Re: Comma-delimited help

[Vernon Hamberg <vhamberg@xxxxxxxxxxx>]

Andy

Go to the archive for this list - link near the bottom of this message - 
and search on "cpytoimpf csv" or some such. I actually searched on "csv 
blank" and got some good threads. This has come up several times.

One of the posts indicated that IBM was thinking of changing this 5 years 
ago. There are supposedly some rules - fixed-length character fields are 
not trimmed, variable-length ones are. But numerics are not trimmed and are 
left-justified.

But I tried this on a V5R1 machine - the VARCHAR was still padded - no 
trimming.

One solution is to use an SQL SELECT statement - this technique has been 
discussed in this list. Do an appropriate search, such as "csv sql", maybe. 
The idea is to convert date, numerics, etc., to character, trim the 
results, & concatenate the lot with commas and quotes as needed. You could 
even do this with a UNION SELECT and put the column names in the first row 
as a bunch of constants. And with a little more work, you could put column 
headings in there, as well, using another UNION SELECT. You could even use 
a UDTF to get this automagically - user-defined table function. I have to 
try that one myself! That would be a call to a system program that'd return 
the whole row, filled with column headings in one swell foop.

HTH
Vern

At 03:41 PM 2/25/2005, you wrote:
> Anyone able to offer a little download help?
>
> I'm trying to download (via ftp) from the 400 to a comma-delimited file
> (to a T&A system), using the following command:
>
> CPYTOIMPF FROMFILE(HS#LIBR/PPPTADN) TOFILE(HS#LIBR/PPPTADNCSV)
>
> The results are like so:
>
> "8732","A","JOHNSON             ","LINDA               ","NODIFF
> ","9066","    "
> "6289","A","SMITH               ","MARIANNE            ","40/40
> ","9596","    "
>
> but I was expecting something like this:
>
> "8732","A","JOHNSON","LINDA","NODIFF","9066","    "
> "6289","A","SMITH","MARIANNE","40/40","9596",""
>
> .... the extra blanks within each set of quotes is killing the import.
> Any clues as to what I should be doing differently would be greatly
> appreciated!
>
>
> Andy