Re: [Haskell-cafe] Parser left recursion

Hi
Another alternative is this Haskell library: https://github.com/paulcc/xsaiga
This is a combinator library which is suitable for mid-scale NLP work,
so handles left recursion and (high amounts of) ambiguity to produce a
packed result (which can be decoded to a list of results if required).
It uses a technique similar to Danielsson's for termination.
The technical details (incl papers) can be found on
http://cs.uwindsor.ca/~hafiz/xsaiga/pub.html, particularly in the IWPT
paper http://cs.uwindsor.ca/~hafiz/pub/iwpt-07.pdf
I'd be interested to see a GLL impl for Haskell, particularly for
comparison with the above. Did Roman C. publish some code for this a
while back?
Paul
On Wed, Feb 27, 2013 at 9:50 AM, Paul Callaghan
Another alternative is this Haskell library: https://github.com/paulcc/xsaiga
This is a combinator library which is suitable for mid-scale NLP work, so handles left recursion and (high amounts of) ambiguity to produce a packed result (which can be decoded to a list of results if required). It uses a technique similar to Danielsson's for termination.
The technical details (incl papers) can be found on http://cs.uwindsor.ca/~hafiz/xsaiga/pub.html, particularly in the IWPT paper http://cs.uwindsor.ca/~hafiz/pub/iwpt-07.pdf
I'd be interested to see a GLL impl for Haskell, particularly for comparison with the above. Did Roman C. publish some code for this a while back?
Paul
On Tue, Feb 26, 2013 at 7:43 PM, Dominique Devriese
wrote: 2013/2/26 Martin Drautzburg
: I wonder if I can enforce the nonNr property somehow, i.e. enforce the rule "will not consider the same nonterminal again without having consumed any input".
You might be interested in this paper:
Danielsson, Nils Anders. "Total parser combinators." ACM Sigplan Notices. Vol. 45. No. 9. ACM, 2010.
Regards, Dominique
_______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe

You can find our experiments here: https://github.com/feuerbach/hagll
But don't set your expectations high :-)
Roman
* Paul Callaghan
Hi
Another alternative is this Haskell library: https://github.com/paulcc/xsaiga
This is a combinator library which is suitable for mid-scale NLP work, so handles left recursion and (high amounts of) ambiguity to produce a packed result (which can be decoded to a list of results if required). It uses a technique similar to Danielsson's for termination.
The technical details (incl papers) can be found on http://cs.uwindsor.ca/~hafiz/xsaiga/pub.html, particularly in the IWPT paper http://cs.uwindsor.ca/~hafiz/pub/iwpt-07.pdf
I'd be interested to see a GLL impl for Haskell, particularly for comparison with the above. Did Roman C. publish some code for this a while back?
Paul
On Wed, Feb 27, 2013 at 9:50 AM, Paul Callaghan
wrote: Another alternative is this Haskell library: https://github.com/paulcc/xsaiga
This is a combinator library which is suitable for mid-scale NLP work, so handles left recursion and (high amounts of) ambiguity to produce a packed result (which can be decoded to a list of results if required). It uses a technique similar to Danielsson's for termination.
The technical details (incl papers) can be found on http://cs.uwindsor.ca/~hafiz/xsaiga/pub.html, particularly in the IWPT paper http://cs.uwindsor.ca/~hafiz/pub/iwpt-07.pdf
I'd be interested to see a GLL impl for Haskell, particularly for comparison with the above. Did Roman C. publish some code for this a while back?
Paul
On Tue, Feb 26, 2013 at 7:43 PM, Dominique Devriese
wrote: 2013/2/26 Martin Drautzburg
: I wonder if I can enforce the nonNr property somehow, i.e. enforce the rule "will not consider the same nonterminal again without having consumed any input".
You might be interested in this paper:
Danielsson, Nils Anders. "Total parser combinators." ACM Sigplan Notices. Vol. 45. No. 9. ACM, 2010.
Regards, Dominique
_______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
_______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
participants (2)
-
Paul Callaghan
-
Roman Cheplyaka