Rexx to C++ Translator
remove....@dev.nul.in.replies
I was just wondering how well received a Rexx to C++ translator might
be?
I have been working on a set of class libraries to facilitate porting of
some legacy Rexx code to C++. However, the only existing translator I
was able to find spits out errors after the first 5 or 6 lines
(RexxTacy).
The Rexx code has been written to be able to run under PC-DOS Rexx, OS/2
Rexx, and AIX Rexx. As such, I was in dire need of a set of libraries
that would facilitate porting non-standard Rexx code (with all of the
conditional processing based on operating environment).
The library as it stands currently is being developped under Watcom C++
10.0a, with support for Rexx string, large number toolkit, and date
procedures.
I was thinking about writing a translator after I've finished this
project, just so that the class library has some kind of an afterlife.
The class library I am writing does not make any provision for 16-bit
operating systems, however. So, it will not (in all likelihood) ever be
ported to PC-DOS Rexx.
You, additionally may also email me at revolu...@hotmail.com with
your responses. (An MVS version won't be forthcoming however, as I'm
not able to get my hands on an MVS machine, or on an MVS C++
compiler...I do have an emulator though...)
F. Scott Ophof
remove....@DEV.NUL.IN.REPLIES wrote:
(what an address... ;-> )
} I was just wondering how well received a Rexx to C++ translator might
} be?
Sounds like a *good* idea!
Rapid development/testing in REXX, and then compilable in C(?)/C++
to various environments.
All one needs then is someone good in C/C++ to smooth out the
inevitable differences between REXX and C/C++ mindsets.
(Those of you who used any version of REXXify to translate EXEC[2]
programs to REXX will know what I mean here.)
I was wondering what had happened to RexxTacy...
Regards,
$$/ F. Scott Ophof <FSO...@BLCLinks.net> Member of RexxLA
Buck, Peter INS
If you can do INTERPRET and VALUE() the rest should be a piece of pie.
Daniel Bruce Lynes
F. Scott Ophof wrote:
> remove....@DEV.NUL.IN.REPLIES wrote:
>
> (what an address... ;-> )
Thanks....I hate those damned spammers...send 'em all to /dev/nul
> } I was just wondering how well received a Rexx to C++ translator might
> } be?
>
> Sounds like a *good* idea!
> Rapid development/testing in REXX, and then compilable in C(?)/C++
> to various environments.
Well, for the time being, I'm just developing the class library for coping
with manual conversions of Rexx to C++ code, as it is really needed on my
current project.
> All one needs then is someone good in C/C++ to smooth out the
> inevitable differences between REXX and C/C++ mindsets.
Differences? Surely you jest? MVS Rexx programmers seem like Commodore
PET line-numbered BASIC programmers to me. This code is hairy at the best
of times.
> I was wondering what had happened to RexxTacy...
I think it died. (Don't quote me on that...)
Daniel Bruce Lynes
Buck, Peter INS wrote:
> If you can do INTERPRET and VALUE() the rest should be a piece of pie.
INTERPRET shouldn't be much of a problem. However, VALUE() is very much
like other Rexx procedures that I've come across in this porting process
that I've begun to hate with a passion. I've really begun to hate Rexx
syntax. :) Well, and mainframe programmers....
Daniel Bruce Lynes
Buck, Peter INS wrote:
> If you can do INTERPRET and VALUE() the rest should be a piece of pie.
The main portion I've been dealing with so far is Rexx's large number
toolkit, and Rexx's string functions. Rexx's string functions are more or
less fully implemented. Rexx large number toolkit has a good headstart,
but by no means close to being finished. I'm currently implementing it so
that it can deal with binary, octal, decimal, and hexadecimal. However,
the option is there to work with base 40 and base 360 (they just happen to
be interesting numeric bases.)
Dennis Bareis
Daniel Bruce Lynes <remove....@dev.nul.in.replies> wrote:
}Buck, Peter INS wrote:
}
}> If you can do INTERPRET and VALUE() the rest should be a piece of pie.
}
}INTERPRET shouldn't be much of a problem. However, VALUE() is very much
}like other Rexx procedures that I've come across in this porting process
}that I've begun to hate with a passion. I've really begun to hate Rexx
}syntax. :) Well, and mainframe programmers....
}
I think you meant "Value() will be easy but Interpret() will be hard or more
likely impossible"! Unless of course you handle interpret by letting rexx
handle it!
If all the rexx functions were available in 'C' then I'd have to start
wondering why you'd write anything but smallish once off programs in rexx (note
I've written a LOT of rexx & 'C' code)....
Just my 2 cents worth,
--------------------------------
Dennis Bareis
Melbourne, AUSTRALIA.
email : d...@anz.com
HomePage: http://www.ozemail.com.au/~dbareis
Contains heaps of OS/2 & DOS freeware I've written (as well as some source code and programmers links)
Examples HTML/REXX preprocessor, REXX compiler, File generation via template...
--------------------------------