Purely Functional Random-Access Lists
(require (planet dvanhorn/ralist:1:5)) |
David Van Horn
(at dvanhorn (dot ccs neu edu))
This is an implementation of purely functional random-access lists that enjoy O(1) first and rest operations and O(log n) list-ref and list-set operations.
Random-access lists implement the sequence interface, so (list? x) implies (sequence? x), and elements of a list may be extracted with any of the for syntactic forms.
This implementation is based on Okasaki, FPCA ’95.
(cons x xs) → list? |
x : any |
xs : list? |
Cons x onto xs.
empty : empty? |
The empty list.
(list x ) → list? |
x : any/c |
Returns a list containing the given xs as its elements.
(list* x xs) → list? |
x : any/c |
xs : list? |
Returns a list containing the give xs as its elements and xs as its tail.
(empty? x) → boolean? |
x : any/c |
Is x the empty list?
(cons? x) → boolean? |
x : any/c |
Is x a cons (a non-empty list)?
(list? x) → boolean? |
x : any/c |
Is x a list?
(first xs) → any |
xs : cons? |
Returns the first element of the list xs.
(rest xs) → list? |
xs : cons? |
Returns the rest of the element of the list xs.
(list-ref xs i) → any/c |
xs : cons? |
i : natural-number/c |
Returns the element of xs at position i. This operation runs in O((min i (log2 (length xs)))).
(list-set xs i v) → cons? |
xs : cons? |
i : natural-number/c |
v : any/c |
Returns a list identical to xs, except v is the ith element. This operation runs in O((min i (log2 (length xs)))).
| ||||||||
xs : cons? | ||||||||
i : natural-number/c | ||||||||
v : any/c |
Returns (values (list-ref xs i) (list-set xs i v)), but is more efficient.
| ||||||||
xs : cons? | ||||||||
i : natural-number/c | ||||||||
proc : procedure? |
Returns (values (list-ref xs i) (list-set xs i (f (list-ref xs i)))), but is more efficient.
(map proc xs) → list? |
proc : procedure? |
xs : list? |
Applies proc to each element of xs from the first element to the last. The result is a list containing each result of proc in order.
(foldr proc init xs) → any |
proc : procedure? |
init : any/c |
xs : list? |
Like foldr but for random-access lists. (Currently accepts only one list).
(foldl proc init xs) → any |
proc : procedure? |
init : any/c |
xs : list? |
Like foldl but for random-access lists. (Currently accepts only one list).
(build-list n proc) → list? |
n : natural-number/c |
proc : procedure? |
Like build-list but produces a random-access list.
(length xs) → natural-number/c |
xs : list? |
Returns the length of the list. This operation runs in O((log2 (length xs))).
(list-tail xs i) → list? |
xs : list? |
i : natural-number/c |
Returns the list after the first i elements of xs. This operation, like its pair counterpart runs in O(i).
(append xs ys) → list? |
xs : list? |
ys : list? |
Returns a list with all the elements of xs and ys in order. Like its pair counterpart, this operation runs in O((length xs)).
(reverse xs) → list? |
xs : list? |
Returns a list with all the elements of xs in reverse order.
(in-list xs) → list? |
xs : list? |
Returns a sequence equivalent to xs. Since lists are sequences, this is a list identity function, but an inlist application can provide better performance when it appears directly in a for clause.