Support modified UTF-8 in command line adaptor

Our modified UTF-8 supports encoding \0 and \n as two bytes, unlike
'normal' modified UTF-8 that only encodes \0.
2 files changed