File size: 5,856 Bytes
3dcad1f |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 |
;;; Continuation-passing style (CPS) intermediate language (IL)
;; Copyright (C) 2013, 2014, 2015 Free Software Foundation, Inc.
;;;; This library is free software; you can redistribute it and/or
;;;; modify it under the terms of the GNU Lesser General Public
;;;; License as published by the Free Software Foundation; either
;;;; version 3 of the License, or (at your option) any later version.
;;;;
;;;; This library is distributed in the hope that it will be useful,
;;;; but WITHOUT ANY WARRANTY; without even the implied warranty of
;;;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
;;;; Lesser General Public License for more details.
;;;;
;;;; You should have received a copy of the GNU Lesser General Public
;;;; License along with this library; if not, write to the Free Software
;;;; Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
;;; Commentary:
;;;
;;; Guile's CPS language is a label->cont mapping, which seems simple
;;; enough. However it's often cumbersome to thread around the output
;;; CPS program when doing non-trivial transformations, or when building
;;; a CPS program from scratch. For example, when visiting an
;;; expression during CPS conversion, we usually already know the label
;;; and the $kargs wrapper for the cont, and just need to know the body
;;; of that cont. However when building the body of that possibly
;;; nested Tree-IL expression we will also need to add conts to the
;;; result, so really it's a process that takes an incoming program,
;;; adds conts to that program, and returns the result program and the
;;; result term.
;;;
;;; It's a bit treacherous to do in a functional style as once you start
;;; adding to a program, you shouldn't add to previous versions of that
;;; program. Getting that right in the context of this program seed
;;; that is threaded through the conversion requires the use of a
;;; pattern, with-cps.
;;;
;;; with-cps goes like this:
;;;
;;; (with-cps cps clause ... tail-clause)
;;;
;;; Valid clause kinds are:
;;;
;;; (letk LABEL CONT)
;;; (setk LABEL CONT)
;;; (letv VAR ...)
;;; (let$ X (PROC ARG ...))
;;;
;;; letk and letv create fresh CPS labels and variable names,
;;; respectively. Labels and vars bound by letk and letv are in scope
;;; from their point of definition onward. letv just creates fresh
;;; variable names for use in other parts of with-cps, while letk binds
;;; fresh labels to values and adds them to the resulting program. The
;;; right-hand-side of letk, CONT, is passed to build-cont, so it should
;;; be a valid production of that language. setk is like letk but it
;;; doesn't create a fresh label name.
;;;
;;; let$ delegates processing to a sub-computation. The form (PROC ARG
;;; ...) is syntactically altered to be (PROC CPS ARG ...), where CPS is
;;; the value of the program being built, at that point in the
;;; left-to-right with-cps execution. That form is is expected to
;;; evaluate to two values: the new CPS term, and the value to bind to
;;; X. X is in scope for the following with-cps clauses. The name was
;;; chosen because the $ is reminiscent of the $ in CPS data types.
;;;
;;; The result of the with-cps form is determined by the tail clause,
;;; which may be of these kinds:
;;;
;;; ($ (PROC ARG ...))
;;; (setk LABEL CONT)
;;; EXP
;;;
;;; $ is like let$, but in tail position. If the tail clause is setk,
;;; then only one value is returned, the resulting CPS program.
;;; Otherwise EXP is any kind of expression, which should not add to the
;;; resulting program. Ending the with-cps with EXP is equivalant to
;;; returning (values CPS EXP).
;;;
;;; It's a bit of a monad, innit? Don't tell anyone though!
;;;
;;; Sometimes you need to just bind some constants to CPS values.
;;; with-cps-constants is there for you. For example:
;;;
;;; (with-cps-constants cps ((foo 34))
;;; (build-term ($values (foo))))
;;;
;;; The body of with-cps-constants is a with-cps clause, or a sequence
;;; of such clauses. But usually you will want with-cps-constants
;;; inside a with-cps, so it usually looks like this:
;;;
;;; (with-cps cps
;;; ...
;;; ($ (with-cps-constants ((foo 34))
;;; (build-term ($values (foo))))))
;;;
;;; which is to say that the $ or the let$ adds the CPS argument for us.
;;;
;;; Code:
(define-module (language cps with-cps)
#:use-module (language cps)
#:use-module (language cps utils)
#:use-module (language cps intmap)
#:export (with-cps with-cps-constants))
(define-syntax with-cps
(syntax-rules (letk setk letv let$ $)
((_ (exp ...) clause ...)
(let ((cps (exp ...)))
(with-cps cps clause ...)))
((_ cps (letk label cont) clause ...)
(let-fresh (label) ()
(with-cps (intmap-add! cps label (build-cont cont))
clause ...)))
((_ cps (setk label cont))
(intmap-add! cps label (build-cont cont)
(lambda (old new) new)))
((_ cps (setk label cont) clause ...)
(with-cps (with-cps cps (setk label cont))
clause ...))
((_ cps (letv v ...) clause ...)
(let-fresh () (v ...)
(with-cps cps clause ...)))
((_ cps (let$ var (proc arg ...)) clause ...)
(call-with-values (lambda () (proc cps arg ...))
(lambda (cps var)
(with-cps cps clause ...))))
((_ cps ($ (proc arg ...)))
(proc cps arg ...))
((_ cps exp)
(values cps exp))))
(define-syntax with-cps-constants
(syntax-rules ()
((_ cps () clause ...)
(with-cps cps clause ...))
((_ cps ((var val) (var* val*) ...) clause ...)
(let ((x val))
(with-cps cps
(letv var)
(let$ body (with-cps-constants ((var* val*) ...)
clause ...))
(letk label ($kargs ('var) (var) ,body))
(build-term ($continue label #f ($const x))))))))
|