c++ - Expression template operator overloading problem with std::vector -

-

i'm working on numerical library uses expression templates. unfortunately encountered problem operator overloads. consider following stripped down example.

#include <vector>  namespace test {     class test {};      template<class a, class b>     class testexpr {};      template<class a, class b>     testexpr<a, b>     operator-(a a, b b)     {         return testexpr<a, b>();     } }  test::test stuff(std::vector<test::test> &v) { return v.back(); }  int main() { }

which gives following error message when compiling gcc 4.4.3 or clang 2.8:

in file included eir_test.cc:2: in file included /usr/include/c++/4.4/vector:64: /usr/include/c++/4.4/bits/stl_vector.h:696:16: error: indirection requires pointer operand       ('testexpr<__gnu_cxx::__normal_iterator<test::test *, std::vector<test::test, std::allocator<test::test> > >, int>' invalid)       { return *(end() - 1); }                ^~~~~~~~~~~~ eir_test.cc:21:12: note: in instantiation of member function 'std::vector<test::test, std::allocator<test::test> >::back' requested here     return v.back();            ^ 1 error generated.

for reason compilers lookup test namespace , find general operator. used form traits magic reduce number of version had make operator. should accept 4 different datatypes (including double , int) lead lot of different combinations.

is there way make work without spelling out combinations every operator?

this because end() returns type class template specialization has 1 argument of type test::test *. when operator- applied in expression end() - 1, argument dependent lookup looks also in namespace of test::test. finds operator- , passes iterator , int.

you fix not accepting , types arguments. example, try accepting (testexpr<a1, b1>, testexpr<a2, b2>) instead. show combinations, possibly there way cut them down using way formulate them?

in opinion, implementation acts way should non-conforming (i think disgusting though). because doing iterator - 1 specified yield iterator previous element , must not crazy that, think. 1 way declare operator non-template accepting iterator type , integer argument (which of iterator's difference_type) directly. way version should preferred.


Comments

Post a Comment

Popular posts from this blog

ASP.NET/SQL find the element ID and update database -

-
basically i've got update query runs when click submit button. i have asp:repeater in page layer inside this <asp:repeater id="dgbookings" runat="server" onitemdatabound="itemdb"> <itemtemplate> <div class="bookingscontent"> <div class="bookingdetails" id="<%# databinder.eval(container.dataitem, "booking_ref") %>"> <p class="infotext"><%# databinder.eval(container.dataitem, "adults") %> <asp:linkbutton onclick="editdetails" text="edit..." runat="server"></asp:linkbutton></p> <p class="infotext"><asp:linkbutton onclick="submitdetails" text="submit..." runat="server"></asp:linkbutton></p> </div> </div> </itemtemplate> <asp:repeater> and outputs ...
Read more

c++ - Compiling static TagLib 1.6.3 libraries for Windows -

-
i having super hard time compiling , using taglib 1.6.3 in qt project. i've tried can think of. taglib claims supported through cmake i'm not having luck. furthermore, i'm confused kinds of files need qt libs! i've built *.a files, *.lib, , *.dll. understand far... believe since i'm working in windows *.lib want. no matter do, end "undefined references" taglib functions try use when try compile qt project. have tried mingw32, msys, visual studio 2008, , cross-compiling windows on linux. turning nothing. what makes less sense me if compile same taglib source qt on mac (g++ think?) works fine! somewhere in windows compilation procedures have going wrong. have been smacking face on desk 30 (on , off) hours trying figure out. since qt uses mingw must compile taglib same compiler? if compile *.lib's visual studio not compatible? are *.a libraries usable in windows? (assuming mingw) i'm still trying handle on c++ stuff, after reading coun...
Read more

PostgreSQL 9.x - pg_read_binary_file & inserting files into bytea -

-
i have been looking everywhere (google, stackoverflow, etc.) documentation on how use postgresql pg_read_binary_file() function. the meaningful thing can find this page in official documentation. every time try use function error. for example: select pg_read_binary_file('/some/path/and/file.gif'); error: absolute path not allowed or select pg_read_binary_file('file.gif'); error: not stat file "file.gif": no such file or directory do need have file in specific directory postgres have access it? if directory? if matters, reason looking @ function because trying insert file database without doing crazy things . as stated @a_horse_with_no_name , @guedes solution ensure file being uploaded on server in pgdata directory. the postgres documentation state file location requirement. additionally, made symlink directory pgdata directory not disturb of postgres data structure. seems working , don't have of above...
Read more