The really odd thing is that there's a great method of doing this that many -- but by no means all -- tools support: dedicated data separator characters. 01 SO

The really odd thing is that there's a great method of doing this that many -- but by no means all -- tools support: dedicated data separator characters. 01 SOH(Start of Header) 02 STX(Start of Text) 03 ETX(End of Text) 04 EOT(End of Trans.) 05 ENQ(Enquiry) 06 ACK(Acknowledgement) 21 NAK(Negative acknowl.) 22 SYN(Synchronous idle) 23 ETB(End of trans. block) 28 FS(File separator) 29 GS(Group separator) 30 RS(Record separator) 31 US(Unit separator) You've got everything you need there to build really useful serialized data pipelines, and it would only be a day's work to implement them as a library in pretty much any language.