Possibility of GLR parser combinators? Spirit-2?
I was briefly reading GLR Parsing in Haskell and found the library based off the paper HaGLR a GLR parser combinator framework and it got me thinking. With the advent of Phoenix-2 (very nice work guys) and FC++ being worked on (to be re-reviewed for boost in the not to distant future?) there could be a possibility of direct & easy port of the haskell code to C++ using either of these libraries as it's foundation and built & used in a simillar fashion to Spirit ala expression templates (maybe an idea for Spirit 2 even). There might be an issue with Phoenix-2 (i don't know enough about the subject to know for sure) since HaGLR is based on the idea of parser combinators i'd assume it takes a simillar approach to the Parsec library using monads, from what i gather from Phoenix-2 docs there currently is no plains to support monads. So what do you guys think? :-)
On 11/15/2005 08:12 AM, Korcan Hussein wrote:
I was briefly reading GLR Parsing in Haskell and found the library based off the paper HaGLR a GLR parser combinator framework and it got me thinking. [snip] using monads, from what i gather from Phoenix-2 docs there currently is no plains to support monads. Could you provide some references to "GLR Parsing", "HaGLR a GLR parser combinator framework", and "monads"?
"Larry Evans" <cppljevans@cox-internet.com> wrote in message news:dlcsjm$991$1@sea.gmane.org...
On 11/15/2005 08:12 AM, Korcan Hussein wrote: Could you provide some references to "GLR Parsing", "HaGLR a GLR parser combinator framework", and "monads"?
GLR parsing in Haskell : http://wiki.di.uminho.pt/wiki/pub/PURe/PurePublications/PUReTR041101.pdf HaGLR: http://wiki.di.uminho.pt/wiki/bin/view/PURe/PUReSoftware Monads: http://en.wikipedia.org/wiki/Monads_in_functional_programming
"Pavel Vozenilek" <pavel_vozenilek@hotmail.com> wrote in message news:dlee2r$99g$1@sea.gmane.org...
"Korcan Hussein" wrote:
FC++ being worked on (to be re-reviewed for boost in the not to distant future?) FC++ was reviewed (cca year ago) with negative result. I am not sure how much of work has been done on it afterwards.
/Pavel
Do you or anyone else have any recollection on why it got a negative result to its review? Michael Goldshteyn
"Michael Goldshteyn" wrote:
FC++ was reviewed (cca year ago) with negative result. I am not sure how much of work has been done on it afterwards.
Do you or anyone else have any recollection on why it got a negative result to its review?
1. low quality and quantity of examples, no tests 2. some parts of the library were redundant with existing libraries 3. people not familiar with functional programming asked for introduction into it, the documentation assumed quite a lot of pre-existing knowledge 4. details in the code, naming conventions The discussion about the library was long and intensive and the decision not to accept it was made after long consideration. It is pity the development stopped afterwards. I am not aware of anyone picking it or creating similar tool. /Pavel
On Thu, 17 Nov 2005 09:13:51 -0600, Michael Goldshteyn wrote:
"Pavel Vozenilek" <pavel_vozenilek@hotmail.com> wrote in message news:dlee2r$99g$1@sea.gmane.org...
"Korcan Hussein" wrote:
FC++ being worked on (to be re-reviewed for boost in the not to distant > future?) FC++ was reviewed (cca year ago) with negative result. I am not sure how much of work has been done on it afterwards.
/Pavel
Do you or anyone else have any recollection on why it got a negative result to its review?
As well as the points Pavel makes, I seem to recall that a major strike against the FC++ library was that it was very slow compared to doing the equivalent thing with existing facilities. S> -- <<< Eagles may soar, but weasels don't get sucked into jet engines >>> 7:21pm up 43 days 2:55, 16 users, load average: 1.40, 1.07, 0.74 Registered Linux User #232457 | LFS ID 11703
participants (1)
-
Korcan Hussein -
Larry Evans -
Michael Goldshteyn -
Pavel Vozenilek -
Spencer Collyer