Rapid API Verification Engine (RAVE)

To keep a large software system development effort on schedule and on budget, it is critical to ensure compliance of all software modules to an Application Programming Interface (API) specification. Currently, the Joint Tactical Radio System (JTRS) discovers API compliance shortfalls in contractor developed software modules on an ad-hoc basis and these defects are reworked only with considerable penalties in schedule slip and cost. This would not happen if every module developer could automatically and independently verify his own component’s compliance prior to delivery, and if the systems integrator could systematically verify a component’s correctness after delivery and prior to acceptance. To succeed, we must move beyond traditional and fallible compliance solutions that rely on simple textual matching. We propose to develop a Rapid API Verification Engine (RAVE) to perform lexical, syntactic, and semantic analysis on API code to infer its true meaning. Key features of our system include a correct and efficient parsing engine built on decades of solid research, a correction suggestion engine that can automatically suggest fixes to common sources of problems, and the use of a model-view-controller framework, which allows us to quickly add multiple user interfaces and meet a broad set of client needs.