gilbert@aimmi.UUCP (Gilbert Cockton) (10/13/86)
My research on dialogue specification has lead me towards the idea of generating/encapsulating `big' structures from/within smaller ones. Two level grammars seem to do the same (and much more, i.e. infinite sets of productions can be specified in the meta-syntax) so I'd like to bone-up (a little!) on 2LGs. There are passing historical references in compiler books, and there's a chapter by Koster in Bauer and Eickel's LNCS 21. However, I've never been able to find a succinct summary of the 2LG arsenal. I don't want to read any of van Wijngaarden's original work either. I just need to understand the bare bones of 2LGs, I've no intention of wrestling with any unsolved problems or open/esoteric research issues kept alive by Formal Language and Automata types. My interest is in the expression of consistency in context free grammars and networks, something to which BNF and syntax diagrams aren't well suited (see work by Phyllis Reisner, Tom Green and Steve Payne). The use of of 2 levels to express context sensitive constraints and anomalies doesn't interest me, as such things should be avoided in Human-Computer interaction languages. So, I'd be grateful for any suggestions on quick and concise references. If there's enough material, I'll send a summary to the net. Thanks in advance -- Gilbert Cockton, Scottish HCI Centre, Ben Line Building, Edinburgh, EH1 1TN JANET: gilbert@uk.ac.hw.aimmi ARPA: gilbert%aimmi.hw.ac.uk@cs.ucl.ac.uk UUCP: ..!{backbone}!aimmi.hw.ac.uk!gilbert