DSXchange

Posted: **Fri Mar 29, 2013 10:29 pm**

Hi,

I have a requirement of creating generic jobs to read the COBOL file using CFF stage and do some basic transformations and load into Tera data. But unfortunately I am not able to use Schema files with CFF stage.

Currently I am using Sequential file to read the file using the schema files and RCP . But I am not sure to pass the transformation dynamically .

Here is an example of my transformation requirements.

Code: Select all


Input Column Names :

1.  COLUMN1
2.  COLUMN2
3.  COLUMN3
4.  COLUMN4
4.  COLUMN4_NULL_IND

And Output Column Names :

1.  COLUMN1
2.  COLUMN2
3.  COLUMN3
4.  If IsNull(COLUMN4_NULL_IND)  Then SetNull() Else COLUMN4 


And for each each file the column names differs.

Kindly someone let me know any solution/logic to achieve this.

Many Thanks,
Anand

Posted: **Sat Mar 30, 2013 5:26 am**

Only the RMM (Read My Mind) stage. You cannot effect transformation without naming the columns in the transformation, with very limited exceptions in the Modify stage (which does not have If..Then..Else capability).

Posted: **Sat Mar 30, 2013 8:50 am**

In teradata, is it going to be bulk load?
Why not achieve this simple transformation in the SQL query?
In this case , like one schema file for each file, you must have a sql query for each table and it must be maintained somewhere.
Not sure of the feasibility. Just a thought

Posted: **Sun Mar 31, 2013 10:47 am**

Yes. This will be bulk load.
Anyway I will try with using customized SQL queries .

_________________
Thanks
Anand

DSXchange

How to pass transformation dynamically

How to pass transformation dynamically