Re: number LR(1) vs LALR(1) Itemsets for a grammar

Mike Spencer <mike.spencer@earthlink.net>
18 Nov 2003 18:53:23 -0500

          From comp.compilers

Related articles
number LR(1) vs LALR(1) Itemsets for a grammar alpanadubey1@yahoo.co.in (2003-11-11)
Re: number LR(1) vs LALR(1) Itemsets for a grammar mike.spencer@earthlink.net (Mike Spencer) (2003-11-18)
| List of all articles for this month |

From: Mike Spencer <mike.spencer@earthlink.net>
Newsgroups: comp.compilers
Date: 18 Nov 2003 18:53:23 -0500
Organization: Compilers Central
References: 03-11-051
Keywords: LALR, LR(1)
Posted-Date: 18 Nov 2003 18:53:23 EST

Alpana wrote:
> I have implemented a LR(1) parse
> table generator. Which, when executed for a C grammar gave me around
> 2350 LR(1) itemsets. There is no way to check this number except doing
> manually which is again not feasible. Does anybody has LR(1) parser
> generator??


You can try Basil: http://www.lazycplusplus.com/basil


It is an LR(1) parser generator, though by default it will generate
extra states only as needed. It will generate a default action in a
state for the most common reduction, and it will merge two states if
the set of actions (shifts and reductions) are the same for the two
states and for all transitions. Option -nda will prevent default
actions and so you should end up with all the LR(1) states.


I'll try it on your C grammar if you like; I too would be interested to
see if our numbers match.


Mike



Post a followup to this message

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