portkey 2018-06-13

if not there, then probably elsewhere in the universe

being optimistic

it’s bizarre: when the closure is erialized, instaparse has not been invoked yet.

so we are passing the grammar as a string

maybe many holes

if strings were corrupted we would have noticed

don't know if they lead to same galaxy though

so how come this difference is?

is an instaparse var mishandled?

with portkey Yoda messing is

parsing the grammar (I repeat: parsing the grammar, not parsing with the parser generated from the grammar)

should yield ([:rule [:nt "S"] [:alt [:cat [:star [:nt "AB"]]]]] [:rule [:nt "AB"] [:alt [:cat [:nt "A"] [:nt "B"]]]] [:rule [:nt "A"] [:alt [:cat [:plus [:string "'a'"]]]]] [:rule [:nt "B"] [:alt [:cat [:plus [:string "'b'"]]]]])

but yields ([:rule [:alt [:cat [:plus [:string "'b'"]]]] [:nt "B"]] [:rule [:alt [:cat [:plus [:string "'a'"]]]] [:nt "A"]] [:rule [:alt [:cat [:nt "B"] [:nt "A"]]] [:nt "AB"]] [:rule [:alt [:cat [:star [:nt "AB"]]]] [:nt "S"]])

portkey seems to establish a new order

Actually I think it’s carbonite.