Re: Converting packed decimal to a char array

You can use the functions you have found but it is more work that you need I suspect. In some cases sprintf might be enough since the additional formatting option D(n,p) is available. The MI built-ins can also provide alternative approaches for some of these functions. For example EDIT_PACKED does a high speed conversion of a packed source to an external character form.

These are all listed in the MI reference http://www-01.ibm.com/support/knowledgecenter/api/content/nl/en-us/ssw_i5_54/books/sc092418.pdf - this seems to be the latest version but not sure. Each of the definitions includes an example of calling the function from C.

Sorry I don’t understand your comment at the end regarding the returned structures. Do you have an RPG definition of them we could look at - it would make it easier to see what you are trying to do. For example you said "The input and outputs are both data buffers that are effectively packed structures” but for us the word “packed” has meaning as referencing a type of numeric field. I’m thinking that when you say “packed” you mean struct with not separators between the fields?

An RPG definition of a struct (known as a DS) incorporating a packed numeric might look like this:

d results ds
d char 10a // 10 character char array
d pkd1 9p 2 // 7.2 packed decimal - occupies 5 bytes
// sign is in rightmost nyble
d zoned1 9s 2 // 7.2 zoned numeric - occupies 10 bytes
// sign is high order bit settings in rightmost byte

If you can show us the structure you are dealing with it would be a lot easier to help you. I’m not a C programmer - I know just enough to be really dangerous - but I know RPG!


Jon Paris

www.partner400.com
www.SystemiDeveloper.com

On Jan 10, 2015, at 5:56 AM, Jevgeni Astanovski <jevgeniast@xxxxxxxxx> wrote:

Thanks, Jon.

Well, by now I found these editing function and got the following:

decimal (15,0) A = 125678D ;
char szTemp[256], szOutput[256], EditMask[256], ZeroFill[1] ;
int n, m, MaskLen, RcvLen ;
Qus_EC_t error_code ;

n = digitsof(A) ;
m = precisionof(A) ;

QECCVTEC(EditMask, &MaskLen, &RcvLen, ZeroFill, "P", " ", n, m,
&error_code) ;

QECEDT(szTemp, RcvLen, &A, "*PACKED ", n, EditMask, MaskLen, "0",
&error_code) ;
memcpy(szOutput, szTemp, RcvLen) ;
szOutput[RcvLen] = 0x00 ;
printf("After conversion A: !%s!\n", szOutput) ;


It produces the following:
After conversion A: ! 125678!
What I expected is a number padded with zeros in the left.
I have to check if the receiver program will accept the changed input...

Is it what you meant in your advise?

You asked what for do I need it.
I have a big set of API-s on iSeries that are called via RPC from PC
environment. The input and outputs are both data buffers that are
effectively packed structures - that is data fields without any
separators with known widths. This is as infrastructure and I cannot
change it. Packed numbers are always returned as char arrays - and it
is also part of the infrastructure....



On Fri, Jan 9, 2015 at 5:47 PM, Jon Paris <jon.paris@xxxxxxxxxxxxxx> wrote:

It would help to know why you are trying to do this (editing the number for example?).

Have you looked at the various editing and conversion capabilities available in the surfaced MI functions (https://publib.boulder.ibm.com/iseries/v5r1/ic2924/books/c0924180.pdf) as well as the C/C++ library extensions? (http://www-01.ibm.com/support/knowledgecenter/#!/ssw_ibm_i_72/rtref/rtrefmain.htm).

Everything you could need is there.


Jon Paris

www.partner400.com
www.SystemiDeveloper.com

On Jan 9, 2015, at 4:00 AM, Jevgeni Astanovski <jevgeniast@xxxxxxxxx> wrote:

Hi all,

spend a lot of time and feeling like I run into a dead cycle with a
problem, that looks trivial.

I have packed decimals in my C programs that I need to convert to a
char array, representing the number (not zero terminated string).

For example:

decimal (15,9) A ;
char szTemp[32], sField[16] ;

sprintf(szTemp, "%016D(15,9)", A) ;
memcpy(sField, szTemp, 16) ;

What I want to achieve is get a function, that receives A and returns sField.
Of course it must support any types of decimal - not only decimal
(15,9) but others like decimal(11,7), decimal (15,0), decimal (5,0)
and so on. And, of course, decimal can be positive or negative...


Any ideas?

Jevgeni.
--
This is the Bare Metal Programming IBM i (AS/400 and iSeries) (C400-L) mailing list
To post a message email: C400-L@xxxxxxxxxxxx
To subscribe, unsubscribe, or change list options,
visit: http://lists.midrange.com/mailman/listinfo/c400-l
or email: C400-L-request@xxxxxxxxxxxx
Before posting, please take a moment to review the archives
at http://archive.midrange.com/c400-l.

--
This is the Bare Metal Programming IBM i (AS/400 and iSeries) (C400-L) mailing list
To post a message email: C400-L@xxxxxxxxxxxx
To subscribe, unsubscribe, or change list options,
visit: http://lists.midrange.com/mailman/listinfo/c400-l
or email: C400-L-request@xxxxxxxxxxxx
Before posting, please take a moment to review the archives
at http://archive.midrange.com/c400-l.

--
This is the Bare Metal Programming IBM i (AS/400 and iSeries) (C400-L) mailing list
To post a message email: C400-L@xxxxxxxxxxxx
To subscribe, unsubscribe, or change list options,
visit: http://lists.midrange.com/mailman/listinfo/c400-l
or email: C400-L-request@xxxxxxxxxxxx
Before posting, please take a moment to review the archives
at http://archive.midrange.com/c400-l.