//
// Suffix Array (Manbar and Myers' O(n (log n)^2))
//
// Description:
// For a string s, tts suffix array is a lexicographically sorted
// list of suffixes of s. For example, for s = "abbab", its SA is
// 0 ab
// 1 abbab
// 2 b
// 3 bab
// 4 bbab
//
// Algorithm:
// Manbar and Myers' doubling algorithm.
// Suppose that suffixes are sorted by its first h characters.
// Then, the comparison of first 2h characters is computed by
// suf(i) <_2h suf(j) == if (suf(i) !=_h suf(j)) suf(i) <_h suf(j)
// else suf(i+h) <_h suf(j+h)
//
// Complexity:
// O(n (log n)^2).
// If we use radix sort instead of standard sort,
// we obtain O(n log n) algorithm. However, it does not improve
// practical performance so much.
//
// Verify:
// SPOJ 6409: SARRAY (80 pt)
//
#include
#include
#include
#include
#include
#include
#include
using namespace std;
#define fst first
#define snd second
#define all(c) ((c).begin()), ((c).end())
struct suffix_array {
int n;
vector x;
suffix_array(const char *s) : n(strlen(s)), x(n) {
vector r(n), t(n);
for (int i = 0; i < n; ++i) r[x[i] = i] = s[i];
for (int h = 1; t[n-1] != n-1; h *= 2) {
auto cmp = [&](int i, int j) {
if (r[i] != r[j]) return r[i] < r[j];
return i+h < n && j+h < n ? r[i+h] < r[j+h] : i > j;
};
sort(all(x), cmp);
for (int i = 0; i+1 < n; ++i) t[i+1] = t[i] + cmp(x[i], x[i+1]);
for (int i = 0; i < n; ++i) r[x[i]] = t[i];
}
}
int operator[](int i) const { return x[i]; }
};
int main() {
char s[100010];
scanf("%s", s);
suffix_array sary(s);
for (int i = 0; i < sary.n; ++i)
printf("%d\n", sary[i]);
}