|
View previous topic :: View next topic
|
| Author |
Message |
rackshit
New User
Joined: 19 Dec 2009
Posts: 16
Location: Mumbai
|
|
|
|
Hi,
i needed a solution for a problem.
I have table1
| Code: |
W-TABLE1 OCCURS 10
INDEXED BY W-TABLE1-IND.
07 W-field1 PIC X(30).
|
and table2
| Code: |
W-TABLE2 OCCURS 15
INDEXED BY W-TABLE2-IND.
07 W-field2 PIC X(20).
|
Now i have to move table1 data to table2 in such a way that 1st 20 of table 1 should be moved to 1st of table2 and the reamining 10 of table1 to the next element in table2, and then start with second record of table1 such that it fits in the reamining 10 of table2, and so on.
Please suggest me a solution for this. I have a bigger table than this which will contains data upto 4000 characters.
I have thought of pseudocode of first moving table1 to a string with length of 600 then cutting the slices and moving into table2. But for my actual length of 4000 i fear the string might be too long.
Thanks in advance.. |
|
| Back to top |
|
 |
dbzTHEdinosauer
Global Moderator
Joined: 20 Oct 2006
Posts: 6966
Location: porcelain throne
|
|
|
|
you are accomplishing nothing by placing the OCCURS clause on the high level element.
typical rookie 'save a couple lines of typing' trick.
define you working-storage as:
| Code: |
01 BIG-AREA.
03 TABLE-1-AREA.
05 W-TABLE1 OCCURS 10
INDEXED BY W-TABLE1-IND.
07 W-FIELD1 PIC X(30).
03 TABLE-1-AREA-RED
REDEFINES
TABLE-1-AREA.
05 W-TABLE1A.
07 W-FIELD1A OCCURS 15
INDEXED BY W-TABLE1A-IND
PIC X(20).
03 TABLE-2-AREA.
05 W-TABLE2 OCCURS 15
INDEXED BY W-TABLE2-IND.
07 W-FIELD2 PIC X(30).
|
then your code is easy:
| Code: |
PERFORM VARYING W-TABLE1A-IND
FROM 1
BY 1
UNTIL W-TABLE1A-IND > 15
SET W-TABLE2-IND TO W-TABLE1A-IND
MOVE W-FIELD1A (W-TABLE1A-IND)
TO W-FIELD2 (W-TABLE2-IND)
END-PERFORM
|
you were trapped by your inability to professionally define Working-Storage |
|
| Back to top |
|
|
Escapa
Senior Member
Joined: 16 Feb 2007
Posts: 1399
Location: IL, USA
|
|
|
|
I suggest you pre process this file in SORT
If you have DFSORT installed at your shop RESIZE operator will do this job very easily. |
|
| Back to top |
|
|
Marso
REXX Moderator
Joined: 13 Mar 2006
Posts: 1353
Location: Israel
|
|
|
|
Levels of W-TABLE1 and W-TABLE2 is missing.
I will assume they are not at level 01:
| Code: |
01 W-TABLE1-AREA.
03 W-TABLE1 OCCURS 10
INDEXED BY W-TABLE1-IND.
07 W-field1 PIC X(30).
01 W-TABLE2-AREA.
03 W-TABLE2 OCCURS 15
INDEXED BY W-TABLE2-IND.
07 W-field2 PIC X(20).
PROCEDURE DIVISION.
...
MOVE W-TABLE1-AREA TO W-TABLE2-AREA !!! |
That should do the trick !!! (and correct me if I am wrong) |
|
| Back to top |
|
|
Escapa
Senior Member
Joined: 16 Feb 2007
Posts: 1399
Location: IL, USA
|
|
|
|
| Quote: |
| That should do the trick !!! |
Yes .. That should.. |
|
| Back to top |
|
|
Marso
REXX Moderator
Joined: 13 Mar 2006
Posts: 1353
Location: Israel
|
|
|
|
| rackshit wrote: |
| I have a bigger table than this which will contains data upto 4000 characters. |
Now we have to see if this solution fits the real definition... |
|
| Back to top |
|
|
Bill Woodger
Moderator Emeritus
Joined: 09 Mar 2011
Posts: 7309
Location: Inside the Matrix
|
|
|
|
As Marso has pointed out, each of the tables is 300 bytes long. Just move one to the other.
This will work with whatever size tables or items the compiler allows, which you can check in the manual but I'm sure they will be sufficient for any requirment that you have. |
|
| Back to top |
|
|
ridgewalker58
New User
Joined: 26 Sep 2008
Posts: 51
Location: New York
|
|
|
|
| When someone CANNOT see what MARSO saw -- I wonder why Mr. Rackshit has a job (I assume) and so many of my friends are out of work. A simple understanding of how our templates (01 group levels and 05, 10, 15 elementary levels) form and reform the data items that we deal with -- doesn't seem to be in Mr. Rackshit's question. |
|
| Back to top |
|
|
Bill Woodger
Moderator Emeritus
Joined: 09 Mar 2011
Posts: 7309
Location: Inside the Matrix
|
|
|
|
| Plus, even you don't need the data simultaneously, you can just use REDEFINES and no MOVEs at all. |
|
| Back to top |
|
|
ridgewalker58
New User
Joined: 26 Sep 2008
Posts: 51
Location: New York
|
|
|
|
WOW -- I guess we get caught up in seeing things ONLY one way.
No moves, just redefines --- saves TIME spent moving data around |
|
| Back to top |
|
|
Bill Woodger
Moderator Emeritus
Joined: 09 Mar 2011
Posts: 7309
Location: Inside the Matrix
|
|
|
|
Still wondering about the purpose of the question. Did a 20-byte table get loaded, then "used" as a 30-byte table (like in a sub-program), then need to get "put back" to it's original state?
Looking back, I see dbz got the redefines first and then (maybe, I'm no mind-reader, can only intuit) decided TS must then want to do something with the 20-byte data, so stuffed it into a new 30x15 table. I should have looked at the code, not just dismiss it when I saw the loop. |
|
| Back to top |
|
|
rackshit
New User
Joined: 19 Dec 2009
Posts: 16
Location: Mumbai
|
|
|
|
Thanks All, for your valuable answers i ve got answer to my question although I had not put the tables in correct way ..
as far as the need is concerned, below is the need:
i have IMS db where one segment has one field with length of 30 and this same field is to be copied to another segment with field having length of 20. And the information which is to be copied/updated is distributed across no. of occurrences of 30 length variable. So first it is unloaded in table1 and then when the information is contiguous i will redefine as said above to table2 and update it in segment2 across successive fields of length 20 (for different occurrences).
at the time of its creation decades back this database had the above layout. the segment having field of 30 length is a kind of lookup table and the segment having field of length 20 is actually data stored.
Still in coding phase, will put problems and solutions of wat m doing/find while i move to compilation and testing.. watch out for this space..
Aim is learning and sharing !! Thanks |
|
| Back to top |
|
|
|
|
