// #pragma GCC optimize("Ofast,unroll-loops") #include <bits/stdc++.h> using namespace std; #define now chrono::steady_clock::now().time_since_epoch().count() #define speedup cin.tie(0)->sync_with_stdio(0) #define bitcount(x) __builtin_popcountll(x) #define lla(x) x.rbegin(), x.rend() #define all(x) x.begin(), x.end() #define Tp template <class T> #define int long long #define pb push_back #define vc vector #define arr array const int inf = 1e18; struct HSH { int b, mod; vc<int> p, pf, sf; HSH (string &s, int n, int _b = 39, int _mod = 1e9 + 7) : b(_b), mod(_mod) { pf.assign(n + 2, 0); sf.assign(n + 2, 0); p.assign(n + 2, 1); s = ' ' + s; for (int i = 1; i <= n; i++) { p[i] = (p[i - 1] * b) % mod; pf[i] = (pf[i - 1] * b + (s[i] - 'a' + 1)) % mod; } for (int i = n; i >= 1; i--) sf[i] = (sf[i + 1] * b + (s[i] - 'a' + 1)) % mod; }; int getp (int l, int r) { return (pf[r] - (pf[l - 1] * p[r - l + 1] % mod) + mod) % mod; } int gets (int l, int r) { return (sf[l] - (sf[r + 1] * p[r - l + 1] % mod) + mod) % mod; } }; inline void solve () { int n; string s; cin >> s; n = s.size(); HSH h (s, n, 39, 1e9 + 7); auto bs = [&] (int &m1, int &m2) -> void { int l = 0, r = n - m2 + 1; while (r - l > 1) { int mid = (l + r) >> 1; if (m1 - mid < 1 || h.getp(m2, m2 + mid) != h.gets(m1 - mid, m1)) r = mid; else l = mid; } m2 += l; m1 -= l; }; int ans = 0; unordered_map<int, int> dp; for (int i = 2; i < n; i++) { int l = i - 1, r = i + 1; if (s[l] == s[r] ) { bs(l, r); while (l < r) { int hash = h.getp(l, r); // cout << l << ' ' << r << ' ' << hash << '\n'; dp[hash] += (r - l + 1); l++; r--; } } } for (int i = 1; i < n; i++) { int l = i, r = i + 1; if (s[l] == s[r]) { bs(l, r); while (l < r) { int hash = h.getp(l, r); // cout << l << ' ' << r << ' ' << hash << '\n'; dp[hash] += (r - l + 1); l++; r--; } } } for (int i = 1; i <= n; i++) dp[h.getp(i, i)]++; for (const auto &[x, val] : dp) ans = max(ans, val); cout << ans << '\n'; } signed main() { speedup; solve(); return 0; } // aa