Generate new column at runtime

[krishna81]

Hi,
I need to generate columns at runtime of job.
job flow like

Seqfile--->TX---->SeqFile

here is my input.

I/p

column1:
ROW1--> aaaa/bbbb/cccc/dddd
ROW2--> eeee/ffff/gggg/hhhh

O/P shoul be

COLUMN1 COLUMN2 COLUMN3 COLUMN4
aaaa bbbb cccc dddd
eeee ffff gggg hhhh

there is no consistancy in delimited fields.


I appreciate for your quick response

kris.

[chulett]

there is no consistency in delimited fields.

What does this mean?

[krishna81]

For example
in the inputfile we have 6 delimited fields in column1
and at second run
in the inputfile we have 5 delimited fields in column1
and at 3rd run
inputfile we have 7 delimited fields in column1.


the expected out put must be

the Outputfile should create whith 6 columns(where each delimited field goes to each column)
and at second run
the Outputfile should create whith 5 columns(where each delimited field goes to each column)
and at 3rd run
the Outputfile should create whith 7 columns(where each delimited field goes to each column)

executing one datastage Job to process these different files.

[Vinodanand]

Hi,

Please check the below link,its not the exact requirement but can help you to think more on the same line,hope it helps !.
http://dsxchange.com/viewtopic.php?t=96 ... c63ff43e05

[chulett]

There's no automated way to do what you want. AFAIK, you will need to treat the record as one long varchar field in job and append the columns and field delimiters based on what you find in the input.