Because the adapter is written in Java, my guess is that the adapter makes proper adjustment based on its host machine.
I do not know what the underlying code is, however, to definitively say if the line separator is hard-coded as \r\n, \n, or if the adapter uses the PrintWriter class, for example, which has the method println().
println() uses the host platform’s own notion of line separator rather than the newline character. While that may work well if you are working on UNIX, once you move to other environments, you may find that you outputted (and formatted) data is goofed.
If your File/IO adapter is the target adapter only, this will not be an issue; your adapter must only map from the UDM into the host machine’s native format.