regex.hpp

来自「CGAL is a collaborative effort of severa」· HPP 代码 · 共 1,634 行 · 第 1/4 页

/*
 *
 * Copyright (c) 1998-2002
 * Dr John Maddock
 *
 * Use, modification and distribution are subject to the 
 * Boost Software License, Version 1.0. (See accompanying file 
 * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt)
 *
 */

 /*
  *   LOCATION:    see http://www.boost.org for most recent version.
  *   FILE         regex.cpp
  *   VERSION      see <boost/version.hpp>
  *   DESCRIPTION: Declares boost::reg_expression<> and associated
  *                functions and classes. This header is the main
  *                entry point for the template regex code.
  */


/* start with C compatibility API */

#ifndef BOOST_RE_REGEX_HPP_INCLUDED
#define BOOST_RE_REGEX_HPP_INCLUDED

#ifndef BOOST_RE_CREGEX_HPP
#include <boost/cregex.hpp>
#endif

#ifdef __cplusplus

// what follows is all C++ don't include in C builds!!

#ifdef BOOST_REGEX_DEBUG
# include <iosfwd>
#endif

#include <new>
#include <cstring>
#ifndef BOOST_REGEX_CONFIG_HPP
#include <boost/regex/config.hpp>
#endif
#ifndef BOOST_REGEX_FWD_HPP
#include <boost/regex_fwd.hpp>
#endif
#ifndef BOOST_REGEX_STACK_HPP
#include <boost/regex/v3/regex_stack.hpp>
#endif
#ifndef BOOST_REGEX_RAW_BUFFER_HPP
#include <boost/regex/v3/regex_raw_buffer.hpp>
#endif
#ifndef BOOST_REGEX_KMP_HPP
#include <boost/regex/v3/regex_kmp.hpp>
#endif
#ifndef BOOST_RE_PAT_EXCEPT_HPP
#include <boost/regex/pattern_except.hpp>
#endif
#ifndef BOOST_REGEX_TRAITS_HPP
#include <boost/regex/regex_traits.hpp>
#endif
#include <boost/type_traits/remove_cv.hpp>
#include <boost/scoped_array.hpp>


namespace boost{

#ifdef __BORLANDC__
   #pragma option push -a8 -b -Vx -Ve -pc -w-8027
#endif

namespace re_detail{

struct re_set_long;
struct re_syntax_base;

} // namespace re_detail

namespace deprecated{
//
// class char_regex_traits_i
// provides case insensitive traits classes (deprecated):
template <class charT>
class char_regex_traits_i : public regex_traits<charT> {};

template<>
class char_regex_traits_i<char> : public regex_traits<char>
{
public:
   typedef char char_type;
   typedef unsigned char uchar_type;
   typedef unsigned int size_type;
   typedef regex_traits<char> base_type;

   char BOOST_REGEX_CALL translate(char c, bool)const
   {
      return static_cast<const regex_traits<char>*>(this)->translate(c, true);
   }
};

#ifndef BOOST_NO_WREGEX
template<>
class char_regex_traits_i<wchar_t> : public regex_traits<wchar_t>
{
public:
   typedef wchar_t char_type;
   typedef unsigned short uchar_type;
   typedef unsigned int size_type;
   typedef regex_traits<wchar_t> base_type;

   wchar_t BOOST_REGEX_CALL translate(wchar_t c, bool)const
   {
      return static_cast<const regex_traits<wchar_t>*>(this)->translate(c, true);
   }
   boost::uint_fast32_t BOOST_REGEX_CALL lookup_classname(const wchar_t* first, const wchar_t* last)const
   {
      boost::uint_fast32_t result = static_cast<const regex_traits<wchar_t>*>(this)->lookup_classname(first, last);
      if((result & base_type::char_class_upper) == base_type::char_class_upper)
         result |= base_type::char_class_alpha;
      return result;
   }
};
#endif
} // namespace deprecated


namespace re_detail{

enum mask_type
{
   mask_take = 1,
   mask_skip = 2,
   mask_any = mask_skip | mask_take,
   mask_all = mask_any
};

struct _narrow_type{};
struct _wide_type{};

template <class charT>
class is_byte;

template<>
class is_byte<char>
{
public:
   typedef _narrow_type width_type;
};

template<>
class is_byte<unsigned char>
{
public:
   typedef _narrow_type width_type;
};

template<>
class is_byte<signed char>
{
public:
   typedef _narrow_type width_type;
};

template <class charT>
class is_byte
{
public:
   typedef _wide_type width_type;
};


//
// compiled structures
//
// the following defs describe the format of the compiled string
//

//
// enum syntax_element_type
// describes the type of a record
enum syntax_element_type
{
   syntax_element_startmark = 0,
   syntax_element_endmark = syntax_element_startmark + 1,
   syntax_element_literal = syntax_element_endmark + 1,
   syntax_element_start_line = syntax_element_literal + 1,
   syntax_element_end_line = syntax_element_start_line + 1,
   syntax_element_wild = syntax_element_end_line + 1,
   syntax_element_match = syntax_element_wild + 1,
   syntax_element_word_boundary = syntax_element_match + 1,
   syntax_element_within_word = syntax_element_word_boundary + 1,
   syntax_element_word_start = syntax_element_within_word + 1,
   syntax_element_word_end = syntax_element_word_start + 1,
   syntax_element_buffer_start = syntax_element_word_end + 1,
   syntax_element_buffer_end = syntax_element_buffer_start + 1,
   syntax_element_backref = syntax_element_buffer_end + 1,
   syntax_element_long_set = syntax_element_backref + 1,
   syntax_element_set = syntax_element_long_set + 1,
   syntax_element_jump = syntax_element_set + 1,
   syntax_element_alt = syntax_element_jump + 1,
   syntax_element_rep = syntax_element_alt + 1,
   syntax_element_combining = syntax_element_rep + 1,
   syntax_element_soft_buffer_end = syntax_element_combining + 1,
   syntax_element_restart_continue = syntax_element_soft_buffer_end + 1
};

#ifdef BOOST_REGEX_DEBUG
// dwa 09/26/00 - This is needed to suppress warnings about an ambiguous conversion
std::ostream& operator<<(std::ostream&, syntax_element_type);
#endif

union offset_type
{
   re_syntax_base* p;
   std::size_t i;
};

//
// struct re_syntax_base
// base class for all syntax types:
struct re_syntax_base
{
   syntax_element_type type;
   offset_type next;
   unsigned int can_be_null;
};

//
// struct re_brace
// marks start or end of (...)
struct re_brace : public re_syntax_base
{
   int index;
};

//
// struct re_literal
// marks a literal string and
// is followed by an array of charT[length]:
struct re_literal : public re_syntax_base
{
   unsigned int length;
};

//
// struct re_long_set
// provides data for sets [...] containing
// wide characters
struct re_set_long : public re_syntax_base
{
   unsigned int csingles, cranges, cequivalents;
   boost::uint_fast32_t cclasses;
   bool isnot;
};

//
// struct re_set
// provides a map of bools for sets containing
// narrow, single byte characters.
struct re_set : public re_syntax_base
{
   unsigned char _map[256];
};

//
// struct re_jump
// provides alternative next destination
struct re_jump : public re_syntax_base
{
   offset_type alt;
   unsigned char _map[256];
};

//
// struct re_repeat
// provides repeat expressions
struct re_repeat : public re_jump
{
   unsigned min, max;
   int id;
   bool leading;
   bool greedy;
   bool singleton;
};


//
// enum re_jump_size_type
// provides compiled size of re_jump
// allowing for trailing alignment
// provide this so we know how many
// bytes to insert
enum re_jump_size_type
{
   re_jump_size = (sizeof(re_jump) + padding_mask) & ~(padding_mask),
   re_repeater_size = (sizeof(re_repeat) + padding_mask) & ~(padding_mask)
};

} // namespace re_detail

//
// class basic_regex
// handles error codes and flags

class BOOST_REGEX_DECL regbase
{
public:
   enum flag_type_
   {
      escape_in_lists = 1,                     // '\' special inside [...]
      char_classes = escape_in_lists << 1,     // [[:CLASS:]] allowed
      intervals = char_classes << 1,           // {x,y} allowed
      limited_ops = intervals << 1,            // all of + ? and | are normal characters
      newline_alt = limited_ops << 1,          // \n is the same as |
      bk_plus_qm = newline_alt << 1,           // uses \+ and \?
      bk_braces = bk_plus_qm << 1,             // uses \{ and \}
      bk_parens = bk_braces << 1,              // uses \( and \)
      bk_refs = bk_parens << 1,                // \d allowed
      bk_vbar = bk_refs << 1,                  // uses \|

      use_except = bk_vbar << 1,               // exception on error
      failbit = use_except << 1,               // error flag
      literal = failbit << 1,                  // all characters are literals
      icase = literal << 1,                    // characters are matched regardless of case
      nocollate = icase << 1,                  // don't use locale specific collation

      basic = char_classes | intervals | limited_ops | bk_braces | bk_parens | bk_refs,
      extended = char_classes | intervals | bk_refs,
      normal = escape_in_lists | char_classes | intervals | bk_refs | nocollate,
      emacs = bk_braces | bk_parens | bk_refs | bk_vbar,
      awk = extended | escape_in_lists,
      grep = basic | newline_alt,
      egrep = extended | newline_alt,
      sed = basic,
      perl = normal
   };
   typedef unsigned int flag_type;

   enum restart_info
   {
      restart_any = 0,
      restart_word = 1,
      restart_line = 2,
      restart_buf = 3,
      restart_continue = 4,
      restart_lit = 5,
      restart_fixed_lit = 6
   };

   flag_type BOOST_REGEX_CALL flags()const
   {
      return _flags;
   }

   regbase();
   regbase(const regbase& b);
protected:
   flag_type _flags;
};

//
// some forward declarations:
namespace re_detail{
template <class iterator, class Allocator>
class _priv_match_data;

#if defined(BOOST_NO_STD_ITERATOR_TRAITS) || defined(BOOST_NO_TEMPLATE_PARTIAL_SPECIALIZATION)

template <class T>
struct regex_iterator_traits 
{
  typedef typename T::iterator_category iterator_category;
  typedef typename T::value_type        value_type;
#if !defined(BOOST_NO_STD_ITERATOR)
  typedef typename T::difference_type   difference_type;
  typedef typename T::pointer           pointer;
  typedef typename T::reference         reference;
#else
  typedef std::ptrdiff_t                difference_type;
  typedef value_type*                   pointer;
  typedef value_type&                   reference;
#endif
};

template <class T>
struct pointer_iterator_traits
{
   typedef std::ptrdiff_t difference_type;
   typedef T value_type;
   typedef T* pointer;
   typedef T& reference;
   typedef std::random_access_iterator_tag iterator_category;
};
template <class T>
struct const_pointer_iterator_traits
{
   typedef std::ptrdiff_t difference_type;
   typedef T value_type;
   typedef const T* pointer;
   typedef const T& reference;
   typedef std::random_access_iterator_tag iterator_category;
};

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