Comments on: Type-safe Scala sequence comprehensions http://blog.tmorris.net/type-safe-scala-sequence-comprehensions/ The weblog of Tony Morris Tue, 06 Jan 2009 12:42:54 +0000 http://wordpress.org/?v=2.6 By: Tony Morris http://blog.tmorris.net/type-safe-scala-sequence-comprehensions/#comment-1698 Tony Morris Fri, 07 Dec 2007 07:18:59 +0000 http://blog.tmorris.net/type-safe-scala-sequence-comprehensions/#comment-1698 Hi Oscar, The example works for Scala 2.5.1, so I'll have to update it for 2.6.0. The correction is actually a much more concise articulation of the intention. In the meantime, you might try replacing all occurrences of c[a] in those cases that the compiler complains with just c. Hi Oscar,
The example works for Scala 2.5.1, so I’ll have to update it for 2.6.0. The correction is actually a much more concise articulation of the intention. In the meantime, you might try replacing all occurrences of c[a] in those cases that the compiler complains with just c.

]]> By: Oscar Picasso http://blog.tmorris.net/type-safe-scala-sequence-comprehensions/#comment-1697 Oscar Picasso Fri, 07 Dec 2007 03:00:19 +0000 http://blog.tmorris.net/type-safe-scala-sequence-comprehensions/#comment-1697 Sorry for the bad formatting, it seems that on submission formatting tags don't work. Maybe this is more readable: type Comprehension[c[_], a] = Functor[c[a], a] with Monad[c[a], a] with Filter[c[a], a] The compiler complains with:c[a] takes no type parameters, expected: one Sorry for the bad formatting, it seems that on submission formatting tags don’t work.

Maybe this is more readable:

type Comprehension[c[_], a] = Functor[c[a], a] with Monad[c[a], a] with Filter[c[a], a]

The compiler complains with:c[a] takes no type parameters, expected: one

]]> By: Oscar Picasso http://blog.tmorris.net/type-safe-scala-sequence-comprehensions/#comment-1696 Oscar Picasso Fri, 07 Dec 2007 02:56:34 +0000 http://blog.tmorris.net/type-safe-scala-sequence-comprehensions/#comment-1696 I tried your example but for the line: type Comprehension[c[_], a] = Functor[c[a], a] with Monad[c[a], a] with Filter[c[a], a]The compiler complains with:c[a] takes no type parameters, expected: one I tried your example but for the line:
type Comprehension[c[_], a] = Functor[c[a], a] with Monad[c[a], a] with Filter[c[a], a]The compiler complains with:c[a] takes no type parameters, expected: one

]]> By: Tony Morris http://blog.tmorris.net/type-safe-scala-sequence-comprehensions/#comment-1695 Tony Morris Thu, 13 Sep 2007 22:23:57 +0000 http://blog.tmorris.net/type-safe-scala-sequence-comprehensions/#comment-1695 G'day Henry, Yes I find it stresses Scala's type inferencing as well, so I have to manually help it along. A good example above is passing the type parameters to foreach explicitly; without them, the code would not compile. G’day Henry,
Yes I find it stresses Scala’s type inferencing as well, so I have to manually help it along. A good example above is passing the type parameters to foreach explicitly; without them, the code would not compile.

]]> By: Henry Ware http://blog.tmorris.net/type-safe-scala-sequence-comprehensions/#comment-1694 Henry Ware Thu, 13 Sep 2007 18:55:07 +0000 http://blog.tmorris.net/type-safe-scala-sequence-comprehensions/#comment-1694 Nice post about an important topic. I suspect the documentation is dated. Higher Order Kinds were just added this summer with 2.5. The crew have written a paper which discusses map, flatMap and filter; it's at Adriaan Moors' site: http://www.cs.kuleuven.be/~adriaan/files/higher.pdf Your approach is more Haskellish than theirs. in an interesting way. I haven't tried your approach (yet), but I find things like this stress Scala's type inference system... sometimes I've been unable to do what I want without losing type information. Nice post about an important topic.

I suspect the documentation is dated. Higher Order Kinds were just added this summer with 2.5. The crew have written a paper which discusses map, flatMap and filter; it’s at Adriaan Moors’ site: http://www.cs.kuleuven.be/~adriaan/files/higher.pdf

Your approach is more Haskellish than theirs. in an interesting way.

I haven’t tried your approach (yet), but I find things like this stress Scala’s type inference system… sometimes I’ve been unable to do what I want without losing type information.

]]>