MetaML is a statically typed functional programming language with special support for program generation. In addition to providing the standard features of contemporary programming languages such as Standard ML, MetaML provides three staging annotations. These staging annotations allow the construction, combination, and execution of object-programs. Our thesis is that MetaML's three staging annotations provide a useful, theoretically sound basis for building program generators. This dissertation reports on our study of MetaML's staging constructs, their use, their implementation, and their formal semantics. Our results include an extended example of where MetaML allows us to produce efficient programs, an explanation of why implementing these constructs in traditional ways can be challenging, two formulations of MetaML's semantics, a type system for MetaML, and a proposal for extending ...

Introduction Today's virtual machines use a layer of software that allows programs compiled in one instruction set to be executed on a processor executing a (different) native instruction set. Virtual machines have become popular in recent years for providing platform independence; however, virtual machines also open many new opportunities for enhancing performance. The co-design of virtual machine software and the underlying hardware microarchitecture will enable enhanced instruction level parallelism and more adaptable performance mechanisms than are possible when hardware and application software are separated by instruction set architectures as is traditionally done. In future high performance computers, a virtual instruction set architecture (V-ISA) will be the level for maintaining architectural compatibility. The V-ISA will be implemented with a virtual machine that blends software and hardware in a symbiotic manner via co-design. The hardware will support an implementationdep

Programmers constantly design, implement, and program in little languages. Two different approaches to the implementation of little languages have evolved. One emphasizes the design of little languages from scratch, using conventional technology to implement interpreters and compilers. The other advances the idea of extending a general-purpose host language; that is, the little language shares the host language’s features (variables, data, loops, functions) where possible; its interpreters and compilers; and even its type soundness theorem. The second approach is often called a language embedding. This paper directs the attention of little language designers to a badly neglected area: the programming environments of little languages. We argue that an embedded little language should inherit not only the host language’s syntactic and semantic structure, but also its programming environment. We illustrate the idea with our DrScheme programming environment and S-XML, a little transformation language for XML trees. DrScheme provides a host of tools for Scheme: a syntax analysis tool, a static debugger, an algebraic stepper, a portable plugin system, and an interactive evaluator. S-XML supports the definition of XML languages using a simple form of schemas, the convenient creation of XML data, and the definition of XML transformations.

. We compare the expressive power of exceptions and continuations when added to a language with local state in the setting of operational semantics. Continuations are shown to be more expressive than exceptions because they can cause a function call to return more than once, whereas exceptions only allow discarding part of the calling context. 1 Introduction Exceptions are part of nearly all modern programming languages, including mainstream ones like Java and C++. Continuations are present only in Scheme and the New Jersey dialect of ML, yet are much more intensely studied by theoreticians and logicians. The relationship between exceptions and continuations is not as widely understood as one would hope, partly because continuations, though in some sense canonical, are more powerful than would at rst appear, and because the control aspect of exceptions can be obscured by intricacies of typing and syntax. We have recently shown that exceptions and continuations, when added to a purely...

. This note introduces Peter Landin's 1965 technical report "A Generalization of Jumps and Labels", which is reprinted in this volume. Its aim is to make that historic paper more accessible to the reader and to help reading it in context. To this end, we explain Landin's control operator J in more contemporary terms, and we recall Burge's solution to a technical problem in Landin's original account. Keywords: J-operator, secd-machine, call/cc, goto, history of programming languages 1. Introduction In the mid-1960's, Peter Landin was exploring functional programming extended with control. Initially, Landin had added the control operator J (for jump) to his language of Applicative Expressions in order to explicate the goto statement from Algol 60 [10]. This general control construct, however, was powerful enough to have independent interest quite beyond the application to Algol: Landin wrote a series of technical reports [11, 12, 13] on functional programming with control, the most si...

The aspect-oriented programming community devotes lots of energy into the provision of complex static language constructs to reason about eventual dynamic properties of a program.

Macros are reflective tools that operate on the representation of programs. Though having been used, and still being useful, for more than thirty years, their semantics and pragmatics are still unclear. This paper proposes a new model to understand the macroexpansion process; this model is based on a reflective tower of macroexpansion engines. 1 Introduction Macros are definitely a forte of Lisp dialects. They confer upon the powerful users a means to adapt their language to their problems and this is, in our mind, one of the key reasons for Lisp's longevity. Despite the various attempts to standardize macros [Ste90, CR91b] and even if most of them are just quite simple abbreviations, they still represent one of the thorniest problems that occur when porting code between dialects of Lisp: a sure sign of their ambiguous nature. Even in Scheme, macros still have a dark side with respect to their scope, their extent and their meaning i.e., their definition language. This paper proposes a...

Recent work proposed defining type-safe macros via interpretation into a multi-stage language. The utility of this approach was illustrated with a language called MacroML, in which all type checking is carried out before macro expansion. Building on this work, the goal of this paper is to develop a macro language that makes it easy for programmers to reason about terms locally. We show that defining the semantics of macros in this manner helps in developing and verifying not only type systems for macro languages but also equational reasoning principles. Because the MacroML calculus is sensetive to renaming of (what appear locally to be) bound variables, we present a calculus of staged notational definitions (SND) that eliminates the renaming problem but retains MacroML's phase distinction. Additionally, SND incorporates the generality of Griffin's account of notational definitions. We exhibit a formal equational theory for SND and prove its soundness.

Monads are widely used in Haskell for modeling computational effects, but defining monads remains a daunting challenge. Since every part of a monad’s definition depends on its computational effects, programmers cannot leverage the common behavior of all monads easily and thus must build from scratch each monad that models a new computational effect. I propose the Unimo framework which allows programmers to define monads and monad transformers in a modular manner. Unimo contains a heavily parameterized observer function which enforces the monad laws, and programmers define a monad by invoking the observer function with arguments that specify the computational effects of the monad. Since Unimo provides the common behavior of all monads in a reusable form, programmers no longer need to rebuild the semantic boilerplate for each monad and can instead focus on the more interesting and rewarding task of modeling the desired computational effects.

Continuations can be used to explain a wide variety of control behaviours, including calling/returning (procedures), raising/handling (exceptions), labelled jumping (goto statements), process switching (coroutines), and backtracking. However, continuations are often manipulated in a highly stylised way, and we show that all of these, bar backtracking, in fact use their continuations linearly ; this is formalised by taking a target language for cps transforms that has both intuitionistic and linear function types.