DSXchange

While connecting to SQL server while using the ODBC connector, it worked when I used the brackets "[" , "]" when the field names had spaces and '#'

Hi All, Even I'm getting the same problem, my source is MF Binary file, when trying to read the value for a column which is of character type in CFF stage with below properties Character set = EBCDIC. The values in MF is given as China! Japan! The same value when coming out of CFF is China] Japan] C...

Do not specify any limit define it as varchar()..!

What Ray is suggesting is that to read the whole row as a single column defined as a varchar datatype..!
So here it doesn't matter what delimiter one specifies.

You can read the values as Char or Varchar and then handle it in the next stage.

Hi,

Is it possible to do the range lookup in a join stage..?

Meaning, if the data volume if large then it's not feasible to use a lookup, so how to execute the concept of range lookup in a join stage...!

With the help of schema files, one can pass the default values

E.g.

record
(
ABC:decimal[15,0];
XYZ:string[max=10] {cycle={value='ABCD'}};
)

In the example above XYZ column is assigned a default value "ABCD"

Yeah...he wanted to add a new column and assign a default value...!

I'm not sure if you could add a new column and assign a default value in modify stage. It expects s mapping from input columns for any new output column. One can add a new column with a default value through a Column Generator stage (or) a Transformer stage.

Make the output column datatype Bigint. It should solve your problem.!

In version 8 of Datastage, this can be generated in transformer itself as a surrogate key in the stage properties that too in parallel mode.

Hi Raja,

As I already suggested the same, it did work out for me.....when I tried this experiment out.

1)Try doing a range lookup on the list of holidays like start date < holiday < End date ... aggregate this count based on the key combination of empid,Start Date and End Date by which one can get the number of holidays in that range. Once you come to know the number of records for a particular pair ...

Keep the target column as Varchar.May i know what is your final target is it a sequential file...?
If so one can directly populate target column into the sequential file as varchar.

I'm not sure if that is a correct way of doing it....as you are reading and updating the table in the same job.One can split the task into two jobs: 1st Job: Extract the table with the where condition i.e. Select * from Table_Name WHERE Table_Name.Column1='XX' AND Table_Name.Column2='CO'; Pass it th...