[Boost-bugs] [Boost C++ Libraries] #6750: istream_iterator in spirit might have defect

Subject: [Boost-bugs] [Boost C++ Libraries] #6750: istream_iterator in spirit might have defect
From: Boost C++ Libraries (noreply_at_[hidden])
Date: 2012-03-31 02:25:15


#6750: istream_iterator in spirit might have defect
---------------------------------------------+------------------------------
 Reporter: Li He <I.demonstrate@…> | Owner: djowel
     Type: Bugs | Status: new
Milestone: To Be Determined | Component: spirit
  Version: Boost 1.49.0 | Severity: Problem
 Keywords: istream_iterator |
---------------------------------------------+------------------------------
 The following program yields a wrong parsing result:
 {{{
 #include <iostream>
 #include <vector>
 #include <algorithm>
 #include <iterator>

 #include <boost/spirit/include/qi.hpp>
 #include <boost/spirit/include/phoenix_core.hpp>
 #include <boost/spirit/include/phoenix_operator.hpp>
 #include <boost/spirit/include/phoenix_stl.hpp>

 namespace spirit = boost::spirit ;
 namespace qi = boost::spirit::qi ;
 namespace ascii = boost::spirit::ascii ;
 namespace phoenix = boost::phoenix ;

 using qi::double_ ;
 using qi::_1 ;
 using phoenix::push_back ;
 using phoenix::ref ;

 int
 main (int argc, char* argv[]) {
   spirit::istream_iterator iter (std::cin), end ;
   std::vector<double> v ;
   bool r = spirit::qi::phrase_parse
     (iter, end,
      double_ [push_back (ref(v), _1)] % ',',
      ascii::space) ;

   if (r) {
     std::cout << "vector: " ;
     std::copy (v.begin(), v.end(),
                std::ostream_iterator<double> (std::cout, ", ")) ;
     std::cout << std::endl ;
   }
   return 0 ;
 }
 }}}

 You may test it with input like "1,2,3,4" and the output is "vector:
 1,2,3,44,". When we switch to std::getline and use the
 std::string::const_iterator, the result turns out fine.

-- 
Ticket URL: <https://svn.boost.org/trac/boost/ticket/6750>
Boost C++ Libraries <http://www.boost.org/>
Boost provides free peer-reviewed portable C++ source libraries.

This archive was generated by hypermail 2.1.7 : 2017-02-16 18:50:09 UTC