International Symposium on Functional and Logic Programming 

About: International Symposium on Functional and Logic Programming is an academic conference. The conference publishes majorly in the area(s): Logic programming & Functional programming. Over the lifetime, 264 publications have been published by the conference receiving 4113 citations.

Code generation via higher-order rewrite systems

Abstract: We present the meta-theory behind the code generation facilities of Isabelle/HOL. To bridge the gap between the source (higher-order logic with type classes) and the many possible targets (functional programming languages), we introduce an intermediate language, Mini-Haskell. To relate the source and the intermediate language, both are given a semantics in terms of higher-order rewrite systems (HRSs). In a second step, type classes are removed from Mini-Haskell programs by means of a dictionary translation; we prove the correctness of this step. Building on equational logic also directly supports a simple but powerful algorithm and data refinement concept.

Probabilistic Inference by Program Transformation in Hakaru (System Description)

Abstract: We present Hakaru, a new probabilistic programming system that allows composable reuse of distributions, queries, and inference algorithms, all expressed in a single language of measures. The system implements two automatic and semantics-preserving program transformations—disintegration, which calculates conditional distributions, and simplification, which subsumes exact inference by computer algebra. We show how these features work together by describing the ideal workflow of a Hakaru user on two small problems. We highlight our composition of transformations and types in design and implementation.

A brief survey of quantum programming languages

Abstract: This article is a brief and subjective survey of quantum programming language research. 1 Quantum Computation Quantum computing is a relatively young subject. It has its beginnings in 1982, when Paul Benioff and Richard Feynman independently pointed out that a quantum mechanical system can be used to perform computations [11, p.12]. Feynman’s interest in quantum computation was motivated by the fact that it is computationally very expensive to simulate quantum physical systems on classical computers. This is due to the fact that such simulation involves the manipulation is extremely large matrices (whose dimension is exponential in the size of the quantum system being simulated). Feynman conceived of quantum computers as a means of simulating nature much more efficiently. The evidence to this day is that quantum computers can indeed perform certain tasks more efficiently than classical computers. Perhaps the best-known example is Shor’s factoring algorithm, by which a quantum computer can find the prime factors of any integer in probabilistic polynomial time [15]. There is no known classical probabilistic algorithm which can solve this problem in polynomial time. In the ten years since the publication of Shor’s result, there has been an enormous surge of research in quantum algorithms and quantum complexity theory. 2 Quantum Programming Languages Quantum physics involves phenomena, such as superposition and entanglement, whose properties are not always intuitive. These same phenomena give quantum computation its power, and are often at the heart of an interesting quantum algorithm. However, there does not yet seem to be a unifying set of principles by which quantum algorithms are developed; each new algorithm seems to rely on a unique set of “tricks” to achieve its particular goal. One of the goals of programming language design is to identify and promote useful “high-level” concepts — abstractions or paradigms which allow humans 2 to think about a problem in a conceptual way, rather than focusing on the details of its implementation. With respect to quantum programming, it is not yet clear what a useful set of abstractions would be. But the study of quantum programming languages provides a setting in which one can explore possible language features and test their usefulness and expressivity. Moreover, the definition of prototypical programming languages creates a unifying formal framework in which to view and analyze existing quantum algorithm. 2.1 Virtual Hardware Models Advances in programming languages are often driven by advances in compiler design, and vice versa. In the case of quantum computation, the situation is complicated by the fact that no practical quantum hardware exists yet, and not much is known about the detailed architecture of any future quantum hardware. To be able to speak of “implementations”, it is therefore necessary to fix some particular, “virtual” hardware model to work with. Here, it is understood that future quantum hardware may differ considerably, but the differences should ideally be transparent to programmers and should be handled automatically by the compiler or operating system. There are several possible virtual hardware models to work with, but fortunately all of them are equivalent, at least in theory. Thus, one may pick the model which fits one’s computational intuitions most closely. Perhaps the most popular virtual hardware model, and one of the easiest to explain, is the quantum circuit model. Here, a quantum circuit is made up from quantum gates in much the same way as a classical logic circuit is made up from logic gates. The difference is that quantum gates are always reversible, and they correspond to unitary transformations over a complex vector space. See e.g. [3] for a succinct introduction to quantum circuits. Of the two basic quantum operations, unitary transformations and measurements, the quantum circuit model emphasizes the former, with measurements always carried out as the very last step in a computation. Another virtual hardware model, and one which is perhaps even better suited for the interpretation of quantum programming languages, is the QRAM model of Knill [9]. Unlike the quantum circuit model, the QRAM models allows unitary transformations and measurements to be freely interleaved. In the QRAMmodel, a quantum device is controlled by a universal classical computer. The quantum device contains a large, but finite number of individually addressable quantum bits, much like a RAM memory chip contains a multitude of classical bits. The classical controller sends a sequence of instructions, which are either of the form “apply unitary transformation U to qubits i and j” or “measure qubit i”. The quantum device carries out these instruction, and responds by making the results of the measurements available. A third virtual hardware model, which is sometimes used in complexity theory, is the quantum Turing machine. Here, measurements are never performed, and the entire operation of the machine, which consists of a tape, head, and finite control, is assumed to be unitary. While this model is theoretically equivalent

“Scrap your boilerplate” reloaded

Abstract: The paper “Scrap your boilerplate” (SYB) introduces a combinator library for generic programming that offers generic traversals and queries. Classically, support for generic programming consists of two essential ingredients: a way to write (type-)overloaded functions, and independently, a way to access the structure of data types. SYB seems to lack the second. As a consequence, it is difficult to compare with other approaches such as PolyP or Generic Haskell. In this paper we reveal the structural view that SYB builds upon. This allows us to define the combinators as generic functions in the classical sense. We explain the SYB approach in this changed setting from ground up, and use the understanding gained to relate it to other generic programming approaches. Furthermore, we show that the SYB view is applicable to a very large class of data types, including generalized algebraic data types.

EasyCheck: test data for free

Abstract: We present a lightweight, automated tool for specification-based testing of declarative programs written in the functional logic programming language Curry and emphasize the usefulness of logic features in its implementation and use. Free variables, nondeterminism and encapsulated search turn out to be elegant and powerful means to express test-data generation.