Archives of the TeradataForum

Message Posted: Sat, 22 Jan 2005 @ 17:16:09 GMT

	<Prev	Next>		<<First	<Prev	Next>	Last>>

....Accused of not being able to load different record types when using VARTEXT i.e csv files.

What I've found :-

TPump and Multiload Can:-

Load records of different record type when the records are of fixed length.

Load VARTEXT files when there is only one record type.

TPump and Multiload Can't

Load VARTEXT files with different record types.

I've been running these on a 390 Mainframe (Channel attached). Perhaps there is a difference if doing this on network attached systems?

Below is a sample file I was trying to load, the first character gives the record type: 0 header, 1 detail(many records), 9 header and | as the delimeter:-

     0|header!2005-01-01
     1|detail_1|detail_2|deatil_3|detail_4
     1|detail_5|detail_6|detail_7|detail_8
     9|trailer|000000004

Can anyone confirm these allegations? If it can't do this them it seems pretty poor.

	<Prev	Next>		<<First	<Prev	Next>	Last>>