DSXchange

Hi All,

Iam moving a CSV file from Windows to UNIX, then reading the CSV file in one of my job and when trying to view data in dataset i'm able to see a box character appended to the value of a column which has a datatype of VARCHAR(8).

The box character is not appended for the values if the string is of length 8.

Eg: For the value "abcdefgh" there is no box character appended, whereas if the value is "abcdef" there is a box character appended.

Is there a way to handle these box characters in Datastage.

Is that varchar column last one in your file ?

Check the record delimiter property in your input.

Yes..... this column is the last one in my file.
Record level final delimiter is end.

rohithmuthyala wrote:Yes..... this column is the last one in my file.
Record level final delimiter is end.

when you transfer file from Windows to Unix or in other way , "Ctrl M" Character will be appended to end of each line. That has to be removed fom the file, otherwise Datastage will not able to read the file.

thanks,
Sri

How ur ftp file from windows to unix? ru using any tools to do this?

ur getting box like character because of ur ftp tool by default selecting
Binary mode

While doing ftp u should select transfer mode is text

I'm using Ipswitch S/W to transfer the file,by changing the mode of transfer the box character won't appear. But if you observe I've asked if there is any way these characters can be handled in Datastage?

Convert(CHAR(13),"",YourField)

Set your record delimiter as "DOS type".

run below command before using the CSv file:

Dos2unix filename filename

3rd solution (heh) - Transfer the file using FTP and select text mode instead of binary. Most ftp solutions automatically flip the end-of-line character if you select text mode.

Any of the above three solutions (sequential file - DOS delimiter, dos2unix or ftp text mode) will work.

Thank you Craig ... It worked...!! :D