How do LL(*) parsers work?

Question

I cannot find any complete description about LL(*) parser, such as ANTLR, on Internet.

I'm wondering what is the difference between an LL(k) parser and an LL(*) one and why they can't support left-recusrive grammars despite their flexibility.

Answer 1

Here is an article (by Terence Parr, the author of antlr) about LL(*) grammar analysis: article with a nice example of what is LL(*) but not LL(k), for any k.

Another good reference (and much more complete) is the "Definitive ANTLR Reference", again by Terence Parr, and the original journal article describing how antlr works [pdf].