DSXchange

I am having two input files as -

input 1 input 2
A A
B B
C C
F D
E

I want output as :
A with data from Input 1 and Input 2
B with data from Input 1 and Input 2
C with data from Input 1 and Input 2
D
E
F with data from Input 1

Or if is not possible then create two outputs

Output 1 - Matched records

with
A with data from Input 1 and Input 2
B with data from Input 1 and Input 2
C with data from Input 1 and Input 2

and output 2 - Unmatched Records
with
D
E
F

Balkrishna wrote:I am having two input files as -

input 1 input 2
A A
B B
C C
F D
E

I want output as :
A with data from Input 1 and Input 2
B with data from Input 1 and Input 2
C with data from Input 1 and Input 2
D
E
F with data from Input 1

Or if is not possible then create two outputs

Output 1 - Matched records

with
A with data from Input 1 and Input 2
B with data from Input 1 and Input 2
C with data from Input 1 and Input 2

and output 2 - Unmatched Records
with
D
E
F

Use a Change Capture and Change Apply Stage in conjunction

Hi Ram,

Can you please explore the same.
It will be a great help to me.

Thanks,
Balkrishna

Balkrishna wrote:I am having two input files as -

input 1 input 2
A A
B B
C C
F D
E

I want output as :
A with data from Input 1 and Input 2
B with data from Input 1 and Input 2
C with data from Input 1 and Input 2
D
E
F with data from Input 1

Or if is not possible then create two outputs

Output 1 - Matched records

with
A with data from Input 1 and Input 2
B with data from Input 1 and Input 2
C with data from Input 1 and Input 2

and output 2 - Unmatched Records
with
D
E
F

ag_ram wrote: Use a Change Capture and Change Apply Stage in conjunction

Does DataStage TX comprise of those stages ?
I am not sure how you will do this with TX but since your on unix, you can easily do this with a simple 4-5 liner shell script.

I suppose you'd have to own Change Capture and Change Apply Stage first. Then you could figure out if you can execute their code from the TX map.

Many ways to do this....

How about merging the files then use the unique function.

You could also merge the files, perform a sort and compare current record to last record read.

Load all the data to a DB then use select unique/distinct.
.
.
.


Problem with trying to work with two distinct files is that you'd probably need to use the LOOKUP or SEARCHUP/DOWN functions. If your files are small it may not be a big deal but the larger the file the more resources and time. Also, each of those functions has some limitations.

Depending on the data, if I used any of the above solutions, I'd probably define each record as a text blob until I had the data I wanted.