习题选自:C++ Primer Plus(第六版) 内容仅供参考,如有错误,欢迎指正 ! C++ decltype和返回类型后置 左右值引用和移动语义 C++11 新的类功能 C++11 Lambda表达式 C++11 包装器function
1. 使用用大括号括起的初始化列表语法重写下述代码。重写后的代码不应使用数组ar:
class Z200
{
private:
int j;
char ch;
double z;
public:
Z200(int jv, char chv, zv) : j(jv), ch(chv), z(zv) {}
...
};
double x = 8.8;
std::string s = "What a bracing effect!";
int k(99);
Z200 zip(200,'Z',0.675);
std::vector<int> ai(5);
int ar[5] = {3, 9, 4, 7, 1};
for (auto pt = ai.begin(), int i = 0; pt != ai.end(); ++pt, ++i)
*pt = ai[i];
重写后代码:
class Z200
{
private:
int j;
char ch;
double z;
public:
Z200(int jv, char chv, zv) : j(jv), ch(chv), z(zv) {}
...
};
double x{8.8}; // = {8.8}
std::string s{"What a bracing effect!"};
int k{99};
Z200 zip{200,'Z',0.675};
std::vector<int> ai{3, 9, 4, 7, 1};
2.2. 在下述简短的程序中,哪些函数调用不对?为什么?对于合法的函数调用,指出其引用参数指向的是什么。
#include <iostream>
using namespace std;
double up(double x) { return 2.0* x;}
void r1(const double &rx) {cout << rx << endl;}
void r2(double &rx) {cout << rx << endl;}
void r3(double &&rx) {cout << rx << endl;}
int main()
{
double w = 10.0;
r1(w);
r1(w+1);
r1(up(w));
r2(w);
r2(w+1);
r2(up(w));
r3(w);
r3(w+1);
r3(up(w));
return 0;
}
r1(w);
----合法,形参rx
指向w
。r1(w+1);
----合法,形参rx
指向一个临时变量,这个变量被初始化为w+1
。r1(up(w));
---合法,形参rx
指向一个临时变量,这个变量被初始化为up(w)
的返回值。r2(w);
---合法,形参rx
指向w
。r2(w+1)
; ---非法,因为w+1
是一个右值。r2(up(w));
---非法,因为up(w)
的返回值是一个右值。r3(w);
---非法,因为右值引用不能指向左值(如w
)。r3(w+1);
---合法,rx
指向表达式w+1
的临时拷贝。r3(up(w));
---合法,rx
指向up(w)
的临时返回值。一般而言,将左值传递给const
左值引用参数的时候,参数将被初始化为左值。将右值传递给函数时,const
左值引用参数将指向右值的临时拷贝。将左值传递给非const
左值引用参数时,参数将被初始化为左值;但非const
左值形参不能接受右值实参。
3. a. 下述简短的程序显示什么?为什么?
#include <iostream>
using namespace std;
double up(double x) { return 2.0 * x; }
void r1(const double &rx) { cout << "const double & rx\n"; }
void r1(double &rx) { cout << "double & rx\n"; }
int main() {
double w = 10.0;
r1(w);
r1(w + 1);
r1(up(w));
return 0;
}
b. 下述简短的程序显示什么?为什么?
#include <iostream>
using namespace std;
double up(double x) { return 2.0 * x; }
void r1(double &rx) { cout << "double & rx\n"; }
void r1(double &&rx) { cout << "double && rx\n"; }
int main() {
double w = 10.0;
r1(w);
r1(w + 1);
r1(up(w));
return 0;
}
c. 下述简短的程序显示什么?为什么?
#include <iostream>
using namespace std;
double up(double x) { return 2.0 * x; }
void r1(const double &rx) { cout << "const double & rx\n"; }
void r1(double &&rx) { cout << "double && rx\n"; }
int main() {
double w = 10.0;
r1(w);
r1(w + 1);
r1(up(w));
return 0;
}
a.
double & rx
const double & rx
const double & rx
非const
左值引用与左值实参匹配,因此r1(w);
调用void r1(double &rx)
。另外两个实参均为右值,const左值引用可以指向他们的拷贝。【将右值传递给函数时,const
左值引用参数将指向右值的临时拷贝。】。
b.
double & rx
double && rx
double && rx
左值引用与左值实参w
匹配,而右值引用与两个右值实参匹配。
c.
const double & rx
double && rx
double && rx
const
左值引用与左值实参w
匹配,而右值引用与两个右值实参匹配。
总之,非const
左值形参与左值实参匹配,非cosnt
右值形参与右值实参匹配;const
左值形参可以与左值或右值实参匹配。如果可供选择的话,编译器优先选择前两种方式。
4. 哪些成员函数是特殊的成员函数?它们特殊的原因是什么?
特殊成员函数:默认构造函数、复制构造函数、移动构造函数、析构函数、复制赋值运算符和移动赋值运算符。这些函数之所以特殊,是因为编译器将根据情况自动提供它们的默认版本。
5. 假设Fizzle类只有如下所示的数据成员:
class Fizzle
{
private:
double bubbles[4000];
...
};
为什么不适合给这个类定义移动构造函数?要让这个类适合定义移动构造函数,应如何修改存储4000个double值的方式?
移动构造函数是在转让数据所有权可行的时候是合适的。但对于标准数组没有转让所有权的机制,因此不适合给该类定义移动构造函数。如果Fizzle
使用指针和动态内存分配来存储这4000个double
值,即可以将数据的地址赋给新指针,以转让其所有权,则适合给Fizzle
定义移动构造函数。
6. 修改下述简短的程序,使其使用lambda表达式而不是f1( )。请不要修改show2( )。
#include <iostream>
template<typename T>
void show2(double x, T& fp) {std::cout << x << " -> " << fp(x) << '\n';}
double f1(double x) { return 1.8*x + 32;}
int main()
{
show2(18.0, f1);
return 0;
}
修改后:
#include <iostream>
template <typename T>
void show2(double x, T& fp) { std::cout << x << " -> " << fp(x) << '\n';}
//void show2(double x, T fp) {std::cout << x << " -> " << fp(x) << '\n';}
int main() {
auto f2 = [](double x) { return 1.8 * x + 32; };
show2(18.0, f2);
//show2(18.0, [](double x){return 1.8*x + 32;});
return 0;
}
7. 修改下述简短而丑陋的程序,使其使用lambda表达式而不是函数符Adder。请不要修改sum( )。
#include <iostream>
#include <array>
const int Size = 5;
template<typename T>
void sum(std::array<double,Size> a, T& fp);
class Adder
{
double tot;
public:
Adder(double q = 0) : tot(q) {}
void operator()(double w) { tot +=w;}
double tot_v () const {return tot;};
};
int main()
{
double total = 0.0;
Adder ad(total);
std::array<double, Size> temp_c = {32.1, 34.3, 37.8, 35.2, 34.7};
sum(temp_c,ad);
total = ad.tot_v();
std::cout << "total: " << ad.tot_v() << '\n';
return 0;
}
template<typename T>
void sum(std::array<double,Size> a, T& fp)
{
for(auto pt = a.begin(); pt != a.end(); ++pt)
{
fp(*pt);
}
}
修改后:
#include <array>
#include <iostream>
const int Size = 5;
template <typename T>
void sum(std::array<double, Size> a, T& fp);
int main() {
double total = 0.0;
std::array<double, Size> temp_c = {32.1, 34.3, 37.8, 35.2, 34.7};
auto f = [&total](double x) { total += x; };
sum(temp_c, f);
std::cout << "total: " << total << '\n';
return 0;
}
template <typename T>
void sum(std::array<double, Size> a, T& fp) {
for (auto pt = a.begin(); pt != a.end(); ++pt) {
fp(*pt);
}
}
1. 下面是一个简短程序的一部分:
int main()
{
using namespace std;
// list of double deduced from list contents
auto q = average_list({15.4, 10.7, 9.0});
cout << q << endl;
// list of int deduced from list contents
cout << average_list({20, 30, 19, 17, 45, 38} ) << endl;
// forced list of double
auto ad = average_list<double>({'A', 70, 65.33});
cout << ad << endl;
return 0;
}
请提供函数average_list( ),让该程序变得完整。它应该是一个模板函数,其中的类型参数指定了用作函数参数的initilize_list模板的类型以及函数的返回类型。
修改后:
#include <algorithm>
#include <initializer_list>
#include <iostream>
using namespace std;
template <typename T>
T average_list(initializer_list<T> l) {
T sum = 0;
if (l.size() == 0) return 0;
for_each(l.begin(), l.end(), [&sum](T t) { sum += t; });
return sum / l.size();
}
int main() {
using namespace std;
// list of double deduced from list contents
auto q = average_list({15.4, 10.7, 9.0});
cout << q << endl;
// list of int deduced from list contents
cout << average_list({20, 30, 19, 17, 45, 38}) << endl;
// forced list of double
auto ad = average_list<double>({'A', 70, 65.33});
cout << ad << endl;
return 0;
}
2. 下面是类Cpmv的声明:
class Cpmv
{
public:
struct Info
{
std::string qcode;
std::string zcode;
};
private:
Info *pi;
public:
Cpmv();
Cpmv(std::string q, std::string z);
Cpmv(const Cpmv & cp);
Cpmv(Cpmv && mv);
~Cpmv();
Cpmv & operator=(const Cpmv & cp);
Cpmv & operator=(Cpmv && mv);
Cpmv operator+(const Cpmv & obj) const;
void Display() const;
};
函数operator+ ( )应创建一个对象,其成员qcode和zcode有操作数的相应成员拼接而成。请提供为移动构造函数和移动赋值运算符实现移动语义的代码。编写一个使用所有这些方法的程序。为方便测试,让各个方法都显示特定的内容,以便知道它们被调用。
代码如下:
#include <iostream>
using namespace std;
class Cpmv {
public:
struct Info {
std::string qcode;
std::string zcode;
};
private:
Info *pi;
public:
Cpmv();
Cpmv(std::string q, std::string z);
Cpmv(const Cpmv &cp);
Cpmv(Cpmv &&mv);
~Cpmv();
Cpmv &operator=(const Cpmv &cp);
Cpmv &operator=(Cpmv &&mv);
Cpmv operator+(const Cpmv &obj) const;
void Display() const;
};
int main() {
Cpmv c1;
Cpmv c2("abc", "123");
Cpmv c3(c2);
c1 = c2;
c1.Display();
Cpmv c4(move(c1));
c4.Display();
Cpmv c5;
c5 = move(c2);
c5.Display();
return 0;
}
Cpmv::Cpmv() {
pi = new Info;
pi->qcode = "";
pi->zcode = "";
cout << "Cpmv() called.\n";
}
Cpmv::Cpmv(std::string q, std::string z) {
pi = new Info;
pi->qcode = q;
pi->zcode = z;
cout << "Cpmv(std::string q, std::string z) called.\n";
}
Cpmv::Cpmv(const Cpmv &cp) {
pi = new Info;
pi->qcode = cp.pi->qcode;
pi->zcode = cp.pi->zcode;
cout << "Cpmv(const Cpmv &cp) called.\n";
}
Cpmv::Cpmv(Cpmv &&mv) {
pi = mv.pi;
mv.pi = nullptr;
cout << "Cpmv(Cpmv &&mv) called.\n";
}
Cpmv::~Cpmv() {
delete pi;
cout << "~Cpmv() called.\n";
}
Cpmv &Cpmv::operator=(const Cpmv &cp) {
cout << "Cpmv &operator=(const Cpmv &cp) called.\n";
if (this == &cp) {
return *this;
}
delete pi;
pi = new Info;
pi->qcode = cp.pi->qcode;
pi->zcode = cp.pi->zcode;
return *this;
}
Cpmv &Cpmv::operator=(Cpmv &&mv) {
cout << "Cpmv &operator=(Cpmv &&mv) called.\n";
if (this == &mv) {
return *this;
}
delete pi;
pi = mv.pi;
mv.pi = nullptr;
return *this;
}
Cpmv Cpmv::operator+(const Cpmv &obj) const {
cout << "Cpmv operator+(const Cpmv &obj) called.\n";
Cpmv cv;
cv.pi->qcode = this->pi->qcode + obj.pi->qcode;
cv.pi->zcode = this->pi->zcode + obj.pi->zcode;
return cv;
}
void Cpmv::Display() const {
cout << "The qcode is " << this->pi->qcode << endl;
cout << "The zcode is " << this->pi->zcode << endl;
}
3. 编写并测试可变参数模板函数sum_value( ),它接受任意长度的参数列表(其中包含数值,但可以是任何类型),并以long double的方式返回这些数值的和。
main.cpp:
#include <iostream>
#include <string>
// definition for 1 parameter
template <typename T>
long double sum_value(const T& value) {
return value;
}
// definition for 2 or more parameters
template <typename T, typename... Args>
long double sum_value(const T& value, const Args&... args) {
return value + sum_value(args...);
}
int main() {
int n = 14;
double x = 2.71828;
std::cout << sum_value(n, x, 'a') << std::endl;
return 0;
}
4. 使用lambda重新编写程序清单16.15。具体地说,使用一个有名称的lambda替换函数outint( ),并将函数符替换为两个匿名lambda表达式。
main.cpp:
#include <algorithm>
#include <iostream>
#include <iterator>
#include <list>
auto outint_l = [](int n) { std::cout << n << " "; };
int main() {
using std::cout;
using std::endl;
using std::list;
int vals[10] = {50, 100, 90, 180, 60, 210, 415, 88, 188, 201};
list<int> yadayada(vals, vals + 10); // range constructor
list<int> etcetera(vals, vals + 10);
// C++11 can use the following instead
// list<int> yadayada = {50, 100, 90, 180, 60, 210, 415, 88, 188, 201};
// list<int> etcetera {50, 100, 90, 180, 60, 210, 415, 88, 188, 201};
cout << "Original lists:\n";
for_each(yadayada.begin(), yadayada.end(), outint_l);
cout << endl;
for_each(etcetera.begin(), etcetera.end(), outint_l);
cout << endl;
yadayada.remove_if(
[](int n) { return n > 100; }); // use a named function object
etcetera.remove_if(
[](int n) { return n > 200; }); // construct a function object
cout << "Trimmed lists:\n";
for_each(yadayada.begin(), yadayada.end(), outint_l);
cout << endl;
for_each(etcetera.begin(), etcetera.end(), outint_l);
cout << endl;
return 0;
}