I want to use spirit parsing some text file. The very basic element in those files is string, float and int. For example:

123L

"hello"

987.4343

I wrote my grammar class:

struct pm_parser : public grammar<pm_parser>
{
    template <typename ScannerT>
    struct definition
    {
        definition(pm_parser const& self)
        {
             factor =
                         int_p[show_int()]>>"L"
                       | '"'>> *((~ch_p('"'))[show_char()]) >>'"'
                       | real_p[show_double()]
                       ;
        }
    rule<ScannerT>      factor;
    rule<ScannerT> const& start() const { return factor; }
    };
};

show_int, show_char, show_double are structs used to show int, char and double value.  When I parse 1L, it works. But when I parse 1, it is parsed twice(int_p and real_p are all works).

How can I define grammar struct to parse text above without errors?