regex.hpp

boost库提供标准的C++ API 配合dev c++使用,功能更加强大

template<>
struct regex_iterator_traits<const wchar_t*> : const_pointer_iterator_traits<wchar_t>{};

#if defined(__SGI_STL_PORT) && defined(__STL_DEBUG)
template<>
struct regex_iterator_traits<std::string::iterator> : pointer_iterator_traits<char>{};
template<>
struct regex_iterator_traits<std::string::const_iterator> : const_pointer_iterator_traits<char>{};
#ifndef BOOST_NO_STD_WSTRING
template<>
struct regex_iterator_traits<std::wstring::iterator> : pointer_iterator_traits<wchar_t>{};
template<>
struct regex_iterator_traits<std::wstring::const_iterator> : const_pointer_iterator_traits<wchar_t>{};
#endif // BOOST_NO_WSTRING
#endif // stport

#else

template <class T>
struct regex_iterator_traits : public std::iterator_traits<T> {};

#endif

template <class I>
struct def_alloc_param_traits
{
   typedef typename regex_iterator_traits<I>::value_type const_value_type;
   typedef typename remove_cv<const_value_type>::type type;
};
template <>
struct def_alloc_param_traits<const char*>
{
   typedef char type;
};
template <>
struct def_alloc_param_traits<const wchar_t*>
{
   typedef wchar_t type;
};

}

template <class iterator, class Allocator =
#if !(defined(BOOST_MSVC) && (BOOST_MSVC <= 1300))
BOOST_DEFAULT_ALLOCATOR(typename re_detail::def_alloc_param_traits<iterator>::type) >
#else
BOOST_DEFAULT_ALLOCATOR(re_detail::def_alloc_param_traits<iterator>::type) >
#endif
class match_results;

//
// class reg_expression
// represents the compiled
// regular expression:
//

#ifdef BOOST_MSVC
#pragma warning(push)
#pragma warning(disable : 4251 4231 4660)
#endif

#ifdef BOOST_REGEX_NO_FWD
template <class charT, class traits = regex_traits<charT>, class Allocator = BOOST_DEFAULT_ALLOCATOR(charT) >
#else
template <class charT, class traits, class Allocator >
#endif
class reg_expression : public regbase
{
public:
   typedef typename traits::size_type traits_size_type;
   typedef typename traits::uchar_type traits_uchar_type;
   typedef typename traits::string_type traits_string_type;
   // typedefs:
   typedef charT char_type;
   typedef traits traits_type;

   // locale_type
   // placeholder for actual locale type used by the
   // traits class to localise *this.
   typedef typename traits::locale_type locale_type;
   // value_type
   typedef charT value_type;
   // reference, const_reference
   typedef charT& reference;
   typedef const charT& const_reference;
   // iterator, const_iterator
   typedef const charT* const_iterator;
   typedef const_iterator iterator;
   // difference_type
   typedef typename Allocator::difference_type difference_type;
   // size_type
   typedef typename Allocator::size_type size_type;   
   // allocator_type
   typedef Allocator allocator_type;
   typedef Allocator alloc_type;
   // flag_type
   typedef regbase::flag_type flag_type;
   
public:
   explicit reg_expression(const Allocator& a = Allocator());
   explicit reg_expression(const charT* p, flag_type f = regbase::normal, const Allocator& a = Allocator());
   reg_expression(const charT* p1, const charT* p2, flag_type f = regbase::normal, const Allocator& a = Allocator());
   reg_expression(const charT* p, size_type len, flag_type f, const Allocator& a = Allocator());
   reg_expression(const reg_expression&);
   ~reg_expression();
   reg_expression& BOOST_REGEX_CALL operator=(const reg_expression&);
   reg_expression& BOOST_REGEX_CALL operator=(const charT* ptr)
   {
      set_expression(ptr, regbase::normal | regbase::use_except);
      return *this;
   }

   //
   // assign:
   reg_expression& assign(const reg_expression& that)
   { return *this = that; }
   reg_expression& assign(const charT* ptr, flag_type f = regbase::normal)
   {
      set_expression(ptr, f | regbase::use_except);
      return *this;
   }

   reg_expression& assign(const charT* first,
                          const charT* last,
                          flag_type f = regbase::normal)
   {
      set_expression(first, last, f | regbase::use_except);
      return *this;
   }
#if !defined(BOOST_NO_MEMBER_TEMPLATES) && !(defined(__IBMCPP__) && (__IBMCPP__ <= 502))

   template <class ST, class SA>
   unsigned int BOOST_REGEX_CALL set_expression(const std::basic_string<charT, ST, SA>& p, flag_type f = regbase::normal)
   { return set_expression(p.data(), p.data() + p.size(), f); }

   template <class ST, class SA>
   explicit reg_expression(const std::basic_string<charT, ST, SA>& p, flag_type f = regbase::normal, const Allocator& a = Allocator())
    : data(a), pkmp(0), error_code_(REG_EMPTY), _expression(0) { set_expression(p, f | regbase::use_except); }

   template <class I>
   reg_expression(I first, I last, flag_type f = regbase::normal, const Allocator& al = Allocator())
    : data(al), pkmp(0), error_code_(REG_EMPTY), _expression(0)
   {
      size_type len = last-first;
      scoped_array<charT> a(new charT[len]);
      std::copy(first, last, a.get());
      set_expression(a.get(), a.get() + len, f | regbase::use_except);
   }

   template <class ST, class SA>
   reg_expression& BOOST_REGEX_CALL operator=(const std::basic_string<charT, ST, SA>& p)
   {
      set_expression(p.c_str(), p.c_str() + p.size(), regbase::normal | regbase::use_except);
      return *this;
   }

   template <class string_traits, class A>
   reg_expression& BOOST_REGEX_CALL assign(
       const std::basic_string<charT, string_traits, A>& s,
       flag_type f = regbase::normal)
   {
      set_expression(s.c_str(), s.c_str() + s.size(), f | regbase::use_except);
      return *this;
   }

   template <class fwd_iterator>
   reg_expression& BOOST_REGEX_CALL assign(fwd_iterator first,
                          fwd_iterator last,
                          flag_type f = regbase::normal)
   {
      size_type len = last-first;
      scoped_array<charT> a(new charT[len]);
      std::copy(first, last, a.get());
      set_expression(a.get(), a.get() + len, f | regbase::use_except);
      return *this;
   }
#else
   unsigned int BOOST_REGEX_CALL set_expression(const std::basic_string<charT>& p, flag_type f = regbase::normal)
   { return set_expression(p.data(), p.data() + p.size(), f | regbase::use_except); }

   reg_expression(const std::basic_string<charT>& p, flag_type f = regbase::normal, const Allocator& a = Allocator())
    : data(a), pkmp(0) { set_expression(p, f | regbase::use_except); }

   reg_expression& BOOST_REGEX_CALL operator=(const std::basic_string<charT>& p)
   {
      set_expression(p.c_str(), p.c_str() + p.size(), regbase::normal | regbase::use_except);
      return *this;
   }

   reg_expression& BOOST_REGEX_CALL assign(
       const std::basic_string<charT>& s,
       flag_type f = regbase::normal)
   {
      set_expression(s.c_str(), s.c_str() + s.size(), f | regbase::use_except);
      return *this;
   }

#endif


   //
   // allocator access:
   Allocator BOOST_REGEX_CALL get_allocator()const;
   //
   // locale:
   locale_type BOOST_REGEX_CALL imbue(locale_type l){ return traits_inst.imbue(l); }
   locale_type BOOST_REGEX_CALL getloc()const{ return traits_inst.getloc(); }
   //
   // flags:
   flag_type BOOST_REGEX_CALL getflags()const
   { return flags(); }
   //
   // str:
   std::basic_string<charT> BOOST_REGEX_CALL str()const
   {
      std::basic_string<charT> result;
      if(this->error_code() == 0)
         result = std::basic_string<charT>(_expression, _expression_len);
      return result;
   }
   //
   // begin, end:
   const_iterator BOOST_REGEX_CALL begin()const
   { return (this->error_code() ? 0 : _expression); }
   const_iterator BOOST_REGEX_CALL end()const
   { return (this->error_code() ? 0 : _expression + _expression_len); }
   //
   // swap:
   void BOOST_REGEX_CALL swap(reg_expression&)throw();
   //
   // size:
   size_type BOOST_REGEX_CALL size()const
   { return (this->error_code() ? 0 : _expression_len); }
   //
   // max_size:
   size_type BOOST_REGEX_CALL max_size()const
   { return UINT_MAX; }
   //
   // empty:
   bool BOOST_REGEX_CALL empty()const
   { return 0 != this->error_code(); }

   unsigned BOOST_REGEX_CALL mark_count()const { return (this->error_code() ? 0 : marks); }
   bool BOOST_REGEX_CALL operator==(const reg_expression&)const;
   bool BOOST_REGEX_CALL operator<(const reg_expression&)const;
   //
   // The following are deprecated as public interfaces
   // but are available for compatibility with earlier versions.
   allocator_type BOOST_REGEX_CALL allocator()const;
   const charT* BOOST_REGEX_CALL expression()const { return (this->error_code() ? 0 : _expression); }
   unsigned int BOOST_REGEX_CALL set_expression(const charT* p, const charT* end, flag_type f = regbase::normal);
   unsigned int BOOST_REGEX_CALL set_expression(const charT* p, flag_type f = regbase::normal) { return set_expression(p, p + traits_type::length(p), f); }
   //
   // this should be private but template friends don't work:
   const traits_type& get_traits()const { return traits_inst; }
   unsigned int BOOST_REGEX_CALL error_code()const
   {
      return error_code_;
   }

private:
   traits_type traits_inst;
   re_detail::raw_storage<Allocator> data;
   unsigned _restart_type;
   unsigned marks;
   int repeats;
   unsigned char* startmap;
   std::size_t _expression_len;
   std::size_t _leading_len;
   const charT* _leading_string;
   std::size_t _leading_string_len;
   re_detail::kmp_info<charT>* pkmp;
   unsigned error_code_;
   charT* _expression;

   void BOOST_REGEX_CALL compile_maps();
   void BOOST_REGEX_CALL compile_map(re_detail::re_syntax_base* node, unsigned char* _map, unsigned int* pnull, unsigned char mask, re_detail::re_syntax_base* terminal = 0)const;
   bool BOOST_REGEX_CALL probe_start(re_detail::re_syntax_base* node, charT c, re_detail::re_syntax_base* terminal)const;
   bool BOOST_REGEX_CALL probe_start_null(re_detail::re_syntax_base* node, re_detail::re_syntax_base* terminal)const;
   void BOOST_REGEX_CALL fixup_apply(re_detail::re_syntax_base* b, unsigned cbraces);
   void BOOST_REGEX_CALL move_offsets(re_detail::re_syntax_base* j, unsigned size);
   re_detail::re_syntax_base* BOOST_REGEX_CALL compile_set(const charT*& first, const charT* last);
   re_detail::re_syntax_base* BOOST_REGEX_CALL compile_set_aux(re_detail::jstack<traits_string_type, Allocator>& singles, re_detail::jstack<traits_string_type, Allocator>& ranges, re_detail::jstack<boost::uint_fast32_t, Allocator>& classes, re_detail::jstack<traits_string_type, Allocator>& equivalents, bool isnot, const re_detail::_narrow_type&);
   re_detail::re_syntax_base* BOOST_REGEX_CALL compile_set_aux(re_detail::jstack<traits_string_type, Allocator>& singles, re_detail::jstack<traits_string_type, Allocator>& ranges, re_detail::jstack<boost::uint_fast32_t, Allocator>& classes, re_detail::jstack<traits_string_type, Allocator>& equivalents, bool isnot, const re_detail::_wide_type&);
   re_detail::re_syntax_base* BOOST_REGEX_CALL compile_set_simple(re_detail::re_syntax_base* dat, unsigned long cls, bool isnot = false);
   unsigned int BOOST_REGEX_CALL parse_inner_set(const charT*& first, const charT* last);

   re_detail::re_syntax_base* BOOST_REGEX_CALL add_simple(re_detail::re_syntax_base* dat, re_detail::syntax_element_type type, unsigned int size = sizeof(re_detail::re_syntax_base));
   re_detail::re_syntax_base* BOOST_REGEX_CALL add_literal(re_detail::re_syntax_base* dat, charT c);
   charT BOOST_REGEX_CALL parse_escape(const charT*& first, const charT* last);
   void BOOST_REGEX_CALL parse_range(const charT*& first, const charT* last, unsigned& min, unsigned& max);
   bool BOOST_REGEX_CALL skip_space(const charT*& first, const charT* last);
   unsigned int BOOST_REGEX_CALL probe_restart(re_detail::re_syntax_base* dat);
   unsigned int BOOST_REGEX_CALL fixup_leading_rep(re_detail::re_syntax_base* dat, re_detail::re_syntax_base* end);
   void BOOST_REGEX_CALL fail(unsigned int err);

protected:
   static int BOOST_REGEX_CALL repeat_count(const reg_expression& e)
   { return e.repeats; }
   static unsigned int BOOST_REGEX_CALL restart_type(const reg_expression& e)
   { return e._restart_type; }
   static const re_detail::re_syntax_base* BOOST_REGEX_CALL first(const reg_expression& e)
   { return (const re_detail::re_syntax_base*)e.data.data(); }
   static const unsigned char* BOOST_REGEX_CALL get_map(const reg_expression& e)
   { return e.startmap; }
   static std::size_t BOOST_REGEX_CALL leading_length(const reg_expression& e)
   { return e._leading_len; }
   static const re_detail::kmp_info<charT>* get_kmp(const reg_expression& e)
   { return e.pkmp; }
   static bool BOOST_REGEX_CALL can_start(charT c, const unsigned char* _map, unsigned char mask, const re_detail::_wide_type&);
   static bool BOOST_REGEX_CALL can_start(charT c, const unsigned char* _map, unsigned char mask, const re_detail::_narrow_type&);
};

#ifdef BOOST_MSVC
#pragma warning (pop)
#endif

template <class charT, class traits, class Allocator>
inline void BOOST_REGEX_CALL reg_expression<charT, traits, Allocator>::swap(reg_expression& that)throw()
{
   // this is not as efficient as it should be,
   // however swapping traits classes is problematic
   // so just use 'brute force' method for now:
   reg_expression<charT, traits, Allocator> e(that);
   that = *this;
   *this = e;
}


//
// class match_results and match_results_base
// handles what matched where

template <class iterator>
struct sub_match
{
   typedef typename re_detail::regex_iterator_traits<iterator>::value_type       value_type;
#if defined(BOOST_NO_STD_ITERATOR_TRAITS) || defined(BOOST_NO_TEMPLATE_PARTIAL_SPECIALIZATION)
   typedef std::ptrdiff_t  difference_type;
#else
   typedef typename re_detail::regex_iterator_traits<iterator>::difference_type       difference_type;
#endif
   typedef iterator                                                  iterator_type;
   
   iterator first;
   iterator second;
   bool matched;

   operator std::basic_string<value_type> ()const
   {
      std::basic_string<value_type> result;
      std::size_t len = boost::re_detail::distance((iterator)first, (iterator)second);
      result.reserve(len);
      iterator i = first;
      while(i != second)
      {
         result.append(1, *i);
         ++i;
      }
      return result;
   }
   #ifdef BOOST_OLD_REGEX_H
   //
   // the following are deprecated, do not use!!
   //
   operator int()const;
   operator unsigned int()const;
   operator short()const
   {
      return (short)(int)(*this);
   }
   operator unsigned short()const
   {
      return (unsigned short)(unsigned int)(*this);
   }
   #endif
   sub_match() { matched = false; }
   sub_match(iterator i) : first(i), second(i), matched(false) {}

   bool operator==(const sub_match& that)const
   {
      return (first == that.first) && (second == that.second) && (matched == that.matched);
   }
   bool BOOST_REGEX_CALL operator !=(const sub_match& that)const
   { return !(*this == that); }

   difference_type BOOST_REGEX_CALL length()const
   {
      difference_type n = boost::re_detail::distance((iterator)first, (iterator)second);
      return n;
   }
};

#ifdef BOOST_OLD_REGEX_H
namespace re_detail{
template <class iterator, class charT>
int do_toi(iterator i, iterator j, char c, int radix)
{
   std::string s(i, j);
   char* p;
   int result = std::strtol(s.c_str(), &p, radix);
#ifndef BOOST_NO_EXCEPTIONS
   if(*p)throw bad_pattern("Bad sub-expression");
#endif
   BOOST_REGEX_NOEH_ASSERT(0 == *p)
   return result;
}

//