"Too many columns in record" - How to resolve this

[Shree0410]

hi guys:
I have the fiollwoing error when I try to read a file in datastage.

read_delimited() - row 1, too many columns in record

I checked the file layout on the Unix box and it exactly matches the column layout of the Datastage job.

Please advise if this could be for a diff reason.

[DSguru2B]

Whats the delimiter you are specifiying for the file. The data must be containing the delimiter thats why you are getting this kind of warning. Whats the quote character that you have specified?

[Shree0410]

Delimiter is pipe and Quote character is 000

[Krazykoolrohit]

change the delimiter, its quite possible that the pipe is coming in your data. change it to any strange ascii value.

[DSguru2B]

Dont change the delimiter if the file is delimited as a pipe, That reaquires a lot of work. Change the quote character to ". Dont put 000. See what happens.

[ameyvaidya]

Could you provide some sample lines from the file in question?

I'd include row 1 too (As it is the one mentioned in the error message).

[Shree0410]

there is no quote character between columns, so I used 000

[kris007]

Did you try using " character as quote character? Give it a shot once.

[Krazykoolrohit]

cant input data contain pipes? how can you rule out that possibility?

[suresh.narasimha]

Check for Line Termination option in ur sequential stage whether it is Unix style or DOS style.

Regards,
Suresh N

[Shree0410]

Thanks Narasimha, somehow it was defaulted to DOS style, I changed it to Unix style and it worked.

[oracledba]

Also there is another reason this error could pop up. This happened to me.
When you have a field name in the source file that you forgot to include in the output columns list then the exact same error pops up.