This is the mail archive of the gcc@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]

Off Topic - COBOL Compiler



> I know this list is not really the place for this, but I would like to know
> if there is any real interest in developing this project.
> 
> I would be interested in developing a free(GPL) COBOL compiler. 

Good idea. :-) It crossed my mind about a half a year ago...

> 
> Are there any individuals/companies/organizations interested in this
> project ?
> 
> Are there any individuals/companies/organizations interested in donating
> source code for this project ? For example a COBOL parser(wishful thinking
> I think).
> 
> Here is a brief outline:
> - A front end compiler based on GCC(EGCS ?). Something like the GNU Pascal.
> - COBOL-85 ISO/ANSI standard. No Object COBOL.

Here is the catch - a COBOL compiler to binary code wouldn't be as
useful as a COBOL->C compiler. Many companies using COBOL code do that 
because they cannot afford to rewrite it in a better language. That's
where COBOL->C would come handy.

I'm dragging through the development of a COBOL->C compiler that
would:

1. Generate *correct* C code that'd work exactly the same way, the old 
   code did.

2. Produce sufficiently clear code to allow people to use the
   resulting C code independently.

It all started as an attempt to apply some Lex/Yacc knowledge I had
acquired from the O'Reilly book. The idea was to make a compiler that
would correctly translate all examples from the 'COBOL for Dummies'
book. Currently, the project is in the phase of stagnation - both
because I'm tight in time and because I'm a bit confused about the
parser generator - looks like COBOL doesn't go along (at least easily) 
with Lex/Yacc. PCCTS could've been an option, yet it is way less
popular than Lex/Yacc.

Finally, the project is developed under GPL and will be donated to FSF 
to become part of the GNU system once it gets more or less usable.

If you think about developing something similar, let's consider
joining our efforts...

Cheers,

-- 
Alexander L. Belikoff
Bloomberg L.P. / BFM Financial Research Ltd.
abel@vallinor4.com, abel@bfr.co.il


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]