Re: [Boost-users] [Range.Ex / String.Algo] Ambiguous call to overload function "find"

5 Jul 2009

      2009/7/3 S.C. Leung 
...
Hi!
I have downloaded Range.Ex uploaded at 26.04.2009 09:09. It seems the
difficulties are still there.
The exact program you posted compiles without errors or warnings on Visual
C++ 9. I installed Boost 1.39 and then copied over the RangeEx files.

[...]
...
009/7/3 Neil Groves 
...
Hello!
On Fri, Jul 3, 2009 at 2:38 AM, 梁绍池  wrote:
...
I use vc 9.0 to compile the following code with boost 1.39.0 and
range_ex.
    #include<string>
    #include
    #include
int main() {
        std::string s = "hello";
        boost::find_first(s, "lo");
        return 0;
    }
The compiler will complain ambiguous call to overload function "find". I
...
found there's also a "find" in boost/range/algorithm.hpp. Even worse when I
change the type of "s" to const std::string, the compiler will resolute
"find" call as the one in boost/range/algorithm.hpp
and reports an error "const_iterator is not a member of
std::_String_const_iterator". Now I just indicated the "find" call in
boost::find_first explicitly. Is there a better way to solve it?
This code compiles cleanly for me. I noticed that you are talking about
"find" rather than "find_first" so I took the liberty of replacing the
"find_first" with "find". This indeed produced a compiler error similar to
the one you describe. However boost::find(s, "lo") does not match the
boost/algorithm/string.hpp version because the second parameter should be a
"Finder object used for searching". Please see:
http://www.boost.org/doc/libs/1_39_0/doc/html/string_algo/reference.html#hea...

It also does not match the RangeEx version because "find" is used to find an
element, as opposed to a sub-sequence, from a range. Therefore I conclude
that either, I have guessed the code snippet that causes the problem
inaccurately, or that the code that produced the compiler error was simply
incorrect.

If I have guessed the code wrongly, would you please supply the exact code
that reproduces the problem. I suspect that find_first is exactly the right
solution. Please let me know either way.

 [...]

Regards,
Neil Groves