This paper presents an implementation and critical analysis of a technique for automated, rigorous scientific program comprehension and error detection. The procedure involves taking a user's existing code, adding semantic declarations for some primitive variables, symbolically executing the user's code, and recognizing semantic concepts from the symbolic expressions generated. This analysis provides high-level, semantic information and detects errors in a user's code. Symbolic execution of a 5k line of code (LOC) scientific code demonstrates implementation of a practical symbolic execution/semantic analysis tool. Despite the promise of this technique, one key challenge is examined, namely ensuring reliability during the symbolic execution/semantic analysis procedure

This paper concerns automated analysis of the meaning or semantics of scientific and engineering code. The procedure involves taking a user's existing code, adding semantic declarations for some primitive variables, and automatically identifying formulae. Parsers encode domain knowledge and recognize formulae in different disciplines including physics, numerical methods, mathematics, and geometry. The parsers will automatically recognize and document some static, semantic concepts and help locate some program semantic errors. Results are shown for three intensively studied codes and seven blind test cases; all test cases are state of the art scientific codes. These techniques may apply to a wider range of scientific codes. If so, the techniques could reduce the time, risk, and effort required to develop and modify scientific codes.