Re: Are Google summer of code 2006 ideas still available?

[Hans Aberg]

On 29 Apr 2007, at 18:09, Joel E. Denny wrote:

> Thanks for the response.  Sorry, my question wasn't clear enough.  I
> thought you originally cited two different problems from merging
> lookaheads.  First, error handling.  Second, token completions in
> interactive parsers.  I'm interested in the latter

Those are two applications, but are caused by the same problem: the  
merging of the states in a way that the valid tokens are pushed to a  
state that may emerge only later, after some reductions have been made.

> though not specifically
> for interactive parsers.

The interactive parser was just an application that that came up on  
the list. It was also discussed in the Usenet newsgroup comp.compilers.

> I'm especially interested in attempts at
> solutions, ...

One way might be to implement a LR(1) that does not merge states in  
such a manner. Another might be to trace in each state the possible  
tokens that may appear after some reductions.

> ...but any literature that cites it at all would be interesting.

All I know is the picture I have in my mind, after discussions in the  
Bison lists. And I can give no guarantee that it is correct. :-) But  
if you consider some examples, such as those you gave in the Bison- 
Patches thread "LR(1) paser generator based on Pager's algorithm",  
you should be able to work it out, I think. :-)

  Hans Aberg