Databases
StackOverflow2013
Postsdbo
POis_convertible for multiple arguments
 

Note that there are some explanatory texts on larger screens.

plurals
  1. POis_convertible for multiple arguments
    primarykey
    data
    text
    <p>Suppose I don't have <code>std::is_convertible</code> for whatever reason and want to implement it myself. The standard says something along these lines:</p> <blockquote> <p>The predicate condition for a template specialization <code>is_convertible&lt;From, To&gt;</code> shall be satisfied if and only if the return expression in the following code would be well-formed, including any implicit conversions to the return type of the function:</p> <pre><code>To f() { return declval&lt;From&gt;(); } </code></pre> </blockquote> <p>Okay, no big deal, I can do it like this (Note argument order opposite to the one in <code>std::is_convertible</code>, this is intentional and irrelevant to the issue):</p> <pre><code>template &lt;typename To_, typename From_&gt; class my_is_convertible { private: template &lt;typename To&gt; struct indirector { indirector(To); }; template &lt;typename To, typename From&gt; struct tag {}; template &lt;typename To, typename From&gt; static auto test(tag&lt;To, From&gt;) -&gt; decltype(indirector&lt;To&gt;(std::declval&lt;From&gt;()), std::true_type()); static auto test(...) -&gt; std::false_type; public: static constexpr bool value = decltype(test(tag&lt;To_, From_&gt;()))::value; }; </code></pre> <p>This seems to work as intended, and as far as I can tell does the same thing.</p> <p>Now I can distinguish between implicit and explicit (or none at all) constructors:</p> <pre><code>struct A {}; struct B {}; struct Test { Test(A); explicit Test(B); }; int main() { std::cout &lt;&lt; my_is_convertible&lt;Test, A&gt;::value; // true std::cout &lt;&lt; my_is_convertible&lt;Test, B&gt;::value; // false return 0; } </code></pre> <p>So far, so good. Now, I want to do the same with multiple argument constructors. This didn't make sense prior to <code>c++11</code> as there was no way to call multiargument constructor implicitly. But now we have brace enclosed initializer list syntax, and <code>explicit</code> keyword on multiargument constructor makes a difference.</p> <p>Let us extend the definition:</p> <blockquote> <p>The predicate condition for a template specialization <code>my_is_convertible_many&lt;To, From...&gt;</code> shall be satisfied if and only if the return expression in the following code would be well-formed, including any implicit conversions to the return type of the function:</p> <pre><code>To f() { return {declval&lt;From&gt;()...}; } </code></pre> </blockquote> <p>To implement it I went the obvious way:</p> <pre><code>template &lt;typename To_, typename... From_&gt; class my_is_convertible_many { private: template &lt;typename To&gt; struct indirector { indirector(To); }; template &lt;typename To, typename... From&gt; struct tag {}; template &lt;typename To, typename... From&gt; static auto test(tag&lt;To, From...&gt;) -&gt; decltype(indirector&lt;To&gt;({std::declval&lt;From&gt;()...}), std::true_type()); static auto test(...) -&gt; std::false_type; public: static constexpr bool value = decltype(test(tag&lt;To_, From_...&gt;()))::value; }; </code></pre> <p>This correctly reports <code>true</code> in the presence of matching implicit constructor and <code>false</code> if the is no matcing constructor. But it fails to compile if there is an explicit matching constructor (at least on gcc 4.8.1):</p> <pre><code>struct A {}; struct B {}; struct C {}; struct Test { Test(A, A); //Test(B, B); explicit Test(C, C); }; int main() { std::cout &lt;&lt; my_is_convertible_many&lt;Test, A, A&gt;::value; // true, correct std::cout &lt;&lt; my_is_convertible_many&lt;Test, B, B&gt;::value; // false, correct std::cout &lt;&lt; my_is_convertible_many&lt;Test, C, C&gt;::value; // error return 0; } </code></pre> <p>The error is about attempting to implicitly call explicit constructor, which on gcc sounds like this:</p> <pre><code>main.cpp: In substitution of 'template&lt;class To, class ... From&gt; static decltype (((my_is_convertible_many&lt;To_, From_&gt;::indirector&lt;To&gt;)({(declval&lt;From&gt;)()...}), std::true_type())) my_is_convertible_many&lt;To_, From_&gt;::test(my_is_convertible_many&lt;To_, From_&gt;::tag&lt;To, From ...&gt;) [with To = To; From = {From ...}; To_ = Test; From_ = {C, C}] [with To = Test; From = {C, C}]': main.cpp:21:73: required from 'constexpr const bool my_is_convertible_many&lt;Test, C, C&gt;::value' main.cpp:37:54: required from here main.cpp:17:97: error: converting to 'Test' from initializer list would use explicit constructor 'Test::Test(C, C)' static auto test(tag&lt;To, From...&gt;) -&gt; decltype(indirector&lt;To&gt;({std::declval&lt;From&gt;()...}), std::true_type()); ^ </code></pre> <p>Which is sensible. I, however, expect this overload of <code>test</code> to sfinae out and the other one to be used instead, thus producing no error.</p> <p>So the question is: Why doesn't this happen, and what can I do about it?</p>
    singulars
    1. This table or related slice is empty.
    1. This table or related slice is empty.
    1. PTQuestion
    1. USyuri kilochek
    1. USyuri kilochek
    plurals
    1. This table or related slice is empty.
    1. This table or related slice is empty.
    1. This table or related slice is empty.
    1. This table or related slice is empty.
    1. VO
      singulars
      1. POis_convertible for multiple arguments
      1. This table or related slice is empty.
      1. VTUpMod
    2. VO
      singulars
      1. POis_convertible for multiple arguments
      1. This table or related slice is empty.
      1. VTUpMod
    3. VO
      singulars
      1. POis_convertible for multiple arguments
      1. This table or related slice is empty.
      1. VTUpMod
    1. COTo me this does not look like `is_convertible` but `is_constructible_with`
      singulars
      1. POis_convertible for multiple arguments
      1. USVinzenz
    2. CO@Vinzenz I just use standard naming. `std::is_constructible` checks whether object of specified type can be constructed explicitly from arguments of specified types. `std::is_convertible` checks if object of one type can be converted to object of another type implicitly (either by constructor call of conversion operator). I am trying to extend the latter, thus the name. This is irrelevant either way.
      singulars
      1. POis_convertible for multiple arguments
      1. USyuri kilochek
    3. COgcc source code says: "When converting from an init list we consider explicit constructors, but actually trying to call one is an error." which sounds like it may be on purpose? Or did it just forget about sfinae? In any case the patch to change it is trivial (1 or 2 lines in cp/call.c). Please file a PR with gcc if you don't get a convincing answer that the code is illegal. By the way, did you check with other compilers?
      singulars
      1. POis_convertible for multiple arguments
      1. USMarc Glisse
 

Querying!

 
Guidance
A row detail

Detail views are divided into sections. All the information in the data section comes from columns in the selected row. The other sections display data from other, related rows.

Related data can be related in a to-one or a to-many fashion. Captions of data related in a to-many fashion link to a list view showing a filtered view of the table.

Try moving around until you find a non-empty to-many entry and click on the label to get to one. You can move back to the root by clicking on the database name in the header.

SQuiL has stopped working due to an internal error.

If you are curious you may find further information in the browser console, which is accessible through the devtools (F12).

Reload