Databases
StackOverflow2013
Postsdbo
PO
 

Note that there are some explanatory texts on larger screens.

plurals
  1. PO
    primarykey
    data
    text
    <h2>The surprise</h2> <p>This is a subtle Clang bug, deeply buried in the Standard. The problem is that in almost all cases, non-type template <em>arguments</em> can be converted to the type of the template <em>parameter</em>. E.g. the expression <code>Int&lt;0&gt;</code> has an <code>int</code> literal argument of value <code>0</code> that is being converted to the type <code>unsigned long long</code> of the template parameter <code>N</code>.</p> <p><strong>14.8.2 Template argument deduction [temp.deduct]/2 2nd bullet</strong></p> <blockquote> <p>-- Non-type arguments must match the types of the corresponding non-type template parameters, <strong>or must be convertible to</strong> the types of the corresponding non-type parameters as specified in 14.3.2, otherwise type deduction fails.</p> </blockquote> <p>Since your class template <code>is_int&lt;T&gt;</code> has a partial specialization, we need to look at</p> <p><strong>14.5.5.1 Matching of class template partial specializations [temp.class.spec.match]</strong></p> <blockquote> <p>1 When a class template is used in a context that requires an instantiation of the class, it is necessary to determine whether the instantiation is to be generated using the primary template or one of the partial specializations. <strong>This is done by matching the template arguments of the class template specialization with the template argument lists of the partial specializations</strong>.</p> <p>2 A partial specialization matches a given actual template argument list if the template arguments of the partial specialization <strong>can be deduced from</strong> the actual template argument list (14.8.2).</p> </blockquote> <p>So it would seem that we can proceed to the earlier quote of 14.8.2/2 2nd bullet and match the second specialization (although in that case an even more complicated overload resolution game would have to be played). </p> <h2>The resolution</h2> <p>However, it turns out (as mentioned by @DyP in the comments) that another clause in the Standard supersedes this:</p> <p><strong>14.8.2.5 Deducing template arguments from a type [temp.deduct.type]</strong></p> <blockquote> <p>17 If, in the declaration of a function template with a non-type template-parameter, the non-type templateparameter is used in an expression in the function parameter-list and, if the corresponding template-argument is deduced, <strong>the template-argument type shall match the type of the template-parameter exactly</strong>, except that a template-argument deduced from an array bound may be of any integral type. [ Example:</p> </blockquote> <pre><code>template&lt;int i&gt; class A { / ... / }; template&lt;short s&gt; void f(A&lt;s&gt;); void k1() { A&lt;1&gt; a; f(a); // error: deduction fails for conversion from int to short f&lt;1&gt;(a); // OK } </code></pre> <p>The upshot is that the partial specialization of <code>is_int</code> cannot be deduced because it does not take the exact same type (<code>unsigned long long</code> vs <code>long long</code>) as the <code>Int</code> class template's formal non-type template parameter.</p> <p>You can resolve this by giving the non-type template parameter <code>N</code> in the partial specialization of <code>is_int</code> the same type as the non-type parameter <code>N</code> in the primary template <code>Int</code>.</p> <pre><code>template &lt;IntType N&gt; // ^^^^^^^^ struct is_int&lt;Int&lt;N&gt;&gt; : std::true_type {}; </code></pre> <p><a href="http://coliru.stacked-crooked.com/a/5d0ad7b77db02034"><strong>Live Example</strong></a>.</p>
    singulars
    1. This table or related slice is empty.
    1. POg++ and clang++ different behaviour with integral template parameter
      singulars
      1. PTQuestion
    1. PTAnswer
    1. USTemplateRex
    1. USTemplateRex
    plurals
    1. This table or related slice is empty.
    1. This table or related slice is empty.
    1. POg++ and clang++ different behaviour with integral template parameter
      singulars
      1. PTQuestion
    1. This table or related slice is empty.
    1. VO
      singulars
      1. PO
      1. This table or related slice is empty.
      1. VTUpMod
    2. VO
      singulars
      1. PO
      1. This table or related slice is empty.
      1. VTUpMod
    3. VO
      singulars
      1. PO
      1. This table or related slice is empty.
      1. VTUpMod
    1. CO@DyP [temp.clss.spec]/3 says that `Int<0>` does not need the dance with function templates because `N` is a non-type parameter in the primary template and can be deduced to `1`. The problem is that `is_int<T>` does not have a non-type template parameter in its primary template, but only in its partial specialization. It's the deduction of `N` in this partialization that requires the function template argument deduction and ends up in 14.8.2.5/17
      singulars
      1. PO
      1. USTemplateRex
    2. CO@DyP indeed it *seems* at first sight, hence the surprise, I tried to clarify in the modified answer. BTW, the process of creating a function type to deduce the parameters of a partially specialized class templates should not be more complicated than writing a function template with the class template specialization as an argument, similar to the overload resolution game (where arguments are deduced for the *other* overloads' parameters, rather than their own parameters).
      singulars
      1. PO
      1. USTemplateRex
    3. CO[Repost from an earlier comment] [core issue 1770](http://www.open-std.org/JTC1/SC22/WG21/docs/cwg_active.html#1770) addresses 14.8.2.5; the proposed resolution makes clear it applies to *all* deductions, not only to function templates (or maybe it already is clear, and I'm just too nit-picky ;)
      singulars
      1. PO
      1. USdyp
 

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