The A SF+S DF Meta-Environment: a Component-Based Language

Development Environment

M.G.J. van den Brand , A. van Deursen , J. Heering
, H.A. de Jong , M. de Jonge , T. Kuipers ,
P. Klint , L. Moonen , P.A. Olivier , Scheerder , J.J. Vinju , E. Visser , and J. Visser


Centrum voor Wiskunde en Informatica (CWI), Kruislaan 413, 1098 SJ Amsterdam, The Netherlands
 Faculty of Philosophy, Utrecht University, Heidelberglaan 8, 3584 CS Utrecht, The Netherlands
Faculty of Mathematics and Computer Science, Utrecht University, Padualaan 14, 2584 CH Utrecht, The
Netherlands

Abstract. The A SF+S DF Meta-Environment is an interactive development environment for the auto-
matic generation of interactive systems for constructing language definitions and generating tools for
them. Over the years, this system has been used in a variety of academic and commercial projects
ranging from formal program manipulation to conversion of COBOL systems. Since the existing im-
plementation of the Meta-Environment started exhibiting more and more characteristics of a legacy
system, we decided to build a completely new, component-based, version. We demonstrate this new
system and stress its open architecture.

1 Introduction
The A SF+S DF Meta-Environment [12] is an interactive development environment for the automatic gen-
eration of interactive systems for constructing language definitions and generating tools for them. A lan-
guage definition typically includes such features as syntax, prettyprinting, typechecking, and execution of
programs in the target language. The A SF+S DF Meta-Environment can help in the following cases:
You have to write a formal specification for some problem and you need interactive support for this.
You are developing your own (application) language and want to create an interactive environment for
it.
You have programs in some existing programming language and you want to analyze or transform
them.
The A SF+S DF formalism [1] [10] allows the definition of syntactic as well as semantic aspects. It can
be used for the definition of languages (for programming, writing specifications, querying databases, text
processing, or other applications). In addition it can be used for the formal specification of a wide variety
of problems. A SF+S DF provides:
A general-purpose algebraic specification formalism based on (conditional) term rewriting.
Modular structuring of specifications.
Integrated definition of lexical, context-free, and abstract syntax.
User-defined syntax, allowing you to write specifications using your own notation.
Complete integration of the definition of syntax, and semantics.
Traversal functions (for writing very concise program transformations), memo functions (for caching
repeated computations), and more.
The A SF+S DF Meta-Environment offers:
Syntax-directed editing of A SF+S DF specifications.
Incremental compilation and testing of specifications.
Compilation of A SF+S DF specifications into dedicated interactive stand-alone environments contain-
ing various tools such as a parser, prettyprinter, syntax-directed editor, debugger, and interpreter or
compiler.
User-defined extensions of the default user-interface.
The design goals of the new implementation to be demonstrated include: openness, reuse, extensibility,
and in particular the possibility to generate complete stand-alone environments for user-defined languages.
 Text Structure Parsetable
Parser
Editor Editor Generator

Graph Tree
TOOLBUS
Browser Repository

ASF+SDF Unparser
Interpreter
Compiler Generator

Fig. 1. Architecture of the A SF+S DF Meta-Environment

2 Technological background
ToolBus A hallmark of legacy systems in general and the old A SF+S DF Meta-Environment in particular
is the entangling of control flow and actual computation. To separate coordination from computation we
use the ToolBus coordination architecture [2], a programmable software bus based on process algebra.
Coordination is expressed by a formal description of the cooperation protocol between components while
computation is expressed in components that may be written in any language. We thus obtain interoper-
ability of heterogeneous components in a (possibly) distributed system.

ATerms Coordination protocol and components have to share data. We use ATerms [4] for this purpose.
These are trees with optional annotations on each node. The annotations are used to store tool-specific
information like text coordinates or color attributes. The implementation of ATerms has two essential prop-
erties: terms are stored using maximal subterm sharing (reducing memory requirements and making deep
equality tests very efficient) and they can be exchanged using a very dense binary encoding that preserves

    
sharing. As a result very large terms (with over nodes) can be processed.

SGLR In our language-centric approach the parser is an essential tool. We use scannerless, generalized-
LR parsing [13]. In this way we can parse arbitrary context-free grammars, an essential property when
combining and parsing large grammars for (dialects of) real-world languages.

Term rewriting A SF+S DF specifications are executed as (conditional) rewrite rules. Both interpretation and
compilation (using the ASF2C compiler [5]) of these rewrite rules are supported. The compiler generates
very efficient C code that implements pattern matching and term traversal. The generated code uses ATerms
as its main data representation, and ensures a minimal use of memory during normalization of terms.

3 Architecture
The architecture of the A SF+S DF Meta-Environment is shown in Figure 1. It consists of a ToolBus that
interconnects the following components:
User interface: the top level user-interface of the system. It consists primarily of a graph browser for
the import graph of the current specification.
Text Editor: a customized version of XEmacs for text editing.
Structure Editor: a syntax-directed editor that closely cooperates with the Text Editor.
Parser: scannerless, generalized-LR parser (SGLR) that is parametrized with a parse table.
Parsetable generator: takes an SDF syntax definition as input and generates a parse table for SGLR.
Tree Repository: stores all terms corresponding to specification modules, parse tables, user-defined
terms, etc.
Compiler: the ASF2C compiler.
Interpreter: executes specifiations by direct interpretation.
Unparser generator: generates prettyprinters.
Fig. 2. The main user-interface of the Meta-Environment is a module browser that provides a graphical and a textual
view of the modules in a specification. A number of operations can be initiated for each module. Here it is shown with
the modules from a small specification of a typechecker for the toy language Pico.

4 Applications of A SF+S DF and the Meta-Environment

There are a number of academic and industrial projects that use either A SF+S DF directly or components
of the Meta-Environment in one way or another. c The applications of A SF+S DF can be split into three
groups:

1. In the field of language prototyping A SF+S DF has been used to describe the syntax and semantics
of domain specific languages, e.g., the language Risla for describing financial products [3]. As an-
other example, the syntax of the algebraic specification language C ASL has been prototyped using
A SF+S DF [7]. B OX [9] [11] is a small domain specific language developed for prettyprinting within
the Meta-Environment.
2. In the field of reverse engineering and system renovation, A SF+S DF is used to analyze and transform
COBOL legacy code [8].
3. As an algebraic specification formalism for specifying language processing tools. In fact, a number of
components of the Meta-Environment itself have been specified using A SF+S DF:
the ASF2C compiler,
the unparser generator, and
parts of the parsetable generator.
For other components, such as the ToolBus and the syntax-directed editor, an A SF+S DF specification
was made for prototyping use only. That specification formed the basis for an optimized, handcrafted
implementation.

Components of the Meta-Environment are used as stand-alone tools in a variety of applications. Exam-
ples are the Stratego compiler [14], the Risla compiler, the Elan environment [6], and a commercial tool
for designing and implementing information systems.

5 Demonstration

We will show a number of applications of the Meta-Environment ranging from a simple typechecking
problem (Figure 2) to syntax-directed editing and transformation of COBOL systems.
6 Obtaining the A SF+S DF Meta-Environment

The A SF+S DF Meta-Environment can be downloaded from:

http://www.cwi.nl/projects/MetaEnv/completa/

Individual components, such as the ATerm library, ToolBus, parser generator, and parser (SGLR) can
be obtained from:  !"#$%&')('*%+,-. .
All components of the A SF+S DF Meta-Environment are available as open source software.

References
1. J.A. Bergstra, J. Heering, and P. Klint, editors. Algebraic Specification. ACM Press/Addison-Wesley, 1989.
2. J.A. Bergstra and P. Klint. The discrete time ToolBus a software coordination architecture. Science of Computer
Programming, 31(2-3):205229, July 1998.
3. M.G.J. van den Brand, A. van Deursen, P. Klint, S. Klusener, and E.A, van den Meulen. Industrial applications of
ASF+SDF. In M. Wirsing and M. Nivat, editors, Algebraic Methodology and Software Technology (AMAST 96),
volume 1101 of LNCS. Springer-Verlag, 1996.
4. M.G.J. van den Brand, H.A. de Jong, P. Klint, and P. Olivier. Efficient Annotated Terms. Software, Practice &
Experience, 30:259291, 2000.
5. M.G.J. van den Brand, P. Klint, and P. A. Olivier. Compilation and memory management for ASF+SDF. In
S. Jahnichen, editor, Compiler Construction (CC 99), volume 1575 of Lecture Notes in Computer Science, pages
198213. Springer-Verlag, 1999.
6. M.G.J. van den Brand and C. Ringeissen. ASF+SDF parsing tools applied to ELAN. In Third International
Workshop on Rewriting Logic and Applications, ENTCS, 2000.
7. M.G.J. van den Brand and J. Scheerder. Development of Parsing Tools for CASL using Generic Language
Technology. In D. Bert, C. Choppy, and P. Mosses, editors, Workshop on Algebraic Development Techniques
(WADT99), volume 1827 of LNCS. Springer-Verlag, 2000.
8. M.G.J. van den Brand, M.P.A. Sellink, and C. Verhoef. Generation of components for software renovation factories
from context-free grammars. Science of Computer Programming, 36:209266, 2000.
9. M.G.J. van den Brand and E. Visser. Generation of formatters for context-free languages. ACM Transactions on
Software Engineering and Methodology, 5:141, 1996.
10. A. van Deursen, J. Heering, and P. Klint, editors. Language Prototyping: An Algebraic Specification Approach,
volume 5 of AMAST Series in Computing. World Scientific, 1996.
11. M. de Jonge. A pretty-printer for every occasion. In I. Ferguson, J. Gray, and L. Scott, editors, Proceedings
of the 2nd International Symposium on Constructing Software Engineering Tools (CoSET2000). University of
Wollongong, Australia, 2000.
12. P. Klint. A meta-environment for generating programming environments. ACM Transactions on Software Engi-
neering and Methodology, 2:176201, 1993.
13. E. Visser. Syntax Definition for Language Prototyping. PhD thesis, University of Amsterdam, 1997.
14. E. Visser, Z. Benaissa, and A. Tolmach. Building Program Optimizers with Rewriting Strategies. In International
Conference on Functional Programming (ICFP98), pages 1326, 1998.