Programming Language Theory: Difference between revisions

Line 12: Line 12:
* <tt>Y ≡ λf. (λx. f (x x)) (λx. f (x x))</tt> (untyped λ-calculus)
* <tt>Y ≡ λf. (λx. f (x x)) (λx. f (x x))</tt> (untyped λ-calculus)
* <tt>Y ≡ S (K (S I I)) (S (S (K S) K) (K (S I I)))</tt> (SKI calculus)
* <tt>Y ≡ S (K (S I I)) (S (S (K S) K) (K (S I I)))</tt> (SKI calculus)
Divergence-free evaluation of the Y-combinator requires call-by-name semantics. Call-by-value semantics can make use of the Θ<sub></sub>v (Turing) or Z-combinators:
Divergence-free evaluation of the Y-combinator requires call-by-name semantics. Call-by-value semantics can make use of the Θ<sub>v</sub> (Turing) or Z-combinators:
* <tt>Θ<sub></sub>v ≡ (λx. λy. (y (λz. x x y z))) (λx. λy. (y (λz. x x y z)))</tt>
* <tt>Θ<sub>v</sub> ≡ (λx. λy. (y (λz. x x y z))) (λx. λy. (y (λz. x x y z)))</tt>
* <tt>Z ≡ λf. (λx. f (λy. x x y)) (λx. f (λy. x x y))</tt>
* <tt>Z ≡ λf. (λx. f (λy. x x y)) (λx. f (λy. x x y))</tt>
The Z-combinator is derived via η-expansion on Y.
The Z-combinator is derived via η-expansion on Y.