I spent some time digging in to the hugs98 implementation to see how it
deals with mixfix. I get it (nice trick on negate, Mark), and it's a good
start, but it's clearly an implementation that is designed to deal with
infix operators only - the mechanism transforms an input tree of arity-2
operators (ignoring negate, which is a special case) into an output tree of
arity-2 operators.

This is a great start, and very useful, but is there an implementation of a
richer mixfix strategy somewhere that I should look at?


shap
_______________________________________________
bitc-dev mailing list
[email protected]
http://www.coyotos.org/mailman/listinfo/bitc-dev

Reply via email to