tiny-cobol-users

Re: [Tiny-cobol-users] STRING problem (DELIMITED BY SPACES)

[Carlos G. <cg...@ya...>]

David
 
String verb that has strage output has no DELIMITED in  first token (by default DELIMITED BY SIZE). DELIMITED BY SPACES is used only in WTEXT1 token (with a value of "12345     ") . It seems that using DELIMITED BY SPACES only in one token does the compiler to use this option in every token of the command.
 
 
         STRING "INSERT INTO TEST " ,  <<<--- No DELIMITED (Default BY SIZE)
             ",",
             W9,
             ",",
             WTEXT1 DELIMITED BY SPACES,
             ");" INTO WRESULT.

Result:
 
INSERT,999,12345);

 
Using DELIMITED BY SIZE in first token gives the correct result. 
 
         STRING "INSERT INTO TEST " DELIMITED BY SIZE,
             ",",
             W9,
             ",",
             WTEXT1 DELIMITED BY SPACES,
             ");" INTO WRESULT.

Result:
 
INSERT INTO TEST ,999,12345);

 
It looks like default DELIMITED BY SIZE doesn´t be use by TC in this case.
 

David Essex <de...@us...> escreveu:
Carlos Gomes wrote:

> In the test program above, the second display (results of String verb
> using DELIMITED BY SPACES) generate a strange output. DEMILITED BY
> SPACES in the fourth string causes a truncation in the first string
> (by default DEMILITED BY SIZE).
> DISPLAY 3, including DELIMITED BY SIZE explicity in the first string,
> generates the expected output.

If I understand you correctly, the second display is generating this output,

2-INSERT,999,12345);

and should be generating this output,

2-INSERT INTO TEST ,999,12345);

correct ?

I've looked at the COBOL 1989 (draft) standard, and I think TC is
generating the correct output '2-INSERT,999,12345);', for the following
reasons.

The figurative constant SPACES (literal) is equivalent to " " (single
space).
Thus the phrase 'DELIMITED BY SPACES' is equivalent to
'DELIMITED BY " "'.
And the literal 'INSERT INTO TEST ' gets truncated to 'INSERT'.

BTW, what are other commercial compilers generating ?


Syntax:
STRING {
{identifier-1 | literal-1}...
DELIMITED [BY]
{identifier-2 | literal-2 | SIZE }
}...
INTO identifier-3

COBOL standard:
14.10.39 STRING statement
...
14.10.39.3 General rules
...
3) When a figurative constant is specified as literal-1 or literal-2, it
refers to an implicit one character data item whose usage shall be the
same as the usage of identifier-3, either display or national.





-------------------------------------------------------
This SF.Net email is sponsored by the 'Do More With Dual!' webinar happening
July 14 at 8am PDT/11am EDT. We invite you to explore the latest in dual
core and dual graphics technology at this free one hour event hosted by HP, 
AMD, and NVIDIA. To register visit http://www.hp.com/go/dualwebinar
_______________________________________________
Tin...@li...
https://lists.sourceforge.net/lists/listinfo/tiny-cobol-users

		
---------------------------------
Yahoo! Acesso Grátis: Internet rápida e grátis. Instale o discador agora!

Re: [Tiny-cobol-users] STRING problem (DELIMITED BY SPACES)

[Fred M. <fr...@mo...>]

David Essex wrote:
> 
> Carlos Gomes wrote:
> 
>  > In the test program above, the second display (results of String verb
>  > using DELIMITED BY SPACES) generate a strange output. DEMILITED BY
>  > SPACES in the fourth string causes a truncation in the first string
>  > (by default DEMILITED BY SIZE).
>  > DISPLAY 3, including DELIMITED BY SIZE explicity in the first string,
>  > generates the expected output.
> 
> If I understand you correctly, the second display is generating this output,
> 
> 2-INSERT,999,12345);
> 
> and should be generating this output,
> 
> 2-INSERT INTO TEST ,999,12345);
> 
> correct ?
> 
> I've looked at the COBOL 1989 (draft) standard, and I think TC is
> generating the correct output '2-INSERT,999,12345);', for the following
> reasons.
> 
> The figurative constant SPACES (literal) is equivalent to " " (single
> space).
> Thus the phrase 'DELIMITED BY SPACES' is equivalent to
> 'DELIMITED BY " "'.
> And the literal 'INSERT INTO TEST ' gets truncated to 'INSERT'.

David, you are correct. I cannot read the original message from Carlos
Gomes because of the message format but I often used DELIMITED BY SIZE
or DELIMITED BY "  " (two space characters between quotes). The latter
works just fine in COBOL in case intermediate spaces are present in a
sending field. 
-- 
Fred Mobach - fr...@mo... - pos...@mo...
Systemhouse Mobach bv - The Netherlands - since 1976
website : http://fred.mobach.nl
Q: servos ad pileum vocare ?
A: servos fenestrae ad pileum rubrem vocare !

Re: [Tiny-cobol-users] STRING problem (DELIMITED BY SPACES)

[David E. <de...@us...>]

Carlos Gomes wrote:

 > In the test program above, the second display (results of String verb
 > using DELIMITED BY SPACES) generate a strange output. DEMILITED BY
 > SPACES in the fourth string causes a truncation in the first string
 > (by default DEMILITED BY SIZE).
 > DISPLAY 3, including DELIMITED BY SIZE explicity in the first string,
 > generates the expected output.

If I understand you correctly, the second display is generating this output,

2-INSERT,999,12345);

and should be generating this output,

2-INSERT INTO TEST ,999,12345);

correct ?

I've looked at the COBOL 1989 (draft) standard, and I think TC is
generating the correct output '2-INSERT,999,12345);', for the following
reasons.

The figurative constant SPACES (literal) is equivalent to " " (single
space).
Thus the phrase 'DELIMITED BY SPACES' is equivalent to
'DELIMITED BY " "'.
And the literal 'INSERT INTO TEST ' gets truncated to 'INSERT'.

BTW, what are other commercial compilers generating ?


Syntax:
STRING {
{identifier-1 | literal-1}...
DELIMITED [BY]
{identifier-2 | literal-2 | SIZE }
}...
INTO identifier-3

COBOL standard:
14.10.39 STRING statement
...
14.10.39.3 General rules
...
3) When a figurative constant is specified as literal-1 or literal-2, it
refers to an implicit one character data item whose usage shall be the
same as the usage of identifier-3, either display or national.

Re: [Tiny-cobol-users] MicroFocus Compatibility option

[David E. <de...@us...>]

Carlos Gomes wrote:

> How do I know MF features associated with option
> --enable-mfcomp when I generate htcobol with this option???

There are only two options which must be explicitly enabled when TC is 
configured.

COMMAND-LINE:
The statement 'ACCEPT identifier-1 FROM COMMAND-LINE' will retrieve the 
command-line arguments and place them in identifier-1, delimited by spaces.
The TC default is to use the UN*X like feature and include the first 
argument, the actual external name (binary) of the calling program.
With the 'mfcomp' option, the external name (binary) is omitted.

example (default):
$prog1.exe a b c -> 'prog1.exe a b c'

example (mfcomp):
$prog1.exe a b c -> 'a b c'


See 'test.code/t17' directory for an example.

EXTERN-FILE:
SELECT ... ASSIGN TO EXTERNAL [ literal-1 | identifier-1 ]

The EXTERNAL clause of the 'SELECT' statement enables the association
of internal file identifiers with external file names, and is achieved 
by the use of environment variables.
This feature mimics the methodology found on main-frames.
The TC default is to use identifier (or literal) and use it to search 
for the environment variable. If the environment variable is found, use 
the name as the actual file name.
With the 'mfcomp' option, the search sequence is more complex and a 
'DD_' prefix is used. It basically mimics the methodology found in MF 
compilers.
See 'fileio.c' file for details.

example (default):
$export SYS002=/home/user123/filein1
...
SELECT filein1 ASSIGN TO EXTERNAL "SYS002"
then ...
filein1=/home/user123/filein1

See 'test.code/t19' directory for an example.

Perhaps a better approach would be to add these as run-time options, 
which could then be defined by the user.

Re: [Tiny-cobol-users] MicroFocus Compatibility option

[Carlos G. <cg...@ya...>]

How do I know  MF features associated with option 
--enable-mfcomp   when I generate htcobol with this option???
 
David Essex <de...@us...> escreveu:
BB wrote:

> How do I invoke the option to map files using the DD_ environment 
> variables?
> 
> You included some kind of compiler option or something?
> 
> (Yes I know I'm losing it)
> 
> Jim Morcombe
> (using someone else's computer)

You have define this option at the configure step.

$./configure --help
...
--enable-mfcomp enable MF compatibility features
...

Since 'DD_' is a MF option,

$./configure --enable-mfcomp

then build the compiler and run-time.



-------------------------------------------------------
This SF.Net email is sponsored by the 'Do More With Dual!' webinar happening
July 14 at 8am PDT/11am EDT. We invite you to explore the latest in dual
core and dual graphics technology at this free one hour event hosted by HP, 
AMD, and NVIDIA. To register visit http://www.hp.com/go/dualwebinar
_______________________________________________
Tin...@li...
https://lists.sourceforge.net/lists/listinfo/tiny-cobol-users

		
---------------------------------
Yahoo! Acesso Grátis: Internet rápida e grátis. Instale o discador agora!

Re: [Tiny-cobol-users] STRING problem (DELIMITED BY SPACES)

[Carlos G. <cg...@ya...>]

David
 
In the test program above, the second display (results of String verb using DELIMITED BY SPACES) generate a strange output. DEMILITED BY SPACES in the fourth string causes a truncation in the first string (by default DEMILITED BY SIZE). 
 
DISPLAY 3, including DELIMITED BY SIZE explicity in the first string, generates the expected output.
 
       IDENTIFICATION DIVISION.
       PROGRAM-ID. SWPROG.
       AUTHOR. SW.
       ENVIRONMENT DIVISION.
       CONFIGURATION SECTION.
       SPECIAL-NAMES.
           DECIMAL-POINT IS COMMA.
       DATA DIVISION.
       WORKING-STORAGE SECTION.
       01  W1.
           03 A1  PIC 9(01).
           03 A2  PIC 9(02).
           03 A3  PIC 9(03).
           03 A4  PIC 9(04).
       01  W2.
           03 B1  PIC 9(01).
           03 B2  PIC 9(02).
           03 B3  PIC 9(03).
       01  WRESULT     PIC X(100) VALUE SPACES.
       01  WTEXT     PIC X(100) VALUE SPACES.
       01  WTEXT1    PIC X(10) VALUE "12345     ".
       01  W9        PIC 999 VALUE 999.
       PROCEDURE DIVISION.
       PARAG.
         MOVE SPACES TO WRESULT.
         STRING "INSERT INTO TEST ",
             ",",
             W9,
             ",",
      *       WTEXT1 DELIMITED BY SPACES,
             ");" INTO WRESULT.
         DISPLAY "1-" WRESULT.
         MOVE SPACES TO WRESULT.
         STRING "INSERT INTO TEST " ,
             ",",
             W9,
             ",",
             WTEXT1 DELIMITED BY SPACES,
             ");" INTO WRESULT.
         DISPLAY "2-" WRESULT.
         MOVE SPACES TO WRESULT.
         STRING "INSERT INTO TEST " DELIMITED BY SIZE,
             ",",
             W9,
             ",",
             WTEXT1 DELIMITED BY SPACES,
             ");" INTO WRESULT.
         DISPLAY "3-" WRESULT.
         MOVE SPACES TO WRESULT.
         STRING "INSERT INTO TEST ",
             ",",
             W9,
             ",",
             WTEXT1 ,
            ");" INTO WRESULT.
         DISPLAY "4-" WRESULT.
         STOP RUN.

 
 
Results:
 
 
1-INSERT INTO TEST ,999,);
2-INSERT,999,12345);
3-INSERT INTO TEST ,999,12345);
4-INSERT INTO TEST ,999,12345     );


David Essex <de...@us...> escreveu:
David Essex wrote:

> Yes, there are several problems here. 
> First the TC grammar does not conform to the COBOL85 standard.
> 
> Syntax:
> STRING {
> {identifier-1 | literal-1}...
> DELIMITED [BY]
> {identifier-2 | literal-2 | SIZE }
> }...
> INTO identifier-3
> ...

OK, I was wrong, and the two books I consulted were wrong.

I'm not sure if this was part of the original COBOL 1985 standard, or a 
later modification, but according to the COBOL 1989 (draft) standard:

STRING ...
The DELIMITED phrase may be omitted only immediately preceding the INTO 
phrase. If it is omitted, DELIMITED BY SIZE is implied.


> Carlos Gomes wrote:
> 
>> I am getting a invalid result when use STRING verb.
>> The sample above has 3 similar STRINGs, using DELIMITED
>> BY SPACES the result is strange!!
>> ...
>> Result in:
>>
>> DISPLAY 1--> INSERT,999,'12345');
>> DISPLAY 2--> INSERT INTO TEST ,999,'12345 ');
>> DISPLAY 3--> INSERT INTO TEST ,999,'12345 ');

I don't see what the problem is.

Exactly what output did you expect ?



-------------------------------------------------------
SF.Net email is sponsored by: Discover Easy Linux Migration Strategies
from IBM. Find simple to follow Roadmaps, straightforward articles,
informative Webcasts and more! Get everything you need to get up to
speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click
_______________________________________________
Tin...@li...
https://lists.sourceforge.net/lists/listinfo/tiny-cobol-users

__________________________________________________
Converse com seus amigos em tempo real com o Yahoo! Messenger 
http://br.download.yahoo.com/messenger/ 

Re: [Tiny-cobol-users] Can i call DB2 database from tinyCOBOL

[David E. <de...@us...>]

Stefan Heublein wrote:

 > ...
 >>TC uses the STDCALL (C call convention) as a default, so the STDCALL
 >>keyword is NOT required.
 >
 > Refering to the following TC grammar from the news group archive,
 > TC is using CDECL as default... :-)
 >
 >>Grammar:
 >>CALL [ STDCALL | WINAPI | CDECL ] name ....

Sorry, my mistake.

It is STDCALL and WINAPI (PASCAL ?) which are the same.

So the STDCALL (WINAPI) keyword IS required when using the MinGW
compiler tool chain.

Re: [Tiny-cobol-users] Bold and underline not working

[David E. <de...@us...>]

Massimo Colani wrote:

 > Will this problem be solved in future releases or
 > will it stay?

I'm currently trying to resolve the problems.

I do have a fix which works in an xterm or rxvt, but does not work 
properly when using a UN*X standard console.
No tests have been perfomed on Win32 using PDcurses.

Also there is a bug with the color set code, resulting in the screen 
background color to be set to what ever the first field background color is.

If you like to the test the code I have, I could update SF CVS.

Re: [Tiny-cobol-users] Bold and underline not working

[Massimo C. <mas...@ya...>]

Buona sera!

Will this problem be solved in future releases or will
it stay?

Massimo

--- David Essex <de...@us...> ha
scritto: 

> Massimo Colani wrote:
> 
>   > I think Roland is right. By configuring Tiny
> Cobol 0.62,
>   > everything goes ok when the test for color_set
> is positive.
>   > When this test fails, the whole make of TC is
> not correct
>   > and the screen layout is bot working.
>   >
>   > Look here, this is a example program I am
> running on a VT 510
>   > genuine terminal, but it also fails on the
> system console
>   > with TERM=linux:
> 
> Yes, there is definitely some thing wrong.
> 
> It appears that the entries and attributes are out
> of sync.
> 
> The problems can be found in version 0.56 and later.
> 
> 
> 
> 
> 
>
-------------------------------------------------------
> SF.Net email is sponsored by: Discover Easy Linux
> Migration Strategies
> from IBM. Find simple to follow Roadmaps,
> straightforward articles,
> informative Webcasts and more! Get everything you
> need to get up to
> speed, fast.
>
http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click
> _______________________________________________
> Tin...@li...
>
https://lists.sourceforge.net/lists/listinfo/tiny-cobol-users
> 



		
___________________________________ 
Yahoo! Messenger: chiamate gratuite in tutto il mondo 
http://it.beta.messenger.yahoo.com

[Tiny-cobol-users] Re: Re: Can i call DB2 database from tinyCOBOL

[<he...@ne...>]

Hi David,

> As of version 0.62, TC does support literals as program identifiers
> (PROGRAM-ID).
Good news. Thank you for the information.

> TC uses the STDCALL (C call convention) as a default, so the STDCALL
> keyword is NOT required.

Refering to the following TC grammar from the news group archive, TC is usi=
ng CDECL as default... :-)
=20
> Grammar:
> CALL [ STDCALL | WINAPI | CDECL ] name ....
=20
> If the call convention option is omitted, the default will be CDECL, or
> the standard C call convention. This is what TC normally uses.
> Note that no entries in the SPECIAL-NAMES clause are required.

Best regards
Stefan


Machen Sie aus 14 Cent spielend bis zu 100 Euro!
Die neue Gaming-Area von Arcor - =FCber 50 Onlinespiele im Angebot.
http://www.arcor.de/rd/emf-gaming-1

Re: [Tiny-cobol-users] MicroFocus Compatibility option

[David E. <de...@us...>]

BB wrote:

> How do I invoke the option to map files using the DD_ environment 
> variables?
> 
> You included some kind of compiler option or something?
> 
> (Yes I know I'm losing it)
> 
> Jim Morcombe
> (using someone else's computer)

You have define this option at the configure step.

$./configure --help
...
--enable-mfcomp enable MF compatibility features
...

Since 'DD_' is a MF option,

$./configure --enable-mfcomp

then build the compiler and run-time.

Re: [Tiny-cobol-users] Can i call DB2 database from tinyCOBOL

[David E. <de...@us...>]

Thanks for the information.


Stefan Heublein wrote:

 > ...
 > - TC doesn't support literals as program identifiers

As of version 0.62, TC does support literals as program identifiers
(PROGRAM-ID).


 > - TC requires the STDCALL keyword for DB2 api calls (Microfocus COBOL:
 > SPECIAL-NAMES. CALL-CONVENTION 74 IS DB2API. CALL DB2API
 > "<api function>")

Unfortunatly, according to the standard, the meaning of the call
convention is implementor-defined.

TC uses the STDCALL (C call convention) as a default, so the STDCALL
keyword is NOT required.

Since the SPECIAL-NAMES section is only partially implimented, so
the CALL-CONVENTION clause is not implimented.


 > - I ran into problems with the precompiler generated "by content"
 > parameter passing to the DB2 api. Converting "by content" to "by
 > reference" solved the problem. Later I found another solution for
 > this problem.
 > When aligning and enlarging the passed structure to word boundary
 > (4 bytes), the "by content" parameter passing can remain unchanged...

Sounds like a boundary (4 bytes) bug of some sort.

Re: [Tiny-cobol-users] Can i call DB2 database from tinyCOBOL

[Stefan H. <he...@ne...>]

Hi,

it's possible to compile and run the IBM supplied DB2 COBOL samples using=
 TC
on Windows.

First of all, you must compile the COBOL source which contains embedded S=
QL
with the DB2 precompiler supplied by IBM (included in the free applicatio=
n
development client package). The precompiler is capable of generating
Microfocus COBOL compatible code. This leads to plain COBOL source (witho=
ut
embedded SQL), which has to be modified for compiling with TC:

- TC doesn't support literals as program identifiers
- TC requires the STDCALL keyword for DB2 api calls (Microfocus COBOL:
SPECIAL-NAMES. CALL-CONVENTION 74 IS DB2API. CALL DB2API "<api function>"=
)
- I ran into problems with the precompiler generated "by content" paramet=
er
passing to the DB2 api. Converting "by content" to "by reference" solved =
the
problem. Later I found another solution for this problem. When aligning a=
nd
enlarging the passed structure to word boundary (4 bytes), the "by conten=
t"
parameter passing can remain unchanged...

Last, but not least, you have to link the TC compiled module to the DB2 a=
pi
libs. These libs are intended for use with Microsoft's Visual C and have =
to
be converted to Mingw usable libs before (see Mingw faqs).

Good luck!

I succeeded to compile and run a COBOL program containing embedded (stati=
c)
SQL and accessing DB2 for z/OS (using DB2 Connect).

Best regards
Stefan
-----Urspr=FCngliche Nachricht-----
Von: tin...@li...
[mailto:tin...@li...]Im Auftrag von
--__--__--

Message: 1
Date: Tue, 5 Jul 2005 21:59:35 -0700 (PDT)
From: nikil chakravarthy <nik...@ya...>
To: tiny-cobol-users <tin...@li...>
Subject: [Tiny-cobol-users] Can i call DB2 database from tinyCOBOL

Hi All,

I am newbie for tinyCOBOL development. I am struggling
with finding information about tinyCOBOL with DB2
database access. Can somebody let me know whether it
is possible to connect to DB2 from tinyCOBOL and send
me some URL's and resources. Your help is very much
appreciated. Looking forward for positive response.

Thanks & Regards,
Nikil

__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around
http://mail.yahoo.com


--__--__--

Message: 2
Date: Wed, 06 Jul 2005 02:35:56 -0400
From: David Essex <de...@us...>
Reply-To: tiny-cobol-users <tin...@li...>
To: tiny-cobol-users <tin...@li...>
Subject: Re: [Tiny-cobol-users] Can i call DB2 database from tinyCOBOL

nikil chakravarthy wrote:

 > I am newbie for tinyCOBOL development. I am struggling
 > with finding information about tinyCOBOL with DB2
 > database access. Can somebody let me know whether it
 > is possible to connect to DB2 from tinyCOBOL and send
 > me some URL's and resources. Your help is very much
 > appreciated. Looking forward for positive response.

TC does not have an SQL pre-processor, thus you can only access any SQL
database using the native API, and not using embedded SQL.

If DB2 includes an SQL pre-processor for COBOL, and I think it does, the
following may (or may not) work.
Pass the embedded SQL COBOL program thru the DB2 SQL pre-processor for
COBOL. And then using TC, compile it's output adding the appropriate DB2
library(s) in the link step.

Re: [Tiny-cobol-users] STRING problem (DELIMITED BY SPACES)

[David E. <de...@us...>]

David Essex wrote:

> Yes, there are several problems here. 
> First the TC grammar does not conform to the COBOL85 standard.
> 
> Syntax:
> STRING {
> {identifier-1 | literal-1}...
> DELIMITED [BY]
> {identifier-2 | literal-2 | SIZE }
> }...
> INTO identifier-3
> ...

OK, I was wrong, and the two books I consulted were wrong.

I'm not sure if this was part of the original COBOL 1985 standard, or a 
later modification, but according to the COBOL 1989 (draft) standard:

STRING ...
The DELIMITED phrase may be omitted only immediately preceding the INTO 
phrase. If it is omitted, DELIMITED BY SIZE is implied.


> Carlos Gomes wrote:
> 
>> I am getting a invalid result when use STRING verb.
>> The sample above has 3 similar STRINGs, using DELIMITED
>> BY SPACES the result is strange!!
>> ...
>> Result in:
>>
>> DISPLAY 1--> INSERT,999,'12345');
>> DISPLAY 2--> INSERT INTO TEST ,999,'12345     ');
>> DISPLAY 3--> INSERT INTO TEST ,999,'12345     ');

I don't see what the problem is.

Exactly what output did you expect ?

[Tiny-cobol-users] MicroFocus Compatibility option

[BB <be...@by...>]

David,

How do I invoke the option to map files using the DD_ environment variables?

You included some kind of compiler option or something?

(Yes I know I'm losing it)

Jim Morcombe
(using someone else's computer)

Re: [Tiny-cobol-users] STRING problem (DELIMITED BY SPACES)

[David E. <de...@us...>]

Yes, there are several problems here.

First the TC grammar does not conform to the COBOL85 standard.

Syntax:
STRING {
{identifier-1 | literal-1}...
DELIMITED [BY]
{identifier-2 | literal-2 | SIZE }
}...
INTO identifier-3

So the following grammar is invalid,

STRING ...
WTEXT1 DELIMITED BY SIZE,
"'", ");"
INTO WRESULT.

STRING ...
");"
INTO WRESULT.

and should be as follows;

STRING ...
WTEXT1 DELIMITED BY SIZE,
"'" ");" DELIMITED BY SIZE,
INTO WRESULT.

STRING ...
");" DELIMITED ...
INTO WRESULT.

The second problem is invalid output, may be (or not) related to the 
grammar problem.
So further investigation will be required.


Carlos Gomes wrote:

 > I am getting a invalid result when use STRING verb.
 > The sample above has 3 similar STRINGs, using DELIMITED
 > BY SPACES the result is strange!!
 > ...
 > WORKING-STORAGE SECTION.
 > 01  WRESULT     PIC X(100) VALUE SPACES.
 > 01  WTEXT     PIC X(100) VALUE SPACES.
 > 01  WTEXT1    PIC X(10) VALUE "12345     ".
 > 01  W9        PIC 999 VALUE 999.
 > PROCEDURE DIVISION.
 > PARAG.
 > STRING "INSERT INTO TEST ",
 >  ",",W9,
 >  ",","'",
 >  WTEXT1 DELIMITED BY SPACES, "'",
 >  ");" INTO WRESULT.
 > DISPLAY WRESULT.
 > MOVE SPACES TO WRESULT.
 > STRING "INSERT INTO TEST ",
 >  ",",W9,
 >  ",","'",
 >  WTEXT1 DELIMITED BY SIZE, "'",
 >  ");" INTO WRESULT.
 > DISPLAY WRESULT.
 > MOVE SPACES TO WRESULT.
 > STRING "INSERT INTO TEST ",
 >  ",",W9,
 >  ",","'",
 >  WTEXT1 , "'",
 >  ");" INTO WRESULT.
 > DISPLAY WRESULT.
 > STOP RUN.
 >
 > Result in:
 >
 > DISPLAY 1--> INSERT,999,'12345');
 > DISPLAY 2--> INSERT INTO TEST ,999,'12345     ');
 > DISPLAY 3--> INSERT INTO TEST ,999,'12345     ');

[Tiny-cobol-users] STRING problem (DELIMITED BY SPACES)

[Carlos G. <cg...@ya...>]

I am getting a invalid result when use STRING verb. The sample above has 3 similar STRINGs,  using DELIMITED BY SPACES the result is strange!!
 
Can anybody help me?
 
       IDENTIFICATION DIVISION.
       PROGRAM-ID. PROG.
       AUTHOR. S.
       ENVIRONMENT DIVISION.
       CONFIGURATION SECTION.
       SPECIAL-NAMES.
           DECIMAL-POINT IS COMMA.
       DATA DIVISION.
       WORKING-STORAGE SECTION.
       01  WRESULT     PIC X(100) VALUE SPACES.
       01  WTEXT     PIC X(100) VALUE SPACES.
       01  WTEXT1    PIC X(10) VALUE "12345     ".
       01  W9        PIC 999 VALUE 999.
       PROCEDURE DIVISION.
       PARAG.
           STRING "INSERT INTO TEST ",
           ",",W9,
           ",","'",
           WTEXT1 DELIMITED BY SPACES, "'",
          ");" INTO WRESULT.
         DISPLAY WRESULT.
          MOVE SPACES TO WRESULT.
           STRING "INSERT INTO TEST ",
           ",",W9,
           ",","'",
           WTEXT1 DELIMITED BY SIZE, "'",
          ");" INTO WRESULT.
         DISPLAY WRESULT.
          MOVE SPACES TO WRESULT.
           STRING "INSERT INTO TEST ",
           ",",W9,
           ",","'",
           WTEXT1 , "'",
          ");" INTO WRESULT.
         DISPLAY WRESULT.
         STOP RUN.

 
Result in:
 
 
DISPLAY 1--> INSERT,999,'12345');
DISPLAY 2--> INSERT INTO TEST ,999,'12345     ');
DISPLAY 3--> INSERT INTO TEST ,999,'12345     ');


		
---------------------------------
Yahoo! Acesso Grátis: Internet rápida e grátis. Instale o discador agora!

Re: [Tiny-cobol-users] Can i call DB2 database from tinyCOBOL

[David E. <de...@us...>]

nikil chakravarthy wrote:

 > I am newbie for tinyCOBOL development. I am struggling
 > with finding information about tinyCOBOL with DB2
 > database access. Can somebody let me know whether it
 > is possible to connect to DB2 from tinyCOBOL and send
 > me some URL's and resources. Your help is very much
 > appreciated. Looking forward for positive response.

TC does not have an SQL pre-processor, thus you can only access any SQL 
database using the native API, and not using embedded SQL.

If DB2 includes an SQL pre-processor for COBOL, and I think it does, the 
following may (or may not) work.
Pass the embedded SQL COBOL program thru the DB2 SQL pre-processor for 
COBOL. And then using TC, compile it's output adding the appropriate DB2 
library(s) in the link step.

[Tiny-cobol-users] Can i call DB2 database from tinyCOBOL

[nikil c. <nik...@ya...>]

Hi All,

I am newbie for tinyCOBOL development. I am struggling
with finding information about tinyCOBOL with DB2
database access. Can somebody let me know whether it
is possible to connect to DB2 from tinyCOBOL and send
me some URL's and resources. Your help is very much
appreciated. Looking forward for positive response.

Thanks & Regards,
Nikil

__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 

Re: [Tiny-cobol-users] Can Jave or C program call COBOL module?

[David E. <de...@us...>]

steven gao wrote:

 > Can Java or c program call cobol modules? in Tinycobol
 > sample files, i find COBOL program file can be easy to
 > call c file program.
 > But COBOL as a legacy system, we often use our new Java
 > or C modules to call COBOL file.are there any methods?

Yes, TC compiled COBOL (sub)programs can call C functions, and C 
programs can call COBOL (sub)programs.
There are several examples in the 'test.code' directories.

Just think of TC compiled COBOL (sub)programs as equivalent to C 
functions, with a return type of INT.

I would suggest using the current CVS version (0.62.13), as it has 
options to define entry points.

AFAIK, no testing has been done using JAVA.

[Tiny-cobol-users] Can Jave or C program call COBOL module?

[steven g. <ste...@gm...>]

Hi,
Can Java or c program call cobol modules? in Tinycobol sample files, i
find COBOL program file can be easy to call c file program.But COBOL
as a legacy system, we often use our new Java or C modules to call
COBOL file.
are there any methods?

Steven

Re: [Tiny-cobol-users] Bold and underline not working

[David E. <de...@us...>]

Massimo Colani wrote:

  > I think Roland is right. By configuring Tiny Cobol 0.62,
  > everything goes ok when the test for color_set is positive.
  > When this test fails, the whole make of TC is not correct
  > and the screen layout is bot working.
  >
  > Look here, this is a example program I am running on a VT 510
  > genuine terminal, but it also fails on the system console
  > with TERM=linux:

Yes, there is definitely some thing wrong.

It appears that the entries and attributes are out of sync.

The problems can be found in version 0.56 and later.

[Tiny-cobol-users] Re: Bold and underline not working

[Massimo C. <mas...@ya...>]

Hello,
 
I think Roland is right. By configuring Tiny Cobol 0.62, everything goes ok when the test for color_set is positive. When this test fails, the whole make of TC is not correct and the screen layout is bot working.
 
Look here, this is a example program I am running on a VT 510 genuine terminal, but it also fails on the system console with TERM=linux:
 

IDENTIFICATION DIVISION.

PROGRAM-ID. TSCR.

DATA DIVISION.

WORKING-STORAGE SECTION.

77 W-SOMETHING PIC X(20).

SCREEN SECTION.

01 S-TESTSCREEN BLANK SCREEN.

03 LINE 10 COLUMN 10 VALUE "Enter something:".

03 LINE 10 COLUMN 27 PIC X(20) USING W-SOMETHING

HIGHLIGHT UNDERLINE.

PROCEDURE DIVISION.

MOVE SPACES TO W-SOMETHING.

DISPLAY S-TESTSCREEN.

ACCEPT S-TESTSCREEN.

DISPLAY SPACES.

DISPLAY "$" W-SOMETHING "$".

STOP RUN.

 

The second term erases the first one and the screen flashes with every key pressed. The output of ./tscr | tee tscr.out is here uuencoded:

begin 600 tscr.out

M&S<;6S$[,C1R&S@;6T@;6S)*&ULQ,#LQ2"`@("`@("`@($5N=&5R('-O;65T

M:&EN9SH;6T@;6S)*&ULQ,#LT-T@;6T@;6S)*&ULQ,#LR-T@;6T@;6S)*&ULQ

M,#LQ2"`@("`@("`@("`@("`@("`@("`@("`@("`@2!M;2!M;,DH;6S$P.S%(

M("`@("`@("`@("`@("`@("`@("`@("`@("!(91M;2!M;,DH;6S$P.S%(("`@

M("`@("`@("`@("`@("`@("`@("`@("!(96P;6T@;6S)*&ULQ,#LQ2"`@("`@

M("`@("`@("`@("`@("`@("`@("`@2&5L;!M;2!M;,DH;6S$P.S%(("`@("`@

M("`@("`@("`@("`@("`@("`@("!(96QL;QM;2!M;,DH;6S$P.S%(("`@("`@

M("`@("`@("`@("`@("`@("`@("!(96QL;R`@("`@("`@("`@("`@(!M;,3`[

M,C=((!M;,3$[,D@D&ULQ,CLR2$AE;&QO("`@("`@("`@("`@("`@&ULQ,SLR

;2"0;6S(T.S%(&UL_,C5H&S<;6S$[,C1R&S@-

`

end

 

I hope this works because my Windows does not seem to like this cut-and-paste.

 

I tried several versions of TC. Like Roland said, the problem occurs since version 0.56 here, too.

 

Massimo

 


		
---------------------------------
Yahoo! Mail: gratis 1GB per i messaggi, antispam, antivirus, POP3

Re: [Tiny-cobol-users] Bold and underline not working

[David E. <de...@us...>]

ROLAND RENNINGER wrote:

 > David Essex wrote:
 >
 >>ROLAND RENNINGER wrote:
 >>
 >>Both the underline and highlight features work on my system.
 >>Both in an xterm and standard Linux terminal.
 >>
 >>The 'ncurses' library does generate the escape sequences.
 >>
 >>Looking at your TC executable library dependencies, I noticed you are
 >>using and old version (3.4) of the 'ncurses' library.
 > ...
 > Well, thanks for your explaination. For the (n)curses library
 > is rooted deep in the operating system, noone here is enthusiastic
 > about replacing it, for TC is the only software having problems
 > with the old version.

If I recall, the objective was to determine if a newer version of the
(n)curses library fixed the problem.

The reason I suggested upgrading the OS, to say the 2.2 kernel (1999),
is because I think that it is the easiest way to test and resolve the
problem.

Just use an old 486 (or above) with 16M RAM and a network card. Then
install a 2.2 system (without X), compile TC and test applications.
This could be achieved in a couple of hours.

The alternative is to upgrade to a new (n)curses library, or have more
than one (n)curses library on the old system (is this possible ?). Since
these options may make you system un-usable, it will require some test
system.
This option would require a lot more work.


 > What we are not able to understand is why nobody wants or can explain
 > why TC <= 0.55 worked, and the newer versions do not work properly any
 > more. That's fact, and we can provide lots of screenshots to prove
 > this. So, something must have changed in the meantime inside the TC
 > source code.

Well here the reasons.

First of all, the people who worked on that portion of the TC Run-Time
are no longer active contributors to TC. Some are not even on TC mailing 
list.

Personally I'm not very familiar with the (n)curses library, or
terminals in general. And any information terminals is very hard to find.

In an attempt to isolate the problem, I've compiled and run some test 
programs using TC version 0.55 and the current CVS version. Both using 
ncurses version 5. While both versions run, the TC 0.55 version seams 
more stable, as it can be used with different TERM settings.

A lot of changes have been made to the curses section of the run-time 
since between 0.55 and 0.62, so it may take some time to try to isolate 
the problem(s).

Re: [Tiny-cobol-users] Bold and underline not working

[ROLAND R. <R.R...@GM...>]

David Essex wrote:
> ROLAND RENNINGER wrote:
> 
> Both the underline and highlight features work on my system. Both in an 
> xterm and standard Linux terminal.
> 
> The 'ncurses' library does generate the escape sequences.
> 
> Looking at your TC executable library dependencies, I noticed you are 
> using and old version (3.4) of the 'ncurses' library.

Well, thanks for your explaination. For the (n)curses library is rooted
deep in the operating system, noone here is enthusiastic about
replacing it, for TC is the only software having problems with the old
version.

What we are not able to understand is why nobody wants or can explain
why TC <= 0.55 worked, and the newer versions do not work properly any
more. That's fact, and we can provide lots of screenshots to prove
this. So, something must have changed in the meantime inside the TC
source code.

Rol'still coding on VMS'and