Databases
StackOverflow2013
Postsdbo
POambiguous template pattern matching when using variadic template arguments of function overloads
 

Note that there are some explanatory texts on larger screens.

plurals
  1. POambiguous template pattern matching when using variadic template arguments of function overloads
    primarykey
    data
    text
    <p>This should be a common situation with variadic templates for example when tree walking children that are variadic template arguments. I find many related questions and answers but either they are about a slightly different thing or they are the same thing and I did not get it. Now to the problem. I have a non variadic tuple like this</p> <pre><code>template &lt;class E, class T1, class T2, class T3, etc...&gt; struct X; </code></pre> <p>and I am overloading functions to have specialized behaviour depending on the first element of such tuple being pointer type or vector pointer type. That works fine but if I pack the template arguments into a single variadic template argument, then the overloads become ambiguous. Here is the error message:</p> <pre><code>variadic.cpp:42:17: error: ambiguous overload for ‘operator&lt;&lt;’ in ‘std::cout &lt;&lt; y’ variadic.cpp:42:17: note: candidates are:... </code></pre> <p>The compiler should prefer <code>X&lt;vector&lt;V*&gt;*,T*...&gt;</code> over <code>X&lt;H*,T*...&gt;</code> when it tries to match <code>vector&lt;double*&gt;*</code> as first element of the tuple.</p> <p>I can disambiguate using enable_if and things work again. However I would like to understand the error and if possible find other means then enable_if. Here is the code:</p> <pre><code>#include &lt;iostream&gt; #include &lt;vector&gt; #include &lt;boost/type_traits/is_fundamental.hpp&gt; #include &lt;boost/utility/enable_if.hpp&gt; using namespace std; template &lt;typename ... T&gt; struct X; template &lt;&gt; struct X &lt;&gt; { }; template &lt;typename H, typename ... T&gt; struct X&lt;H*,T*...&gt; : public X&lt;T*...&gt; { H* value; X(H* value, T*... args) : value(value), X&lt;T*...&gt;(args...) { } }; template &lt;typename H, typename ... T&gt; #ifdef DO_NOT_WANNA_SEE_THE_BUG typename boost::enable_if&lt;boost::is_fundamental&lt;H&gt;, std::ostream&gt;::type&amp; operator&lt;&lt;(std::ostream&amp; stream, X&lt;H*,T*...&gt; const &amp; x) #else std::ostream&amp; operator&lt;&lt;(std::ostream&amp; stream, X&lt;H*,T*...&gt; const &amp; x) #endif { return stream &lt;&lt; "specialized scalar pointer"; } template &lt;typename V, typename ... T&gt; std::ostream&amp; operator&lt;&lt;(std::ostream&amp; stream, X&lt;vector&lt;V*&gt;*,T*...&gt; const &amp; x) { return stream &lt;&lt; "specialized vector pointer"; } int main() { double a,b; vector&lt;double *&gt; v; X&lt;double*,double*&gt; x (&amp;a,&amp;b); X&lt;vector&lt;double*&gt;*, double*&gt; y (&amp;v, &amp;b); cout &lt;&lt; x &lt;&lt; endl; cout &lt;&lt; y &lt;&lt; endl; // this line is ambiguous according to gcc 4.6 and later } </code></pre>
    singulars
    1. PO
      singulars
      1. PTAnswer
    1. This table or related slice is empty.
    1. PTQuestion
    1. US0x499602D2
    1. USPatrick Fromberg
    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. PO
      singulars
      1. PTAnswer
    1. VO
      singulars
      1. POambiguous template pattern matching when using variadic template arguments of function overloads
      1. This table or related slice is empty.
      1. VTUpMod
    2. VO
      singulars
      1. POambiguous template pattern matching when using variadic template arguments of function overloads
      1. This table or related slice is empty.
      1. VTUpMod
    3. VO
      singulars
      1. POambiguous template pattern matching when using variadic template arguments of function overloads
      1. This table or related slice is empty.
      1. VTUpMod
    1. CONot that this helps at all, but Apple LLVM 4.2 (clang) chews this right up without a problem. I'm curious to know if gcc has the same issue without variadic `T*...` (which from what I see isn't needed), and rather you used `T...`. It makes no sense why it should matter, but then again, gcc is puking, and clang isn't so... [see it live on ideone.com](http://ideone.com/ja1ZSb)
      singulars
      1. POambiguous template pattern matching when using variadic template arguments of function overloads
      1. USWhozCraig
    2. COFollowup: I just noticed ideone.com uses gcc 4.8, just fyi.
      singulars
      1. POambiguous template pattern matching when using variadic template arguments of function overloads
      1. USWhozCraig
    3. CO@WhozCraig. Perfect answer. You are right. I wanted to restrict everything to pointers but of course it is enough to do that for the Head. It is not required for the packed Tail.Your suggestion works for old gcc 4.6 also. Your remark on LLVM is also very interesting. Many Thanks!
      singulars
      1. POambiguous template pattern matching when using variadic template arguments of function overloads
      1. USPatrick Fromberg
 

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