---|---|
Pages | 462-466 |
Number of pages | 5 |
DOIs | |
Publication status | Published - 2013 |
Keywords
- Legacy code
- knowledge extraction
- reverse engineering
- symbolic execution