← 返回题目列表

AtCoder · APG4b_ah

AH. 4.03。模板

中文机器翻译题面 查看原题
page_url:https://atcoder.jp/contests/apg4b/tasks/APG4b_ah 任务ID:3780

上一页 | 下一页

关键点

  • tenpureto是用来将结构体和函数的“类型”进行泛化的功能
  • “执行类似处理但使用的类型不同的函数”函数模板可以使用[函数模板]来进行汇总
  • 函数模板的声明
template <typename テンプレート引数>
返り値の型 関数名(引数の型 引数...(必要な分だけ書く)) {
  // 処理内容
}
  • 作为函数模板定义的函数的调用
関数名<テンプレート引数>(引数1, 引数2, ...);
  • 当想将结构体的内容对类型进行泛化时克拉斯滕普雷特可以使用[函数模板]来进行汇总
  • 结构体模板的声明
template <typename テンプレート引数>
struct 構造体名 {
  (構造体の内容)
};
  • 使用通过结构体模板所定义的结构体
構造体名<テンプレート引数>

tenpureto

tenpureto是将结构体和函数的“类型”进行泛化的功能。

函数模板

函数模板使用[函数模板]可以将“执行类似处理但使用的类型不同的函数”进行汇总。

例如,到目前为止处理过的STL的min函数以及max函数是使用函数模板定义的。 以此int也适用于类型double类型也可以以同样的方式使用。

下面,我们将解释``某个数值。 我将以x并将其平方作为示例来解释函数。 待处理的号码int类型可以写成如下。 假设您还需要一个对

// xの二乗を返す
int square_int(int x) {
  return x * x;
}

这里 如果你知道double类型的数字进行平方的函数。

// xの二乗を返す
double square_double(double x) {
  return x * x;
}

这些函数可以按如下方式使用。

#include <bits/stdc++.h>
using namespace std;

// xの二乗を返す
int square_int(int x) {
  return x * x;
}

// xの二乗を返す
double square_double(double x) {
  return x * x;
}

int main() {
  int a = 3;
  double b = 1.2;

  cout << square_int(a) << endl;
  cout << square_double(b) << endl;
}
実行結果
9
1.44

两个函数(int和double) 函数内部的描述完全一样。

通过使用函数模板功能,可以将这种“不同类型的相同处理”组合在一起。 具体来说,定义该函数如下。

// xの二乗を返す (関数テンプレート版)
template <typename T>
T square(T x) {
  return x * x;
}

调用该函数的一方如下。

#include <bits/stdc++.h>
using namespace std;

// xの二乗を返す (関数テンプレート版)
template <typename T>
T square(T x) {
  return x * x;
}

int main() {
  int a = 3;
  double b = 1.2;

  cout << square<int>(a) << endl;  // int版のsquare関数の呼び出し
  cout << square<double>(b) << endl;  // double版のsquare関数の呼び出し
}
実行結果
9
1.44

函数模板的声明

声明函数模板:

template <typename テンプレート引数>
返り値の型 関数名(引数の型 引数...(必要な分だけ書く)) {
  // 処理内容
}

模板可以用作返回值类型和参数类型。

还可以采用多个模板参数。

template <typename テンプレート引数1, typename テンプレート引数2, ...(必要な分だけ書く)>
返り値の型 関数名(引数の型 引数...(必要な分だけ書く)) {
  // 処理内容
}

结构成员函数也可以使用模板来定义。

使用模板调用函数

要调用使用模板定义的函数,请编写如下。

関数名<テンプレート引数>(引数1, 引数2, ...);

仅传递定义函数时指定的模板参数数量。

関数名<テンプレート引数1, テンプレート引数2, ...(必要な分だけ書く)>(引数1, 引数2, ...);

但是,如果可以推断参数的类型,则可以省略指定模板参数。

#include <bits/stdc++.h>
using namespace std;

// xの二乗を返す (関数テンプレート版)
template <typename T>
T square(T x) {
  return x * x;
}

int main() {
  int a = 3;
  double b = 1.2;

  // テンプレート引数の省略
  cout << square(a) << endl;  // int版のsquare関数の呼び出し
  cout << square(b) << endl;  // double版のsquare関数の呼び出し
}
実行結果
9
1.44

结构中的模板

定义结构时也可以使用模板。 例如,这允许您创建一个概括成员变量类型的结构。

创建一个这样的模板 它被称为克拉斯滕普雷特。

到目前为止我已经处理过了。 STL 容器(例如vector呀 您可以前往map)是使用类模板定义的。 例如vector是vector<int>作为int类型数组, 您现在可以使用vector<double>作为double类型的数组。

以下程序使用模板概括坐标值类型。 这是Point结构的示例。

#include <bits/stdc++.h>
using namespace std;

// クラステンプレートの宣言
template <typename T>
struct Point {
  T x;
  T y;
  void print() {
    cout << "(" << x << ", " << y << ")" << endl;
  }
};

int main() {
  // int型用のPoint構造体
  Point<int> p1 = { 0, 1 };
  p1.print(); // (0, 1)

  // double型用のPoint構造体
  Point<double> p2 = { 2.3, 4.5 };
  p2.print(); // (2.3, 4.5)
}
実行結果
(0, 1)
(2.3, 4.5)

结构体模板的声明

声明类模板:

template <typename テンプレート引数>
struct 構造体名 {
  (構造体の内容)
};

模板参数可以用作类型名称。

还可以采用多个模板参数。

template <typename テンプレート引数1, テンプレート引数2, ...(必要な分だけ書く)>
struct 構造体名 {
  (構造体の内容)
};

将结构与模板一起使用

要使用通过类模板定义的结构,请指定模板参数,如下所示。

構造体名<テンプレート引数>

在上面的示例程序中, 对于类模板Point它用作Point<int>呀 您可以前往Point<double>。

与函数模板不同,类模板的模板参数不能省略。

常量模板

函数模板和类模板都可以使用整数类型常量作为模板参数。 当您想要根据常量的使用位置切换常量时,常量模板非常有用。

作为一个具体示例, 用于获取tuple的元素get功能,bitset使用常量模板定义。

以下是在函数模板中采用常量模板参数的示例。

#include <bits/stdc++.h>
using namespace std;

// タプルのINDEX1番目とINDEX2番目を交換する関数
template <int INDEX1, int INDEX2>
void tuple_swap(tuple<int, int, int> &x) {
  swap(get<INDEX1>(x), get<INDEX2>(x));
}

int main() {
  tuple<int, int, int> x = make_tuple(1, 2, 3);

  tuple_swap<0, 2>(x);  // 1番目と3番目を交換
  cout << get<0>(x) << ", " << get<1>(x) << ", " << get<2>(x) << endl;

  tuple_swap<0, 1>(x);  // 1番目と2番目を交換
  cout << get<0>(x) << ", " << get<1>(x) << ", " << get<2>(x) << endl;
}
実行結果
3, 2, 1
2, 3, 1

以下是在类模板中采用常量模板参数的示例。

#include <bits/stdc++.h>
using namespace std;

// MODで割ったあまりをxに保持するような構造体
template <int MOD>
struct Modulo {
  int x;
  Modulo() {}
  Modulo(int v) {
    x = v % MOD;
  }
  // 足し算
  Modulo plus(Modulo<MOD> other) {
    Modulo<MOD> result;
    result.x = (x + other.x) % MOD;
    return result;
  }
};

int main() {
  Modulo<10> a(7), b(5);
  auto c = a.plus(b);
  cout << c.x << endl; // (5 + 7) % 10 == 2
}
実行結果
2

注意点

即使使用模板声明的函数或结构存在拼写错误或其他错误,如果程序中未使用它,也可能不会导致编译错误。 这是因为未使用的模板在编译期间会被忽略。欲了解更多信息,请阅读“关于详细故事的具体化”。

下面是函数模板中的拼写错误示例。

#include <bits/stdc++.h>
using namespace std;

struct X {
  void f() {
    cout << "X" << endl;
  }
};
struct Y {
  void f() {
    cout << "Y" << endl;
  }
};

template <typename T>
void call_f(T v) {
  v.g();  // fと書くところをタイプミスでgと書いてしまったとする
}

int main() {
  // 何もしない
}
実行結果

函数call_f原本准备调用函数f,但由于拼写错误,它正在尝试调用函数g。 虽然有错字,但是程序中并没有使用这个模板,所以编译运行正常。

如下, 当我尝试使用call_f时,我第一次收到编译错误。

#include <bits/stdc++.h>
using namespace std;

struct X {
  void f() {
    cout << "X" << endl;
  }
};
struct Y {
  void f() {
    cout << "Y" << endl;
  }
};

template <typename T>
void call_f(T v) {
  v.g();  // fと書くところをタイプミスでgと書いてしまったとする
}

int main() {
  X x;
  Y y;
  call_f(x); // Xのfが呼ばれてほしい
  call_f(y); // Yのfが呼ばれてほしい
}
标准错误输出
./Main.cpp: In instantiation of ‘void call_f(T) [with T = X]’:
./Main.cpp:23:11:   required from here
./Main.cpp:17:3: error: ‘struct X’ has no member named ‘g’
   v.g();  // fと書くところをタイプミスでgと書いてしまったとする
   ^
./Main.cpp: In instantiation of ‘void call_f(T) [with T = Y]’:
./Main.cpp:24:11:   required from here
./Main.cpp:17:3: error: ‘struct Y’ has no member named ‘g’

在编写诸如“一点一点地编写,编译它,检查它......”之类的代码时,您需要小心。


详细信息

模板具体化

模板是一个有用的功能,可以概括类似的处理,但它们实际上只是“自动生成函数和结构的蓝图”。 例如,函数模板是“自动生成特定于每种类型的函数的功能。”

当使用通过模板定义的函数或结构时,实际上会生成实体(函数或结构),然后使用。 以这种方式从模板创建实体称为模板实例化。

模板实例化由编译器自动完成。 当程序中使用模板时,它会根据其使用方式进行实例化,因此 如果程序中从未使用过模板,则不会实例化该模板。

square函数示例

作为函数模板的具体示例引入。 我们以square函数为例。

#include <bits/stdc++.h>
using namespace std;

/*
// xの二乗を返す
int square_int(int x) {
  return x * x;
}

// xの二乗を返す
double square_double(double x) {
  return x * x;
}
*/

// xの二乗を返す (関数テンプレート版)
template <typename T>
T square(T x) {
  return x * x;
}

int main() {
  int a = 3;
  double b = 1.2;

  cout << square<int>(a) << endl;  // int版のsquare関数の呼び出し
  cout << square<double>(b) << endl;  // double版のsquare関数の呼び出し
}
実行結果
9
1.44

首先, 简单地声明square的函数模板不会生成任何函数。

在程序中 当像square<int>(a)这样使用时,编译器会告诉你int为 生成square関数」(square_int函数)。 同样地 当像square<double>(b)一样使用时,double的函数( 创建square_double函数)。 square<int>(a)和 我讨厌square<double>(b)各自具体化,因此程序执行时实际调用的函数是不同的。 与

但是,square<int>(2)和 我讨厌square<int>(3)相同 对于调用int类型,将调用相同的函数。 square<int>(2)通过int对类型执行物化, 对于square<int>(3),最好想象调用已经实现的东西。

竞争性编程中经常使用的模板示例

作为竞技编程中经常使用的函数模板, 介绍chmax/chmin。

#include <bits/stdc++.h>
using namespace std;

// aよりもbが大きいならばaをbで更新する
// (更新されたならばtrueを返す)
template <typename T>
bool chmax(T &a, const T& b) {
  if (a < b) {
    a = b;  // aをbで更新
    return true;
  }
  return false;
}
// aよりもbが小さいならばaをbで更新する
// (更新されたならばtrueを返す)
template <typename T>
bool chmin(T &a, const T& b) {
  if (a > b) {
    a = b;  // aをbで更新
    return true;
  }
  return false;
}

//-------------------------------

int f(int n) {
  // (適当な関数)
  return n * n - 8 * n + 3;
}

int main() {
  // f(1), f(2), ..., f(10)の最小値と最大値を求める
  int ans_min = 1000000000;
  int ans_max = 0;
  for (int i = 1; i <= 10; i++) {
    chmin(ans_min, f(i));  // 最小値を更新
    chmax(ans_max, f(i));  // 最大値を更新
  }
  cout << ans_min << " " << ans_max << endl;
}
実行結果
-13 23

chmin当找到最小值时,chmax在计算最大值时很有用。

类型名称和类别

在模板声明中typename部分是 即使你把它写成class,它也有完全相同的含义。

// template <typename T>と書くのと同じ
template <class T>
T square(T x) {
  return x * x;
}
// template <typename T>と書くのと同じ
template <class T>
struct Point {
  T x;
  T y;
  void print() {
    cout << "(" << x << ", " << y << ")" << endl;
  }
};

这里不介绍的功能

以下是本页面未介绍的一些与模板相关的功能示例。 如果您想了解更多,请查看一下。

  • 模板专业化
    • 通过专门化,可以说,“我想使用模板进行概括,但我想为特定类型提供不同的内容。”
  • 模板的显式实例化
  • 可变长度模板
  • 变量模板
  • 别名模板
  • SFINAE规则
  • 等

上一页 | 下一页