Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]


Groups > comp.compilers > #879 > unrolled thread

Ambiguous non-terminal in GLR

Started byMaxim Reznik <reznikmm@front.ru>
First post2013-04-01 11:16 +0300
Last post2013-04-01 11:16 +0300
Articles 1 — 1 participant

Back to article view | Back to comp.compilers


Contents

  Ambiguous non-terminal in GLR Maxim Reznik <reznikmm@front.ru> - 2013-04-01 11:16 +0300

#879 — Ambiguous non-terminal in GLR

FromMaxim Reznik <reznikmm@front.ru>
Date2013-04-01 11:16 +0300
SubjectAmbiguous non-terminal in GLR
Message-ID<13-04-002@comp.compilers>
Hi, all

I'm playing with GLR and have a question.

When GLR parser reduces some text to the same non-terminal in two or
more ways it merges parse subtrees. Rekers uses 'symbol nodes' for this.

I this not each non-terminal could cause a merge. Knowing in advance
what non-terminals never merge will greatly simplify parse tree
construction.

For instance in Elkhound Technical Report the author implemented C++
grammar for GLR parser. He describes it:

    The grammar currently has 37 shift/reduce conflicts, 47
reduce/reduce conflicts and 8 ambiguous nonterminal.

How can I separate ambiguous and unambiguous nonterminal for a given
CFG? Where can I read about this?


Thank you
--
Maxim Reznik

[toc] | [standalone]


Back to top | Article view | comp.compilers


csiph-web