On 05/01/2010 11:44, Nicolas Pouillard wrote:
> Reusing the work done in the Yi [1][2] editor for the Haskell syntax should
> be pretty straightforward. Very long and painful however due to the complexity
> of the grammar of a real language.
>
> [1]: http://www.haskell.org/haskellwiki/Yi
> [2]: http://www.cse.chalmers.se/~bernardy/FunctionalIncrementalParsing.pdf

Thanks for the links. The paper is a very interesting reading indeed. 
Its main focus is on incrementality (not reparsing the whole buffer at 
every keystroke). I'm not so sure how important it is in the context of 
the current discussion though: I guess that with an efficient parsing 
technology and modern computers, parsing even a big buffer at every 
keystroke should be fast enough. Trivial optimizations like storing the 
internal state of the parser at some point could also be used if needed.

I'm more concerned about the error recovery aspect; the paper suggests 
the use of annotated error recovery rules, but writing them for a 
grammar like OCaml's does not seem an easy task at all.


Alain