[antlr-interest] ANTLR 3 parser generating algorithm questions

[Nikolay Ognyanov]

I read through Terence Parr's book, ANTLR site and archives of this list 
but was not
able to find  answers to some questions related to ANTLR 3 parser generating
algorithm. Can somebody please answer the following?

    * Is there an accessible formal description (other than source code)
      of ANTLR 3 parser
      generating algorithm?
    * Is ANTLR 3 capable (apart from normal space and time limitations)
      of generating
      a parser for every LL(k) grammar? Or for every strong LL(k) grammar?
          o If not then what exactly are the limitations?
          o If yes then how exactly is lookahead beyond next symbol handled?
                + By computing in advance complete lookahead sets
                  wherever needed?
                + By generating DFAs which effectively compute at run
                  time (parts of)
                  lookahead sets needed?
                + Somehow else? How?
    * What exactly are the limitations of ANTLR 3 LL(*) parser
      generating strategy?
      Is it capable of handling wcwey regular lookahead language or are
      there some
      stronger limitations?
    * What exactly are the limitations for use of recursion in lexer
      grammars?


Regards
Nikolay

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.antlr.org/pipermail/antlr-interest/attachments/20090916/979d8e21/attachment.html