#apache #archlinux #asterisk #centos #debian #gentoo #haskell #kde #kubuntu #lisp #math #mysql #perl #python #ruby-lang #rubyonrails #suse #ubuntu #vim #wikipedia 0 1 2 3 4 5 6 7 8 9
Top Prev 3205 3206 3207 3208 3209 3210 3211 3212 3213 3214 3215 3216 3217 3218 3219 3220 3221 3222 3223 Next
#haskell
<fasta> I don't like the module system if that's the only way to do it.
<EvilTerran> could you import Data.List () ?
<nmessenger> I usually import the non-Prelude functions from Data.List
<fasta> EvilTerran: what would that accomplish?
<Botje> :t let but = ((not .) . flip elem) in but[1..5]
<lambdabot> forall a. (Num a, Enum a) => a -> Bool
<fasta> EvilTerran: AFAIK, that imports nothing from Data.List
<EvilTerran> hang on... are they imported or re-implemented in the prelude?
<fasta> EvilTerran: imported, probably
<EvilTerran> in which case, i thought that might hide them again. just a thought.
<EvilTerran> probably wrong.
<nmessenger> It'd be nice if there were an "if I get a name clash between these two modules, prefer module A" feature.
<EvilTerran> why do you want to exclude the prelude list functions, anyway?
<fasta> EvilTerran: the names clash with names I want to use.
<dcoutts> so far so good, no parse failures while parsing linux (using default configuration)
<mwc> fasta, isn't it possible to import the prelude qualified?
<fasta> mwc: how does that solve the problem?
<fasta> mwc: it doesn't, unless I am widely mistaken.
<fasta> wildly*
<mwc> fasta, well, the prelude is only implicitly imported iff you don't explicitly import it
<mwc> so you could do import Prelude hiding (length)
<mwc> if you want to use Length for your own purposes
<fasta> mwc: That's not my problem.
<fasta> mwc: never mind it
<mwc> Sounded like you wanted to use names from the prelude
<mwc> <fasta> EvilTerran: the names clash with names I want to use.
<fasta> mwc: I want to use the union of Data.List and Prelude without some names.
<fasta> mwc: that idea cannot be expressed without writing a code generator
<dons> dcoutts: good sign!
<mwc> dcoutts, c2hs stress testing?
<dcoutts> yep
<fasta> mwc: cannot meaning without typing out every identifier by hand
<mwc> well that's impressive then
<dcoutts> yay, it finished, 2 parse failures
<dcoutts> both the same issue I think
<dcoutts> yep, same issue
<dcoutts> redeclaration of typedef'ed names in function parameter lists
<dcoutts> typedef char foo; void bar (int foo) { foo = foo + 1; }
<dons> Igloo: http://urchin.earth.li/~ian/cabal/base64-string/
<dons> 404?
<allbery_b> ah, C. a language only K&R could love :)
<dcoutts> allbery_b, yep, though some non-K&R extensions are even worse :-)
<kolmodin> dcoutts: still fighting __attribute__, must be a real hard one
<dcoutts> kolmodin, I'm currently ignoring them all in the lexer
<kolmodin> heh, ok
<dcoutts> so the parser never sees them at all
<dcoutts> but that means we can't get the info out of them like packing/alignment attributes
<kolmodin> right. but you{-or anyone else-} can always come back to it
<kolmodin> does the other stuff work?
<kolmodin> I mean, after having cursed __attribute__ for a week, it might be time to move on before you throw out your computer throguh the window
Previous Page Next Page