No Arabic abstract
Context-Oriented Programming (COP) is a programming paradigm to encourage modularization of context-dependent software. Key features of COP are layers---modules to describe context-dependent behavioral variations of a software system---and their dynamic activation, which can modify the behavior of multiple objects that have already been instantiated. Typechecking programs written in a COP language is difficult because the activation of a layer can even change objects interfaces. Inoue et al. have informally discussed how to make JCop, an extension of Java for COP by Appeltauer et al., type-safe. In this article, we formalize a small COP language called ContextFJ$_{<:}$ with its operational semantics and type system and show its type soundness. The language models main features of the type-safe version of JCop, including dynamically activated first-class layers, inheritance of layer definitions, layer subtyping, and layer swapping.
We study the fundamental efficiency of delimited control. Specifically, we show that effect handlers enable an asymptotic improvement in runtime complexity for a certain class of functions. We consider the generic count problem using a pure PCF-like base language $lambda_b$ and its extension with effect handlers $lambda_h$. We show that $lambda_h$ admits an asymptotically more efficient implementation of generic count than any $lambda_b$ implementation. We also show that this efficiency gap remains when $lambda_b$ is extended with mutable state. To our knowledge this result is the first of its kind for control operators.
XML database query languages such as XQuery employ regular expression types with structural subtyping. Subtyping systems typically have two presentations, which should be equivalent: a declarative version in which the subsumption rule may be used anywhere, and an algorithmic version in which the use of subsumption is limited in order to make typechecking syntax-directed and decidable. However, the XQuery standard type system circumvents this issue by using imprecise typing rules for iteration constructs and defining only algorithmic typechecking, and another extant proposal provides more precise types for iteration constructs but ignores subtyping. In this paper, we consider a core XQuery-like language with a subsumption rule and prove the completeness of algorithmic typechecking; this is straightforward for XQuery proper but requires some care in the presence of more precise iteration typing disciplines. We extend this result to an XML update language we have introduced in earlier work.
Object-oriented scripting languages such as JavaScript or Python gain in popularity due to their flexibility. Still, the growing code bases written in the languages call for methods that make possible to automatically control the properties of the programs that ensure their stability in the running time. We propose a type system, called Lucretia, that makes possible to control the object structure of languages with reflection. Subject reduction and soundness of the type system with respect to the semantics of the language is proved.
Session types are used to describe communication protocols in distributed systems and, as usual in type theories, session subtyping characterizes substitutability of the communicating processes. We investigate the (un)decidability of subtyping for session types in asynchronously communicating systems. We first devise a core undecidable subtyping relation that is obtained by imposing limitations on the structure of types. Then, as a consequence of this initial undecidability result, we show that (differently from what stated or conjectured in the literature) the three notions of asynchronous subtyping defined so far for session types are all undecidable. Namely, we consider the asynchronous session subtyping by Mostrous and Yoshida for binary sessions, the relation by Chen et al. for binary sessions under the assumption that every message emitted is eventually consumed, and the one by Mostrous et al. for multiparty session types. Finally, by showing that two fragments of the core subtyping relation are decidable, we evince that further restrictions on the structure of types make our core subtyping relation decidable.
Programming languages serve a dual purpose: to communicate programs to computers, and to communicate programs to humans. Indeed, it is this dual purpose that makes programming language design a constrained and challenging problem. Inheritance is an essential aspect of that second purpose: it is a tool to improve communication. Humans understand new concepts most readily by first looking at a number of concrete examples, and later abstracting over those examples. The essence of inheritance is that it mirrors this process: it provides a formal mechanism for moving from the concrete to the abstract.