I wish to produce fixed files with two columns of 5 bytes each (Char(5), Decimal(5)).
Example of what I wish to produce:
" a 1"
" b 2" etc.
I am reading in comma delimited input file as folows:
"a,1"
"b,2" etc.
I also wish to have my output be in EBCDIC, so I am using CFF for target. I have simply been unable to get the space padding to work.
1. I have attempted to use the 'Pad char' record option on the CFF to perform the padding for me using the default of space. I've also tried specifying \x20.
2. I have attempted to use the 'pad char' property on the input link of the sequential file stage (Edit Column Meta Data) and have set this to various values including space. This did not work.
3. I have changed $APT_STRING_PADCHAR to various values include 0x20, none of which had an impact on my output.
My sequential file tests always have the following output:
The 'a' value is padded with 4 NUL bytes, and the '1' value is padded by four zeros and two spaces.H:\Test\Source>od -x Sample_Padding_Out2.txt
0000000000 0061 0000 2000 3020 3030 3130 0A2E
0000000016
My CFF tests always have the following output:
Here all the padding is 100% x00 bytes.H:\Test\Source>od -x Sample_Padding_Out1.txt
0000000000 0061 0000 0000 1C00
0000000010
My understanding is that the Pad char and APT_STRING_PADCHAR should be padding with spaces instead of the NULL bytes, or at least for my CHAR(5) data. Is this a known bug? Does anybody have any suggestions?
Thanks,
Greg