莫名其妙的可变参数模板练习
Baffling variadic templates exercise
我给自己设定了这个任务来帮助学习可变参数模板。函数 add_and_cat()
应该首先接受一对,然后是可变数量的整数或字符串或混合。当它递归地遇到这些时,它应该从 pair.first 中减去每个 int,并将每个字符串连接到 pair.second 上。函数 pp 只是打印对。
它似乎在很多情况下都有效,但在其他情况下我收到一个编译错误,声称没有匹配的函数(在底部)......尽管据我所知似乎只有一个非常明显的候选者:/我玩过并试图推理出来但是......还没有运气。我错过了什么?
------------------------ 代码:---------------- ----------
#include <iostream>
void pp(std::pair<int,std::string> printme) {
std::cout << printme.first << " : " << printme.second << "\n";
}
//base case int
//must be int or std::string for now
void add_and_cat(std::pair<int, std::string>& store, int i) {
store.first -= i;
}
//base case string
void add_and_cat(std::pair<int, std::string>& store, std::string s) {
store.second += s;
}
//full case for int
template<typename ...Ts>
void add_and_cat(std::pair<int, std::string>& store, int i, Ts... rest) {
store.first -= i;
add_and_cat(store, rest...);
}
//full case for string
template<typename ...Ts>
void add_and_cat(std::pair<int, std::string>& store, std::string s, Ts... rest) {
store.second += s;
add_and_cat(store, rest...);
}
int main()
{
std::pair<int, std::string> p{0,"START"};
//add_and_cat(p, 1, 2, 3, 4); pp(p); //fine
//add_and_cat(p, 3, 4, 5, 6); pp(p); //fine
//add_and_cat(p, "A", "B", "C", "D"); pp(p); //fine
//add_and_cat(p, "D", "E", "F", "G"); pp(p); //fine
//add_and_cat(p, "A", 1, "B"); pp(p); //fine
//add_and_cat(p, 1, "A", 1, "B"); pp(p); //compile error
//add_and_cat(p, 1, 2, "A",3); pp(p); //compile error
//add_and_cat(p, "A", 1, 2, "B"); pp(p); //fine
//add_and_cat(p, "A", 1, 2, "B","C"); pp(p); //compile error
//add_and_cat(p, 1, 2, "B","C"); pp(p); //compile error
return 0;
}
---------------------------- 错误:------------ --------------
/mnt/c/Users/Tim/Nextcloud/playground/seqan3/source/hello_world.cpp: In instantiation of ‘std::pair<int, std::__cxx11::basic_string<char> > add_and_cat(std::pair<int, std::__cxx11::basic_string<char> >&, int, Ts ...) [with Ts = {const char*, int, const char*}]’:
/mnt/c/Users/Tim/Nextcloud/playground/seqan3/source/hello_world.cpp:273:45: required from here
/mnt/c/Users/Tim/Nextcloud/playground/seqan3/source/hello_world.cpp:250:20: error: no matching function for call to ‘add_and_cat(std::pair<int, std::__cxx11::basic_string<char> >&, const char*&, int&, const char*&)’
250 | return add_and_cat(store, rest...);
| ~~~~~~~~~~~^~~~~~~~~~~~~~~~
/mnt/c/Users/Tim/Nextcloud/playground/seqan3/source/hello_world.cpp:233:29: note: candidate: ‘std::pair<int, std::__cxx11::basic_string<char> > add_and_cat(std::pair<int, std::__cxx11::basic_string<char> >&, int)’
233 | std::pair<int, std::string> add_and_cat(std::pair<int, std::string>& store, int i) {
| ^~~~~~~~~~~
/mnt/c/Users/Tim/Nextcloud/playground/seqan3/source/hello_world.cpp:233:29: note: candidate expects 2 arguments, 4 provided
/mnt/c/Users/Tim/Nextcloud/playground/seqan3/source/hello_world.cpp:240:29: note: candidate: ‘std::pair<int, std::__cxx11::basic_string<char> > add_and_cat(std::pair<int, std::__cxx11::basic_string<char> >&, std::string)’
240 | std::pair<int, std::string> add_and_cat(std::pair<int, std::string>& store, std::string s) {
| ^~~~~~~~~~~
/mnt/c/Users/Tim/Nextcloud/playground/seqan3/source/hello_world.cpp:240:29: note: candidate expects 2 arguments, 4 provided
/mnt/c/Users/Tim/Nextcloud/playground/seqan3/source/hello_world.cpp:247:29: note: candidate: ‘template<class ... Ts> std::pair<int, std::__cxx11::basic_string<char> > add_and_cat(std::pair<int, std::__cxx11::basic_string<char> >&, int, Ts ...)’
247 | std::pair<int, std::string> add_and_cat(std::pair<int, std::string>& store, int i, Ts... rest) {
| ^~~~~~~~~~~
/mnt/c/Users/Tim/Nextcloud/playground/seqan3/source/hello_world.cpp:247:29: note: template argument deduction/substitution failed:
/mnt/c/Users/Tim/Nextcloud/playground/seqan3/source/hello_world.cpp:250:28: note: cannot convert ‘rest#0’ (type ‘const char*’) to type ‘int’
250 | return add_and_cat(store, rest...);
| ^~~~
非常感谢您的帮助!
当 int
重载必须调用 string
重载时失败,即当 int
参数在 string
参数之前时。您必须先声明该函数,然后才能调用它:
#include <iostream>
//...
// declare "full case for string" called in "full case for int"
template<typename ...Ts>
void add_and_cat(std::pair<int, std::string>& store, std::string s, Ts... rest);
//full case for int
template<typename ...Ts>
void add_and_cat(std::pair<int, std::string>& store, int i, Ts... rest) {
store.first -= i;
add_and_cat(store, rest...);
}
//full case for string
template<typename ...Ts>
void add_and_cat(std::pair<int, std::string>& store, std::string s, Ts... rest) {
store.second += s;
add_and_cat(store, rest...);
}
int main()
{
std::pair<int, std::string> p{0,"START"};
add_and_cat(p, 1, "A", 1, "B"); pp(p); // no compile error
add_and_cat(p, 1, 2, "A",3); pp(p); // no compile error
add_and_cat(p, "A", 1, 2, "B","C"); pp(p); // no compile error
add_and_cat(p, 1, 2, "B","C"); pp(p); // no compile error
}
请注意,您可以使用 fold expression:
来避免递归
void add_one(std::pair<int,std::string>& store,int x){ store.first -=x; }
void add_one(std::pair<int,std::string>& store,const std::string& x){ store.second +=x; }
template <typename ...Ts>
void add_and_cat2(std::pair<int,std::string>& store,Ts... t){
(add_one(store,t),...);
}
我给自己设定了这个任务来帮助学习可变参数模板。函数 add_and_cat()
应该首先接受一对
它似乎在很多情况下都有效,但在其他情况下我收到一个编译错误,声称没有匹配的函数(在底部)......尽管据我所知似乎只有一个非常明显的候选者:/我玩过并试图推理出来但是......还没有运气。我错过了什么?
------------------------ 代码:---------------- ----------
#include <iostream>
void pp(std::pair<int,std::string> printme) {
std::cout << printme.first << " : " << printme.second << "\n";
}
//base case int
//must be int or std::string for now
void add_and_cat(std::pair<int, std::string>& store, int i) {
store.first -= i;
}
//base case string
void add_and_cat(std::pair<int, std::string>& store, std::string s) {
store.second += s;
}
//full case for int
template<typename ...Ts>
void add_and_cat(std::pair<int, std::string>& store, int i, Ts... rest) {
store.first -= i;
add_and_cat(store, rest...);
}
//full case for string
template<typename ...Ts>
void add_and_cat(std::pair<int, std::string>& store, std::string s, Ts... rest) {
store.second += s;
add_and_cat(store, rest...);
}
int main()
{
std::pair<int, std::string> p{0,"START"};
//add_and_cat(p, 1, 2, 3, 4); pp(p); //fine
//add_and_cat(p, 3, 4, 5, 6); pp(p); //fine
//add_and_cat(p, "A", "B", "C", "D"); pp(p); //fine
//add_and_cat(p, "D", "E", "F", "G"); pp(p); //fine
//add_and_cat(p, "A", 1, "B"); pp(p); //fine
//add_and_cat(p, 1, "A", 1, "B"); pp(p); //compile error
//add_and_cat(p, 1, 2, "A",3); pp(p); //compile error
//add_and_cat(p, "A", 1, 2, "B"); pp(p); //fine
//add_and_cat(p, "A", 1, 2, "B","C"); pp(p); //compile error
//add_and_cat(p, 1, 2, "B","C"); pp(p); //compile error
return 0;
}
---------------------------- 错误:------------ --------------
/mnt/c/Users/Tim/Nextcloud/playground/seqan3/source/hello_world.cpp: In instantiation of ‘std::pair<int, std::__cxx11::basic_string<char> > add_and_cat(std::pair<int, std::__cxx11::basic_string<char> >&, int, Ts ...) [with Ts = {const char*, int, const char*}]’:
/mnt/c/Users/Tim/Nextcloud/playground/seqan3/source/hello_world.cpp:273:45: required from here
/mnt/c/Users/Tim/Nextcloud/playground/seqan3/source/hello_world.cpp:250:20: error: no matching function for call to ‘add_and_cat(std::pair<int, std::__cxx11::basic_string<char> >&, const char*&, int&, const char*&)’
250 | return add_and_cat(store, rest...);
| ~~~~~~~~~~~^~~~~~~~~~~~~~~~
/mnt/c/Users/Tim/Nextcloud/playground/seqan3/source/hello_world.cpp:233:29: note: candidate: ‘std::pair<int, std::__cxx11::basic_string<char> > add_and_cat(std::pair<int, std::__cxx11::basic_string<char> >&, int)’
233 | std::pair<int, std::string> add_and_cat(std::pair<int, std::string>& store, int i) {
| ^~~~~~~~~~~
/mnt/c/Users/Tim/Nextcloud/playground/seqan3/source/hello_world.cpp:233:29: note: candidate expects 2 arguments, 4 provided
/mnt/c/Users/Tim/Nextcloud/playground/seqan3/source/hello_world.cpp:240:29: note: candidate: ‘std::pair<int, std::__cxx11::basic_string<char> > add_and_cat(std::pair<int, std::__cxx11::basic_string<char> >&, std::string)’
240 | std::pair<int, std::string> add_and_cat(std::pair<int, std::string>& store, std::string s) {
| ^~~~~~~~~~~
/mnt/c/Users/Tim/Nextcloud/playground/seqan3/source/hello_world.cpp:240:29: note: candidate expects 2 arguments, 4 provided
/mnt/c/Users/Tim/Nextcloud/playground/seqan3/source/hello_world.cpp:247:29: note: candidate: ‘template<class ... Ts> std::pair<int, std::__cxx11::basic_string<char> > add_and_cat(std::pair<int, std::__cxx11::basic_string<char> >&, int, Ts ...)’
247 | std::pair<int, std::string> add_and_cat(std::pair<int, std::string>& store, int i, Ts... rest) {
| ^~~~~~~~~~~
/mnt/c/Users/Tim/Nextcloud/playground/seqan3/source/hello_world.cpp:247:29: note: template argument deduction/substitution failed:
/mnt/c/Users/Tim/Nextcloud/playground/seqan3/source/hello_world.cpp:250:28: note: cannot convert ‘rest#0’ (type ‘const char*’) to type ‘int’
250 | return add_and_cat(store, rest...);
| ^~~~
非常感谢您的帮助!
当 int
重载必须调用 string
重载时失败,即当 int
参数在 string
参数之前时。您必须先声明该函数,然后才能调用它:
#include <iostream>
//...
// declare "full case for string" called in "full case for int"
template<typename ...Ts>
void add_and_cat(std::pair<int, std::string>& store, std::string s, Ts... rest);
//full case for int
template<typename ...Ts>
void add_and_cat(std::pair<int, std::string>& store, int i, Ts... rest) {
store.first -= i;
add_and_cat(store, rest...);
}
//full case for string
template<typename ...Ts>
void add_and_cat(std::pair<int, std::string>& store, std::string s, Ts... rest) {
store.second += s;
add_and_cat(store, rest...);
}
int main()
{
std::pair<int, std::string> p{0,"START"};
add_and_cat(p, 1, "A", 1, "B"); pp(p); // no compile error
add_and_cat(p, 1, 2, "A",3); pp(p); // no compile error
add_and_cat(p, "A", 1, 2, "B","C"); pp(p); // no compile error
add_and_cat(p, 1, 2, "B","C"); pp(p); // no compile error
}
请注意,您可以使用 fold expression:
来避免递归void add_one(std::pair<int,std::string>& store,int x){ store.first -=x; }
void add_one(std::pair<int,std::string>& store,const std::string& x){ store.second +=x; }
template <typename ...Ts>
void add_and_cat2(std::pair<int,std::string>& store,Ts... t){
(add_one(store,t),...);
}