View previous topic :: View next topic
|
Author |
Message |
raja_natural
New User
Joined: 07 Apr 2010 Posts: 2 Location: Mumbai
|
|
|
|
Hi,
I have below 3 input Input files
file 1: VB LRECL= 509
1-4 : ID (integer)
9-13 : Rec-type
14-23: Act-date
File 2: VB LRECL = 535
27-30 ID(integer)
35-39 rec-type
40-49 act-date
File 3: FB LRECL = 58
user_id 4
rec 6
PS-DATE 10
STRT-DATE 10
END-DATE 10
DTL-TYPE 10
TRANS 4
SRC 4
i wanted to copy all the files to a common output file.
i want to concatenate all the 3 files and create common output file.
file 1:
1111 AA 2015-10-01
2222 BB 2015-11-01
File 2:
3333 CC 2015-02-01
4444 CC 2015-01-16
5555 CC 2015-01-10 2015-01-11 2015-01-12 type1 ACCD XXXX
6666 CC 2015-01-10 2015-01-11 2015-01-12 type2 ABCX XXXX
7777 CC 2015-01-10 2015-01-11 2015-01-12 type3 AXXD XXXX
Output file should be :
1111 AA 2015-10-01
2222 BB 2015-11-01
3333 CC 2015-02-01
4444 CC 2015-01-16
5555 DD 2015-01-10 type1 ACCD XXXX
6666 DD 2015-01-11 type2 ABCX XXXX
7777 EE 2015-01-12 type3 AXXD XXXX
is it possible to combine all these steps and achieve the output result in one step.
Thanks in advance for your help. |
|
Back to top |
|
|
Bill Woodger
Moderator Emeritus
Joined: 09 Mar 2011 Posts: 7309 Location: Inside the Matrix
|
|
|
|
What happened to File 3 in your data?
You can concatenate the two VB files. |
|
Back to top |
|
|
raja_natural
New User
Joined: 07 Apr 2010 Posts: 2 Location: Mumbai
|
|
|
|
Sorry for the typo Bill.
File 3 : data
5555 CC 2015-01-10 2015-01-11 2015-01-12 type1 ACCD XXXX
6666 CC 2015-01-10 2015-01-11 2015-01-12 type2 ABCX XXXX
7777 CC 2015-01-10 2015-01-11 2015-01-12 type3 AXXD XXXX |
|
Back to top |
|
|
Bill Woodger
Moderator Emeritus
Joined: 09 Mar 2011 Posts: 7309 Location: Inside the Matrix
|
|
|
|
Because of the mix of VB and FB you can't concatenate directly.
You could use SORT to convert (FTOV) your FB to VB, with VLTRIM if you have large numbers of some trailing value.
Then concatenate in a simple COPY.
If desperate for one step, there is ICETOOL.
Even JOINKEYS (concatenate the two VBs on F1, and put some lower value in a temporary extension in the JNF1CNTL, and the FB for F2, and put some higher value. "Match" on the temporary values (there won't be any) and use presence or otherwise to know where to build the output from (make the REFORMAT variable-length, and that's how it works for the "conversion" in this). |
|
Back to top |
|
|
|