Originally posted by: SystemAdmin

I am working on a scenario where I have got two file's:
file 1
HDR file with multiple PO's

& file 2
DTL file with multiple dtl lines for each PO

I am trying to now work on getting multiple output files each having unique PO HDR+ corresponding PO details.

I am presently getting one o/p file with each PO & its corresponding details.

How do i now separate it to create multiple o/p files depending on number of PO's??? (probably using PUT function)
#IBM-Websphere-Transformation-Extender
#DataExchange
#IBMSterlingTransformationExtender

Originally posted by: janhess

For each header use put to create a file and use something from the header to identify the file name.
#IBMSterlingTransformationExtender
#DataExchange
#IBM-Websphere-Transformation-Extender

Originally posted by: jvanboga

I like janhess idea. Could also execute a runmap for each hdr + dtl group. each method has adv/disadv.
#IBM-Websphere-Transformation-Extender
#IBMSterlingTransformationExtender
#DataExchange

Originally posted by: SystemAdmin

could u elaborate on the runmap for each hdr+dtl group ???
#DataExchange
#IBMSterlingTransformationExtender
#IBM-Websphere-Transformation-Extender

Originally posted by: jvanboga

essentially, you'd pass your header plus all of your related detail to a runmap. You'd also override the runmap output file name with your run command.

There should be an explanation on how to do that in one of the pdf's. Check out the Design Studio Intro, Command Server and Map Designer guides. You might also search your map (mms) files for either "run(" or "run (". Looking at them will give you some ideas.
#IBMSterlingTransformationExtender
#IBM-Websphere-Transformation-Extender
#DataExchange

Originally posted by: SystemAdmin

i am getting multiple o/p files using PUT func , but the records are not being grouped...

for eg

file 1-rec1
file 2-rec2
file 3-rec3
file 4-rec1

i want only 3 o/p files to be created & rec1 to be added to file 1....so file 1 should have 2 records....
any suggestions about this grouping ???
#DataExchange
#IBM-Websphere-Transformation-Extender
#IBMSterlingTransformationExtender

Originally posted by: SystemAdmin

Hi ASK,

I think in your case you better create a second map that has two input cards, one containing 1 header, and one containing s Details. You then call this map with RUN() for each Hdr you have, and you pass the following inputs to your second map:

• the Hdr
• all corresponding Details to this Hdr (with extract)

Here is an idea of how your RUN function will look like :

=VALID(RUN ( "Create_One_File", ECHOIN( 1, PACKAGE (Hdr) ) + ECHOIN( 2, PACKAGE ( EXTRACT (...all Dtl corresponding to given Hdr...) ) + " -OF1 File_xxxHdrxxx.txt"),
FAIL ("Error running map Create_One_File " + LastErrorCode() + " " + LasterrorMsg() ) )

The option -OF1 means you override the adapter of output card one with "File" and specify the filename with something coming out of your Hdr.

Hope this will help :)
François
#IBM-Websphere-Transformation-Extender
#DataExchange
#IBMSterlingTransformationExtender

Originally posted by: SystemAdmin

Hi François
I am unable to get exactly what you are trying to explain?
could you please elaborate clearly.
Thanks for your help.
regards
ASK
#IBM-Websphere-Transformation-Extender
#DataExchange
#IBMSterlingTransformationExtender

Originally posted by: janhess

I think you need something like this.

Input card 1 - Header records + POs
Input Card 2 - details records

output 1 file structure header, POs ,Details for each PO
use functional map to generate details by using Extract function

Output 2 have Item occurring S and use PUT with the append parameter (I think filename+)(might need to be through a functional map) and use part of the header that's coommon to the multiple headers that you want to group in one file.
#IBMSterlingTransformationExtender
#IBM-Websphere-Transformation-Extender
#DataExchange

Originally posted by: SystemAdmin

i require multiple files for each PO, not just the headers...
i/p file1 i/p file2
HDR for PO1 DTL for PO1
HDR for PO2 DTL for PO2
HDR for PO3 DTL for PO3

o/p to have multiple files as

file1 file2 file3
HDR for PO1 HDR for PO2 HDR for PO3
DTL for PO1 DTL for PO2 DTL for PO3
#DataExchange
#IBM-Websphere-Transformation-Extender
#IBMSterlingTransformationExtender

Originally posted by: janhess

I think that's what I was trying to say. The only thing you need to worry about is if there are multiple headers that you need to put into the same file where you don't want to output the additional headers.
#DataExchange
#IBMSterlingTransformationExtender
#IBM-Websphere-Transformation-Extender

Originally posted by: jvanboga

janhess is on the money. output card2 should produce a output file for each group of hdr+dtl
#IBMSterlingTransformationExtender
#IBM-Websphere-Transformation-Extender
#DataExchange

Originally posted by: SystemAdmin

i have created a sample scenario and attaching it here...

my ip file is
AK,2485226918
SB,6786440699
SM,4047254258
AK,9372454169

my o/p which iam getting is
all records in diffn files....so 4 o/p files getting created instead of 3.

as record AK is repeated, it should be grouped....

let me know where iam going wrong ???
#IBM-Websphere-Transformation-Extender
#DataExchange
#IBMSterlingTransformationExtender

Originally posted by: janhess

here's an example of producing the output files from your file in the body of your post. You can figure out the extract yourself.
#DataExchange
#IBMSterlingTransformationExtender
#IBM-Websphere-Transformation-Extender

Originally posted by: SystemAdmin

hey, the attached example is not working ???
#IBMSterlingTransformationExtender
#DataExchange
#IBM-Websphere-Transformation-Extender

Originally posted by: janhess

You need 2 subfolders called logs and trace. Otherwise change the map to put the logs & trace where you want them.
#IBM-Websphere-Transformation-Extender
#IBMSterlingTransformationExtender
#DataExchange

Originally posted by: janhess

I've done the work for you again.
#IBM-Websphere-Transformation-Extender
#IBMSterlingTransformationExtender
#DataExchange

Originally posted by: SystemAdmin

how about doing it this way ......i have tried to use control break while reading the data in type tree itself.....
#DataExchange
#IBM-Websphere-Transformation-Extender
#IBMSterlingTransformationExtender

Originally posted by: janhess

That will work if your input file is sorted. However, my solution will work for an unsorted file.
You could sort the data in the first output card if it is unsorted and use your method.
#DataExchange
#IBMSterlingTransformationExtender
#IBM-Websphere-Transformation-Extender

Originally posted by: janhess

I think this may be more efficient.
#IBM-Websphere-Transformation-Extender
#IBMSterlingTransformationExtender
#DataExchange

Originally posted by: janhess

Ignore the last attachment, I forgot to save the map.
#IBM-Websphere-Transformation-Extender
#DataExchange
#IBMSterlingTransformationExtender

Originally posted by: SystemAdmin

you r absolutely right....my method will not work on unsorted file....how do i then sort it?can i use sortup/dwn in the extract ??? the last attachment u sent does not create the req o/p for unsorted file ....
#IBM-Websphere-Transformation-Extender
#IBMSterlingTransformationExtender
#DataExchange

Originally posted by: jvanboga

I must have missed something here. thought you were trying to merge two files. one had header info, the other the detail.

If that's correct and you have small - med sized datasets you could use this formula. It can be a resource pig but it would work.
#DataExchange
#IBM-Websphere-Transformation-Extender
#IBMSterlingTransformationExtender

Originally posted by: SystemAdmin

U did not miss something....just a modification in the o/p that we get, we need to put it in multiple files for each HDR & its corresponding DTL set......

#DataExchange
#IBMSterlingTransformationExtender
#IBM-Websphere-Transformation-Extender

Originally posted by: SystemAdmin

i have done a slight modification to ur work & we now get the req o/p...
#DataExchange
#IBMSterlingTransformationExtender
#IBM-Websphere-Transformation-Extender

Originally posted by: jvanboga

I'd create an output tree defined something like

file
.output_group (0:s)
..hdr
..dtl (1:s)

Then a rule at the output_group level similar to janhess's or like the below (this is not very efficient if your file is large but will work fine for small - mid size files)

f_map_hdr (hdr rec, dtl file)

then in that f_map, rule 1 hdr:rec = hdr:rec, dtl rule = if (dtl condition matchs hdr rec then dtl:rec)

then output card 2 puts each output_group or runs a map for each output_group creating your files.
#IBM-Websphere-Transformation-Extender
#DataExchange
#IBMSterlingTransformationExtender

Originally posted by: jvanboga

btw.... not familiar with the rar file extention. what should i open it with?
#IBMSterlingTransformationExtender
#IBM-Websphere-Transformation-Extender
#DataExchange

Originally posted by: janhess

Winrar. You need to download it. It's like winzip but they charge for creating RAR files. But the example doesn't help. I think I answered the output problem 2 posts ago.
#DataExchange
#IBM-Websphere-Transformation-Extender
#IBMSterlingTransformationExtender