[kde-linux] 20090622KL -- Kobol & MySQL

[John Culleton]

On Tuesday 23 June 2009 01:36:15 pm Bruce MacArthur wrote:
> On Tuesday 23 June 2009 10:21, John Culleton wrote:
> > On Monday 22 June 2009 08:12:25 pm Bruce MacArthur wrote:
> > > Greetings --
> > >
> > > Does anyone know of a way to work with existing MySQL
> > > databases from within KOBOL (or COBOL) programs?  Does REKALL
> > > simplify this -- or just change the complexities?  What else
> > > might be helpful -- either to have or to know?  Thanks for
> > > any assistance possible!!!
> >
> > Not familiar with KOBOL or REKALL. In general COBOL has its own
> > file system including an indexed file system with the
> > possibility of multiple keys per record, a primary which must
> > be unique and alternate keys that need not be unique.  I
> > suggest downloading your data from MySql  in the form of a
> > report  and then reading the lines of that report into a COBOL
> > program that will create a suitable indexed sequential file.
> >
> > Suggest Open Cobol instead of a commercial version.  Also, use
> > the mailing list
> > "open-cobol-list" <open-cobol-list at lists.sourceforge.net>
> > instead of this one for COBOL questions.
>
> Hello, John --
> KOBOL and REKALL are commercial products from "The Kompany". 
> KOBOL includes an IDE and the compiler produces native code
> directly -- rather than "translating" to C and then compiling. 
> It is suitable for both Windows and Linux; I'm not sure about
> Mac.  You take one single source-code deck and simply compile it
> on the respective system.
>
> I am "leary" of the translate-and-compile approach, partly
> because I have never seen the evidence that implicit
> redefinitions work in C. For instance, consider the following
> Data Division code --
>
> 01  DATE-SALE                    PIC X(10).
> 01  DATE-SALE-RDEF          REDEFINES DATE-SALE.
>       05  DATE-SALE-Y           PIC 9(04).
>       05  FILLER                     PIC X(01).
>       05  DATE-SALE-M          PIC 9(02).
>       05  FILLER                     PIC X(01).
>       05  DATE-SALE-D          PIC 9(02).
>
> If C cannot do this "natively", then how can it be made to handle
> the task in any kind of "translation"?  (Of course, if it CAN do
> this, and I am just that ignorant of C, then there is NO problem
> at all!)
>
> I certainly do understand the COBOL and ISAM notion.  Keys are no
> problem to me.  But what if I want to define a MySQL database and
> then load it from other data, using a COBOL / KOBOL program? 
> This is why professional COBOL allows a program to make
> simultaneous use of both genuine databases AND "native" files.
>
> THANK YOU for pointing me to OpenCOBOL -- I had not been aware of
> it previously.  I have examined some of their web-site.  I was
> slightly frustrated because there was no visible PLUS workable
> (!) User Manual. I found quite a few bits and pieces about
> compiler options and the like -- which can be VERY good to know!
> -- but no manual as-such.
>
> In spite of these "issues", OpenCOBOL may well be my best option.
>  I will continue my investigations.  THANKS AGAIN.
>
> By the way, thank you, also, for pointing me to that specific
> list.  I agree that it will be much more appropriate.  I mailed
> this list because of its general character and the fact that I
> will be using COBOL within KDE-Linux.  I suspect that the
> mailing-list you cited is pretty-much limited specifically to the
> OpenCOBOL product, and is NOT aimed at COBOL in general. 
> Therefore, I am guessing that my general inquiry would not
> "really" belong there either!  Am I wrong again?

They are OC only but aren't anal about it. Tiny COBOL is also good 
but its creator, Rildo Pragana,  has gone on to other work. The OC 
manual is old, but basically it tells you what the command line 
switches do and how to link subprograms, COBOL to COBOL COBOL to C 
and so on.  You are assumed to have a COBOL manual or text lying 
around :<) OC itself is pretty much standards driven, unlike most 
of the commercial products which as always are looking for product 
differentiation and unique features. It were always thus in the 
world of COBOL, from the 1960s on. 

There are two  COBOL news groups also, comp.lang.cobol and 
alt.cobol.   Try http://groups.google.com

-- 
John Culleton
Create Book Covers with Scribus/e-book $5.95
http://www.booklocker.com/books/4055.html