I have a csv input file that contains single actual ASCII 'nul' characters (0x00 or '\0') in the fields where no data is present; BODS file-import reads these as 0x00 values. These field values can be seen in a SQL select when the field is formatted by the bintohex() operator (yielding a '00' string).
It appears that the 'NULL indicator' handling in the file-format should be the way to recognize these, but the Reference Guide isn't clear as to how to code a non-printable character. I tried entering \0 and NULL into the parameter, but these did not match the actual nul character. How should a nul character (or indeed any non-printable) be specified?
I'm also investigating how to remove these in a query transform, but haven't yet found the equivalent of a per-column CASE statement or other method to recognize the 0x00 value and replace it with NULL; a replace_substr() function does not allow a NULL as the replacement value?
BODS is at 4.2 SP5 (designer version tag 14.2.5.800)
Thanks in Advance,
Donn