A Generic Static Analysis Framework for Domain-specific Languages

Software used to monitor and control operations within an automation system is defined using domain-specific languages. Latent errors in the control code, if left undetected, can lead to unexpected system failures compromising the safety and the security of the automation system. Traditional analysis techniques are insufficient to detect such errors as they do not cater specifically to the underlying domain-specific language. However, given the diversity of different automation domains, there is no standard platform for analysis of these languages. This paper proposes a generic static analysis framework for domain-specific languages used in the automation domain. The analysis approach exhaustively detects runtime errors in control code and ensures compliance to good programming practices. These runtime errors and coding violations are checked against abstract syntax trees and control flow graphs derived from the code. Data Flow Analysis (DFA), Abstract interpretation and pattern-based matching techniques are used to identify domain specific errors and coding violations for control languages.