Re: C to Ada83 Translation

Re: C to Ada83 Translation

[Boston Technical]

Del Fredricks (d-lfred@Primenet.Com) wrote:
: Has anyone attempted to write a straight through translator to 
: convert C, Kernighan and Ritchie, into Ada 83?

: I know, I know, not preferable, but in this case we have been painted 
: into a corner and need a rope to swing out.

: If so, can you recall what the actual effort turned out to be?  Also, is 
: there any such C to Ada translators which can be had, even for a 
: reasonable price?


: --
: ------------------------------------------------------------------------------
: Del & Laurie Fredricks     ___ __     ____  ___       ___ ___
: d-lfred@primenet.com      /__)/__) / / / / /_  /\  / /_    /
:                          /   / \  / / / / /__ /  \/ /___  /


Dear Sir:

In answer to your question, may I suggest SIEMENS' XperCASE family of tools.
These are PC-Based, Windows driven tools that require only 1.5MB of RAM 
per Module, and 1.5 MB of HD Space per Module.

These programs are being used by companies such as EDS, TRW, and the Jet 
Propulsion Labs, to do translation to and from ADA.  They were designed 
as re-engineering and reverse-engineering tools, and have been so well 
received that EDS, through a European subsidiary (MBP) is the European 
distributor of the product.

Further, XperCASE is reviewed in April 17, 1994 issue of PC WEEK, very 
favorably.  Additionally, it was written up in IEEE Computer, in July 
1994, Applications Trends Magazine (unsure of date), and Computer 
Reseller News (unsure of date).

You would need 2 modules of XperCASE.  One copy of XperCASE C/C++ PACK, 
at $1595., and one copy of XperCASE SPX (non-language specific), at $695.

You would read your code into the C/C++ PACK, which would break it into 
Structure Diagrams.  You than save the SDs, and inport them into SPX.  
That program than converts the SDs into Ada format, and you generate the 
ADA code through an ADA Compiler that you would link to SPX's Module.

Obviously there is some user involvement in this translation, but 
XperCASE was designed to standardize and simplify said translations.

I hope this information is helpful.  For further info, you can email us 
at BosTech@world,std.com.

Sincerely:

Charles Lavine
Boston Technical Distribution Corp.
617-248-8989

Re: C to Ada83 Translation

[Larry Kilgallen, LJK Software]

In article <3gre40$8tm@news.primenet.com>, d-lfred@Primenet.Com (Del Fredricks) writes:

> Has anyone attempted to write a straight through translator to 
> convert C, Kernighan and Ritchie, into Ada 83?

I think it would be great to have such a tool, even though the result
would not be "good" Ada.  Managers who would hesitate conversion to
Ada by conventional techniques certainly hesitate due to the amount
of time which would be spent on mechanical tasks, rather than effort
which might be spent on making the code more reliable.

My wife, who does not use Ada, commented that such a tool might be
useful even to sites which never use the resulting binary output.
Sort of a "super-Lint". (She does not use C either, but was forced
to once.)

Larry Kilgallen

Re: C to Ada83 Translation

[Thomas Wilsher]

>In article <3gre40$8tm@news.primenet.com>, d-lfred@Primenet.Com (Del Fredricks) writes:

>> Has anyone attempted to write a straight through translator to 
>> convert C, Kernighan and Ritchie, into Ada 83?

On my NeXT cube at home I have an executable named c2ada,
which translates C *headers* to the ada equivalent. 
I believe that this program was released in binary form only,
and I don't have the author's name here. 

Although much less ambitious in scope than what the original poster
suggested, it seems that this program would still be very useful.
It seems this program could be extended to generate Ada 
wrappers around C functions autmatically, the function bodies themselves
 being implemented in C and being imported. (I am assuming something like
GNAT's 'pragma import' here). That would allow you to get a system 
up and running 
fairly quickly from existing C code, and then one could go in
and rewrite the function bodies in Ada over time.
Any comments on the feasability of such a scheme? (I am not an Ada programmer)
Also, has anyone any pointers to the c2ada program, is it still 
publically available (maybe even with source, at this point), and is it any good?


--thomas