Version 5.0 The LEDA User Manual

Sorting and Searching
void L.sort(int (∗cmp)(const E& , const E& ), int dd = 1)
sorts the items of L using the ordering defined
by the compare function cmp : E×E −→ int, with
cmp(a, b)
⎧
⎨ < 0, if a < b
= 0, if a = b
⎩
> 0, if a > b
More precisely, if (in 1 , . . . , in n ) and
(out 1 , . . . , out n ) denote the values of L
before and after the call of sort, then
cmp(L[out j ], L[out j+1 ]) ≤ 0 for 1 ≤ j < n
and there is a permutation π of [1..n] such that
out i = in πi for 1 ≤ i ≤ n .
void L.sort( ) sorts the items of L using the default ordering
of type E, i.e., the linear order defined by function
int compare(const E&, const E&). If E is a
user-defined type, you have to provide the compare
function (see Section 2.3).
void L.merge sort(int (∗cmp)(const E& , const E& ))
sorts the items of L using merge sort and the ordering
defined by cmp. The sort is stable, i.e., if
x = y and 〈x〉 is before 〈y〉 in L then 〈x〉 is before
〈y〉 after the sort. L.merge sort( ) is more efficient
than L.sort( ) if L contains large pre-sorted intervals.
void L.merge sort( ) as above, but uses the default ordering of type E.
If E is a user-defined type, you have to provide the
compare function (see Section 2.3).
void L.bucket sort(int i, int j, int (∗b)(const E& ))
void L.bucket sort(int (∗b)(const E& ))
sorts the items of L using bucket sort, where b
maps every element x of L to a bucket b(x) ∈ [i..j].
If b(x) < b(y) then 〈x〉 appears before 〈y〉 after the
sort. If b(x) = b(y), the relative order of x and y
before the sort is retained, thus the sort is stable.
sorts list into increasing order as prescribed by
b Precondition: b is an integer-valued function on
E.