On Fri, 24 Sep 2004, Malcolm Wallace wrote: > Hugs doesn't support mutually-recursive modules at all. Ghc and nhc98 > support them only if you hand-write a .hi-boot file to bootstrap the > compilation. I would guess that better support from the mainstream > implementations is unlikely, because it would be a large amount of > effort for a situation which occurs only very rarely, and for which > there is a relatively easy workaround.
Namely? The only workaround I know of is putting all type declarations and functions that depend on each other into one module, import them in the modules they logically belong to and document that the objects should be imported from the redirecting modules rather than from the modules they are defined in. I know of two projects where different tasks had to be merged into one module because of the lack of mutually recursive imports, namely functionalMetaPost and Haskore. It's interesting how other languages solve this problem. In Modula-3 it is solved by explicit module interfaces and partial revelation. One can define type T in module A and type T in module B very abstractly as pointers to something and the complete data structures (which may contain references to either interface) are usually revealed in the implementations of the modules. I'm curious how Oberon solves it, because it doesn't need explicit interfaces but derives them from the implementation files. Probably one can say that Oberon extracts something like a .hiboot file from every module file automatically. Why can't GHC and Hugs go this way? _______________________________________________ Haskell-Cafe mailing list [EMAIL PROTECTED] http://www.haskell.org/mailman/listinfo/haskell-cafe
