DSXchange

Hi,

In my source csv file, I am getting some nulls without quotes ...DataStage is writing it to a seq file as a "?" instead nulls..Why is this happening..? I am using a after-job subroutine shell script to scrub "?' to NULLS..but I would like to know the reason and try to resolve it in the beginning...
any clues..?

Thanks

I dont see any warnings or errors in my log too...
I dont even realize what kind of problems it may lead like lookup failures etc...any ideas, Pls let me know...

Do you have any NULL pad char?

Do you mean Default NULL padding...I leave it blank...
Here is how it is happening...
File1 is read and written to File2 -- File 1 has some nulls..written as nulls in File2.. so far fine
Again, in next job, File 2 is read and written to File 3--- here the Nulls in File 2 are wriiten as "?" in File 3...
in the same way, in some other scenarios:
nulls without quotes in csv files are read as nulls but written as "?" in o/p files..looks weired..
any idea from this...pls?

Sai,

And to my surprise...not all nulls in all fields do this...only few fields..
Do I need to configure NLS or something....?

Thankx

Have you given the quotes char? It may mis-interpret it to into '?'

Could you please check by passing srNVL(columnname,"") in which ever column you have Null values.

Hope this helps.

Thanks & Regards
Mahesh

Mahesh,

What does this function do..?
Does this enclose the column in quotes ?...

I think it is to replace NULL with quotes.

In short, something like
NULL_To_Default(YourField, Default_For_NULL)

and code being

If IsNULL(YourField) Then Default_For_NULL Else YourField

Thus the value "" will be substituted for Default_For_NULL

As you problem is that there a some Null values which come out as ? , in order to avoid this you could also use a srNVL(COLUMN_NAME," ") which actually puts in an empty value.

Hope this helps.

As Sainath pointed out this will put in a Empty Value.

Regards
Mahesh