Re: Compiler project needed (David Starner)
22 Feb 2000 09:28:04 -0500

          From comp.compilers

Related articles
Compiler project needed (Per Olesen) (2000-02-22)
Re: Compiler project needed (2000-02-22)
Re: Compiler project needed (2000-02-22)
Re: Compiler project needed (Charles E. Bortle, Jr.) (2000-02-22)
Re: Compiler project needed (2000-02-22)
Re: Compiler project needed (2000-02-23)
Re: Compiler project needed (Stephen Sulzer) (2000-02-23)
Re: Compiler project needed (Juergen Kahrs) (2000-02-23)
Re: Compiler project needed (Daniel C. Wang) (2000-02-27)
[8 later articles]
| List of all articles for this month |

From: (David Starner)
Newsgroups: comp.compilers
Date: 22 Feb 2000 09:28:04 -0500
Organization: Oklahoma State University
References: 00-02-112
Keywords: design

On 22 Feb 2000 00:02:35 -0500,
  Per Olesen <> wrote:
>I'm studying computer science on a Danish university and I'm going to
>write a compiler as a project in a course I'm taking.

Some ideas, assuming you want it to be interesting to someone else
after you're done.

*** From ***
Modula 3 seems to be a language in search of a nice portable compiler.
If you can make CVSup compile easier or on more platforms, it will be
appreciated. (A M3toAda compiler is my current project for my course.)

Interesting (not currently useful) procedural languages are Algol 60,
Algol 68, Algol W, and PL/I. Jovial is another one with intersting
history, but finding data on it is hard. Red, Blue(?) and Yellow(?)
lost to Green for the title of Ada - implementing one of them could
provide an intersting historical perspective on what we won and lost
when Green won. (Again data is hard to find.) (If someone could point
us to an Internet site with information on Jovial or the 3 would-be
Ada's, I'd be greatly appreciative.)

Standard ML might be another intersting language to compile.

Looking through Advanced Programming Language Design (by Raphael
A. Finkel) (a very cool book) lists CSP, DP, Gedanken, Io, and Post as
unimplemented (with cavets on CSP and Post.) Of course, there's
reasons most of those have not been implemented, usually having to do
with being unusual and hard to implement.

*** To ***
JVM and C are both currently popular targets, for good reason. MMIX
might be another good target, depending on how pragmatic you
are. (MMIX is Knuth's new theoritical assembly language for TAOCP.) I
would think Ada or C++ or the likes would be nice to compile to if you
needed exception support and other funky stuff.

A cool hack would be to compile upwards; e.g. compile C into ML or
Lisp, or something similarly audacious. (A reasonable implementation
would certainly impress me.)

The GNU Task for Programming Languages
has some suggestions. These break down into two groups - the procedural
languages (Fortran 90, Modula 3, Cobol, PL/I, etc.) as GCC frontends, and
the more interesting Foo (Perl, Python, Javascript, TCL, etc.) -> Scheme

David Starner -

Post a followup to this message

Return to the comp.compilers page.
Search the comp.compilers archives again.