A Two-Phase Static Analysis for Reliable Adaptation

A type system for first-class layers with inheritance, subtyping, and swapping

“…Subjective-C [23], an extension of Objective-C with COP, adopts imperative context activation with imperative context relationship description, which supports various kinds of declarations of dependency between layers, such as implication, requirement, and exclusion. Context Petri Nets [11,12] is a context-oriented extension of Petri Nets, and helps formalization of description of context dependencies in Subjective-C. M L CoDa [19,20] provides two kinds of components; one for declarative description of context dependencies and the other for functional computation. It also provides a type and effect system and a loading-time verification mechanism that detects failures in adaptation.…”

A Context-Oriented Extension of F#

“…for Actually, the code above is F#, a dialect of ML. Since we wanted to keep the F# parser unmodified, the syntax is slightly different from the one used in [9] and recalled in Section 2. This is only an implementation detail, because a simple macro-expansion suffices to translate the original syntax in the intermediate notation used in this section.…”

“…In addition to the formal aspects of ML CoDa studied in [9], a main feature of our approach is that a single and fairly small set of constructs is sufficient enough for becoming a practical programming language, as shown in the present paper. Indeed, ML CoDa can easily be embedded in a real programming eco-system, in our case .NET, so preserving compatibility with its future extensions and with legacy code.…”

“…In previous papers [10,9] we followed this line of research from a foundational perspective and we proposed a programming language core, called ML CoDa , specifically designed for adaptation and equipped with a clear formal semantics. It has two components: a logic constituent for programming the context and a functional one for computing.…”

“…Its higher-order facilities are essential to exchange the bundle of functionalities required to manage adaptivity changes. Moreover, ML CoDa offers a further support through a static analysis that guarantees programs to always be able to adapt in every context [9].…”