Hi. On Wednesday, January 16, 2013 9:56:39 PM UTC+2, David Nolen wrote: > > Ok. But at that point can't you just process the type environment with > Clojure code?
But I get them out with replacements, is there any way to avoid this? > Or would you like those fresh vars to reify differently, as in you want > something other than _N? > The types with fresh vars will be function signatures and can be used as such. It will be even easier to use them with lvars than with _* symbols. At least in this case. Br, Timo > > On Wednesday, January 16, 2013, Timo Westkämper wrote: > >> Hi. >> >> On Wednesday, January 16, 2013 9:27:37 PM UTC+2, David Nolen wrote: >>> >>> Not sure I follow. What is there to further optimize? Is there something >>> that you're trying to do with q that you can clarify further? >>> >> >> The final output is map which includes form / type mappings. And the >> types can be lvars. I believe in the final output lvars are replaced by _0, >> _1 etc symbols. So there is some final normalization in the end. >> >> Is this correct, or did I misinterpret something? >> >> Forgive my vague terminology. >> >> Timo >> >> >>> >>> Thanks, >>> David >>> >>> >>> On Wed, Jan 16, 2013 at 2:19 PM, Timo Westkämper >>> <timo.we...@mysema.com>wrote: >>> >>>> Hi >>>> >>>> I have been able to improve the performance of the core.logic based >>>> type inference in symbol quite a lot based on David's suggestions >>>> https://github.com/timowest/**symbol/blob/master/src/symbol/**types.clj<https://github.com/timowest/symbol/blob/master/src/symbol/types.clj> >>>> >>>> The biggest change was to use maps for the type environment >>>> >>>> I wonder if it is possible to optimize / customize how the final map is >>>> generated >>>> >>>> defn new-env >>>> [env form] >>>> (first (run 1 [q] >>>> (fresh [type] (typedo env form type q))))) >>>> >>>> q will be a map in this case. >>>> >>>> I believe it would be ok to just give the map out as it is, and >>>> transform lvars that are included on demand. >>>> >>>> In other words, I have optimized goals for map population and querying >>>> and I also want to opimize the final extraction of the type environment. >>>> >>>> Is this possible? >>>> >>>> Br, >>>> Timo Westkämper >>>> >>>> -- >>>> You received this message because you are subscribed to the Google >>>> Groups "Clojure" group. >>>> To post to this group, send email to clo...@googlegroups.com >>>> Note that posts from new members are moderated - please be patient with >>>> your first post. >>>> To unsubscribe from this group, send email to >>>> clojure+u...@**googlegroups.com >>>> For more options, visit this group at >>>> http://groups.google.com/**group/clojure?hl=en<http://groups.google.com/group/clojure?hl=en> >>> >>> >>> -- >> You received this message because you are subscribed to the Google >> Groups "Clojure" group. >> To post to this group, send email to clojure@googlegroups.com >> Note that posts from new members are moderated - please be patient with >> your first post. >> To unsubscribe from this group, send email to >> clojure+unsubscr...@googlegroups.com >> For more options, visit this group at >> http://groups.google.com/group/clojure?hl=en > > -- You received this message because you are subscribed to the Google Groups "Clojure" group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
