BABEL'01 Proceedings
BABEL'01 Proceedings |
|
This page contains the proceedings of BABEL'01, the First International Workshop on Multi-Language Infrastructure and Interoperability.
Document readers are available from the Additional Resources page. See the Contributors page for information on the authors of the materials contained herein.
| Preface by: | Nick Benton Andrew Kennedy |
Recent years have seen a resurgence of interest in multi-language tools and intermediate languages, and in interoperability between programs and components written in different programming languages. Shared infrastructure such as code generators, analysis tools and garbage collectors can greatly ease the task of producing a high-quality implementation of a new programming language, whilst being able to interoperate easily with code written in existing languages is essential for such an implementation to be useful in practice. The BABEL workshop brought together researchers and developers working on multi-language integration.
BABEL was sponsored by ACM SIGPLAN and took place on September 8th, 2001 as part of PLI'01, the Colloquium on Principles, Logics, and Implementations of High-Level Programming Languages.
The programme committee selected 10 papers from 22 submissions. Each paper was reviewed by three or more committee members or their subreferees.
The members of the BABEL'01 programme committee were:
| Nick Benton (chair) | (Microsoft Research, Cambridge) |
| Fergus Henderson | (University of Melbourne) |
| Andrew Kennedy | (Microsoft Research, Cambridge) |
| Greg Morrisett | (Cornell University) |
| Martin Odersky | (Ecole Polytechnique Fédérale de Lausanne) |
| John Reppy | (Bell Labs) |
| Andrew Tolmach | (Portland State University) |
| David Wakeling | (University of Exeter) |
| Author: | Zhong Shao |
Sun's Java architecture introduced a safe virtual machine (VM) in which an ensemble of software components developed independently could smoothly interoperate. The goal of Microsoft's Common Language Runtime (CLR) is to generalize this approach and allow components in many source languages to interoperate safely. CLR supports flexible interoperation by compiling various source languages into a common intermediate language and by using a unified type system. However, the type system in CLR (and Java VM) enforces only conventional type safety in an object-oriented system. Therefore, higher-level specifications (e.g., resource bounds, generalized access control, formal software protocols) cannot be enforced. Because conventional type systems are too inflexible for real applications, developers often bypass the type system, producing code that steps outside the managed part of the VM; such components cannot be verified.
At Yale we have been developing typed common intermediate languages (named FLINT) that can support safely not only the standard object-oriented model, but also higher-order generic (polymorphic) programming and Java-style reflection (introspection). Unlike CLR, our type system is independent of any particular programming model, yet it is capable of expressing all valid propositions and proofs in higher-order predicate logic (so it can be used to capture and verify advanced program properties). The rich type system of FLINT makes it possible to typecheck both compiler intermediate code and low level machine code; this allows typechecking to take place at any phase of compilation, even after optimizations and register allocation. It also leads to a smaller and more extensible VM because low-level native routines that would otherwise be in VM can now be verified and moved into a certified library. This talk describes our vision of the FLINT system, outline our approach to its design, and survey the technologies that can be brought to support its implementation.
| Authors: | Kathleen Fisher Riccardo Pucella John Reppy |
 View PDF |
 Download PostScript |
Practical implementations of high-level languages must provide access to libraries and system services that have APIs specified in a low-level language (usually C). An important characteristic of such mechanisms is the foreign-interface policy that defines how to bridge the semantic gap between the high-level language and C. For example, IDL-based tools generate code to marshal data into and out of the high-level representation according to user annotations. The design space of foreign-interface policies is large and there are pros and cons to each approach. Rather than commit to a particular policy, we choose to focus on the problem of supporting a gamut of interoperability policies. In this paper, we describe a framework for language interoperability that is expressive enough to support very efficient implementations of a wide range of different foreign-interface policies. We describe two tools that implement substantially different policies on top of our framework and present benchmarks that demonstrate their efficiency.
| Author: | Leif Kornstaedt |
| View PDF |
Download PostScript |
This paper reports practical experience in implementing Alice, an extension of Standard ML, on top of an existing implementation of Oz. This approach yields a high-quality implementation with little effort. The combination is an advanced programming system for both Oz and Alice, which offers more than either language on its own.
| Author: | Matthias Blume |
| View PDF |
Download PostScript |
We present a new foreign-function interface for SML/NJ. It is based on the idea of data-level interoperability – the ability of ML programs to inspect as well as manipulate C data structures directly. The core component of this work is an encoding of the almost complete C type system in ML types. The encoding makes extensive use of a "folklore" typing trick, taking advantage of ML's polymorphism, its type constructors, its abstraction mechanisms, and even functors. A small low-level component which deals with C struct and union declarations as well as program linkage is hidden from the programmer's eye by a simple program-generator tool that translates C declarations to corresponding ML glue code.
| Author: | Don Syme |
| View PDF |
Download PostScript |
This paper describes several extensions to the .NET Common Intermediary Language (CIL), each of which is designed to enable easier implementation of typed high-level programming languages on the .NET platform, and to promote closer integration and interoperability between these languages. In particular we aim for easier interoperability between components whose interfaces are expressed using function types, discriminated unions and parametric polymorphism, regardless of the languages in which these components are implemented. We show that it is possible to add these constructs to an existing, "real world" intermediary language and that this allows corresponding subsets of constructs to be compiled uniformly, which in turn will allow programmers to use these constructs seamlessly between different languages. In this paper we discuss the motivations for our extensions, which are together called Extended IL (ILX), and describe them via examples. In this setting, many of the traditional responsibilities of the backend of a compiler must be moved to ILX and the execution environment, in particular those related to representation choices and low-level optimizations. We have modified a Haskell compiler to generate this language, and have implemented an assembler that translates the extensions to regular or polymorphic CIL code.
| Authors: | Tyson Dowd Fergus Henderson Peter Ross |
| View PDF |
Download PostScript |
The .NET Common Language Runtime (CLR) offers a new opportunity to experiment with multi-language interoperation, and provides a relatively rare chance to explore deep interoperation of a wide range of programming language paradigms. This article describes how the logic/functional programming language Mercury is compiled to the CLR. We describe the problems we have encountered with generating code for the CLR, give some preliminary benchmark results, and suggest some possible improvements to the CLR regarding separate compilation, verifiability, tail calls, and efficiency.
| Authors: | Mark Shields Simon Peyton Jones |
| View PDF |
Download PostScript |
Haskell has a sophisticated mechanism for overloading identifiers with multiple definitions at distinct types. Object-oriented programming has a similar notion of overriding and overloading for methods names. Unfortunately, it is not possible to encode object-oriented overloading directly using Haskell overloading. This deficiency becomes particularly tiresome when Haskell programs wish to call methods imported from an object-oriented library.
We present two refinements of Haskell's type class system: Closed classes and overlapping instances. We demonstrate how we may exploit the refined system so as to be able to encode object-oriented classes within Haskell. This encoding allows us to mimic, within Haskell, the overloading resolution rules employed by object-oriented languages without the need for additional type annotations or name mangling. As a consequence, object-oriented class libraries are very convenient to import and use within Haskell.
| Author: | Fermín Reig |
| View PDF |
Download PostScript |
This paper identifies high-level program properties that can be discovered by static analysis in a compiler front end, and that are useful for classical low-level optimizations. We suggest how intermediate language code could be annotated to convey these properties to the code generator.
| Author: | Jürg Gutknecht |
| View PDF |
Download PostScript |
Active Oberon is a substantial evolution of the programming language Oberon. It distinguishes itself by a novel object model and by its integration into the .NET language interoperability framework.
The three concepts characterizing Active Oberon are: (a) active object types, (b) a single and unifying notion of abstraction called definition, and (c) a static module construct. These concepts are in fact powerful combinations of concepts: Active objects integrate active behavior with reactive message handling, definitions unify the units of usage, implementation and inheritance, and modules represent both package and static object. The rigid concept of class hierarchy has been sacrificed in Active Oberon to a more flexible concept of aggregation that is based on a generalized IMPLEMENTS relation. The relations IMPORTS and REFINES are used to specify static module dependencies and to derive new definitions from existing ones respectively.
This article is a report on a work in progress. We divide our presentation into three parts: (a) A short recall of the history of programming languages developed at the ETH, (b) an extensive conceptual overview of Active Oberon's object model called the Active Object System (AOS), (c) a discussion of the mapping of the AOS into the Common Type System (CTS) exposed by .NET.
| Authors: | Peter Housel Christian Stork Vivek Haldar Niall Dalton Michael Franz |
| View PDF |
Download PostScript |
The Java Virtual Machine is primarily designed for transporting Java programs. As a consequence, when JVM bytecodes are used to transport programs in other languages, the result becomes less acceptable the more the source language diverges from Java. Microsoft's .NET transport format fares better in this respect because it has a more flexible type system and instruction set, but it is not extensible, and (for example) has no provision for supporting explicit programmer-specified parallelism. Both platforms have difficulty making transported programs run efficiently.
This paper discusses first steps towards mobile code representations that are independent (in the sense that the representation can be appropriately parameterized) of the source language (e.g., Java), intermediate representation (e.g., bytecode), and target architecture (e.g., x86). We call this kind of parameterizable framework language-agnostic.
We present two techniques which provide parts of the envisioned language-agnostic functionality. Compressed abstract syntax trees as a wire format provide for a very dense encoding of programs at a high level of abstraction. We show how to parameterize the compression algorithm in a modular fashion with knowledge beyond the purely syntactical level. This leads to the notion of well-formedness by construction. The second technique defines the semantics of programs by mapping from abstract syntax trees to a typed core calculus representation. Based on this representation it becomes possible to use portable definitions of security policies and to execute programs written in different source languages, even if a more efficient trusted native compiler is not available on the target platform.
| Authors: | Michel Schinz Martin Odersky |
| View PDF |
Download PostScript |
A problem that often has to be solved by compilers for functional languages targeting the Java Virtual Machine is the elimination of tail calls. This paper explains how we solved it in our Funnel compiler and presents some experimental results about the impact our technique has on both performance and size of the compiled programs.
© 2001-2002 Microsoft Corporation. All rights reserved. Terms of Use.