Delimited continuation

From Wikipedia, the free encyclopedia

In programming languages, a delimited continuation, composable continuation or partial continuation, is a "slice" of a continuation frame that has been reified into a function. Unlike regular continuations, delimited continuations return a value, and thus may be reused and composed. Control delimiters, the basis of delimited continuations, were introduced by Felleisen[1] though early allusions to composable and delimited continuations can be found in Carolyn Talcott's Stanford 1984 dissertation, Felleisen and Friedman's PARL 1987 paper,[2] and Felleisen's 1987 dissertation.[3]

History

Delimited continuations were first introduced by Felleisen in 1988[1] with an operator called {\mathcal  {F}}, first introduced in a tech report in 87,[2] along with a prompt construct \#. The operator was designed to be a generalization of control operators that had been described in the literature such as call/cc from Scheme, ISWIM's J operator, John C. Reynolds' escape operator, and others. Subsequently, many competing delimited control operators were invented by the programming languages research community such as prompt & control,[4] shift & reset,[5] cupto,[6] fcontrol, and others.

Examples

Various operators for delimited continuations have been proposed in the research literature.[7]

One proposal[5] offers two control operators: shift and reset. The reset operator sets the limit for the continuation while the shift operator captures or reifies the current continuation up to the innermost enclosing reset. For example, consider the following snippet in Scheme:

 (* 2 (reset (+ 1 (shift k (k 5)))))

The reset delimits the continuation that shift captures (named by k in this example). When this snippet is executed, the use of shift will bind k to the continuation (+ 1 []) where [] represents the part of the computation that is to be filled with a value. This directly corresponds to the code that surrounds the shift up to the reset. Since the body of shift (i.e., (k 5)) immediately invokes the continuation, this code is equivalent to the following:

 (* 2 (+ 1 5))

In general, these operators can encode more interesting behavior by, for example, returning the captured continuation as a value or invoking it multiple times. The shift operator passes the captured continuation to the code in its body, which can either invoke the continuation, produce it as a result, or ignore it entirely. Whatever result that shift produces is provided to the innermost reset, discarding the continuation in between the reset and shift. However, if the continuation is invoked, then it effectively re-installs the continuation after returning to the reset. When the entire computation within reset is completed, the result is returned by the delimited continuation.[8] For example, in this Scheme code:

(reset (* 2 (shift k CODE)))

whenever CODE invokes (k N), (* 2 N) is evaluated and returned.

This is equivalent to the following:

 (let ((k (lambda (x) (* 2 x)))) CODE)

Furthermore, once the entire computation within shift is completed, the continuation is discarded, and execution restarts outside reset. Therefore,

(reset (* 2 (shift k (k (k 4)))))

invokes (k 4) first (which returns 8), and then (k 8) (which returns 16). At this point, the shift expression has terminated, and the rest of the reset expression is discarded. Therefore, the final result is 16.

Everything that happens outside the reset expression is hidden, i.e. not influenced by the control transfer. For example, this returns 17:

(+ 1 (reset (* 2 (shift k (k (k 4))))))

Delimited continuations were first described independently by Felleisen et al.[9] and Johnson.[10] They have since been used in a large number of domains, particularly in defining new control operators; see Queinnec[11] for a survey.

Let's take a look at a more complicated example. Let null be the empty list:

(reset
  (begin
    (shift k (cons 1 (k (void)))) ;; (1)
    null))

Since the context captured by shift consists of (begin [*] null) (where [*] is the hole for parameter injection), the first call of k inside shift evaluates to null, and the body of shift determines the value of the expression, we get (1) as a result.

Making this example more complicated, add a line:

(reset
  (begin
    (shift k (cons 1 (k (void))))
    (shift k (cons 2 (k (void))))
    null))

If we comment out the first shift, we already know the result, it is (2); so we can as well rewrite the expression like this:

(reset
  (begin
    (shift k (cons 1 (k (void))))
    (list 2)))

This is pretty familiar, and can be rewritten as (cons 1 (list 2)), that is, (list 1 2).

We can define yield using this trick:

(define (yield x) (shift k (cons x (k (void)))))

and use it in building lists:

(reset (begin
         (yield 1)
         (yield 2)
         (yield 3)
         null))    ;; (list 1 2 3)

If we replace cons with stream-cons, we can build lazy streams:

 (define (stream-yield x) (shift k (stream-cons x (k (void)))))
 (define lazy-example
   (reset (begin
           (stream-yield 1)
           (stream-yield 2)
           (stream-yield 3)
           stream-null)))

We can generalize this and convert lists to stream, in one fell swoop:

(define (list->stream xs)
  (reset (begin
           (for-each stream-yield xs)
           stream-null)))

In a more complicated example below the continuation can be safely wrapped into a body of a lambda, and be used as such:

(define (for-each->stream-maker for-each) 
  (stream-lambda (collection) 
    (reset (begin 
             (for-each (lambda (element) 
                         (shift k 
                           (stream-cons element (k 'ignored)))) 
                       collection) 
             stream-null))))

The part between reset and shift includes control functions like lambda and for-each; this is impossible to rephrase using lambdas.

Delimited continuations are also useful in linguistics: see Continuations in linguistics for details.

References

  1. 1.0 1.1 Matthias Felleisen (1980). "The theory and practice of first-class prompts". Principles of Programming Languages: 180–190. doi:10.1145/73560.73576. ISBN 0-89791-252-7. 
  2. 2.0 2.1 Felleisen; Friedman; Duba; Merrill (1987). Beyond Continuations (Technical report). Indiana University. 87-216. 
  3. Matthias Felleisen (1987). The Calculi of Lambda-v-CS Conversion: A Syntactic Theory of Control and State in Imperative Higher-Order Programming Languages (Thesis). 
  4. Sitaram, Dorai; Felleisen, Matthias (1990). "Control Delimiters and their Hierarchies". Lisp and Symbolic Computation. 
  5. 5.0 5.1 Olivier Danvy and Andre Filinski (1990). "Abstracting Control". LISP and Functional Programming: 151–160. doi:10.1145/91556.91622. ISBN 0-89791-368-X. 
  6. Gunter; Rémy; Riecke (1995). "A generalization of exceptions and control in ML-like languages". Functional programming languages and computer architecture. 
  7. See for instance the operators offered by the racket/control Racket library ; the following examples can run in Racket using (require racket/control)
  8. Gasbichler, Martin; Sperber, Michael (2002). Final Shift for Call/cc: Direct Implementation of Shift and Reset. 
  9. Felleisen, Matthias; Friedman, Daniel P.; Duba, Bruce; Marrill, John (February 1987). Beyond continuations. Technical Report 216. Computer Science Department, Indiana University. 
  10. Johnson, Gregory F. (June 1987). "GL: a denotational testbed with continuations and partial continuations". Proc. SIGPLAN '87 Symposium on Interpreters and Interpretive Techniques. pp. 218–225. 
  11. Queinnec, Christian (April 1994). A library of high-level control operators. École Polytechnique and INRIA-Rocquencourt. 

External links

This article is issued from Wikipedia. The text is available under the Creative Commons Attribution/Share Alike; additional terms may apply for the media files.