Re: sequential file formats
- From: "William M. Klein" <wmklein@xxxxxxxxxxxxxxxxx>
- Date: Thu, 28 Jul 2005 00:57:14 GMT
Your requests for "hex dumps" of files isn't getting many replies. I know that
it isn't something that I am about to try and do. What you really SHOULD be
asking for (IMHO) is references to online documentation. This will (again in my
opinion) be a much better way of finding out how different compilers and
run-times handle different "things".
For example,
Check out how all the different systems handle:
COMP-1, COMP-2, COMP-3, COMP-4, COMP-5, and COMP-6 (depending upon which
they have)
COMP
Other floating-point (if available)
Binary (big-endian/little-endian among other things)
sign-nibbles (Packed-Decimal and Usage Display)
UTF-16, DBCS, other - for USAGE NATIONAL (if supported)
ASCII vs EBCDIC (or other) for USAGE DISPLAY
Do they support BIT data? (or Boolean)
Do they support unsigned packed-decimal
also, references to "compiler options/directives" that impact sequential file
formats. See (for a good starting place)
http://www.talsystems.com/tsihome_h...oads/C2IEEE.htm
and
"Introduction to COBOL Numeric Data Types
Michael Mattias has created a text and graphics based introduction to COBOL
numeric data types. It was designed for intermediate to advanced programmers not
familiar with the COBOL language or COBOL numeric data storage. A Windows Write
or compatible word processor is recommended, but an ASCII text file format is
also included). The PKZIP format file includes an executable program file as
well as PB 3.2 source code. This file is Copyrighted but free for personal use.
A compiler is NOT required to use this product.
http://www.flexus.com/ftp/cobdata.zip
--
Bill Klein
wmklein <at> ix.netcom.com
"theador" <Theador@xxxxxxxxx> wrote in message
news:1122506771.018520.173960@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
> epc8@xxxxxxxx wrote:
>
>> Then dealing with the differences in record structure will be the
>> *least* of your problems. Coping with the myriad ways of encoding the
>> actual data is much more difficult.
>>
>> The easiest and most reliable way of moving data from COBOL X to COBOL
>> Y is via text files in which all of the digits in a number and its sign
>> (if present) are explicitly represented.
>
> Do you have some specific examples of different ways that different
> compilers encode data when writing to a sequential file? I would be
> especially interested in hex dumps of the differences rather than just
> a discussion. I am not doubting you, I would just like to see the
> actual examples. One that I am aware of is that IBM COBOL on AIX
> encodes the sign digit of COMP3 data differently than microfocus or IBM
> COBOL on zOS. I dont currently have access to those compilers so I am
> afraid I can not provide the hex dumps.
>
> Theador.
>
.
- Follow-Ups:
- Re: sequential file formats
- From: theador
- Re: sequential file formats
- References:
- sequential file formats
- From: theador
- Re: sequential file formats
- From: William M. Klein
- Re: sequential file formats
- From: theador
- Re: sequential file formats
- From: epc8
- Re: sequential file formats
- From: theador
- sequential file formats
- Prev by Date: Re: Diffence between two CICS READ ; FILE & DATASET
- Next by Date: Re: sequential file formats
- Previous by thread: Re: sequential file formats
- Next by thread: Re: sequential file formats
- Index(es):
Relevant Pages
|