Action Machines: A Framework for Encoding and Composing Partial Behaviors

Abstract

We describe action machines, a framework for encoding and composing partial behavioral descriptions. Action machines encode behavior as a variation of labeled transition systems where the labels are observable activities of the described artifact and the states capture full data models. Labels may also have structure, and both labels and states may be partial with a symbolic representation of the unknown parts. Action machines may stem from software models or programs, and can be composed in a variety of ways to synthesize new behaviors. The composition operators described here include synchronized and interleaving parallel composition, sequential composition, and alternating simulation. We use action machines in analysis processes such as model checking and model-based testing. The current main application is in the area of model-based conformance testing, where our approach addresses practical problems users at Microsoft have in applying model-based testing technology.

Interaction Coverage Meets Path Coverage by SMT Constraint Solving

“…For the general approach on how such models are executed and explored by Spec Explorer, we refer readers to the other publications about this technology ( [10,11,13,14]). For this exposition it should suffice to know that model exploration yields a transition system where each state represents the model state (in this case the file server content).…”

“…Our solution is integrated into the Spec Explorer model-based testing tool [10,11,13] version 2, and is driven by application requirements in protocol testing for Microsoft's interoperability program [12]. In protocol testing, actions represent messages sent via the network, which often have many different parameters.…”

Protocol Modeling with Model Program Composition

“…Various approaches, including explicit state exploration [30] as well as symbolic reachability analysis [36], may be applied. The main difference compared to composition of action machines [23] is that composition of model programs is syntactic, whereas composition of action machines is defined in the style of natural semantics using inference rules and symbolic computation that incorporates the notion of computable approximations of subsumption checking between symbolic states. The computable approximations reflect the power of the underlying decision procedures that are being used and are an integral part of the composition, using a three-valued logic.…”

Composition of Model Programs

“…Various approaches, including explicit state exploration as well as exploration with symbolic labels and states, may be applied. For example, action machines [8] rely on symbolic techniques. The main difference compared to composition of action machines is that composition of model programs is syntactic, whereas composition of action machines is defined in the style of natural semantics using inference rules and symbolic computation that incorporates the notion of computable approximations of subsumption checking between symbolic states.…”

“…More about model-based testing applications and further motivation for the composition of model programs can be found in [4,8,17,16]. The most recent work related to model programs where composition is discussed from a practical perspective is the forthcoming textbook [11].…”