How is P0522R0 breaking code?

后端 未结 2 399
一生所求
一生所求 2020-12-01 23:53

Today I was reading the C++17 support page of clang. I\'ve notice something odd. The feature Matching template template parameters to compatible arguments (P0522R0)

相关标签:
2条回答
  • 2020-12-02 00:11

    A more common scenario is when some code wants to inspect the template arguments with a set of partial specializations, e.g.:

    template<class> struct Foo;
    
    template<template<class> class X, class T> 
    struct Foo<X<T>> { /* ... */ };
    
    template<template<class, class> class X, class T, class U> 
    struct Foo<X<T, U>> { /* ... */ };
    
    // etc., etc.
    

    Foo<std::vector<int>> is now ill-formed without an appropriate partial ordering fix.

    0 讨论(0)
  • 2020-12-02 00:26

    You can have code like this:

    template<template<typename> typename>
    struct Foo {};
    
    template<typename, typename = void>
    struct Bar {};
    
    Foo<Bar> unused;
    

    Without the defect resolution, unused would be ill-formed, because foo takes a template with only one template parameter, and not two. If you relied on this (maybe for SFINAE):

    template<template<typename> typename>
    void foo();
    
    template<template<typename, typename> typename>
    void foo();
    
    template<typename, typename = void>
    struct Bar {};
    
    int main() {
        foo<Bar>(); // ambiguous after resolution!
    }
    

    Then the call would fail! The problem is that there was no corresponding change to partial ordering, and so both candidate functions have the same viability, and the call is ambiguous.

    0 讨论(0)
提交回复
热议问题