What is the trick of SLL(k) grammars subset LL(k) grammars for k>1

This is the sequel of me asking trivial questions and getting very
smart answers I do not understand which makes me yell at people ;)


Today's problem is how come every SLL(1) grammar is an LL(1) grammar
*and* the other way round, while this is not true for k>1. For k>1
SLL(k) grammars are a subset of LL(k). An example of an LL(2) grammar
which is not SLL(2) taken from Grune/Jacobs (this time lower case
letters are terminals):


S -> aAaa | bAba
A -> b | lambda


I have a problem with the proof given in Grune/Jacobs. I have no
objections and understand every step taken, but still have no
*intuition* why all this is true.


Thanks for any hints. Any - no - this is no homework, I am just an
all-purpose programmer who really is interested :)


Oliver

Return to the comp.compilers page.
Search the comp.compilers archives again.