DSXchange

Sudheer,

could you explain the problem a bit more. You said you have a comma delimiter and that is working. What is the format of the rows that are different? Do the rows have the value "..." for just one column?

One possibility is that you have incomplete rows (rows with delimiters less than expected. If this is the case, when you try to view data on the csv file, you'll get an error message. The problem can be solved by setting (double click on the sequential file / outputs / format) the Missing Column Action as "Pad with empty string".

Thanks for the reply,
I am getting some junk data in a csv file.
I am cleaning the file in unix and then calling it in datastage.

thanks,
Sudheer

Sudheer

What you can do is read your csv file record as one single string and then inside your transformer check the total commas you expect for good records. Have two links as ouput. One being the good ones which you wanted to use and the other for the rejected records.