DSXchange

Hi,

I have a table with 400 columns and i may expect 200 columns with null data ( say 10% of cases ) , and my job will get aboret in dev / sit / production if no of warnings > 50 , if i want to handle this at one go instead of writing for each and every column, how to achieve this?

Any easier method to achieve the same.

Thanks,
friend.kak@gmail.com

Hi

suppress the warnings from director .

Thanks
sanjay

friend.kak@gmail.com wrote:Hi,

I have a table with 400 columns and i may expect 200 columns with null data ( say 10% of cases ) , and my job will get aboret in dev / sit / production if no of warnings > 50 , if i want to handle this at one go instead of writing for each and every column, how to achieve this?

Any easier method to achieve the same.

Thanks,
friend.kak@gmail.com

I also faced similar problem ...i think we have to edit the extended properties of each and every field ...This has been discussed in the forum sometime back ...

Hi

suppress the warnings from director .

Thanks
sanjay

friend.kak@gmail.com wrote:Hi,

I have a table with 400 columns and i may expect 200 columns with null data ( say 10% of cases ) , and my job will get aboret in dev / sit / production if no of warnings > 50 , if i want to handle this at one go instead of writing for each and every column, how to achieve this?

Any easier method to achieve the same.

Thanks,
friend.kak@gmail.com

Oops..i thought that its a sequential file ....i didnt see the table ...

nagarjuna wrote:I also faced similar problem ...i think we have to edit the extended properties of each and every field ...This has been discussed in the forum sometime back ...

Still, same advice applies. Do it for every field that needs it, 4 or 400. Or perhaps back up in the source database instead, with something like NVL for Oracle. In any case, handle it appropriately regardless of the number.

Research the Derivation Substitution utility in the Transformer stage.

You can modify multiple columns in one go - if they have similar / identical derivations.

Thanks for replies so far, but still I need a smart method of handling this at one go , may be handling at source side as suggested , or may be a routine which can help on this..... Yet to decide on this implementation. Thanks, friend.kak@gmail.com

Thanks for replies so far, but still I need a smart method of handling this at one go , may be handling at source side as suggested , or may be a routine which can help on this..... Yet to decide on this implementation. Thanks, friend.kak@gmail.com

No Short Cut, Need to use Derivation Substitution as suggested by Ray.

Did you try to handle the nulls into a schema file ?

Hi All,

I have tried to handle nulls for decimal and varchar field using schema files. It works wonderfully.
Just assign default="<value"> and then run an update for that "<value>"

Regards
Sreeni

Research "derivation substitution" here or in the manuals.

Hi Sreeni,

Thanks for info, Do you mean to say to create schema file like this and try.

col1: nullable timestamp {null_field=""} ;
col2: nullable timestamp {null_field=""} ;

...
col400: nullable timestamp {null_field=""}

Do you mean the same? if this is the same , then we are doing Null handling here itelf. i tried this. Thanks. friend.kak@gmail.com