We present a verified compiler to an idealized assembly language from a small, untyped functional language with mutable references and exceptions. The compiler is programmed in the Coq proof assistant and has a proof of total correctness with respect to bigstep operational semantics for the source and target languages. Compilation is staged and includes standard phases like translation to continuation-passing style and closure conversion, as well as a common subexpression elimination optimization. In this work, our focus has been on discovering and using techniques that make our proofs easy to engineer and maintain. While most programming language work with proof assistants uses very manual proof styles, all of our proofs are implemented as adaptive programs in Coq’s tactic language, making it possible to reuse proofs unchanged as new language features are added. In this paper, we focus especially on phases of compilation that rearrange the structure of syntax with nested variable binders. That aspect has been a key challenge area in past compiler verification projects, with much more effort expended in the statement and proof of binder-related lemmas than is found in standard penciland-paper proofs. We show how to exploit the representation technique of parametric higher-order abstract syntax to avoid the need to prove any of the usual lemmas about binder manipulation, often leading to proofs that are actually shorter than their pencil-andpaper analogues. Our strategy is based on a new approach to encoding operational semantics which delegates all concerns about substitution to the meta language, without using features incompatible with general-purpose type theories like Coq’s logic.

Abstract. Combining higher-order abstract syntax and (co)-induction in a logical framework is well known to be problematic. Previous work [ACM02] described the implementation of a tool called Hybrid, within Isabelle HOL, syntax, and reasoned about using tactical theorem proving and principles of (co)induction. Moreover, it is definitional, which guarantees consistency within a classical type theory. The idea is to have a de Bruijn representation of syntax, while offering tools for reasoning about them at the higher level. In this paper we describe how to use it in a multi-level reasoning fashion, similar in spirit to other meta-logics such as Linc and Twelf. By explicitly referencing provability in a middle layer called a specification logic, we solve the problem of reasoning by (co)induction in the presence of non-stratifiable hypothetical judgments, which allow very elegant and succinct specifications of object logic inference rules. We first demonstrate the method on a simple example, formally proving type soundness (subject reduction) for a fragment of a pure functional language, using a minimal intuitionistic logic as the specification logic. We then prove an analogous result for a continuation-machine presentation of the operational semantics of the same language, encoded this time in an ordered linear logic that serves as the specification layer. This example demonstrates the ease with which we can incorporate new specification logics, and also illustrates a significantly

We show that, in a parametric model of polymorphism, the type ∀α.((α → α) → α) → (α → α → α) → α is isomorphic to closed de Bruijn terms. That is, the type of closed higher-order abstract syntax terms is isomorphic to a concrete representation. To demonstrate the proof we have constructed a model of parametric polymorphism inside the Coq proof assistant. The proof of the theorem requires parametricity over Kripke relations. We also investigate some variants of this representation.

Given the complexity of the metatheoretic reasoning involved with current programming languages and their type systems, techniques for mechanical formalization and checking of the metatheory have received much recent attention. In previous work, we advocated a combination of locally nameless representation and cofinite quantification as a lightweight style for carrying out such formalizations in the Coq proof assistant. As part of the presentation of that methodology, we described a number of operations associated with variable binding and listed a number of properties, called “infrastructure lemmas, ” about those operations that needed to be shown. The proofs of these infrastructure lemmas are generally straightforward, given a specification of the binding structure of the language. In this work, we present LNgen, a prototype tool for automatically generating these definitions, lemmas, and proofs from Ott-like language specifications. Furthermore, the tool also generates a recursion scheme for defining functions over syntax, which was not available in our previous work. We also show the soundness and completeness of our tool’s output. For untyped lambda terms, we prove the adequacy of our representation with respect to a fully concrete representation, and we argue that the representation is complete—that we generate the right set of lemmas—with respect to Gordon and Melham’s “Five Axioms of Alpha-Conversion. ” Finally, we claim that our recursion scheme is simpler to work with than either Gordon and Melham’s recursion scheme or the recursion scheme of Nominal Logic. 1.

There are two approaches to formalizing the syntax of typed object languages in a proof assistant or programming language. The extrinsic approach is to first define a type that encodes untyped object expressions and then make a separate definition of typing judgements over the untyped terms. The intrinsic approach is to make a single definition that captures well-typed object expressions, so ill-typed expressions cannot even be expressed. Intrinsic encodings are attractive and naturally enforce the requirement that metalanguage operations on object expressions, such as substitution, respect object types. The price is that the metalanguage types of intrinsic encodings and operations involve non-trivial dependency, adding significant complexity. This paper describes intrinsic-style formalizations of both simply-typed and polymorphic languages, and basic syntactic operations thereon, in the Coq proof assistant. The Coq types encoding object-level variables (de Bruijn indices) and terms are indexed by both type and typing environment. One key construction is the boot-strapping of definitions and lemmas about the action of substitutions in terms of similar ones for a simpler notion of renamings. In the simply-typed case, this yields definitions that are free of any use of type equality coercions. In the polymorphic case, some substitution operations do still require type coercions, which we at least partially tame by uniform use of heterogeneous equality.

is a joint program with Carleton University, administered by the Ottawa-

Why polymorphism?

Abstract—We present a point of view concerning HOAS (Higher-Order Abstract Syntax) and an extensive exercise in HOAS along this point of view. The point of view is that HOAS can be soundly and fruitfully regarded as a definitional extension on top of FOAS (First-Order Abstract Syntax). As such, HOAS is not only an encoding technique, but also a higher-order view of a first-order reality. A rich collection of concepts and proof principles is developed inside the standard mathematical universe to give technical life to this point of view. The exercise consists of a new proof of Strong Normalization for System F. HOAS makes our proof considerably more direct than previous proofs. The concepts and results presented here have been formalized in the theorem prover Isabelle/HOL.

Abstract There are two approaches to formalizing the syntax of typed object languages in a proof assistant or programming language. The extrinsic approach is to first define a type that encodes untyped object expressions and then make a separate definition of typing judgements over the untyped terms. The intrinsic approach is to make a single definition that captures well-typed object expressions, so ill-typed expressions cannot even be expressed. Intrinsic encodings are attractive and naturally enforce the requirement that metalanguage operations on object expressions, such as substitution, respect object types. The price is that the metalanguage types of intrinsic encodings and operations involve non-trivial dependency, adding significant complexity. This paper describes intrinsic-style formalizations of both simply-typed and polymorphic languages, and basic syntactic operations thereon, in the Coq proof assistant. The Coq types encoding object-level variables (de Bruijn indices) and terms are indexed by both type and typing environment. One key construction is the boot-strapping of definitions and lemmas about the action of substitutions in terms of similar ones for a simpler notion of renamings. In the simply-typed case, this yields definitions that are free of any use of type equality coercions. In the polymorphic case, some substitution operations do still require type coercions, which we at least partially tame by uniform use of heterogeneous equality.

LF is a dependent type theory in which many other formal systems can be conveniently embedded. However, correct use of LF relies on nontrivial metatheoretic developments such as proofs of correctness of decision procedures for LF’s judgments. Although detailed informal proofs of these properties have been published, they have not been formally verified in a theorem prover. We have formalized these properties within Isabelle/HOL using the Nominal Datatype Package, closely following a recent article by Harper and Pfenning. In the process, we identified and resolved a gap in one of the proofs and a small number of minor lacunae in others. We also formally derive a version of the type checking algorithm from which Isabelle/HOL can generate executable code. Besides its intrinsic interest, our formalization provides a foundation for studying the adequacy of LF encodings, the correctness of Twelf-style metatheoretic reasoning, and the metatheory of extensions to LF.