achim d. brucker

Dear all,
before I start with a crude hack :-): is there a nice way to use the
parser-structures generated by ml-lexx/ml-yacc (from the sml/NJ 110.56
distribution ) without using the CM?
Moreover, is possible to use these sources with poly/ml?

Achim

rob arthan

I don't know, but if you can't, then the SLRP parser generator that comes as
part of the OpenProofPower development kit is an alternative to ml-yacc
that works with SML/NJ, Poly/ML and doesn't depend on CM. It generates code
that should work with any Standard ML compiler and it has been used (after
a few adjustments to the build scripts etc.) with Moscow ML.

The user guide for SLRP is at:

http://www.lemma-one.com/ProofPower/doc/usr032.pdf

If there is a demand for it, then I'd be very happy to unbundle it from the
ProofPower-specific things to make it more accessible for general purpose
use (at the moment it comes bundled with a miscellany shell scripts that
you need to build and/or develop ProofPower but not to use SLRP itself.
SLRP is available open source under the GPL. Let me know if you need more
info.

Regards,

Rob.

matthias blume

Rob Arthan <rda@lemma-one.com> writes:


ML-yacc in SML/NJ does not depend on CM. Using it with CM is simply a
lot more convenient.

There should be no problem using it with other compilers. The shell
command ml-yacc is a standalone program which takes a grammar file as
input and generates two SML source files as output. You can run
ml-yacc by hand and include the resulting two files in your project.

The only minor difficulty will be that you also need to include files
from the ML-Yacc library (in the SML/NJ source distribution under
ml-yacc/lib), because ml-yacc-generated code refers to those.

Notice that SML/NJ now also comes with an LL(k) parser generator
modeled after ANTLR. The new tool is called "ml-antlr". You might
want to give that a whirl, too. The sources is under ml-lpt, and
besides ml-antlr there is a new version of ml-lex called ml-ulex, too.
See the manual under ml-lpt/doc for details. (You need to latex
manual.tex.)

Matthias

heiner

In article <elckke$k10$1@wolfberry.srv.cs.cmu.edu>,
Rob Arthan <rda@lemma-one.com> writes:

I for one would appreciate an unbundeled parser generator.
--
Heiner Marxen http://www.drb.insel.de/~heiner/

achim d. brucker

Hi,
Matthias Blume <find@my.address.elsewhere> schrieb:

of course, I already used the command-line versions. I got stuck when
trying to re-implement all the stuff contained in ml-yacc
library. Surely, I looked into the sources, but I somehow missed it
(my fault) ...

Nevertheless, I can now confirm that using the files from the ml-yacc
library together with poly/ML 5.0 (beta) works great! Thanks a lot.
It a very good feeling to know, that our implementation will support
both, sml/NJ and the next version of poly/ML.

Very nice. Sadly we already converted a larger ANTLR grammar (for OCL)
file to ml-yacc/ml-lex a half year ago :-) Thus, our
parser/type-checker will use ml-yacc/ml-lex, at least for the near
future.

Achim