I am following the general strategy described in Sam Tobin-Hochstadt's work. He is the person behind typed-racket. His Phd dissertation[1] has an overview of the area of gradual/soft/refinement typing in dynamic languages. It's a good place to start, with a pretty gentle introduction to the motivations and issues involved (though I'd imagine you'd need a rough understanding of sequent calculus notation to follow his description of the semantics of his type system).
[1] http://www.ccs.neu.edu/home/samth/ On Sep 27, 12:43 am, Paul Koerbitz <paul.koerb...@gmail.com> wrote: > Hi Nathan! > > I am very intrigued by your approach. I would love to contribute, my problem > is that I don't know the first thing about type inference systems (as in how > they work on the inside). Do you have a good reference here? I'll take a > look at what you've done, maybe bother you with some questions .... Thanks > for the work! > > cheers > Paul > > > > > > > > On Fri, Sep 23, 2011 at 04:47, Nathan Sorenson <n...@sfu.ca> wrote: > > Yes I am very hopeful progress is made on that front! I've been having > > great success with 'Constraint Handling Rules' (CHR), which serves as > > my basis for solving type constraints. > >https://github.com/nsorenson/Clojure-CHR > > contains my very preliminary crack at implementing this system. > > > The trick with dynamic languages such as Clojure is that the functions > > are extremely lax in what they accept--without a fairly advanced > > inference system, you will simply infer every variable to be type > > Object. I've been using CHR disjunctive branches to offer what's > > called "occurrence typing"--a system where you may assign different > > types to variables depending on the structure of the code. For > > instance, if you have a (map? x) as a predicate on an if clause, you > > know in the 'then' branch that x can be typed to Map. > > > Nothing to show yet, and I'm not sure how well it will work on large > > programs, but my first few little experiments are looking promising. > > > On Sep 22, 3:28 am, Javier Neira Sanchez <atreyu....@gmail.com> wrote: > > > i am surprised nobody mentioned gradual/optional typing< > >http://lambda-the-ultimate.org/node/1707>, > > > typed racket <http://docs.racket-lang.org/ts-guide/> and the possible > > type > > > checker to be built by *some clever hacker* on core.logic< > >https://github.com/clojure/core.logic>some day. > > > -- > > 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
