Oleg Kiselyov

Finally tagless, partially evaluated: Tagless staged interpreters for simpler typed languages

We have built the first family of tagless interpretations for a higher-order typed object language in a typed metalanguage (Haskell or ML) that require no dependent types, generalized algebraic data types, or postprocessing to eliminate tags. The statically type-preserving interpretations include an evaluator, a compiler (or staged evaluator), a partial evaluator, and call-by-name and call-by-value continuation-passing style (CPS) transformers. Our principal technique is to encode de Bruijn or higher-order abstract syntax using combinator functions rather than data constructors. In other words, we represent object terms not in an initial algebra but using the coalgebraic structure of the λ-calculus. Our representation also simulates inductive maps from types to types, which are required for typed partial evaluation and CPS transformations. Our encoding of an object term abstracts uniformly over the family of ways to interpret it, yet statically assures that the interpreters never get stuck. This family of interpreters thus demonstrates again that it is useful to abstract over higher-kinded types.

Strongly typed heterogeneous collections

A heterogeneous collection is a datatype that is capable of storing data of different types, while providing operations for look-up, update, iteration, and others. There are various kinds of heterogeneous collections, differing in representation, invariants, and access operations. We describe HLIST - a Haskell library for strongly typed heterogeneous collections including extensible records. We illustrate HLISTs benefits in the context of type-safe database access in Haskell. The HLIST library relies on common extensions of Haskell 98. Our exploration raises interesting issues regarding Haskells type system, in particular, avoidance of overlapping instances, and reification of type equality and type unification.

Backtracking, interleaving, and terminating monad transformers: (functional pearl)

We design and implement a library for adding backtracking computations to any Haskell monad. Inspired by logic programming, our library provides, in addition to the operations required by the MonadPlus interface, constructs for fair disjunctions, fair conjunctions, conditionals, pruning, and an expressive top-level interface. Implementing these additional constructs is easy in models of backtracking based on streams, but not known to be possible in continuation-based models. We show that all these additional constructs can be generically and monadically realized using a single primitive msplit. We present two implementations of the library: one using success and failure continuations; and the other using control operators for manipulating delimited continuations.

Embedded Probabilistic Programming

Two general techniques for implementing a domain-specific language (DSL) with less overhead are the finally-tagless embedding of object programs and the direct-style representation of side effects. We use these techniques to build a DSL for probabilistic programming , for expressing countable probabilistic models and performing exact inference and importance sampling on them. Our language is embedded as an ordinary OCaml library and represents probability distributions as ordinary OCaml programs. We use delimited continuations to reify probabilistic programs as lazy search trees, which inference algorithms may traverse without imposing any interpretive overhead on deterministic parts of a model. We thus take advantage of the existing OCaml implementation to achieve competitive performance and ease of use. Inference algorithms can easily be embedded in probabilistic programs themselves.

A methodology for generating verified combinatorial circuits

High-level programming languages offer significant expressivity but provide little or no guarantees about resource use. Resource-bounded languages --- such as hardware-description languages --- provide strong guarantees about the runtime behavior of computations but often lack mechanisms that allow programmers to write more structured, modular, and reusable programs. To overcome this basic tension in language design, recent work advocated the use of Resource-aware Programming (RAP) languages, which take into account the natural distinction between the development platform and the deployment platform for resource-constrained software.This paper investigates the use of RAP languages for the generation of combinatorial circuits. The key challenge that we encounter is that the RAP approach does not safely admit a mechanism to express a posteriori (post-generation) optimizations. The paper proposes and studies the use of abstract interpretation to overcome this problem. The approach is illustrated using an in-depth analysis of the Fast Fourier Transform (FFT). The generated computations are comparable to those generated by FFTW.

Delimited dynamic binding

Dynamic binding and delimited control are useful together in many settings, including Web applications, database cursors, and mobile code. We examine this pair of language features to show that the semantics of their interaction is ill-defined yet not expressive enough for these uses.We solve this open and subtle problem. We formalise a typed language DB+DC that combines a calculus DB of dynamic binding and a calculus DC of delimited control. We argue from theoretical and practical points of view that its semantics should be based on delimited dynamic binding: capturing a delimited continuation closes over part of the dynamic environment, rather than all or none of it; reinstating the captured continuation supplements the dynamic environment, rather than replacing or inheriting it. We introduce a type- and reduction-preserving translation from DB + DC to DC, which proves that delimited control macro-expresses dynamic binding. We use this translation to implement DB+DC in Scheme, OCaml, and Haskell.We extend DB + DC with mutable dynamic variables and a facility to obtain not only the latest binding of a dynamic variable but also older bindings. This facility provides for stack inspection and (more generally) folding over the execution context as an inductive data structure.

Extensible effects: an alternative to monad transformers

We design and implement a library that solves the long-standing problem of combining effects without imposing restrictions on their interactions (such as static ordering). Effects arise from interactions between a client and an effect handler (interpreter); interactions may vary throughout the program and dynamically adapt to execution conditions. Existing code that relies on monad transformers may be used with our library with minor changes, gaining efficiency over long monad stacks. In addition, our library has greater expressiveness, allowing for practical idioms that are inefficient, cumbersome, or outright impossible with monad transformers. Our alternative to a monad transformer stack is a single monad, for the coroutine-like communication of a client with its handler. Its type reflects possible requests, i.e., possible effects of a computation. To support arbitrary effects and their combinations, requests are values of an extensible union type, which allows adding and, notably, subtracting summands. Extending and, upon handling, shrinking of the union of possible requests is reflected in its type, yielding a type-and-effect system for Haskell. The library is lightweight, generalizing the extensible exception handling to other effects and accurately tracking them in types.

Shifting the stage: staging with delimited control

It is often hard to write programs that are efficient yet reusable. For example, an efficient implementation of Gaussian elimination should be specialized to the structure and known static properties of the input matrix. The most profitable optimizations, such as choosing the best pivoting or memoization, cannot be expected of even an advanced compiler because they are specific to the domain, but expressing these optimizations directly makes for ungainly source code. Instead, a promising and popular way to reconcile efficiency with reusability is for a domain expert to write code generators. Two pillars of this approach are types and effects. Typed multilevel languages such as MetaOCaml ensure _safety_: a well-typed code generator neither goes wrong nor generates code that goes wrong. Side effects such as state and control ease correctness: an effectful generator can resemble the textbook presentation of an algorithm, as is familiar to domain experts, yet insert let for memoization and if for bounds-checking, as is necessary for efficiency. However, adding effects blindly renders multilevel types unsound. We introduce the first two-level calculus with control effects and a sound type system. We give small-step operational semantics as well as a continuation-passing style (CPS) translation. For soundness, our calculus restricts the code generators effects to the scope of generated binders. Even with this restriction, we can finally write efficient code generators for dynamic programming and numerical methods in direct style, like in algorithm textbooks, rather than in CPS or monadic style.

Soutei, a logic-based trust-management system

We describe the design and implementation of a trust-management system Soutei, a dialect of Binder, for access control in distributed systems. Soutei policies and credentials are written in a declarative logic-based security language and thus constitute distributed logic programs. Soutei policies are modular, concise, and readable. They support policy verification, and, despite the simplicity of the language, express role- and attribute-based access control lists, and conditional delegation. We describe the real-world deployment of Soutei into a publish-subscribe web service with distributed and compartmentalized administration, emphasizing the often overlooked aspect of authorizing the creation of resources and the corresponding policies. Soutei brings Binder from a research prototype into the real world. Supporting large, truly distributed policies required non-trivial changes to Binder, in particular mode-restriction and goal-directed top-down evaluation. To improve the robustness of our evaluator, we describe a fair and terminating backtracking algorithm.

Multi-stage programming with functors and monads: eliminating abstraction overhead from generic code

With Gaussian Elimination as a representative family of numerical and symbolic algorithms, we use multi-stage programming, monads and Ocamls advanced module system to demonstrate the complete elimination of the abstraction overhead while avoiding any inspection of the generated code. We parameterize our Gaussian Elimination code to a great extent (over domain, matrix representations, determinant tracking, pivoting policies, result types, etc) at no run-time cost. Because the resulting code is generated just right and not changed afterwards, we enjoy MetaOCamls guaranty that the generated code is well-typed. We further demonstrate that various abstraction parameters (aspects) can be made orthogonal and compositional, even in the presence of name-generation for temporaries and other bindings and “interleaving” of aspects. We also show how to encode some domain-specific knowledge so that “clearly wrong” compositions can be statically rejected by the compiler when processing the generator rather than the generated code.