#961721 (C++17) No.2690 A present from B (Hard)

提出ソース
結果

問題	No.2690 A present from B (Hard)
ユーザー	👑 p-adic
提出日時	2024-03-19 12:33:25
言語	C++17 (gcc 12.3.0 + boost 1.83.0)
結果	AC
実行時間	1,041 ms / 2,000 ms
コード長	54,663 bytes
コンパイル時間	4,014 ms
コンパイル使用メモリ	250,440 KB
実行使用メモリ	6,820 KB
最終ジャッジ日時	2024-09-30 05:26:06
合計ジャッジ時間	13,830 ms
ジャッジサーバーID （参考情報）	judge2 / judge4
このコードへのチャレンジ
（要ログイン）
テストケース

テストケース表示
入力	結果	実行時間実行使用メモリ
testcase_00	AC	1 ms 6,820 KB
testcase_01	AC	2 ms 6,816 KB
testcase_02	AC	2 ms 6,816 KB
testcase_03	AC	2 ms 6,816 KB
testcase_04	AC	2 ms 6,816 KB
testcase_05	AC	2 ms 6,816 KB
testcase_06	AC	1,030 ms 6,816 KB
testcase_07	AC	1,041 ms 6,816 KB
testcase_08	AC	1,024 ms 6,816 KB
testcase_09	AC	15 ms 6,816 KB
testcase_10	AC	16 ms 6,820 KB
testcase_11	AC	14 ms 6,820 KB
testcase_12	AC	2 ms 6,816 KB
testcase_13	AC	2 ms 6,820 KB
testcase_14	AC	2 ms 6,816 KB
testcase_15	AC	2 ms 6,820 KB
testcase_16	AC	2 ms 6,816 KB
testcase_17	AC	597 ms 6,820 KB
testcase_18	AC	376 ms 6,816 KB
testcase_19	AC	195 ms 6,816 KB
testcase_20	AC	572 ms 6,816 KB
testcase_21	AC	132 ms 6,820 KB
testcase_22	AC	108 ms 6,820 KB
testcase_23	AC	352 ms 6,820 KB
testcase_24	AC	335 ms 6,816 KB
testcase_25	AC	89 ms 6,816 KB
testcase_26	AC	576 ms 6,816 KB
testcase_27	AC	885 ms 6,816 KB
testcase_28	AC	877 ms 6,820 KB
testcase_29	AC	871 ms 6,816 KB
権限があれば一括ダウンロードができます
ソースコード

raw source code
#ifndef INCLUDE_MODE
  #define INCLUDE_MODE
  // #define REACTIVE
  // #define USE_GETLINE
#endif

#ifdef INCLUDE_MAIN

IN VO Solve()
{
  CIN( int , N , M );
  CIN_A( int , A , M );
  CEXPR( int , infty , 2e5 );
  GeometricProgressionMinDualSqrtDecomposition dp{ Module<int,int>() , infty , { -1 , 1 } , id<int>( N ) };
  FOREQINV( j , M - 1 , 0 ){
    int dpp_minus_2 = A[j]-2 >= 0 ? dp[A[j]-2] : -1;
    int dpp_minus_1 = dp[A[j]-1];
    int dpp_0 = dp[A[j]];
    int dpp_1 = A[j]+1 < N ? dp[A[j]+1] : -1;
    dp.Set( A[j]-1 , dpp_0 );
    dp.Set( A[j] , dpp_minus_1 );
    if( A[j]-2 >= 0 && dpp_0 > dpp_minus_2 + 1 ){
      dp.Set( A[j]-1 , dpp_minus_2 + 1 );
    }
    if( A[j]+1 < N && dpp_minus_1 > dpp_1 + 1 ){
      dp.Set( A[j] , dpp_1 + 1 );
    }
    int dp_minus_1 = dp[A[j]-1];
    int dp_0 = dp[A[j]];
    if( dpp_minus_1 > dp_minus_1 ){
      dp.IntervalSetMin( 0 , A[j] - 2 , 0 , dp_minus_1 + ( A[j]-1 ) );
      dp.IntervalSetMin( A[j] , N - 1 , 1 , dp_minus_1 - ( A[j]-1 ) );
    }
    if( dpp_0 > dp_0 ){
      dp.IntervalSetMin( 0 , A[j] - 1 , 0 , dp_0 + A[j] );
      dp.IntervalSetMin( A[j] + 1 , N - 1 , 1 , dp_0 - A[j] );
    }
  }
  FOR( i , 1 , N ){
    cout << dp[i] << " \n"[i==N-1];
  }
}
REPEAT_MAIN(1);

#else // INCLUDE_MAIN

#ifdef INCLUDE_SUB

// COMPAREに使用。圧縮時は削除する。
ll Naive( ll N , ll M , ll K )
{
  ll answer = N + M + K;
  return answer;
}

// COMPAREに使用。圧縮時は削除する。
ll Answer( ll N , ll M , ll K )
{
  // START_WATCH;
  ll answer = N + M + K;

  // // TLに準じる乱択や全探索。デフォルトの猶予は100.0[ms]。
  // CEXPR( double , TL , 2000.0 );
  // while( CHECK_WATCH( TL ) ){

  // }
  return answer;
}

// 圧縮時は中身だけ削除する。
IN VO Experiment()
{
  // CEXPR( int , bound , 10 );
  // FOREQ( N , 0 , bound ){
  //   FOREQ( M , 0 , bound ){
  //     FOREQ( K , 0 , bound ){
  //   	COUT( N , M , K , ":" , Naive( N , M , K ) );
  //     }
  //   }
  //   // cout << Naive( N ) << ",\n"[N==bound];
  // }
}

// 圧縮時は中身だけ削除する。
IN VO SmallTest()
{
  // CEXPR( int , bound , 10 );
  // FOREQ( N , 0 , bound ){
  //   FOREQ( M , 0 , bound ){
  //     FOREQ( K , 0 , bound ){
  //   	COMPARE( N , M , K );
  //     }
  //   }
  // }
}

// 圧縮時は中身だけ削除する。
IN VO RandomTest()
{
  // CEXPR( int , bound_N , 1e5 ); CIN_ASSERT( N , 1 , bound_N );
  // CEXPR( ll , bound_M , 1e18 ); CIN_ASSERT( M , 1 , bound_M );
  // CEXPR( ll , bound_K , 1e9 ); CIN_ASSERT( K , 1 , bound_K );
  // COMPARE( N , M , N );
}

#define INCLUDE_MAIN
#include __FILE__

#else // INCLUDE_SUB

#ifdef INCLUDE_LIBRARY

/*

C-x 3 C-x o C-x C-fによるファイル操作用

BFS (5KB)
c:/Users/user/Documents/Programming/Mathematics/Geometry/Graph/BreadthFirstSearch/compress.txt

CoordinateCompress (3KB)
c:/Users/user/Documents/Programming/Mathematics/SetTheory/DirectProduct/CoordinateCompress/compress.txt

DFSOnTree (11KB)
c:/Users/user/Documents/Programming/Mathematics/Geometry/Graph/DepthFirstSearch/Tree/a.hpp

Divisor (4KB)
c:/Users/user/Documents/Programming/Mathematics/Arithmetic/Prime/Divisor/compress.txt

IntervalAddBIT (9KB)
c:/Users/user/Documents/Programming/Mathematics/SetTheory/DirectProduct/AffineSpace/BIT/IntervalAdd/compress.txt

Polynomial (21KB)
c:/Users/user/Documents/Programming/Mathematics/Polynomial/compress.txt

UnionFind (3KB)
c:/Users/user/Documents/Programming/Mathematics/Geometry/Graph/UnionFindForest/compress.txt

*/

// VVV 常設でないライブラリは以下に挿入する。

template <typename R , typename U>
class VirtualRSet :
  virtual public UnderlyingSet<U>
{

public:
  virtual U Action( const R& r , const U& u ) = 0;
  inline U Power( const U& u , const R& r );
  inline U ScalarProduct( const R& r , const U& u );

};

template <typename U , typename MAGMA>
class RegularRSet :
  virtual public VirtualRSet<U,U> ,
  public MAGMA
{

public:
  inline RegularRSet( MAGMA magma );
  inline U Action( const U& r , const U& u );

};
template <typename MAGMA> RegularRSet( MAGMA magma ) -> RegularRSet<inner_t<MAGMA>,MAGMA>;

template <typename R , typename U , typename O_U>
class AbstractRSet :
  virtual public VirtualRSet<R,U>
{

private:
  O_U m_o_U;
  
public:
  // 型推論のために余計な引数dummy0,dummy1が必要
  inline AbstractRSet( const R& dummy0 , const U& dummy1 , O_U o_U );
  inline U Action( const R& r , const U& u );

};

template <typename R , typename U , typename O_U , typename GROUP>
class AbstractModule :
  public AbstractRSet<R,U,O_U> ,
  public GROUP
{

public:
  // 型推論のために余計な引数dummyが必要
  inline AbstractModule( const R& dummy , O_U o_U , GROUP M );

};
template <typename R , typename O_U , typename GROUP> AbstractModule( const R& dummy , O_U o_U , GROUP M ) -> AbstractModule<R,inner_t<GROUP>,O_U,GROUP>;

template <typename R , typename U>
class Module :
  virtual public VirtualRSet<R,U> ,
  public AdditiveGroup<U>
{

private:
  inline U Action( const R& r , const U& u );
  
};

template <typename R , typename MAGMA> inline RegularRSet<R,MAGMA>::RegularRSet( MAGMA magma ) : MAGMA( move( magma ) ) {}
template <typename R , typename U , typename O_U> inline AbstractRSet<R,U,O_U>::AbstractRSet( const R& dummy0 , const U& dummy1 , O_U o_U ) : m_o_U( move( o_U ) ) { static_assert( is_invocable_r_v<U,O_U,R,U> ); }
template <typename R , typename U , typename O_U , typename GROUP> inline AbstractModule<R,U,O_U,GROUP>::AbstractModule( const R& dummy , O_U o_U , GROUP M ) : AbstractRSet<R,U,O_U>( dummy , M.One() , move( o_U ) ) , GROUP( move( M ) ) { static_assert( is_same_v<U,inner_t<GROUP>> ); }

template <typename U , typename MAGMA> inline U RegularRSet<U,MAGMA>::Action( const U& r , const U& u ) { return this->Product( r , u ); }
template <typename R , typename U , typename O_U> inline U AbstractRSet<R,U,O_U>::Action( const R& r , const U& u ) { return m_o_U( r , u ); }
template <typename R , typename U> inline U Module<R,U>::Action( const R& r , const U& u ) { return r * u; }

template <typename R , typename U> inline U VirtualRSet<R,U>::Power( const U& u , const R& r ) { return Action( r , u ); }
template <typename R , typename U> inline U VirtualRSet<R,U>::ScalarProduct( const R& r , const U& u ) { return Action( r , u ); }

IN CE int Sqrt(CRI N)NE{if(N <= 1){RE 1;}int left = 0;int right = N;WH(left + 1 < right){int m =(left + right)/ 2;(m <=(N - 1)/ m?left:right)= m;}RE right;}

// 入力の範囲内で要件
// (1) LがRの基点付き左作用構造（例えば基点付きマグマの正則左加群構造）である。
// (2) XがUの「Lの基点がUの恒等変換に対応する左L作用構造」である。
// を満たす場合にのみサポート。
// Mに可換性を課す場合は一点作用がO(1)のCommutativeDualSqrtDecompositionが使用可。

// 配列による初期化O(N)

// 一点取得O(1)

// 一点代入O(N^{1/2})（ただし状況次第でO(1)）
// X.Action()による一点作用はなし（一点代入か区間作用を使う）
// X.Action()による区間作用O(N^{1/2})
template <typename R , typename PT_MAGMA , typename U , typename R_SET>
class DualSqrtDecomposition
{

protected:
  PT_MAGMA m_L;
  R_SET m_X;
  int m_N;
  int m_N_sqrt;
  int m_N_d;
  int m_N_m;
  vector<U> m_a;
  vector<R> m_b;

public:
  inline DualSqrtDecomposition( PT_MAGMA L , R_SET X , vector<U> a = {} );
  inline DualSqrtDecomposition( PT_MAGMA L , R_SET X , vector<U> a , const int& N_sqrt );

  template <typename...Args> inline void Initialise( Args&&... args );
  inline void Set( const int& i , const U& u );
  // ArgはL.Actionの第1引数として有効な型。
  template <typename Arg> inline void IntervalAct( const int& i_start , const int& i_final , const Arg& r );

  // 参照返しでないことに注意。
  inline U operator[]( const int& i );
  inline U Get( const int& i );

protected:
  // 作用の遅延評価を解消する。
  inline void Update( const int& d );
  
};
template <typename PT_MAGMA , typename R_SET , typename...Args> DualSqrtDecomposition( PT_MAGMA M , R_SET X , Args&&... args ) -> DualSqrtDecomposition<inner_t<PT_MAGMA>,PT_MAGMA,inner_t<R_SET>,R_SET>;

template <typename R , typename PT_MAGMA , typename U , typename R_SET> inline DualSqrtDecomposition<R,PT_MAGMA,U,R_SET>::DualSqrtDecomposition( PT_MAGMA L , R_SET X , vector<U> a ) : m_L( move( L ) ) , m_X( move( X ) ) , m_N( a.size() ) , m_N_sqrt( Sqrt( m_N ) ) , m_N_d( ( m_N + m_N_sqrt - 1 ) / m_N_sqrt ) , m_N_m( m_N_d * m_N_sqrt ) , m_a( move( a ) ) , m_b( m_N_d , m_L.Point() ) { static_assert( is_same_v<R,inner_t<PT_MAGMA>> && is_same_v<U,inner_t<R_SET>> ); }
template <typename R , typename PT_MAGMA , typename U , typename R_SET> inline DualSqrtDecomposition<R,PT_MAGMA,U,R_SET>::DualSqrtDecomposition( PT_MAGMA L , R_SET X , vector<U> a , const int& N_sqrt ) : m_L( move( L ) ) , m_X( move( X ) ) , m_N( a.size() ) , m_N_sqrt( N_sqrt ) , m_N_d( ( m_N + m_N_sqrt - 1 ) / m_N_sqrt ) , m_N_m( m_N_d * m_N_sqrt ) , m_a( move( a ) ) , m_b( m_N_d , m_L.Point() ) { static_assert( is_same_v<R,inner_t<PT_MAGMA>> && is_same_v<U,inner_t<R_SET>> ); }

template <typename R , typename PT_MAGMA , typename U , typename R_SET> template <typename...Args> inline void DualSqrtDecomposition<R,PT_MAGMA,U,R_SET>::Initialise( Args&&... args ) { *this = DualSqrtDecomposition<R,PT_MAGMA,U,R_SET>( move( m_L ) , move( m_X ) , forward<decay_t<Args>>( args )... ); }
template <typename R , typename PT_MAGMA , typename U , typename R_SET> inline void DualSqrtDecomposition<R,PT_MAGMA,U,R_SET>::Set( const int& i , const U& u ) { U& m_ai = m_a[i]; if( m_ai != u ){ Update( i / m_N_sqrt ); m_ai = u; } }

template <typename R , typename PT_MAGMA , typename U , typename R_SET> template <typename Arg> inline void DualSqrtDecomposition<R,PT_MAGMA,U,R_SET>::IntervalAct( const int& i_start , const int& i_final , const Arg& r )
{

  const int i_min = max( i_start , 0 );
  const int i_ulim = min( i_final + 1 , m_N );
  const int d_0 = ( i_min + m_N_sqrt - 1 ) / m_N_sqrt;
  const int d_1 = max( d_0 , i_ulim / m_N_sqrt );
  const int i_0 = min( d_0 * m_N_sqrt , i_ulim );
  const int i_1 = max( i_0 , d_1 * m_N_sqrt );
  const R r_copy = m_L.Action( r , m_L.Point() );

  if( d_0 > 0 ){

    Update( d_0 - 1 );

  }
  
  for( int i = i_min ; i < i_0 ; i++ ){

    U& m_ai = m_a[i];
    m_ai = m_X.Action( r_copy , m_ai );

  }
  
  for( int d = d_0 ; d < d_1 ; d++ ){

    R& m_bd = m_b[d];
    m_bd = m_L.Action( r , m_bd );

  }

  if( d_1 < m_N_d ){

    Update( d_1 );

  }
  
  for( int i = i_1 ; i < i_ulim ; i++ ){

    U& m_ai = m_a[i];
    m_ai = m_X.Action( r_copy , m_ai );

  }

  return;
  
}

template <typename R , typename PT_MAGMA , typename U , typename R_SET> inline U DualSqrtDecomposition<R,PT_MAGMA,U,R_SET>::operator[]( const int& i ) { assert( 0 <= i && i < m_N ); return m_X.Action( m_b[i / m_N_sqrt] , m_a[i] ); }
template <typename R , typename PT_MAGMA , typename U , typename R_SET> inline U DualSqrtDecomposition<R,PT_MAGMA,U,R_SET>::Get( const int& i ) { return operator[]( i ); }

template <typename R , typename PT_MAGMA , typename U , typename R_SET> inline void DualSqrtDecomposition<R,PT_MAGMA,U,R_SET>::Update( const int& d )
{

  R& m_bd = m_b[d];
  const R& point = m_L.Point();

  if( m_bd != point ){

    const int j_min = d * m_N_sqrt;
    const int j_ulim = min( j_min + m_N_sqrt , m_N );

    for( int j = j_min ; j < j_ulim ; j++ ){

      U& m_aj = m_a[j];
      m_aj = m_X.Action( m_bd , m_aj );

    }

    m_bd = point;

  }
  
  return;

}

// 自由U加群の構造射と演算の合成
template <typename U>
class GeometricProgressionMinComposition :
  virtual public PointedSet<vector<U>> ,
  virtual public VirtualRSet<pair<int,U>,vector<U>>
{
  
public:
  inline GeometricProgressionMinComposition( const int& size , const U& infty );
  vector<U> Action( const pair<int,U>& y0 , const vector<U>& y1 );

};

// 自由U加群の左作用構造
template <typename U , typename N_MODULE>
class GeometricProgressionMinAction :
  virtual public VirtualRSet<vector<U>,pair<U,int>>
{
  
private:
  N_MODULE m_M;
  const vector<U>* m_p_memory;

public:
  inline GeometricProgressionMinAction( N_MODULE M , const vector<U>& memory );
  pair<U,int> Action( const vector<U>& y , const pair<U,int>& u );

};

template <typename U , typename N_MODULE>
using BaseOfGeometricProgressionMinDualSqrtDecomposition = DualSqrtDecomposition<vector<U>,GeometricProgressionMinComposition<U>,pair<U,int>,GeometricProgressionMinAction<U,N_MODULE>>;

// 入力の範囲内で要件
// (1) Mがbool operator<(const U&,const U&)に関してUの全順序N加群構造である。
// を満たす場合にのみサポート。
// クエリ中の等比数列(r^i y)_iの公比rの種類数をCとおく。

// 零初期化O(N)
// 配列による初期化O(N)

// 一点取得O(C)

// 一点代入O(N^{1/2}C)（ただし状況次第でO(1)）
// 等比数列(r^i y)_iによる区間min作用O(N^{1/2} log_2 C)
template <typename U , typename N_MODULE>
class GeometricProgressionMinDualSqrtDecomposition :
  public BaseOfGeometricProgressionMinDualSqrtDecomposition<U,N_MODULE>
{

private:
  vector<U> m_memory;
  
public:
  template <typename...Args> inline GeometricProgressionMinDualSqrtDecomposition( N_MODULE M , const U& infty , vector<U> memory_r , vector<U> a = {} , Args&&... args );

  template <typename...Args> inline void Initialise( Args&&... args );
  inline void Set( const int& i , U u );
  inline void IntervalAct( const int& i_start , const int& i_final , const vector<U>& f ) = delete;
  inline void IntervalSetMin( const int& i_start , const int& i_final , const int& r_num , const U& y );

  inline U operator[]( const int& i );
  inline U Get( const int& i );

private:
  static vector<pair<U,int>> GetGraph( vector<U> a );
  
};
template <typename N_MODULE , typename...Args> GeometricProgressionMinDualSqrtDecomposition( N_MODULE M , Args&&... args ) -> GeometricProgressionMinDualSqrtDecomposition<inner_t<N_MODULE>,N_MODULE>;


template <typename U> inline GeometricProgressionMinComposition<U>::GeometricProgressionMinComposition( const int& size , const U& infty ) : PointedSet<vector<U>>( vector( size , infty ) ) {}
template <typename U , typename N_MODULE> inline GeometricProgressionMinAction<U,N_MODULE>::GeometricProgressionMinAction( N_MODULE M , const vector<U>& memory ) : m_M( move( M ) ) , m_p_memory( &memory ) { static_assert( is_same_v<U,inner_t<N_MODULE>> ); }

template <typename U>
vector<U> GeometricProgressionMinComposition<U>::Action( const pair<int,U>& y0 , const vector<U>& y1 )
{

  vector<U> answer = y1;
  U& answer_r = answer[y0.first];
  answer_r = min( answer_r , y0.second );
  return answer;

}

template <typename U , typename N_MODULE>
pair<U,int> GeometricProgressionMinAction<U,N_MODULE>::Action( const vector<U>& y , const pair<U,int>& u )
{
    
  const int size = y.size();
  U temp = u.first;
  
  for( int i = 0 ; i < size ; i++ ){

    temp = min( temp , m_M.Product( m_M.Power( ( *m_p_memory )[i] , u.second ) , y.at( i ) ) );

  }

  return { move( temp ) , u.second };

}

template <typename U , typename N_MODULE> template <typename...Args> inline GeometricProgressionMinDualSqrtDecomposition<U,N_MODULE>::GeometricProgressionMinDualSqrtDecomposition( N_MODULE M , const U& infty , vector<U> memory_r , vector<U> a , Args&&... args ) : BaseOfGeometricProgressionMinDualSqrtDecomposition<U,N_MODULE>( GeometricProgressionMinComposition<U>( memory_r.size() , infty ) , GeometricProgressionMinAction<U,N_MODULE>( move( M ) , m_memory ) , GetGraph( move( a ) ) , forward<decay_t<Args>>( args )... ) , m_memory( move( memory_r ) ) {}

template <typename U , typename N_MODULE> template <typename...Args> inline void GeometricProgressionMinDualSqrtDecomposition<U,N_MODULE>::Initialise( Args&&... args ) { *this = GeometricProgressionMinDualSqrtDecomposition<U,N_MODULE>( move( this->m_M , forward<decay_t<Args>>( args )... ) ); }

template <typename U , typename N_MODULE> inline void GeometricProgressionMinDualSqrtDecomposition<U,N_MODULE>::Set( const int& i , U u ) { BaseOfGeometricProgressionMinDualSqrtDecomposition<U,N_MODULE>::Set( i , { move( u ) , i } ); }

template <typename U , typename N_MODULE> inline void GeometricProgressionMinDualSqrtDecomposition<U,N_MODULE>::IntervalSetMin( const int& i_start , const int& i_final , const int& r_num , const U& y ) { assert( 0 <= r_num && r_num < int( m_memory.size() ) ); BaseOfGeometricProgressionMinDualSqrtDecomposition<U,N_MODULE>::IntervalAct( i_start , i_final , pair<int,U>{ r_num , y } ); }

template <typename U , typename N_MODULE> inline U GeometricProgressionMinDualSqrtDecomposition<U,N_MODULE>::operator[]( const int& i ) { return BaseOfGeometricProgressionMinDualSqrtDecomposition<U,N_MODULE>::operator[]( i ).first; }
template <typename U , typename N_MODULE> inline U GeometricProgressionMinDualSqrtDecomposition<U,N_MODULE>::Get( const int& i ) { return operator[]( i ); }

template <typename U , typename N_MODULE> inline vector<pair<U,int>> GeometricProgressionMinDualSqrtDecomposition<U,N_MODULE>::GetGraph( vector<U> a ) { const int size = a.size(); vector<pair<U,int>> answer( size ); for( int i = 0 ; i < size ; i++ ){ answer[i] = { move( a[i] ) , i }; } return answer; }

// AAA 常設でないライブラリは以上に挿入する。

#define INCLUDE_SUB
#include __FILE__

#else // INCLUDE_LIBRARY

#ifdef DEBUG
  #define _GLIBCXX_DEBUG
  #define REPEAT_MAIN( BOUND ) START_MAIN; signal( SIGABRT , &AlertAbort ); AutoCheck( exec_mode , use_getline ); CEXPR( int , bound_test_case_num , BOUND ); int test_case_num = 1; if( exec_mode == solve_mode ){ if CE( bound_test_case_num > 1 ){ CERR( "テストケースの個数を入力してください。" ); SET_ASSERT( test_case_num , 1 , bound_test_case_num ); } } else { if( exec_mode == experiment_mode ){ Experiment(); } else if( exec_mode == small_test_mode ){ SmallTest(); } else if( exec_mode == random_test_mode ){ CERR( "ランダムテストを行う回数を指定してください。" ); SET_LL( test_case_num ); REPEAT( test_case_num ){ RandomTest(); } } RE 0; } FINISH_MAIN
  #define DEXPR( LL , BOUND , VALUE1 , VALUE2 ) CEXPR( LL , BOUND , VALUE2 )
  #define ASSERT( A , MIN , MAX ) CERR( "ASSERTチェック： " , ( MIN ) , ( ( MIN ) <= A ? "<=" : ">" ) , A , ( A <= ( MAX ) ? "<=" : ">" ) , ( MAX ) ); AS( ( MIN ) <= A && A <= ( MAX ) )
  #define SET_ASSERT( A , MIN , MAX ) if( exec_mode == solve_mode ){ SET_LL( A ); ASSERT( A , MIN , MAX ); } else if( exec_mode == random_test_mode ){ CERR( #A , " = " , ( A = GetRand( MIN , MAX ) ) ); } else { AS( false ); }
  #define SOLVE_ONLY ST_AS( __FUNCTION__[0] == 'S' )
  #define CERR( ... ) VariadicCout( cerr , __VA_ARGS__ ) << endl
  #define COUT( ... ) VariadicCout( cout << "出力： " , __VA_ARGS__ ) << endl
  #define CERR_A( A , N ) OUTPUT_ARRAY( cerr , A , N ) << endl
  #define COUT_A( A , N ) cout << "出力： "; OUTPUT_ARRAY( cout , A , N ) << endl
  #define CERR_ITR( A ) OUTPUT_ITR( cerr , A ) << endl
  #define COUT_ITR( A ) cout << "出力： "; OUTPUT_ITR( cout , A ) << endl
#else
  #pragma GCC optimize ( "O3" )
  #pragma GCC optimize ( "unroll-loops" )
  #pragma GCC target ( "sse4.2,fma,avx2,popcnt,lzcnt,bmi2" )
  #define REPEAT_MAIN( BOUND ) START_MAIN; CEXPR( int , bound_test_case_num , BOUND ); int test_case_num = 1; if CE( bound_test_case_num > 1 ){ SET_ASSERT( test_case_num , 1 , bound_test_case_num ); } FINISH_MAIN
  #define DEXPR( LL , BOUND , VALUE1 , VALUE2 ) CEXPR( LL , BOUND , VALUE1 )
  #define ASSERT( A , MIN , MAX ) AS( ( MIN ) <= A && A <= ( MAX ) )
  #define SET_ASSERT( A , MIN , MAX ) SET_LL( A ); ASSERT( A , MIN , MAX )
  #define SOLVE_ONLY 
  #define CERR( ... ) 
  #define COUT( ... ) VariadicCout( cout , __VA_ARGS__ ) << ENDL
  #define CERR_A( A , N ) 
  #define COUT_A( A , N ) OUTPUT_ARRAY( cout , A , N ) << ENDL
  #define CERR_ITR( A ) 
  #define COUT_ITR( A ) OUTPUT_ITR( cout , A ) << ENDL
#endif
#ifdef REACTIVE
  #define ENDL endl
#else
  #define ENDL "\n"
#endif
#ifdef USE_GETLINE
  #define SET_LL( A ) { GETLINE( A ## _str ); A = stoll( A ## _str ); }
  #define GETLINE_SEPARATE( SEPARATOR , ... ) SOLVE_ONLY; string __VA_ARGS__; VariadicGetline( cin , SEPARATOR , __VA_ARGS__ )
  #define GETLINE( ... ) SOLVE_ONLY; GETLINE_SEPARATE( '\n' , __VA_ARGS__ )
#else
  #define SET_LL( A ) cin >> A
  #define CIN( LL , ... ) SOLVE_ONLY; LL __VA_ARGS__; VariadicCin( cin , __VA_ARGS__ )
  #define SET_A( A , N ) SOLVE_ONLY; FOR( VARIABLE_FOR_SET_A , 0 , N ){ cin >> A[VARIABLE_FOR_SET_A]; }
  #define CIN_A( LL , A , N ) VE<LL> A( N ); SET_A( A , N );
#endif
#include <bits/stdc++.h>
using namespace std;
#define ATT __attribute__( ( target( "sse4.2,fma,avx2,popcnt,lzcnt,bmi2" ) ) )
#define START_MAIN int main(){ ios_base::sync_with_stdio( false ); cin.tie( nullptr )
#define FINISH_MAIN REPEAT( test_case_num ){ if CE( bound_test_case_num > 1 ){ CERR( "testcase " , VARIABLE_FOR_REPEAT_test_case_num , ":" ); } Solve(); CERR( "" ); } }
#define START_WATCH chrono::system_clock::time_point watch = chrono::system_clock::now()
#define CURRENT_TIME static_cast<double>( chrono::duration_cast<chrono::microseconds>( chrono::system_clock::now() - watch ).count() / 1000.0 )
#define CHECK_WATCH( TL_MS ) ( CURRENT_TIME < TL_MS - 100.0 )
#define CEXPR( LL , BOUND , VALUE ) CE LL BOUND = VALUE
#define SET_A_ASSERT( A , N , MIN , MAX ) FOR( VARIABLE_FOR_SET_A , 0 , N ){ SET_ASSERT( A[VARIABLE_FOR_SET_A] , MIN , MAX ); }
#define CIN_ASSERT( A , MIN , MAX ) decldecay_t( MAX ) A; SET_ASSERT( A , MIN , MAX )
#define CIN_A_ASSERT( A , N , MIN , MAX ) vector<decldecay_t( MAX )> A( N ); SET_A_ASSERT( A , N , MIN , MAX )
#define FOR( VAR , INITIAL , FINAL_PLUS_ONE ) for( decldecay_t( FINAL_PLUS_ONE ) VAR = INITIAL ; VAR < FINAL_PLUS_ONE ; VAR ++ )
#define FOREQ( VAR , INITIAL , FINAL ) for( decldecay_t( FINAL ) VAR = INITIAL ; VAR <= FINAL ; VAR ++ )
#define FOREQINV( VAR , INITIAL , FINAL ) for( decldecay_t( INITIAL ) VAR = INITIAL ; VAR + 1 > FINAL ; VAR -- )
#define AUTO_ITR( ARRAY ) auto itr_ ## ARRAY = ARRAY .BE() , end_ ## ARRAY = ARRAY .EN()
#define FOR_ITR( ARRAY ) for( AUTO_ITR( ARRAY ) , itr = itr_ ## ARRAY ; itr_ ## ARRAY != end_ ## ARRAY ; itr_ ## ARRAY ++ , itr++ )
#define REPEAT( HOW_MANY_TIMES ) FOR( VARIABLE_FOR_REPEAT_ ## HOW_MANY_TIMES , 0 , HOW_MANY_TIMES )
#define SET_PRECISION( DECIMAL_DIGITS ) cout << fixed << setprecision( DECIMAL_DIGITS )
#define OUTPUT_ARRAY( OS , A , N ) FOR( VARIABLE_FOR_OUTPUT_ARRAY , 0 , N ){ OS << A[VARIABLE_FOR_OUTPUT_ARRAY] << (VARIABLE_FOR_OUTPUT_ARRAY==N-1?"":" "); } OS
#define OUTPUT_ITR( OS , A ) { auto ITERATOR_FOR_OUTPUT_ITR = A.BE() , EN_FOR_OUTPUT_ITR = A.EN(); bool VARIABLE_FOR_OUTPUT_ITR = ITERATOR_FOR_COUT_ITR != END_FOR_COUT_ITR; WH( VARIABLE_FOR_OUTPUT_ITR ){ OS << *ITERATOR_FOR_COUT_ITR; ( VARIABLE_FOR_OUTPUT_ITR = ++ITERATOR_FOR_COUT_ITR != END_FOR_COUT_ITR ) ? OS : OS << " "; } } OS
#define RETURN( ... ) SOLVE_ONLY; COUT( __VA_ARGS__ ); RE
#define COMPARE( ... ) auto naive = Naive( __VA_ARGS__ ); auto answer = Answer( __VA_ARGS__ ); bool match = naive == answer; COUT( "(" , #__VA_ARGS__ , ") == (" , __VA_ARGS__ , ") : Naive == " , naive , match ? "==" : "!=" , answer , "== Answer" ); if( !match ){ RE; }

// 圧縮用
#define TE template
#define TY typename
#define US using
#define ST static
#define AS assert
#define IN inline
#define CL class
#define PU public
#define OP operator
#define CE constexpr
#define CO const
#define NE noexcept
#define RE return 
#define WH while
#define VO void
#define VE vector
#define LI list
#define BE begin
#define EN end
#define SZ size
#define LE length
#define PW Power
#define MO move
#define TH this
#define CRI CO int&
#define CRUI CO uint&
#define CRL CO ll&
#define VI virtual 
#define ST_AS static_assert
#define reMO_CO remove_const
#define is_COructible_v is_constructible_v
#define rBE rbegin
#define reSZ resize

// 型のエイリアス
#define decldecay_t(VAR)decay_t<decltype(VAR)>
TE <TY F,TY...Args> US ret_t = decltype(declval<F>()(declval<Args>()...));
TE <TY T> US inner_t = TY T::type;
US uint = unsigned int;
US ll = long long;
US ull = unsigned long long;
US ld = long double;
US lld = __float128;
TE <TY INT> US T2 = pair<INT,INT>;
TE <TY INT> US T3 = tuple<INT,INT,INT>;
TE <TY INT> US T4 = tuple<INT,INT,INT,INT>;
US path = pair<int,ll>;

// 入出力用
#define DF_OF_COUT_FOR_VE(V)TE <CL Traits,TY Arg> IN basic_ostream<char,Traits>& OP<<(basic_ostream<char,Traits>& os,CO V<Arg>& arg){auto BE = arg.BE(),EN = arg.EN();auto IT = BE;WH(IT != EN){(IT == BE?os:os << " ")<< *IT;IT++;}RE os;}
TE <CL Traits> IN basic_istream<char,Traits>& VariadicCin(basic_istream<char,Traits>& is){RE is;}
TE <CL Traits,TY Arg,TY... ARGS> IN basic_istream<char,Traits>& VariadicCin(basic_istream<char,Traits>& is,Arg& arg,ARGS&... args){RE VariadicCin(is >> arg,args...);}
TE <CL Traits> IN basic_istream<char,Traits>& VariadicGetline(basic_istream<char,Traits>& is,CO char& separator){RE is;}
TE <CL Traits,TY Arg,TY... ARGS> IN basic_istream<char,Traits>& VariadicGetline(basic_istream<char,Traits>& is,CO char& separator,Arg& arg,ARGS&... args){RE VariadicGetline(getline(is,arg,separator),separator,args...);}
DF_OF_COUT_FOR_VE(VE);
DF_OF_COUT_FOR_VE(LI);
DF_OF_COUT_FOR_VE(set);
DF_OF_COUT_FOR_VE(unordered_set);
TE <CL Traits,TY Arg1,TY Arg2> IN basic_ostream<char,Traits>& OP<<(basic_ostream<char,Traits>& os,CO pair<Arg1,Arg2>& arg){RE os << arg.first << " " << arg.second;}
TE <CL Traits,TY Arg> IN basic_ostream<char,Traits>& VariadicCout(basic_ostream<char,Traits>& os,CO Arg& arg){RE os << arg;}
TE <CL Traits,TY Arg1,TY Arg2,TY... ARGS> IN basic_ostream<char,Traits>& VariadicCout(basic_ostream<char,Traits>& os,CO Arg1& arg1,CO Arg2& arg2,CO ARGS&... args){RE VariadicCout(os << arg1 << " ",arg2,args...);}


// 算術用
TE <TY T> CE T PositiveBaseRS(CO T& a,CO T& p){RE a >= 0?a % p:p - 1 -((-(a + 1))% p);}
TE <TY T> CE T RS(CO T& a,CO T& p){RE PositiveBaseRS(a,p < 0?-p:p);}
TE <TY T> CE T PositiveBaseQuotient(CO T& a,CO T& p){RE(a - PositiveBaseRS(a,p))/ p;}
TE <TY T> CE T Quotient(CO T& a,CO T& p){RE p < 0?PositiveBaseQuotient(-a,-p):PositiveBaseQuotient(a,p);}

#define POWER( ANSWER , ARGUMENT , EXPONENT )				\
  ST_AS( ! is_same<decldecay_t( ARGUMENT ),int>::value && ! is_same<decldecay_t( ARGUMENT ),uint>::value ); \
  decldecay_t( ARGUMENT ) ANSWER{ 1 };					\
  {									\
    decldecay_t( ARGUMENT ) ARGUMENT_FOR_SQUARE_FOR_POWER = ( ARGUMENT ); \
    decldecay_t( EXPONENT ) EXPONENT_FOR_SQUARE_FOR_POWER = ( EXPONENT ); \
    WH( EXPONENT_FOR_SQUARE_FOR_POWER != 0 ){				\
      if( EXPONENT_FOR_SQUARE_FOR_POWER % 2 == 1 ){			\
	ANSWER *= ARGUMENT_FOR_SQUARE_FOR_POWER;			\
      }									\
      ARGUMENT_FOR_SQUARE_FOR_POWER *= ARGUMENT_FOR_SQUARE_FOR_POWER;	\
      EXPONENT_FOR_SQUARE_FOR_POWER /= 2;				\
    }									\
  }									\

#define POWER_MOD( ANSWER , ARGUMENT , EXPONENT , MODULO )		\
  ll ANSWER{ 1 };							\
  {									\
    ll ARGUMENT_FOR_SQUARE_FOR_POWER = ( ( ARGUMENT ) % ( MODULO ) ) % ( MODULO ); \
    ARGUMENT_FOR_SQUARE_FOR_POWER < 0 ? ARGUMENT_FOR_SQUARE_FOR_POWER += ( MODULO ) : ARGUMENT_FOR_SQUARE_FOR_POWER; \
    decldecay_t( EXPONENT ) EXPONENT_FOR_SQUARE_FOR_POWER = ( EXPONENT ); \
    WH( EXPONENT_FOR_SQUARE_FOR_POWER != 0 ){				\
      if( EXPONENT_FOR_SQUARE_FOR_POWER % 2 == 1 ){			\
	ANSWER = ( ANSWER * ARGUMENT_FOR_SQUARE_FOR_POWER ) % ( MODULO ); \
      }									\
      ARGUMENT_FOR_SQUARE_FOR_POWER = ( ARGUMENT_FOR_SQUARE_FOR_POWER * ARGUMENT_FOR_SQUARE_FOR_POWER ) % ( MODULO ); \
      EXPONENT_FOR_SQUARE_FOR_POWER /= 2;				\
    }									\
  }									\

#define FACTORIAL_MOD( ANSWER , ANSWER_INV , INVERSE , MAX_INDEX , CE_LENGTH , MODULO ) \
  ll ANSWER[CE_LENGTH];							\
  ll ANSWER_INV[CE_LENGTH];						\
  ll INVERSE[CE_LENGTH];						\
  {									\
    ll VARIABLE_FOR_PRODUCT_FOR_FACTORIAL = 1;				\
    ANSWER[0] = VARIABLE_FOR_PRODUCT_FOR_FACTORIAL;			\
    FOREQ( i , 1 , MAX_INDEX ){						\
      ANSWER[i] = ( VARIABLE_FOR_PRODUCT_FOR_FACTORIAL *= i ) %= ( MODULO ); \
    }									\
    ANSWER_INV[0] = ANSWER_INV[1] = INVERSE[1] = VARIABLE_FOR_PRODUCT_FOR_FACTORIAL = 1; \
    FOREQ( i , 2 , MAX_INDEX ){						\
      ANSWER_INV[i] = ( VARIABLE_FOR_PRODUCT_FOR_FACTORIAL *= INVERSE[i] = ( MODULO ) - ( ( ( ( MODULO ) / i ) * INVERSE[ ( MODULO ) % i ] ) % ( MODULO ) ) ) %= ( MODULO ); \
    }									\
  }									\

// 二分探索用
// EXPRESSIONがANSWERの広義単調関数の時、EXPRESSION >= CO_TARGETの整数解を格納。
#define BS( ANSWER , MINIMUM , MAXIMUM , EXPRESSION , DESIRED_INEQUALITY , CO_TARGET , INEQUALITY_FOR_CHECK , UPDATE_U , UPDATE_L , UPDATE_ANSWER ) \
  ST_AS( ! is_same<decldecay_t( CO_TARGET ),uint>::value && ! is_same<decldecay_t( CO_TARGET ),ull>::value ); \
  ll ANSWER = MINIMUM;							\
  {									\
    ll L_BS = MINIMUM;							\
    ll U_BS = MAXIMUM;							\
    ANSWER = UPDATE_ANSWER;						\
    ll EXPRESSION_BS;							\
    CO ll CO_TARGET_BS = ( CO_TARGET );			\
    ll DIFFERENCE_BS;							\
    WH( L_BS < U_BS ){						\
      DIFFERENCE_BS = ( EXPRESSION_BS = ( EXPRESSION ) ) - CO_TARGET_BS; \
      CERR( "二分探索中：" , "L_BS =" , L_BS , "<=" , #ANSWER , "=" , ANSWER , "<=" , U_BS , "= U_BS :" , #EXPRESSION , "=" , EXPRESSION_BS , DIFFERENCE_BS > 0 ? ">" : DIFFERENCE_BS < 0 ? "<" : "=" , CO_TARGET_BS , "=" , #CO_TARGET ); \
      if( DIFFERENCE_BS INEQUALITY_FOR_CHECK 0 ){			\
	U_BS = UPDATE_U;						\
      } else {								\
	L_BS = UPDATE_L;						\
      }									\
      ANSWER = UPDATE_ANSWER;						\
    }									\
    if( L_BS > U_BS ){							\
      CERR( "二分探索失敗：" , "L_BS =" , L_BS , ">" , U_BS , "= U_BS :" , #ANSWER , ":=" , #MAXIMUM , "+ 1 =" , MAXIMUM + 1  ); \
      CERR( "二分探索マクロにミスがある可能性があります。変更前の版に戻してください。" ); \
      ANSWER = MAXIMUM + 1;						\
    } else {								\
      CERR( "二分探索終了：" , "L_BS =" , L_BS , "<=" , #ANSWER , "=" , ANSWER , "<=" , U_BS , "= U_BS" ); \
      CERR( "二分探索が成功したかを確認するために" , #EXPRESSION , "を計算します。" ); \
      CERR( "成功判定が不要な場合はこの計算を削除しても構いません。" );	\
      EXPRESSION_BS = ( EXPRESSION );					\
      CERR( "二分探索結果：" , #EXPRESSION , "=" , EXPRESSION_BS , ( EXPRESSION_BS > CO_TARGET_BS ? ">" : EXPRESSION_BS < CO_TARGET_BS ? "<" : "=" ) , CO_TARGET_BS ); \
      if( EXPRESSION_BS DESIRED_INEQUALITY CO_TARGET_BS ){		\
	CERR( "二分探索成功：" , #ANSWER , ":=" , ANSWER );		\
      } else {								\
	CERR( "二分探索失敗：" , #ANSWER , ":=" , #MAXIMUM , "+ 1 =" , MAXIMUM + 1 ); \
	CERR( "単調でないか、単調増加性と単調減少性を逆にしてしまったか、探索範囲内に解が存在しません。" ); \
	ANSWER = MAXIMUM + 1;						\
      }									\
    }									\
  }									\

// 単調増加の時にEXPRESSION >= CO_TARGETの最小解を格納。
#define BS1( ANSWER , MINIMUM , MAXIMUM , EXPRESSION , CO_TARGET ) BS( ANSWER , MINIMUM , MAXIMUM , EXPRESSION , >= , CO_TARGET , >= , ANSWER , ANSWER + 1 , ( L_BS + U_BS ) / 2 )
// 単調増加の時にEXPRESSION <= CO_TARGETの最大解を格納。
#define BS2( ANSWER , MINIMUM , MAXIMUM , EXPRESSION , CO_TARGET ) BS( ANSWER , MINIMUM , MAXIMUM , EXPRESSION , <= , CO_TARGET , > , ANSWER - 1 , ANSWER , ( L_BS + 1 + U_BS ) / 2 )
// 単調減少の時にEXPRESSION >= CO_TARGETの最大解を格納。
#define BS3( ANSWER , MINIMUM , MAXIMUM , EXPRESSION , CO_TARGET ) BS( ANSWER , MINIMUM , MAXIMUM , EXPRESSION , >= , CO_TARGET , < , ANSWER - 1 , ANSWER , ( L_BS + 1 + U_BS ) / 2 )
// 単調減少の時にEXPRESSION <= CO_TARGETの最小解を格納。
#define BS4( ANSWER , MINIMUM , MAXIMUM , EXPRESSION , CO_TARGET ) BS( ANSWER , MINIMUM , MAXIMUM , EXPRESSION , <= , CO_TARGET , <= , ANSWER , ANSWER + 1 , ( L_BS + U_BS ) / 2 )
// t以下の値が存在すればその最大値のiterator、存在しなければend()を返す。
TE <TY T> IN TY set<T>::iterator MaximumLeq(set<T>& S,CO T& t){CO auto EN = S.EN();if(S.empty()){RE EN;}auto IT = S.upper_bound(t);RE IT == EN?S.find(*(S.rBE())):IT == S.BE()?EN:--IT;}
// t未満の値が存在すればその最大値のiterator、存在しなければend()を返す。
TE <TY T> IN TY set<T>::iterator MaximumLt(set<T>& S,CO T& t){CO auto EN = S.EN();if(S.empty()){RE EN;}auto IT = S.lower_bound(t);RE IT == EN?S.find(*(S.rBE())):IT == S.BE()?EN:--IT;}
// t以上の値が存在すればその最小値のiterator、存在しなければend()を返す。
TE <TY T> IN TY set<T>::iterator MinimumGeq(set<T>& S,CO T& t){RE S.lower_bound(t);}
// tより大きい値が存在すればその最小値のiterator、存在しなければend()を返す。
TE <TY T> IN TY set<T>::iterator MinimumGt(set<T>& S,CO T& t){RE S.upper_bound(t);}

// 尺取り法用
// VAR_TPAがINITからUPDATEを繰り返しCONTINUE_CONDITIONを満たす限り、ON_CONDITIONを判定して
// trueならON、falseならOFFとなる。直近のONの区間を[VAR_TPA_L,VAR_TPA_R)で管理する。
#define TPA( VAR_TPA , INIT , UPDATE , CONTINUE_CONDITION , ON_CONDITION , ONON , ONOFF , OFFON , OFFOFF , FINISH ) \
  {									\
    auto VAR_TPA = INIT;						\
    auto VAR_TPA ## _L = VAR_TPA;					\
    auto VAR_TPA ## _R = VAR_TPA;					\
    bool on_TPA = false;						\
    int state_TPA = 3;							\
    WH( CONTINUE_CONDITION ){						\
      bool on_TPA_next = ON_CONDITION;					\
      state_TPA = ( ( on_TPA ? 1 : 0 ) << 1 ) | ( on_TPA_next ? 1 : 0 ); \
      CERR( "尺取り中： [L,R) = [" , VAR_TPA ## _L , "," , VAR_TPA ## _R , ") ," , #VAR_TPA , "=" , VAR_TPA , "," , ( ( state_TPA >> 1 ) & 1 ) == 1 ? "on" : "off" , " ->" , ( state_TPA & 1 ) == 1 ? "on" : "off" ); \
      if( state_TPA == 0 ){						\
	OFFOFF; VAR_TPA ## _L = VAR_TPA ## _R = VAR_TPA; UPDATE;	\
      } else if( state_TPA == 1 ){					\
	OFFON; VAR_TPA ## _L = VAR_TPA; UPDATE; VAR_TPA ## _R = VAR_TPA; \
      } else if( state_TPA == 2 ){					\
	ONOFF; VAR_TPA ## _L = VAR_TPA ## _R = VAR_TPA; UPDATE;		\
      } else {								\
	ONON; UPDATE; VAR_TPA ## _R = VAR_TPA;				\
      }									\
      on_TPA = on_TPA_next;						\
    }									\
    CERR( "尺取り終了： [L,R) = [" , VAR_TPA ## _L , "," , VAR_TPA ## _R , ") ," , #VAR_TPA , "=" , VAR_TPA ); \
    FINISH;								\
  }									\

// データ構造用
TE <TY T,TE <TY...> TY V> IN auto OP+(CO V<T>& a0,CO V<T>& a1)-> decldecay_t((declval<V<T>>().push_back(declval<T>()),a0)){if(a0.empty()){RE a1;}if(a1.empty()){RE a0;}AS(a0.SZ()== a1.SZ());V<T> AN{};for(auto IT0 = a0.BE(),IT1 = a1.BE(),EN0 = a0.EN();IT0 != EN0;IT0++,IT1++){AN.push_back(*IT0 + *IT1);}RE AN;}
TE <TY T,TY U> IN pair<T,U> OP+(CO pair<T,U>& t0,CO pair<T,U>& t1){RE{t0.first + t1.first,t0.second + t1.second};}
TE <TY T,TY U,TY V> IN tuple<T,U,V> OP+(CO tuple<T,U,V>& t0,CO tuple<T,U,V>& t1){RE{get<0>(t0)+ get<0>(t1),get<1>(t0)+ get<1>(t1),get<2>(t0)+ get<2>(t1)};}
TE <TY T,TY U,TY V,TY W> IN tuple<T,U,V,W> OP+(CO tuple<T,U,V,W>& t0,CO tuple<T,U,V,W>& t1){RE{get<0>(t0)+ get<0>(t1),get<1>(t0)+ get<1>(t1),get<2>(t0)+ get<2>(t1),get<3>(t0)+ get<3>(t1)};}
TE <TY T> IN T Addition(CO T& t0,CO T& t1){RE t0 + t1;}
TE <TY T> IN T Xor(CO T& t0,CO T& t1){RE t0 ^ t1;}
TE <TY T> IN T MU(CO T& t0,CO T& t1){RE t0 * t1;}
TE <TY T> IN CO T& Zero(){ST CO T z{};RE z;}
TE <TY T> IN CO T& One(){ST CO T o = 1;RE o;}TE <TY T> IN T AdditionInv(CO T& t){RE -t;}
TE <TY T> IN T Id(CO T& v){RE v;}
TE <TY T> IN T Min(CO T& a,CO T& b){RE a < b?a:b;}
TE <TY T> IN T Max(CO T& a,CO T& b){RE a < b?b:a;}

// グラフ用
TE <TY T,TE <TY...> TY V> IN auto Get(CO V<T>& a){RE[&](CRI i = 0){RE a[i];};}
TE <TY T = int> IN VE<T> id(CRI SZ){VE<T> AN(SZ);FOR(i,0,SZ){AN[i]= i;}RE AN;}

// グリッド問題用
int H,W,H_minus,W_minus,HW;
VE<string> wall_str;VE<VE<bool> > non_wall;
char walkable = '.',unwalkable = '#';
IN T2<int> EnumHW(CRI v){RE{v / W,v % W};}
IN int EnumHW_inv(CO T2<int>& ij){auto&[i,j]= ij;RE i * W + j;}
CO string direction[4]={"U","R","D","L"};
IN int DirectionNumberOnGrid(CRI i,CRI j,CRI k,CRI h){RE i<k?2:i>k?0:j<h?1:j>h?3:(AS(false),-1);}
IN int DirectionNumberOnGrid(CRI v,CRI w){auto[i,j]=EnumHW(v);auto[k,h]=EnumHW(w);RE DirectionNumberOnGrid(i,j,k,h);}
IN int ReverseDirectionNumberOnGrid(CRI n){AS(0<=n&&n<4);RE(n+2)%4;}
IN VE<int> EdgeOnGrid(CRI v){VE<int>AN{};auto[i,j]=EnumHW(v);if(i>0&&wall_str[i-1][j]==walkable){AN.push_back(EnumHW_inv({i-1,j}));}if(i+1<H&&wall_str[i+1][j]==walkable){AN.push_back(EnumHW_inv({i+1,j}));}if(j>0&&wall_str[i][j-1]==walkable){AN.push_back(EnumHW_inv({i,j-1}));}if(j+1<W&&wall_str[i][j+1]==walkable){AN.push_back(EnumHW_inv({i,j+1}));}RE AN;}
IN VE<path> WeightedEdgeOnGrid(CRI v){VE<path>AN{};auto[i,j]=EnumHW(v);if(i>0&&wall_str[i-1][j]==walkable){AN.push_back({EnumHW_inv({i-1,j}),1});}if(i+1<H&&wall_str[i+1][j]==walkable){AN.push_back({EnumHW_inv({i+1,j}),1});}if(j>0&&wall_str[i][j-1]==walkable){AN.push_back({EnumHW_inv({i,j-1}),1});}if(j+1<W&&wall_str[i][j+1]==walkable){AN.push_back({EnumHW_inv({i,j+1}),1});}RE AN;}
IN VO SetWallStringOnGrid(CRI i,VE<string>& S){if(S.empty()){S.reSZ(H);}cin>>S[i];AS(int(S[i].SZ())==W);}
IN VO SetWallOnGrid(CRI i,VE<VE<bool>>& b){if(b.empty()){b.reSZ(H,VE<bool>(W));}auto&S_i=wall_str[i];auto&b_i=b[i];FOR(j,0,W){b_i[j]=S_i[j]==walkable?false:(AS(S_i[j]==unwalkable),true);}}

// デバッグ用
#ifdef DEBUG
  IN VO AlertAbort( int n ) { CERR( "abort関数が呼ばれました。assertマクロのメッセージが出力されていない場合はオーバーフローの有無を確認をしてください。" ); }
  VO AutoCheck( int& exec_mode , CO bool& use_getline );
  IN VO Solve();
  IN VO Experiment();
  IN VO SmallTest();
  IN VO RandomTest();
  ll GetRand( CRL Rand_min , CRL Rand_max );
  IN VO BreakPoint( CRI LINE ) {}
  int exec_mode;
  CEXPR( int , solve_mode , 0 );
  CEXPR( int , sample_debug_mode , 1 );
  CEXPR( int , submission_debug_mode , 2 );
  CEXPR( int , library_search_mode , 3 );
  CEXPR( int , experiment_mode , 4 );
  CEXPR( int , small_test_mode , 5 );
  CEXPR( int , random_test_mode , 6 );
  #ifdef USE_GETLINE
    CEXPR( bool , use_getline , true );
  #else
    CEXPR( bool , use_getline , false );
  #endif
#else
  ll GetRand( CRL Rand_min , CRL Rand_max ) { ll answer = time( NULL ); RE answer * rand() % ( Rand_max + 1 - Rand_min ) + Rand_min; }
#endif

// VVV 常設ライブラリは以下に挿入する。
// Map (1KB)
// c:/Users/user/Documents/Programming/Mathematics/Function/Map/compress.txt
CL is_ordered{PU:is_ordered()= delete;TE <TY T> ST CE auto Check(CO T& t)-> decltype(t < t,true_type());ST CE false_type Check(...);TE <TY T> ST CE CO bool value = is_same_v< decltype(Check(declval<T>())),true_type >;};
TE <TY T , TY U>US Map = conditional_t<is_COructible_v<unordered_map<T,int>>,unordered_map<T,U>,conditional_t<is_ordered::value<T>,map<T,U>,VO>>;

// Algebra (4KB)
// c:/Users/user/Documents/Programming/Mathematics/Algebra/compress.txt
#define DC_OF_CPOINT(POINT)IN CO U& POINT()CO NE
#define DC_OF_POINT(POINT)IN U& POINT()NE
#define DF_OF_CPOINT(POINT)TE <TY U> IN CO U& VirtualPointedSet<U>::POINT()CO NE{RE Point();}
#define DF_OF_POINT(POINT)TE <TY U> IN U& VirtualPointedSet<U>::POINT()NE{RE Point();}
TE <TY U>CL UnderlyingSet{PU:US type = U;};TE <TY U>CL VirtualPointedSet:VI PU UnderlyingSet<U>{PU:VI CO U& Point()CO NE = 0;VI U& Point()NE = 0;DC_OF_CPOINT(Unit);DC_OF_CPOINT(Zero);DC_OF_CPOINT(One);DC_OF_CPOINT(Infty);DC_OF_POINT(init);DC_OF_POINT(root);};TE <TY U>CL PointedSet:VI PU VirtualPointedSet<U>{PU:U m_b_U;IN PointedSet(CO U& b_u = U());IN CO U& Point()CO NE;IN U& Point()NE;};TE <TY U>CL VirtualNSet:VI PU UnderlyingSet<U>{PU:VI U Transfer(CO U& u)= 0;IN U Inverse(CO U& u);};TE <TY U,TY F_U>CL AbstractNSet:VI PU VirtualNSet<U>{PU:F_U m_f_U;IN AbstractNSet(F_U f_U);IN U Transfer(CO U& u);};TE <TY U>CL VirtualMagma:VI PU UnderlyingSet<U>{PU:VI U Product(CO U& u0,CO U& u1)= 0;IN U Sum(CO U& u0,CO U& u1);};TE <TY U = ll>CL AdditiveMagma:VI PU VirtualMagma<U>{PU:IN U Product(CO U& u0,CO U& u1);};TE <TY U = ll>CL MultiplicativeMagma:VI PU VirtualMagma<U>{PU:IN U Product(CO U& u0,CO U& u1);};TE <TY U,TY M_U>CL AbstractMagma:VI PU VirtualMagma<U>{PU:M_U m_m_U;IN AbstractMagma(M_U m_U);IN U Product(CO U& u0,CO U& u1);};
TE <TY U> IN PointedSet<U>::PointedSet(CO U& b_U):m_b_U(b_U){}TE <TY U> IN CO U& PointedSet<U>::Point()CO NE{RE m_b_U;}TE <TY U> IN U& PointedSet<U>::Point()NE{RE m_b_U;}DF_OF_CPOINT(Unit);DF_OF_CPOINT(Zero);DF_OF_CPOINT(One);DF_OF_CPOINT(Infty);DF_OF_POINT(init);DF_OF_POINT(root);TE <TY U,TY F_U> IN AbstractNSet<U,F_U>::AbstractNSet(F_U f_U):m_f_U(MO(f_U)){ST_AS(is_invocable_r_v<U,F_U,U>);}TE <TY U,TY F_U> IN U AbstractNSet<U,F_U>::Transfer(CO U& u){RE m_f_U(u);}TE <TY U> IN U VirtualNSet<U>::Inverse(CO U& u){RE Transfer(u);}TE <TY U,TY M_U> IN AbstractMagma<U,M_U>::AbstractMagma(M_U m_U):m_m_U(MO(m_U)){ST_AS(is_invocable_r_v<U,M_U,U,U>);}TE <TY U> IN U AdditiveMagma<U>::Product(CO U& u0,CO U& u1){RE u0 + u1;}TE <TY U> IN U MultiplicativeMagma<U>::Product(CO U& u0,CO U& u1){RE u0 * u1;}TE <TY U,TY M_U> IN U AbstractMagma<U,M_U>::Product(CO U& u0,CO U& u1){RE m_m_U(u0,u1);}TE <TY U> IN U VirtualMagma<U>::Sum(CO U& u0,CO U& u1){RE Product(u0,u1);}

TE <TY U>CL VirtualMonoid:VI PU VirtualMagma<U>,VI PU VirtualPointedSet<U>{};TE <TY U = ll>CL AdditiveMonoid:VI PU VirtualMonoid<U>,PU AdditiveMagma<U>,PU PointedSet<U>{};TE <TY U = ll>CL MultiplicativeMonoid:VI PU VirtualMonoid<U>,PU MultiplicativeMagma<U>,PU PointedSet<U>{PU:IN MultiplicativeMonoid(CO U& e_U);};TE <TY U,TY M_U>CL AbstractMonoid:VI PU VirtualMonoid<U>,PU AbstractMagma<U,M_U>,PU PointedSet<U>{PU:IN AbstractMonoid(M_U m_U,CO U& e_U);};
TE <TY U> IN MultiplicativeMonoid<U>::MultiplicativeMonoid(CO U& e_U):PointedSet<U>(e_U){}TE <TY U,TY M_U> IN AbstractMonoid<U,M_U>::AbstractMonoid(M_U m_U,CO U& e_U):AbstractMagma<U,M_U>(MO(m_U)),PointedSet<U>(e_U){}

TE <TY U>CL VirtualGroup:VI PU VirtualMonoid<U>,VI PU VirtualPointedSet<U>,VI PU VirtualNSet<U>{};TE <TY U = ll>CL AdditiveGroup:VI PU VirtualGroup<U>,PU AdditiveMonoid<U>{PU:IN U Transfer(CO U& u);};TE <TY U,TY M_U,TY I_U>CL AbstractGroup:VI PU VirtualGroup<U>,PU AbstractMonoid<U,M_U>,PU AbstractNSet<U,I_U>{PU:IN AbstractGroup(M_U m_U,CO U& e_U,I_U i_U);};
TE <TY U,TY M_U,TY I_U> IN AbstractGroup<U,M_U,I_U>::AbstractGroup(M_U m_U,CO U& e_U,I_U i_U):AbstractMonoid<U,M_U>(MO(m_U),e_U),AbstractNSet<U,I_U>(MO(i_U)){}TE <TY U> IN U AdditiveGroup<U>::Transfer(CO U& u){RE -u;}

// Graph (5KB)
// c:/Users/user/Documents/Programming/Mathematics/Geometry/Graph/compress.txt
TE <TY T,TY R1,TY R2,TY E>CL VirtualGraph:VI PU UnderlyingSet<T>{PU:VI R1 Enumeration(CRI i)= 0;IN R2 Enumeration_inv(CO T& t);TE <TY PATH> IN R2 Enumeration_inv(CO PATH& p);IN VO Reset();VI CRI SZ()CO NE = 0;VI E& edge()NE = 0;VI ret_t<E,T> Edge(CO T& t)= 0;VI IN R2 Enumeration_inv_Body(CO T& t)= 0;};TE <TY T,TY R1,TY R2,TY E>CL EdgeImplimentation:VI PU VirtualGraph<T,R1,R2,E>{PU:int m_SZ;E m_edge;IN EdgeImplimentation(CRI SZ,E edge);IN CRI SZ()CO NE;IN E& edge()NE;IN ret_t<E,T> Edge(CO T& t);};TE <TY E>CL Graph:PU EdgeImplimentation<int,CRI,CRI,E>{PU:IN Graph(CRI SZ,E edge);IN CRI Enumeration(CRI i);TE <TY F> IN Graph<F> GetGraph(F edge)CO;IN CRI Enumeration_inv_Body(CRI t);};TE <TY T,TY Enum_T,TY Enum_T_inv,TY E>CL EnumerationGraph:PU EdgeImplimentation<T,ret_t<Enum_T,int>,ret_t<Enum_T_inv,T>,E>{PU:Enum_T m_enum_T;Enum_T_inv m_enum_T_inv;IN EnumerationGraph(CRI SZ,Enum_T enum_T,Enum_T_inv enum_T_inv,E edge);IN ret_t<Enum_T,int> Enumeration(CRI i);TE <TY F> IN EnumerationGraph<T,Enum_T,Enum_T_inv,F> GetGraph(F edge)CO;IN ret_t<Enum_T_inv,T> Enumeration_inv_Body(CO T& t);};TE <TY Enum_T,TY Enum_T_inv,TY E> EnumerationGraph(CRI SZ,Enum_T enum_T,Enum_T_inv enum_T_inv,E edge)-> EnumerationGraph<decldecay_t(declval<Enum_T>()(0)),Enum_T,Enum_T_inv,E>;TE <TY T,TY E>CL MemorisationGraph:PU EdgeImplimentation<T,T,CRI,E>{PU:int m_LE;VE<T> m_memory;Map<T,int> m_memory_inv;IN MemorisationGraph(CRI SZ,E edge);IN T Enumeration(CRI i);IN VO Reset();TE <TY F> IN MemorisationGraph<T,F> GetGraph(F edge)CO;IN CRI Enumeration_inv_Body(CO T& t);};TE <TY E> MemorisationGraph(CRI SZ,E edge)-> MemorisationGraph<decldecay_t(declval<E>()().back()),E>;TE <TY E> MemorisationGraph(CRI SZ,E edge)-> MemorisationGraph<decldecay_t(get<0>(declval<E>()().back())),E>;
TE <TY T,TY R1,TY R2,TY E> IN EdgeImplimentation<T,R1,R2,E>::EdgeImplimentation(CRI SZ,E edge):m_SZ(SZ),m_edge(MO(edge)){ST_AS(is_COructible_v<T,R1> && is_COructible_v<int,R2> && is_invocable_v<E,T>);}TE <TY E> IN Graph<E>::Graph(CRI SZ,E edge):EdgeImplimentation<int,CRI,CRI,E>(SZ,MO(edge)){}TE <TY T,TY Enum_T,TY Enum_T_inv,TY E> IN EnumerationGraph<T,Enum_T,Enum_T_inv,E>::EnumerationGraph(CRI SZ,Enum_T enum_T,Enum_T_inv enum_T_inv,E edge):EdgeImplimentation<T,ret_t<Enum_T,int>,ret_t<Enum_T_inv,T>,E>(SZ,MO(edge)),m_enum_T(MO(enum_T)),m_enum_T_inv(MO(enum_T_inv)){}TE <TY T,TY E> IN MemorisationGraph<T,E>::MemorisationGraph(CRI SZ,E edge):EdgeImplimentation<T,T,CRI,E>(SZ,MO(edge)),m_LE(),m_memory(),m_memory_inv(){ST_AS(is_invocable_v<E> && is_invocable_v<E,T>);}TE <TY E> IN CRI Graph<E>::Enumeration(CRI i){RE i;}TE <TY T,TY Enum_T,TY Enum_T_inv,TY E> IN ret_t<Enum_T,int> EnumerationGraph<T,Enum_T,Enum_T_inv,E>::Enumeration(CRI i){RE m_enum_T(i);}TE <TY T,TY E> IN T MemorisationGraph<T,E>::Enumeration(CRI i){AS(0 <= i && i < m_LE);RE m_memory[i];}TE <TY T,TY R1,TY R2,TY E> IN R2 VirtualGraph<T,R1,R2,E>::Enumeration_inv(CO T& t){RE Enumeration_inv_Body(t);}TE <TY T,TY R1,TY R2,TY E> TE <TY PATH> IN R2 VirtualGraph<T,R1,R2,E>::Enumeration_inv(CO PATH& p){RE Enumeration_inv_Body(get<0>(p));}TE <TY E> IN CRI Graph<E>::Enumeration_inv_Body(CRI i){RE i;}TE <TY T,TY Enum_T,TY Enum_T_inv,TY E> IN ret_t<Enum_T_inv,T> EnumerationGraph<T,Enum_T,Enum_T_inv,E>::Enumeration_inv_Body(CO T& t){RE m_enum_T_inv(t);}TE <TY T,TY E> IN CRI MemorisationGraph<T,E>::Enumeration_inv_Body(CO T& t){if(m_memory_inv.count(t)== 0){AS(m_LE < TH->SZ());m_memory.push_back(t);RE m_memory_inv[t]= m_LE++;}RE m_memory_inv[t];}TE <TY T,TY R1,TY R2,TY E> VO VirtualGraph<T,R1,R2,E>::Reset(){}TE <TY T,TY E> IN VO MemorisationGraph<T,E>::Reset(){m_LE = 0;m_memory.clear();m_memory_inv.clear();}TE <TY T,TY R1,TY R2,TY E> IN CRI EdgeImplimentation<T,R1,R2,E>::SZ()CO NE{RE m_SZ;}TE <TY T,TY R1,TY R2,TY E> IN E& EdgeImplimentation<T,R1,R2,E>::edge()NE{RE m_edge;}TE <TY T,TY R1,TY R2,TY E> IN ret_t<E,T> EdgeImplimentation<T,R1,R2,E>::Edge(CO T& t){RE m_edge(t);}TE <TY E> TE <TY F> IN Graph<F> Graph<E>::GetGraph(F edge)CO{RE Graph<F>(TH->SZ(),MO(edge));}TE <TY T,TY Enum_T,TY Enum_T_inv,TY E> TE <TY F> IN EnumerationGraph<T,Enum_T,Enum_T_inv,F> EnumerationGraph<T,Enum_T,Enum_T_inv,E>::GetGraph(F edge)CO{RE EnumerationGraph<T,Enum_T,Enum_T_inv,F>(TH->SZ(),m_enum_T,m_enum_T_inv,MO(edge));}TE <TY T,TY E> TE <TY F> IN MemorisationGraph<T,F> MemorisationGraph<T,E>::GetGraph(F edge)CO{RE MemorisationGraph<T,F>(TH->SZ(),MO(edge));}

// ConstexprModulo (7KB)
// c:/Users/user/Documents/Programming/Mathematics/Arithmetic/Mod/ConstexprModulo/compress.txt
#define RP Represent
#define DeRP Derepresent
CEXPR(uint,P,998244353);
TE <uint M,TY INT> CE INT RS(INT n)NE{RE MO(n < 0?((((++n)*= -1)%= M)*= -1)+= M - 1:n < INT(M)?n:n %= M);}TE <TY INT> CE INT& RSP(INT& n)NE{CE CO uint trunc =(1 << 23)- 1;INT n_u = n >> 23;n &= trunc;INT n_uq =(n_u / 7)/ 17;n_u -= n_uq * 119;n += n_u << 23;RE n < n_uq?n += P - n_uq:n -= n_uq;}
#define DC_OF_CM_FOR_MOD(OPR)CE bool OP OPR(CO Mod<M>& n)CO NE
#define DC_OF_AR_FOR_MOD(OPR,EX)CE Mod<M> OP OPR(Mod<M> n)CO EX;
#define DF_OF_CM_FOR_MOD(OPR)TE <uint M> CE bool Mod<M>::OP OPR(CO Mod<M>& n)CO NE{RE m_n OPR n.m_n;}
#define DF_OF_AR_FOR_MOD(OPR,EX,LEFT,OPR2)TE <uint M> CE Mod<M> Mod<M>::OP OPR(Mod<M> n)CO EX{RE MO(LEFT OPR2 ## = *TH);}TE <uint M,TY T> CE Mod<M> OP OPR(T n0,CO Mod<M>& n1)EX{RE MO(Mod<M>(MO(n0))OPR ## = n1);}
TE <uint M>CL Mod{PU:uint m_n;CE Mod()NE;CE Mod(CO Mod<M>& n)NE;CE Mod(Mod<M>&& n)NE;TE <TY T> CE Mod(T n)NE;CE Mod<M>& OP=(Mod<M> n)NE;CE Mod<M>& OP+=(CO Mod<M>& n)NE;CE Mod<M>& OP-=(CO Mod<M>& n)NE;CE Mod<M>& OP*=(CO Mod<M>& n)NE;IN Mod<M>& OP/=(Mod<M> n);TE <TY INT> CE Mod<M>& OP<<=(INT n);TE <TY INT> CE Mod<M>& OP>>=(INT n);CE Mod<M>& OP++()NE;CE Mod<M> OP++(int)NE;CE Mod<M>& OP--()NE;CE Mod<M> OP--(int)NE;DC_OF_CM_FOR_MOD(==);DC_OF_CM_FOR_MOD(!=);DC_OF_CM_FOR_MOD(<);DC_OF_CM_FOR_MOD(<=);DC_OF_CM_FOR_MOD(>);DC_OF_CM_FOR_MOD(>=);DC_OF_AR_FOR_MOD(+,NE);DC_OF_AR_FOR_MOD(-,NE);DC_OF_AR_FOR_MOD(*,NE);DC_OF_AR_FOR_MOD(/,);TE <TY INT> CE Mod<M> OP^(INT EX)CO;TE <TY INT> CE Mod<M> OP<<(INT n)CO;TE <TY INT> CE Mod<M> OP>>(INT n)CO;CE Mod<M> OP-()CO NE;CE Mod<M>& SignInvert()NE;IN Mod<M>& Invert();TE <TY INT> CE Mod<M>& PW(INT EX);CE VO swap(Mod<M>& n)NE;CE CO uint& RP()CO NE;ST CE Mod<M> DeRP(CO uint& n)NE;ST IN CO Mod<M>& Inverse(CO uint& n);ST IN CO Mod<M>& Factorial(CO uint& n);ST IN CO Mod<M>& FactorialInverse(CO uint& n);ST IN Mod<M> Combination(CO uint& n,CO uint& i);ST IN CO Mod<M>& zero()NE;ST IN CO Mod<M>& one()NE;TE <TY INT> CE Mod<M>& PositivePW(INT EX)NE;TE <TY INT> CE Mod<M>& NonNegativePW(INT EX)NE;TE <TY T> CE Mod<M>& Ref(T&& n)NE;ST CE uint& Normalise(uint& n)NE;};
US MP = Mod<P>;
TE <uint M> CL Mod;TE <uint M>CL COantsForMod{PU:COantsForMod()= delete;ST CE CO uint g_memory_bound =
#ifdef DEBUG
1e3;
#else
1e6;
#endif
ST CE CO uint g_memory_LE = M < g_memory_bound?M:g_memory_bound;ST CE uint g_M_minus = M - 1;ST CE uint g_M_minus_2 = M - 2;ST CE uint g_M_minus_2_neg = 2 - M;};
TE <uint M> CE Mod<M>::Mod()NE:m_n(){}TE <uint M> CE Mod<M>::Mod(CO Mod<M>& n)NE:m_n(n.m_n){}TE <uint M> CE Mod<M>::Mod(Mod<M>&& n)NE:m_n(MO(n.m_n)){}TE <uint M> TE <TY T> CE Mod<M>::Mod(T n)NE:m_n(RS<M>(MO(n))){ST_AS(is_COructible_v<uint,decay_t<T> >);}TE <uint M> CE Mod<M>& Mod<M>::OP=(Mod<M> n)NE{RE Ref(m_n = MO(n.m_n));}TE <uint M> CE Mod<M>& Mod<M>::OP+=(CO Mod<M>& n)NE{RE Ref(Normalise(m_n += n.m_n));}TE <uint M> CE Mod<M>& Mod<M>::OP-=(CO Mod<M>& n)NE{RE Ref(m_n < n.m_n?(m_n += M)-= n.m_n:m_n -= n.m_n);}TE <uint M> CE Mod<M>& Mod<M>::OP*=(CO Mod<M>& n)NE{RE Ref(m_n = RS<M>(ull(m_n)* n.m_n));}TE <> CE MP& MP::OP*=(CO MP& n)NE{ull m_n_copy = m_n;RE Ref(m_n = MO((m_n_copy *= n.m_n)< P?m_n_copy:RSP(m_n_copy)));}TE <uint M> IN Mod<M>& Mod<M>::OP/=(Mod<M> n){RE OP*=(n.Invert());}TE <uint M> TE <TY INT> CE Mod<M>& Mod<M>::OP<<=(INT n){AS(n >= 0);RE *TH *= Mod<M>(2).NonNegativePW(MO(n));}TE <uint M> TE <TY INT> CE Mod<M>& Mod<M>::OP>>=(INT n){AS(n >=0);WH(n-- > 0){((m_n & 1)== 0?m_n:m_n += M)>>= 1;}RE *TH;}TE <uint M> CE Mod<M>& Mod<M>::OP++()NE{RE Ref(m_n < COantsForMod<M>::g_M_minus?++m_n:m_n = 0);}TE <uint M> CE Mod<M> Mod<M>::OP++(int)NE{Mod<M> n{*TH};OP++();RE n;}TE <uint M> CE Mod<M>& Mod<M>::OP--()NE{RE Ref(m_n == 0?m_n = COantsForMod<M>::g_M_minus:--m_n);}TE <uint M> CE Mod<M> Mod<M>::OP--(int)NE{Mod<M> n{*TH};OP--();RE n;}DF_OF_CM_FOR_MOD(==);DF_OF_CM_FOR_MOD(!=);DF_OF_CM_FOR_MOD(>);DF_OF_CM_FOR_MOD(>=);DF_OF_CM_FOR_MOD(<);DF_OF_CM_FOR_MOD(<=);DF_OF_AR_FOR_MOD(+,NE,n,+);DF_OF_AR_FOR_MOD(-,NE,n.SignInvert(),+);DF_OF_AR_FOR_MOD(*,NE,n,*);DF_OF_AR_FOR_MOD(/,,n.Invert(),*);TE <uint M> TE <TY INT> CE Mod<M> Mod<M>::OP^(INT EX)CO{RE MO(Mod<M>(*TH).PW(MO(EX)));}TE <uint M> TE <TY INT> CE Mod<M> Mod<M>::OP<<(INT n)CO{RE MO(Mod<M>(*TH)<<= MO(n));}TE <uint M> TE <TY INT> CE Mod<M> Mod<M>::OP>>(INT n)CO{RE MO(Mod<M>(*TH)>>= MO(n));}TE <uint M> CE Mod<M> Mod<M>::OP-()CO NE{RE MO(Mod<M>(*TH).SignInvert());}TE <uint M> CE Mod<M>& Mod<M>::SignInvert()NE{RE Ref(m_n > 0?m_n = M - m_n:m_n);}TE <uint M> IN Mod<M>& Mod<M>::Invert(){AS(m_n != 0);uint m_n_neg;RE m_n < COantsForMod<M>::g_memory_LE?Ref(m_n = Inverse(m_n).m_n):((m_n_neg = M - m_n)< COantsForMod<M>::g_memory_LE)?Ref(m_n = M - Inverse(m_n_neg).m_n):PositivePW(uint(COantsForMod<M>::g_M_minus_2));}TE <uint M> TE <TY INT> CE Mod<M>& Mod<M>::PositivePW(INT EX)NE{Mod<M> PW{*TH};EX--;WH(EX != 0){(EX & 1)== 1?*TH *= PW:*TH;EX >>= 1;PW *= PW;}RE *TH;}TE <uint M> TE <TY INT> CE Mod<M>& Mod<M>::NonNegativePW(INT EX)NE{RE EX == 0?Ref(m_n = 1):Ref(PositivePW(MO(EX)));}TE <uint M> TE <TY INT> CE Mod<M>& Mod<M>::PW(INT EX){bool neg = EX < 0;AS(!(neg && m_n == 0));RE neg?PositivePW(MO(EX *= COantsForMod<M>::g_M_minus_2_neg)):NonNegativePW(MO(EX));}TE <uint M> CE VO Mod<M>::swap(Mod<M>& n)NE{std::swap(m_n,n.m_n);}TE <uint M> IN CO Mod<M>& Mod<M>::Inverse(CO uint& n){AS(n < COantsForMod<M>::g_memory_LE);ST Mod<M> memory[COantsForMod<M>::g_memory_LE]={zero(),one()};ST uint LE_curr = 2;WH(LE_curr <= n){memory[LE_curr].m_n = M - memory[M % LE_curr].m_n * ull(M / LE_curr)% M;LE_curr++;}RE memory[n];}TE <uint M> IN CO Mod<M>& Mod<M>::Factorial(CO uint& n){AS(n < COantsForMod<M>::g_memory_LE);ST Mod<M> memory[COantsForMod<M>::g_memory_LE]={one(),one()};ST uint LE_curr = 2;WH(LE_curr <= n){(memory[LE_curr]= memory[LE_curr - 1])*= LE_curr;LE_curr++;}RE memory[n];}TE <uint M> IN CO Mod<M>& Mod<M>::FactorialInverse(CO uint& n){ST Mod<M> memory[COantsForMod<M>::g_memory_LE]={one(),one()};ST uint LE_curr = 2;WH(LE_curr <= n){(memory[LE_curr]= memory[LE_curr - 1])*= Inverse(LE_curr);LE_curr++;}RE memory[n];}TE <uint M> IN Mod<M> Mod<M>::Combination(CO uint& n,CO uint& i){RE i <= n?Factorial(n)* FactorialInverse(i)* FactorialInverse(n - i):zero();}TE <uint M> CE CO uint& Mod<M>::RP()CO NE{RE m_n;}TE <uint M> CE Mod<M> Mod<M>::DeRP(CO uint& n)NE{Mod<M> n_copy{};n_copy.m_n = n;RE n_copy;}TE <uint M> IN CO Mod<M>& Mod<M>::zero()NE{ST CE CO Mod<M> z{};RE z;}TE <uint M> IN CO Mod<M>& Mod<M>::one()NE{ST CE CO Mod<M> o{1};RE o;}TE <uint M> TE <TY T> CE Mod<M>& Mod<M>::Ref(T&& n)NE{RE *TH;}TE <uint M> CE uint& Mod<M>::Normalise(uint& n)NE{RE n < M?n:n -= M;}TE <uint M> IN Mod<M> Inverse(CO Mod<M>& n){RE MO(Mod<M>(n).Invert());}TE <uint M> CE Mod<M> Inverse_CE(Mod<M> n)NE{RE MO(n.NonNegativePW(M - 2));}TE <uint M,TY INT> CE Mod<M> PW(Mod<M> n,INT EX){RE MO(n.PW(MO(EX)));}TE <uint M> CE VO swap(Mod<M>& n0,Mod<M>& n1)NE{n0.swap(n1);}TE <uint M> IN string to_string(CO Mod<M>& n)NE{RE to_string(n.RP())+ " + " + to_string(M)+ "Z";}TE <uint M,CL Traits> IN basic_istream<char,Traits>& OP>>(basic_istream<char,Traits>& is,Mod<M>& n){ll m;is >> m;n = m;RE is;}TE <uint M,CL Traits> IN basic_ostream<char,Traits>& OP<<(basic_ostream<char,Traits>& os,CO Mod<M>& n){RE os << n.RP();}
// AAA 常設ライブラリは以上に挿入する。

#define INCLUDE_LIBRARY
#include __FILE__

#endif // INCLUDE_LIBRARY

#endif // INCLUDE_SUB

#endif // INCLUDE_MAIN
yukicoder

結果

テストケース

ソースコード
