|Simula to C++ translator email@example.com (1996-08-15)|
|Re: Simula to C++ translator firstname.lastname@example.org (1996-08-16)|
|Re: Simula to C++ translator email@example.com (stanley (s.t.h.) chow) (1996-08-16)|
|Re: Simula to C++ translator firstname.lastname@example.org (Peter Brueckner) (1996-08-19)|
|From:||email@example.com (Dennis Yelle)|
|Date:||16 Aug 1996 11:31:18 -0400|
|Keywords:||C++, translator, OOP|
>I am presently trying to find the best way to migrate a large system
>written mainly in Simula from an old and soon unsupported platform to
>a new platform.
>3) Construct a translator tool. This could either be specific to the needs of
> this particular system, a full compiler with a strange output language, or
> something in between. A full compiler is, of course, the most exciting
>We pick item 3. Now, how do I construct the translator? I could write it from
>scratch using tools like lex and yacc, or employ some higher-level tool like,
>say, the Cornell Synthesiser Generator. The translator must run in OS/2, NT
>or UNIX. It does not have to be particularly fast or have a nice user
I think you made your first mistake when you decided to do "A full compiler"
instead of a translator specific to your needs.
It looks like you made this choice because it was "the most exciting choice".
This is not a good reason. In fact, I think it is a bad reason.
On the other hand, if you start out doing a translator tool, when
you get done you will have something that quacks like a compiler.
But my gut feeling is that it might actually work this year.
If you start out aiming for a "full compiler" your project might
get canceled before you get done.
I am wondering if maybe there is a 4th option that you
have overlooked. Maybe something that sounds stupid like
simulating the old platform on the new platform?
firstname.lastname@example.org (Dennis Yelle)
Return to the
Search the comp.compilers archives again.