Coulmn name in file

nani0907 · Post by **nani0907** » Fri Jan 30, 2009 6:26 am

Hi all,

I need to populate the column name if the it has values .else no values has present then empty file should be created. By setting the option in sequential file for column names first= true doesnot help me as if there are no values,only column name is generated. Could any body help me how to achieved this by derivations.

keshav0307 · Post by **keshav0307** » Fri Jan 30, 2009 6:30 am

please try to be very clear on your requirement.
do you mean if there are no record then empty file should be generated else the first row should be the column names??

nani0907 · Post by **nani0907** » Fri Jan 30, 2009 7:37 am

If there are no records then empty files should be generated.

if we have records then we should aslo pupulate column names along with records.

Thanks

Mike · Post by **Mike** » Fri Jan 30, 2009 7:43 am

Out of the box... you don't get conditional column headers. If that's your requirement, you have to design and code it yourself...

Mike

keshav0307 · Post by **keshav0307** » Fri Jan 30, 2009 7:51 am

OK , Create two files in the job, one file for only header records with column names only, another file with records only and then in after job subroutine, append the second file into first file.

nani0907 · Post by **nani0907** » Fri Jan 30, 2009 7:56 am

If there are no values in second file, the first file will populate the column names.but my case is that if no records, column names also should not be created.

keshav0307 · Post by **keshav0307** » Fri Jan 30, 2009 8:15 am

that was just an idea.
you can have a condition , if second file size is greater then Zero then only append. Else delete the first file and then touch <file name>

Mike · Post by **Mike** » Fri Jan 30, 2009 8:33 am

Lots of potential ways to skin that cat. An easy one that comes to mind is to "after job" count the lines in the file and if the count is one (i.e. headers only), then empty it out.

Mike

mk_ds09 · Post by **mk_ds09** » Fri Jan 30, 2009 10:14 am

One more way you can achieve that is.
in properties set column names 'yes', it will give you column names in the output file.

In after job subroutine.
If wc -l output <= 1 then touch output_file.

that will create empty file if there is only one record with the columns.

hope this helps..

----------------
MK