Re: Interactive parsing with Bison

[Hans Aberg]

On 8 Jul 2006, at 03:13, Richard Stallman wrote:

>                                               and we didn't think people
>     would be generating parsers with millions of states.
>
> Do people really do that?  I am surprised.  Why?

State blowup will also occur if Bison gets an LR(1) option. The need  
for this comes from surprising places: better error correction, and  
interactive parsers that want to display a correct set of lookahead  
tokens (LALR(1) can apply further actions, in the case of an error  
token arrives in the input). But also for experimentation with GLR.

In general, when computers become more powerful, programmers will try  
to make use of it, that is the experience.

  Hans Aberg