Row Splitter

[Kirtikumar]

Hi,

I am using Row splitter in one DS job. According to the document, any columns that are defined on both, the input link and output link, are passed straight through the stage.

But jobs is as follows:

Code: Select all

Seq. File ---> Row Splitter ---> Seq. file

The input link has 2 columns :

Code: Select all

RECTYPE      Char  1
DATA         Char  300

I am splitting the column DATA and has been mentioned in the stage. The output link has meta data :

Code: Select all

RECTYPE      Char  1
FNAME        Char  150
LNAME        Char  150

According to document the RECTYPE should carried as it is and DATA columns should be split.
But here, RECTYPE is not carried forward. Though all RECTYPE in input seq. file are populated, in output file these fields are blank.

Any inputs will be appreciated.

[yannish]

Hi Kirti,

try moving the pass through columns to the end of the output column definitions and in the Output/Format-page check "Missing columns action" to be None. That should work.


Cheers,

Janne
Finland

PS. The sun has been seen here too

[Precious]

try moving the pass through columns to the end of the output column definitions and in the Output/Format-page check "Missing columns action" to be None. That should work

I have tried the suggested solution, and it works. :D


Is there any reason that the pass thru columns have to be at the end? Is there no way to implement this with the pass thru columns at the beginning?

Thanx

[Kirtikumar]

Hi Janne,

I am working in DS 7.1 Server. I checked Output-Format page. But this option was not there. Is this option available only in 7.5?

Regards,
Kirtikumar.

PS: Its summer here and temp has been around 40.

Hi Kirti,

try moving the pass through columns to the end of the output column definitions and in the Output/Format-page check "Missing columns action" to be None. That should work.


Cheers,

Janne
Finland

PS. The sun has been seen here too

[yannish]

It should be there. I just checked my DS 7.1. On the bottom left corner...

Precious: At least I haven't managed to put the passthrough columns in the begining of the output columns. I think it is the way row splitter works that it first splits the column specified and only after it runs out of data in that column it starts to look for other columns. But then again it's quite simple transformer stage to reorder the columns so I don't see big problem there. But I have to say it would be a nice feature to have the mapping of columns made a bit easier

Janne