We don't need to auto-detect the formats of caption values; they are
already formatted by the time they arrive in to ScDPOutput.