What encoding is used when calling an Oracle PL/SQL procedure
ColdFusion is Java based so I guess the string variables are encoded using UTF-16.
When going from UTF-16(2 byte encoding) to a 1 byte encoding(e.g. WE8ISO8859P1) as is the case in my example, the MSB is ignored and only the LSB is considered.
This would explain the observed behavior:I noticed that some of the bytes just seem to get swallowed when sent to the DB, e.g. the left single quotation mark (0x2018 in UTF-16) becomes just 0x18 in the DB.