|GMD Karlsruhe Toolbox - Cocktail email@example.com (Josef Grosch) (1992-03-05)|
|GMD Karlsruhe Toolbox - Cocktail firstname.lastname@example.org (Josef Grosch) (1992-09-08)|
|From:||Josef Grosch <email@example.com>|
|Keywords:||tools, FTP, parse, lex|
|Organization:||GMD Forschungsstelle Karlsruhe|
|Date:||Thu, 5 Mar 1992 14:39:11 GMT|
A new version (9202) of the GMD Toolbox for Compiler Construction -
also known as Cocktail (Compiler-Compiler-Toolbox Karlsruhe) -
is now available via anonymous ftp:
TOOLBOX(1) GMD-Forschungsstelle-Karlsruhe TOOLBOX(1)
toolbox - tool box for compiler construction
Toolbox is a set of program generators or compiler
construction tools for nearly all phases of a compiler. The
compiler construction tools support the automatic generation
of compilers for imperative programming languages. The
design goals for this tool box were practical usability,
significantly reduced construction effort for compilers, and
high quality of the generated compilers. Especially with
respect to efficiency the tools are competitive to
programming by hand. Currently the tools can generate
compiler modules in the target languages C and Modula-2.
First realistic applications demonstrate the excellent
performance of the tools and show that the tools allow the
construction of production quality compilers.
Rex generator for lexical analyzers
Lalr LALR(1) parser generator
Ell LL(1) parser generator
Ast generator for abstract syntax trees
Ag generator for attribute evaluators
Puma transformation of abstract syntax trees
Reuse library of reusable modules
C (ANSI + K&R) or Modula-2
C (ANSI + K&R), C++, or Modula-2
DEC Station / ULTRIX
VAX / ULTRIX or BSD UNIX 4.2
SUN 3 or SUN 4 / SunOS
PCS Cadmus / MUNIX
rusmv1.rus.uni-stuttgart.de = 188.8.131.52
Medium: DC 300/600 data cartridge, TK 50, or Exabyte in tar format
source programs in Modula-2 as well as in C
documentation in troff- und Postscript-format
GMD Forschungsstelle an der Universitaet Karlsruhe
TAPE HANDLING FEE
Source licence: 500 DM or 250 US $
- The Toolbox contains a new tool for the transformation of attributed trees
called 'puma'. It is based on pattern-matching. It replaces its predecessor
'estra' and comes with documentation in English.
- The tool for abstract syntax trees 'ast' has been extended from single to
multiple inheritance. So-called "subunits" allow the implementation of one
abstract tree by several compilation units. The concept of "views" makes
it possible to derive from a common specification several abstract syntax
trees which represent subsets.
- The attribute grammar tool 'ag' has analogously been extended to process
object-oriented attribute grammars with multiple inheritance. It supports the
generation of several attribute evaluators that run one after the other.
- The error handling module for the parser generators 'lalr' and 'ell' are
independent of the parser module or the grammar. The manual for these parser
generators has been completely rewritten.
- The Modula-2 to C translator 'mtc' has a new code-generator. This brings
big efficiency improvements: 30% smaller program, 10% faster, 75% less
dynamic memory consumption. It generates ANSI C as well as K&R C.
- The sources of all tools are in ANSI C as well as in K&R C.
- All tools generate modules in the target languages C (ANSI + K&R), C++,
- The interface to the operating system has been redesigned. It allows to
switch IO operations between UNIX system calls and C library calls.
This should assure much better portability.
Return to the
Search the comp.compilers archives again.