[thirdparty/gcc.git] / libstdc++-v3 / include / bits / regex_compiler.tcc

// class template regex -*- C++ -*-

// Copyright (C) 2013-2014 Free Software Foundation, Inc.
//
// This file is part of the GNU ISO C++ Library.  This library is free
// software; you can redistribute it and/or modify it under the
// terms of the GNU General Public License as published by the
// Free Software Foundation; either version 3, or (at your option)
// any later version.

// This library is distributed in the hope that it will be useful,
// but WITHOUT ANY WARRANTY; without even the implied warranty of
// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
// GNU General Public License for more details.

// Under Section 7 of GPL version 3, you are granted additional
// permissions described in the GCC Runtime Library Exception, version
// 3.1, as published by the Free Software Foundation.

// You should have received a copy of the GNU General Public License and
// a copy of the GCC Runtime Library Exception along with this program;
// see the files COPYING3 and COPYING.RUNTIME respectively.  If not, see
// <http://www.gnu.org/licenses/>.

/**
 *  @file bits/regex_compiler.tcc
 *  This is an internal header file, included by other library headers.
 *  Do not attempt to use it directly. @headername{regex}
 */

// FIXME make comments doxygen format.

// This compiler refers to "Regular Expression Matching Can Be Simple And Fast"
// (http://swtch.com/~rsc/regexp/regexp1.html"),
// but doesn't strictly follow it.
//
// When compiling, states are *chained* instead of tree- or graph-constructed.
// It's more like structured programs: there's if statement and loop statement.
//
// For alternative structure(say "a|b"), aka "if statement", two branchs should
// be constructed. However, these two shall merge to an "end_tag" at the end of
// this operator:
//
//                branch1
//              /        \
// => begin_tag            end_tag =>
//              \        /
//                branch2
//
// This is the difference between this implementation and that in Russ's
// article.
//
// That's why we introduced dummy node here ------ "end_tag" is a dummy node.
// All dummy node will be eliminated at the end of compiling process.

namespace std _GLIBCXX_VISIBILITY(default)
{
namespace __detail
{
_GLIBCXX_BEGIN_NAMESPACE_VERSION

  template<typename _FwdIter, typename _TraitsT>
    _Compiler<_FwdIter, _TraitsT>::
    _Compiler(_FwdIter __b, _FwdIter __e,
	      const _TraitsT& __traits, _FlagT __flags)
    : _M_flags((__flags
		& (regex_constants::ECMAScript
		   | regex_constants::basic
		   | regex_constants::extended
		   | regex_constants::grep
		   | regex_constants::egrep
		   | regex_constants::awk))
	       ? __flags
	       : __flags | regex_constants::ECMAScript),
    _M_traits(__traits),
    _M_ctype(std::use_facet<_CtypeT>(_M_traits.getloc())),
    _M_scanner(__b, __e, _M_flags, _M_traits.getloc()),
    _M_nfa(_M_flags)
    {
      _StateSeqT __r(_M_nfa, _M_nfa._M_start());
      __r._M_append(_M_nfa._M_insert_subexpr_begin());
      this->_M_disjunction();
      if (!_M_match_token(_ScannerT::_S_token_eof))
	__throw_regex_error(regex_constants::error_paren);
      __r._M_append(_M_pop());
      _GLIBCXX_DEBUG_ASSERT(_M_stack.empty());
      __r._M_append(_M_nfa._M_insert_subexpr_end());
      __r._M_append(_M_nfa._M_insert_accept());
      _M_nfa._M_eliminate_dummy();
    }

  template<typename _FwdIter, typename _TraitsT>
    void
    _Compiler<_FwdIter, _TraitsT>::
    _M_disjunction()
    {
      this->_M_alternative();
      while (_M_match_token(_ScannerT::_S_token_or))
	{
	  _StateSeqT __alt1 = _M_pop();
	  this->_M_alternative();
	  _StateSeqT __alt2 = _M_pop();
	  auto __end = _M_nfa._M_insert_dummy();
	  __alt1._M_append(__end);
	  __alt2._M_append(__end);
	  _M_stack.push(_StateSeqT(_M_nfa,
				   _M_nfa._M_insert_alt(__alt1._M_start,
							__alt2._M_start, false),
				   __end));
	}
    }

  template<typename _FwdIter, typename _TraitsT>
    void
    _Compiler<_FwdIter, _TraitsT>::
    _M_alternative()
    {
      if (this->_M_term())
	{
	  _StateSeqT __re = _M_pop();
	  this->_M_alternative();
	  __re._M_append(_M_pop());
	  _M_stack.push(__re);
	}
      else
	_M_stack.push(_StateSeqT(_M_nfa, _M_nfa._M_insert_dummy()));
    }

  template<typename _FwdIter, typename _TraitsT>
    bool
    _Compiler<_FwdIter, _TraitsT>::
    _M_term()
    {
      if (this->_M_assertion())
	return true;
      if (this->_M_atom())
	{
	  this->_M_quantifier();
	  return true;
	}
      return false;
    }

  template<typename _FwdIter, typename _TraitsT>
    bool
    _Compiler<_FwdIter, _TraitsT>::
    _M_assertion()
    {
      if (_M_match_token(_ScannerT::_S_token_line_begin))
	_M_stack.push(_StateSeqT(_M_nfa, _M_nfa._M_insert_line_begin()));
      else if (_M_match_token(_ScannerT::_S_token_line_end))
	_M_stack.push(_StateSeqT(_M_nfa, _M_nfa._M_insert_line_end()));
      else if (_M_match_token(_ScannerT::_S_token_word_bound))
	// _M_value[0] == 'n' means it's negtive, say "not word boundary".
	_M_stack.push(_StateSeqT(_M_nfa, _M_nfa.
	      _M_insert_word_bound(_M_value[0] == 'n')));
      else if (_M_match_token(_ScannerT::_S_token_subexpr_lookahead_begin))
	{
	  auto __neg = _M_value[0] == 'n';
	  this->_M_disjunction();
	  if (!_M_match_token(_ScannerT::_S_token_subexpr_end))
	    __throw_regex_error(regex_constants::error_paren);
	  auto __tmp = _M_pop();
	  __tmp._M_append(_M_nfa._M_insert_accept());
	  _M_stack.push(
	      _StateSeqT(
		_M_nfa,
		_M_nfa._M_insert_lookahead(__tmp._M_start, __neg)));
	}
      else
	return false;
      return true;
    }

  template<typename _FwdIter, typename _TraitsT>
    void
    _Compiler<_FwdIter, _TraitsT>::
    _M_quantifier()
    {
      bool __neg = (_M_flags & regex_constants::ECMAScript);
      auto __init = [this, &__neg]()
	{
	  if (_M_stack.empty())
	    __throw_regex_error(regex_constants::error_badrepeat);
	  __neg = __neg && _M_match_token(_ScannerT::_S_token_opt);
	};
      if (_M_match_token(_ScannerT::_S_token_closure0))
	{
	  __init();
	  auto __e = _M_pop();
	  _StateSeqT __r(_M_nfa, _M_nfa._M_insert_alt(_S_invalid_state_id,
						      __e._M_start, __neg));
	  __e._M_append(__r);
	  _M_stack.push(__r);
	}
      else if (_M_match_token(_ScannerT::_S_token_closure1))
	{
	  __init();
	  auto __e = _M_pop();
	  __e._M_append(_M_nfa._M_insert_alt(_S_invalid_state_id, __e._M_start,
					     __neg));
	  _M_stack.push(__e);
	}
      else if (_M_match_token(_ScannerT::_S_token_opt))
	{
	  __init();
	  auto __e = _M_pop();
	  auto __end = _M_nfa._M_insert_dummy();
	  _StateSeqT __r(_M_nfa, _M_nfa._M_insert_alt(_S_invalid_state_id,
						      __e._M_start, __neg));
	  __e._M_append(__end);
	  __r._M_append(__end);
	  _M_stack.push(__r);
	}
      else if (_M_match_token(_ScannerT::_S_token_interval_begin))
	{
	  if (_M_stack.empty())
	    __throw_regex_error(regex_constants::error_badrepeat);
	  if (!_M_match_token(_ScannerT::_S_token_dup_count))
	    __throw_regex_error(regex_constants::error_badbrace);
	  _StateSeqT __r(_M_pop());
	  _StateSeqT __e(_M_nfa, _M_nfa._M_insert_dummy());
	  long __min_rep = _M_cur_int_value(10);
	  bool __infi = false;
	  long __n;

	  // {3
	  if (_M_match_token(_ScannerT::_S_token_comma))
	    if (_M_match_token(_ScannerT::_S_token_dup_count)) // {3,7}
	      __n = _M_cur_int_value(10) - __min_rep;
	    else
	      __infi = true;
	  else
	    __n = 0;
	  if (!_M_match_token(_ScannerT::_S_token_interval_end))
	    __throw_regex_error(regex_constants::error_brace);

	  __neg = __neg && _M_match_token(_ScannerT::_S_token_opt);

	  for (long __i = 0; __i < __min_rep; ++__i)
	    __e._M_append(__r._M_clone());

	  if (__infi)
	    {
	      auto __tmp = __r._M_clone();
	      _StateSeqT __s(_M_nfa,
			     _M_nfa._M_insert_alt(_S_invalid_state_id,
						  __tmp._M_start, __neg));
	      __tmp._M_append(__s);
	      __e._M_append(__s);
	    }
	  else
	    {
	      if (__n < 0)
		__throw_regex_error(regex_constants::error_badbrace);
	      auto __end = _M_nfa._M_insert_dummy();
	      // _M_alt is the "match more" branch, and _M_next is the
	      // "match less" one. Switch _M_alt and _M_next of all created
	      // nodes. This is a hacking but IMO works well.
	      std::stack<_StateIdT> __stack;
	      for (long __i = 0; __i < __n; ++__i)
		{
		  auto __tmp = __r._M_clone();
		  auto __alt = _M_nfa._M_insert_alt(__tmp._M_start,
						    __end, __neg);
		  __stack.push(__alt);
		  __e._M_append(_StateSeqT(_M_nfa, __alt, __tmp._M_end));
		}
	      __e._M_append(__end);
	      while (!__stack.empty())
		{
		  auto& __tmp = _M_nfa[__stack.top()];
		  __stack.pop();
		  swap(__tmp._M_next, __tmp._M_alt);
		}
	    }
	  _M_stack.push(__e);
	}
    }

  template<typename _FwdIter, typename _TraitsT>
    bool
    _Compiler<_FwdIter, _TraitsT>::
    _M_atom()
    {
      if (_M_match_token(_ScannerT::_S_token_anychar))
	_M_stack.push(_StateSeqT(_M_nfa,
				_M_nfa._M_insert_matcher
				(_AnyMatcher<_TraitsT>(_M_traits))));
      else if (_M_try_char())
	_M_stack.push(_StateSeqT(_M_nfa,
				 _M_nfa._M_insert_matcher
				 (_CharMatcher<_TraitsT>(_M_value[0],
								 _M_traits,
								 _M_flags))));
      else if (_M_match_token(_ScannerT::_S_token_backref))
	_M_stack.push(_StateSeqT(_M_nfa, _M_nfa.
				 _M_insert_backref(_M_cur_int_value(10))));
      else if (_M_match_token(_ScannerT::_S_token_quoted_class))
	{
	  _GLIBCXX_DEBUG_ASSERT(_M_value.size() == 1);
	  _BMatcherT __matcher(_M_ctype.is(_CtypeT::upper, _M_value[0]),
			       _M_traits, _M_flags);
	  __matcher._M_add_character_class(_M_value);
	  _M_stack.push(_StateSeqT(_M_nfa,
		_M_nfa._M_insert_matcher(std::move(__matcher))));
	}
      else if (_M_match_token(_ScannerT::_S_token_subexpr_no_group_begin))
	{
	  _StateSeqT __r(_M_nfa, _M_nfa._M_insert_dummy());
	  this->_M_disjunction();
	  if (!_M_match_token(_ScannerT::_S_token_subexpr_end))
	    __throw_regex_error(regex_constants::error_paren);
	  __r._M_append(_M_pop());
	  _M_stack.push(__r);
	}
      else if (_M_match_token(_ScannerT::_S_token_subexpr_begin))
	{
	  _StateSeqT __r(_M_nfa, _M_nfa._M_insert_subexpr_begin());
	  this->_M_disjunction();
	  if (!_M_match_token(_ScannerT::_S_token_subexpr_end))
	    __throw_regex_error(regex_constants::error_paren);
	  __r._M_append(_M_pop());
	  __r._M_append(_M_nfa._M_insert_subexpr_end());
	  _M_stack.push(__r);
	}
      else if (!_M_bracket_expression())
	return false;
      return true;
    }

  template<typename _FwdIter, typename _TraitsT>
    bool
    _Compiler<_FwdIter, _TraitsT>::
    _M_bracket_expression()
    {
      bool __neg =
	_M_match_token(_ScannerT::_S_token_bracket_neg_begin);
      if (!(__neg || _M_match_token(_ScannerT::_S_token_bracket_begin)))
	return false;
      _BMatcherT __matcher(__neg, _M_traits, _M_flags);
      while (!_M_match_token(_ScannerT::_S_token_bracket_end))
	_M_expression_term(__matcher);
      _M_stack.push(_StateSeqT(_M_nfa,
			       _M_nfa._M_insert_matcher(std::move(__matcher))));
      return true;
    }

  template<typename _FwdIter, typename _TraitsT>
    void
    _Compiler<_FwdIter, _TraitsT>::
    _M_expression_term(_BMatcherT& __matcher)
    {
      if (_M_match_token(_ScannerT::_S_token_collsymbol))
	__matcher._M_add_collating_element(_M_value);
      else if (_M_match_token(_ScannerT::_S_token_equiv_class_name))
	__matcher._M_add_equivalence_class(_M_value);
      else if (_M_match_token(_ScannerT::_S_token_char_class_name))
	__matcher._M_add_character_class(_M_value);
      else if (_M_try_char()) // [a
	{
	  auto __ch = _M_value[0];
	  if (_M_try_char())
	    {
	      if (_M_value[0] == '-') // [a-
		{
		  if (_M_try_char()) // [a-z]
		    {
		      __matcher._M_make_range(__ch, _M_value[0]);
		      return;
		    }
		  // If the dash is the last character in the bracket
		  // expression, it is not special.
		  if (_M_scanner._M_get_token()
		      != _ScannerT::_S_token_bracket_end)
		    __throw_regex_error(regex_constants::error_range);
		}
	      __matcher._M_add_char(_M_value[0]);
	    }
	  __matcher._M_add_char(__ch);
	}
      else
	__throw_regex_error(regex_constants::error_brack);
    }

  template<typename _FwdIter, typename _TraitsT>
    bool
    _Compiler<_FwdIter, _TraitsT>::
    _M_try_char()
    {
      bool __is_char = false;
      if (_M_match_token(_ScannerT::_S_token_oct_num))
	{
	  __is_char = true;
	  _M_value.assign(1, _M_cur_int_value(8));
	}
      else if (_M_match_token(_ScannerT::_S_token_hex_num))
	{
	  __is_char = true;
	  _M_value.assign(1, _M_cur_int_value(16));
	}
      else if (_M_match_token(_ScannerT::_S_token_ord_char))
	__is_char = true;
      return __is_char;
    }

  template<typename _FwdIter, typename _TraitsT>
    bool
    _Compiler<_FwdIter, _TraitsT>::
    _M_match_token(_TokenT token)
    {
      if (token == _M_scanner._M_get_token())
	{
	  _M_value = _M_scanner._M_get_value();
	  _M_scanner._M_advance();
	  return true;
	}
      return false;
    }

  template<typename _FwdIter, typename _TraitsT>
    int
    _Compiler<_FwdIter, _TraitsT>::
    _M_cur_int_value(int __radix)
    {
      long __v = 0;
      for (typename _StringT::size_type __i = 0;
	   __i < _M_value.length(); ++__i)
	__v =__v * __radix + _M_traits.value(_M_value[__i], __radix);
      return __v;
    }

  template<typename _TraitsT>
    bool
    _BracketMatcher<_TraitsT>::operator()(_CharT __ch) const
    {
      bool __ret = false;
      if (_M_traits.isctype(__ch, _M_class_set)
	  || _M_char_set.count(_M_translate(__ch))
	  || _M_equiv_set.count(_M_traits.transform_primary(&__ch, &__ch+1)))
	__ret = true;
      else
	{
	  _StringT __s = _M_get_str(_M_flags & regex_constants::collate
				    ? _M_translate(__ch) : __ch);
	  for (auto& __it : _M_range_set)
	    if (__it.first <= __s && __s <= __it.second)
	      {
		__ret = true;
		break;
	      }
	}
      if (_M_is_non_matching)
	return !__ret;
      else
	return __ret;
    }

_GLIBCXX_END_NAMESPACE_VERSION
} // namespace __detail
} // namespace
Commit	Line	Data
6cb784b6 TS	1	// class template regex -- C++ --
6cb784b6 TS	2
aa118a03	3	// Copyright (C) 2013-2014 Free Software Foundation, Inc.
6cb784b6 TS	4	//
	5	// This file is part of the GNU ISO C++ Library. This library is free
	6	// software; you can redistribute it and/or modify it under the
	7	// terms of the GNU General Public License as published by the
	8	// Free Software Foundation; either version 3, or (at your option)
	9	// any later version.
	10
	11	// This library is distributed in the hope that it will be useful,
	12	// but WITHOUT ANY WARRANTY; without even the implied warranty of
	13	// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
	14	// GNU General Public License for more details.
	15
	16	// Under Section 7 of GPL version 3, you are granted additional
	17	// permissions described in the GCC Runtime Library Exception, version
	18	// 3.1, as published by the Free Software Foundation.
	19
	20	// You should have received a copy of the GNU General Public License and
	21	// a copy of the GCC Runtime Library Exception along with this program;
	22	// see the files COPYING3 and COPYING.RUNTIME respectively. If not, see
	23	// <http://www.gnu.org/licenses/>.
	24
	25	/**
	26	* @file bits/regex_compiler.tcc
	27	* This is an internal header file, included by other library headers.
	28	* Do not attempt to use it directly. @headername{regex}
	29	*/
	30
b21abcee	31	// FIXME make comments doxygen format.
7c812a2a TS	32
	33	// This compiler refers to "Regular Expression Matching Can Be Simple And Fast"
	34	// (http://swtch.com/~rsc/regexp/regexp1.html"),
	35	// but doesn't strictly follow it.
	36	//
	37	// When compiling, states are chained instead of tree- or graph-constructed.
	38	// It's more like structured programs: there's if statement and loop statement.
	39	//
	40	// For alternative structure(say "a\|b"), aka "if statement", two branchs should
	41	// be constructed. However, these two shall merge to an "end_tag" at the end of
	42	// this operator:
	43	//
	44	// branch1
	45	// / \
	46	// => begin_tag end_tag =>
	47	// \ /
	48	// branch2
	49	//
	50	// This is the difference between this implementation and that in Russ's
	51	// article.
	52	//
	53	// That's why we introduced dummy node here ------ "end_tag" is a dummy node.
	54	// All dummy node will be eliminated at the end of compiling process.
	55
6cb784b6 TS	56	namespace std _GLIBCXX_VISIBILITY(default)
	57	{
	58	namespace __detail
	59	{
	60	_GLIBCXX_BEGIN_NAMESPACE_VERSION
	61
68e69ce2 JW	62	template<typename _FwdIter, typename _TraitsT>
68e69ce2 JW	63	_Compiler<_FwdIter, _TraitsT>::
33fbbb76	64	_Compiler(_FwdIter __b, _FwdIter __e,
e280b6ff	65	const _TraitsT& __traits, _FlagT __flags)
c2669da9 TS	66	: _M_flags((__flags
	67	& (regex_constants::ECMAScript
	68	\| regex_constants::basic
	69	\| regex_constants::extended
	70	\| regex_constants::grep
	71	\| regex_constants::egrep
	72	\| regex_constants::awk))
	73	? __flags
	74	: __flags \| regex_constants::ECMAScript),
	75	_M_traits(__traits),
68e69ce2	76	_M_ctype(std::use_facet<_CtypeT>(_M_traits.getloc())),
ab1c993b	77	_M_scanner(__b, __e, _M_flags, _M_traits.getloc()),
c2669da9	78	_M_nfa(_M_flags)
6cb784b6	79	{
7c812a2a TS	80	_StateSeqT __r(_M_nfa, _M_nfa._M_start());
	81	__r._M_append(_M_nfa._M_insert_subexpr_begin());
	82	this->_M_disjunction();
	83	if (!_M_match_token(_ScannerT::_S_token_eof))
	84	__throw_regex_error(regex_constants::error_paren);
	85	__r._M_append(_M_pop());
	86	_GLIBCXX_DEBUG_ASSERT(_M_stack.empty());
	87	__r._M_append(_M_nfa._M_insert_subexpr_end());
	88	__r._M_append(_M_nfa._M_insert_accept());
	89	_M_nfa._M_eliminate_dummy();
6cb784b6 TS	90	}
6cb784b6 TS	91
68e69ce2	92	template<typename _FwdIter, typename _TraitsT>
6cb784b6	93	void
68e69ce2	94	_Compiler<_FwdIter, _TraitsT>::
6cb784b6 TS	95	_M_disjunction()
	96	{
	97	this->_M_alternative();
7c812a2a	98	while (_M_match_token(_ScannerT::_S_token_or))
6cb784b6	99	{
7c812a2a TS	100	_StateSeqT __alt1 = _M_pop();
	101	this->_M_alternative();
	102	_StateSeqT __alt2 = _M_pop();
	103	auto __end = _M_nfa._M_insert_dummy();
	104	__alt1._M_append(__end);
	105	__alt2._M_append(__end);
	106	_M_stack.push(_StateSeqT(_M_nfa,
	107	_M_nfa._M_insert_alt(__alt1._M_start,
7b86458e	108	__alt2._M_start, false),
7c812a2a	109	__end));
6cb784b6 TS	110	}
	111	}
	112
68e69ce2	113	template<typename _FwdIter, typename _TraitsT>
6cb784b6	114	void
68e69ce2	115	_Compiler<_FwdIter, _TraitsT>::
6cb784b6 TS	116	_M_alternative()
	117	{
	118	if (this->_M_term())
	119	{
7c812a2a	120	_StateSeqT __re = _M_pop();
6cb784b6	121	this->_M_alternative();
7c812a2a	122	__re._M_append(_M_pop());
6cb784b6 TS	123	_M_stack.push(__re);
6cb784b6 TS	124	}
7c812a2a TS	125	else
7c812a2a TS	126	_M_stack.push(_StateSeqT(_M_nfa, _M_nfa._M_insert_dummy()));
6cb784b6 TS	127	}
6cb784b6 TS	128
68e69ce2	129	template<typename _FwdIter, typename _TraitsT>
6cb784b6	130	bool
68e69ce2	131	_Compiler<_FwdIter, _TraitsT>::
6cb784b6 TS	132	_M_term()
	133	{
	134	if (this->_M_assertion())
	135	return true;
	136	if (this->_M_atom())
	137	{
	138	this->_M_quantifier();
	139	return true;
	140	}
	141	return false;
	142	}
	143
68e69ce2	144	template<typename _FwdIter, typename _TraitsT>
6cb784b6	145	bool
68e69ce2	146	_Compiler<_FwdIter, _TraitsT>::
6cb784b6 TS	147	_M_assertion()
6cb784b6 TS	148	{
7c812a2a	149	if (_M_match_token(_ScannerT::_S_token_line_begin))
7d9d2185	150	_M_stack.push(_StateSeqT(_M_nfa, _M_nfa._M_insert_line_begin()));
7c812a2a	151	else if (_M_match_token(_ScannerT::_S_token_line_end))
7d9d2185	152	_M_stack.push(_StateSeqT(_M_nfa, _M_nfa._M_insert_line_end()));
7c812a2a	153	else if (_M_match_token(_ScannerT::_S_token_word_bound))
7b86458e TS	154	// _M_value[0] == 'n' means it's negtive, say "not word boundary".
	155	_M_stack.push(_StateSeqT(_M_nfa, _M_nfa.
	156	_M_insert_word_bound(_M_value[0] == 'n')));
7c812a2a	157	else if (_M_match_token(_ScannerT::_S_token_subexpr_lookahead_begin))
7b86458e TS	158	{
	159	auto __neg = _M_value[0] == 'n';
	160	this->_M_disjunction();
	161	if (!_M_match_token(_ScannerT::_S_token_subexpr_end))
	162	__throw_regex_error(regex_constants::error_paren);
	163	auto __tmp = _M_pop();
	164	__tmp._M_append(_M_nfa._M_insert_accept());
	165	_M_stack.push(
	166	_StateSeqT(
	167	_M_nfa,
	168	_M_nfa._M_insert_lookahead(__tmp._M_start, __neg)));
	169	}
7c812a2a TS	170	else
	171	return false;
	172	return true;
6cb784b6 TS	173	}
6cb784b6 TS	174
68e69ce2	175	template<typename _FwdIter, typename _TraitsT>
6cb784b6	176	void
68e69ce2	177	_Compiler<_FwdIter, _TraitsT>::
6cb784b6 TS	178	_M_quantifier()
6cb784b6 TS	179	{
c2669da9	180	bool __neg = (_M_flags & regex_constants::ECMAScript);
7b86458e	181	auto __init = [this, &__neg]()
6cb784b6 TS	182	{
	183	if (_M_stack.empty())
	184	__throw_regex_error(regex_constants::error_badrepeat);
7b86458e TS	185	__neg = __neg && _M_match_token(_ScannerT::_S_token_opt);
	186	};
	187	if (_M_match_token(_ScannerT::_S_token_closure0))
	188	{
	189	__init();
7c812a2a TS	190	auto __e = _M_pop();
7c812a2a TS	191	_StateSeqT __r(_M_nfa, _M_nfa._M_insert_alt(_S_invalid_state_id,
7b86458e	192	__e._M_start, __neg));
7c812a2a	193	__e._M_append(__r);
6cb784b6	194	_M_stack.push(__r);
6cb784b6	195	}
7c812a2a	196	else if (_M_match_token(_ScannerT::_S_token_closure1))
6cb784b6	197	{
7b86458e	198	__init();
7c812a2a	199	auto __e = _M_pop();
7b86458e TS	200	__e._M_append(_M_nfa._M_insert_alt(_S_invalid_state_id, __e._M_start,
7b86458e TS	201	__neg));
7c812a2a	202	_M_stack.push(__e);
6cb784b6	203	}
7c812a2a	204	else if (_M_match_token(_ScannerT::_S_token_opt))
6cb784b6	205	{
7b86458e	206	__init();
7c812a2a TS	207	auto __e = _M_pop();
	208	auto __end = _M_nfa._M_insert_dummy();
	209	_StateSeqT __r(_M_nfa, _M_nfa._M_insert_alt(_S_invalid_state_id,
7b86458e	210	__e._M_start, __neg));
7c812a2a TS	211	__e._M_append(__end);
7c812a2a TS	212	__r._M_append(__end);
6cb784b6	213	_M_stack.push(__r);
6cb784b6	214	}
7c812a2a	215	else if (_M_match_token(_ScannerT::_S_token_interval_begin))
6cb784b6	216	{
c2669da9 TS	217	if (_M_stack.empty())
c2669da9 TS	218	__throw_regex_error(regex_constants::error_badrepeat);
6cb784b6 TS	219	if (!_M_match_token(_ScannerT::_S_token_dup_count))
6cb784b6 TS	220	__throw_regex_error(regex_constants::error_badbrace);
7c812a2a TS	221	_StateSeqT __r(_M_pop());
7c812a2a TS	222	_StateSeqT __e(_M_nfa, _M_nfa._M_insert_dummy());
6cb43087	223	long __min_rep = _M_cur_int_value(10);
c2669da9	224	bool __infi = false;
6cb43087	225	long __n;
c2669da9	226
7c812a2a	227	// {3
6cb784b6	228	if (_M_match_token(_ScannerT::_S_token_comma))
7c812a2a	229	if (_M_match_token(_ScannerT::_S_token_dup_count)) // {3,7}
c2669da9 TS	230	__n = _M_cur_int_value(10) - __min_rep;
	231	else
	232	__infi = true;
	233	else
	234	__n = 0;
6cb784b6 TS	235	if (!_M_match_token(_ScannerT::_S_token_interval_end))
6cb784b6 TS	236	__throw_regex_error(regex_constants::error_brace);
c2669da9 TS	237
	238	__neg = __neg && _M_match_token(_ScannerT::_S_token_opt);
	239
6cb43087	240	for (long __i = 0; __i < __min_rep; ++__i)
c2669da9 TS	241	__e._M_append(__r._M_clone());
	242
	243	if (__infi)
	244	{
	245	auto __tmp = __r._M_clone();
	246	_StateSeqT __s(_M_nfa,
	247	_M_nfa._M_insert_alt(_S_invalid_state_id,
	248	__tmp._M_start, __neg));
	249	__tmp._M_append(__s);
	250	__e._M_append(__s);
	251	}
	252	else
	253	{
	254	if (__n < 0)
	255	__throw_regex_error(regex_constants::error_badbrace);
	256	auto __end = _M_nfa._M_insert_dummy();
	257	// _M_alt is the "match more" branch, and _M_next is the
	258	// "match less" one. Switch _M_alt and _M_next of all created
	259	// nodes. This is a hacking but IMO works well.
	260	std::stack<_StateIdT> __stack;
6cb43087	261	for (long __i = 0; __i < __n; ++__i)
c2669da9 TS	262	{
	263	auto __tmp = __r._M_clone();
	264	auto __alt = _M_nfa._M_insert_alt(__tmp._M_start,
	265	__end, __neg);
	266	__stack.push(__alt);
	267	__e._M_append(_StateSeqT(_M_nfa, __alt, __tmp._M_end));
	268	}
	269	__e._M_append(__end);
	270	while (!__stack.empty())
	271	{
	272	auto& __tmp = _M_nfa[__stack.top()];
	273	__stack.pop();
	274	swap(__tmp._M_next, __tmp._M_alt);
	275	}
	276	}
7c812a2a	277	_M_stack.push(__e);
6cb784b6 TS	278	}
	279	}
	280
68e69ce2	281	template<typename _FwdIter, typename _TraitsT>
6cb784b6	282	bool
68e69ce2	283	_Compiler<_FwdIter, _TraitsT>::
6cb784b6 TS	284	_M_atom()
	285	{
	286	if (_M_match_token(_ScannerT::_S_token_anychar))
7c812a2a TS	287	_M_stack.push(_StateSeqT(_M_nfa,
7c812a2a TS	288	_M_nfa._M_insert_matcher
bb038ece	289	(_AnyMatcher<_TraitsT>(_M_traits))));
7c812a2a TS	290	else if (_M_try_char())
	291	_M_stack.push(_StateSeqT(_M_nfa,
	292	_M_nfa._M_insert_matcher
bb038ece	293	(_CharMatcher<_TraitsT>(_M_value[0],
7c812a2a TS	294	_M_traits,
	295	_M_flags))));
	296	else if (_M_match_token(_ScannerT::_S_token_backref))
	297	_M_stack.push(_StateSeqT(_M_nfa, _M_nfa.
	298	_M_insert_backref(_M_cur_int_value(10))));
	299	else if (_M_match_token(_ScannerT::_S_token_quoted_class))
6cb784b6	300	{
7c812a2a TS	301	_GLIBCXX_DEBUG_ASSERT(_M_value.size() == 1);
	302	_BMatcherT __matcher(_M_ctype.is(_CtypeT::upper, _M_value[0]),
	303	_M_traits, _M_flags);
	304	__matcher._M_add_character_class(_M_value);
	305	_M_stack.push(_StateSeqT(_M_nfa,
7d9d2185	306	_M_nfa._M_insert_matcher(std::move(__matcher))));
6cb784b6	307	}
7c812a2a	308	else if (_M_match_token(_ScannerT::_S_token_subexpr_no_group_begin))
6cb784b6	309	{
7c812a2a TS	310	_StateSeqT __r(_M_nfa, _M_nfa._M_insert_dummy());
	311	this->_M_disjunction();
	312	if (!_M_match_token(_ScannerT::_S_token_subexpr_end))
	313	__throw_regex_error(regex_constants::error_paren);
	314	__r._M_append(_M_pop());
	315	_M_stack.push(__r);
6cb784b6	316	}
7c812a2a	317	else if (_M_match_token(_ScannerT::_S_token_subexpr_begin))
6cb784b6	318	{
7c812a2a	319	_StateSeqT __r(_M_nfa, _M_nfa._M_insert_subexpr_begin());
6cb784b6 TS	320	this->_M_disjunction();
	321	if (!_M_match_token(_ScannerT::_S_token_subexpr_end))
	322	__throw_regex_error(regex_constants::error_paren);
7c812a2a TS	323	__r._M_append(_M_pop());
7c812a2a TS	324	__r._M_append(_M_nfa._M_insert_subexpr_end());
6cb784b6	325	_M_stack.push(__r);
6cb784b6	326	}
7c812a2a TS	327	else if (!_M_bracket_expression())
	328	return false;
	329	return true;
6cb784b6 TS	330	}
6cb784b6 TS	331
68e69ce2	332	template<typename _FwdIter, typename _TraitsT>
6cb784b6	333	bool
68e69ce2	334	_Compiler<_FwdIter, _TraitsT>::
6cb784b6 TS	335	_M_bracket_expression()
6cb784b6 TS	336	{
33fbbb76 TS	337	bool __neg =
	338	_M_match_token(_ScannerT::_S_token_bracket_neg_begin);
	339	if (!(__neg \|\| _M_match_token(_ScannerT::_S_token_bracket_begin)))
e280b6ff	340	return false;
33fbbb76	341	_BMatcherT __matcher(__neg, _M_traits, _M_flags);
7c812a2a TS	342	while (!_M_match_token(_ScannerT::_S_token_bracket_end))
7c812a2a TS	343	_M_expression_term(__matcher);
7d9d2185 JW	344	_M_stack.push(_StateSeqT(_M_nfa,
7d9d2185 JW	345	_M_nfa._M_insert_matcher(std::move(__matcher))));
6cb784b6 TS	346	return true;
	347	}
	348
68e69ce2	349	template<typename _FwdIter, typename _TraitsT>
6cb784b6	350	void
68e69ce2	351	_Compiler<_FwdIter, _TraitsT>::
6cb784b6 TS	352	_M_expression_term(_BMatcherT& __matcher)
	353	{
	354	if (_M_match_token(_ScannerT::_S_token_collsymbol))
7c812a2a TS	355	__matcher._M_add_collating_element(_M_value);
	356	else if (_M_match_token(_ScannerT::_S_token_equiv_class_name))
	357	__matcher._M_add_equivalence_class(_M_value);
	358	else if (_M_match_token(_ScannerT::_S_token_char_class_name))
	359	__matcher._M_add_character_class(_M_value);
	360	else if (_M_try_char()) // [a
e280b6ff	361	{
33fbbb76 TS	362	auto __ch = _M_value[0];
33fbbb76 TS	363	if (_M_try_char())
e280b6ff	364	{
7c812a2a	365	if (_M_value[0] == '-') // [a-
e280b6ff	366	{
33fbbb76 TS	367	if (_M_try_char()) // [a-z]
	368	{
	369	__matcher._M_make_range(__ch, _M_value[0]);
	370	return;
	371	}
	372	// If the dash is the last character in the bracket
	373	// expression, it is not special.
	374	if (_M_scanner._M_get_token()
	375	!= _ScannerT::_S_token_bracket_end)
e280b6ff	376	__throw_regex_error(regex_constants::error_range);
e280b6ff	377	}
33fbbb76	378	__matcher._M_add_char(_M_value[0]);
e280b6ff	379	}
33fbbb76	380	__matcher._M_add_char(__ch);
e280b6ff	381	}
7c812a2a TS	382	else
7c812a2a TS	383	__throw_regex_error(regex_constants::error_brack);
6cb784b6 TS	384	}
6cb784b6 TS	385
68e69ce2	386	template<typename _FwdIter, typename _TraitsT>
33fbbb76	387	bool
68e69ce2	388	_Compiler<_FwdIter, _TraitsT>::
33fbbb76 TS	389	_M_try_char()
	390	{
	391	bool __is_char = false;
	392	if (_M_match_token(_ScannerT::_S_token_oct_num))
	393	{
	394	__is_char = true;
	395	_M_value.assign(1, _M_cur_int_value(8));
	396	}
	397	else if (_M_match_token(_ScannerT::_S_token_hex_num))
	398	{
	399	__is_char = true;
	400	_M_value.assign(1, _M_cur_int_value(16));
	401	}
	402	else if (_M_match_token(_ScannerT::_S_token_ord_char))
	403	__is_char = true;
	404	return __is_char;
	405	}
	406
68e69ce2	407	template<typename _FwdIter, typename _TraitsT>
7c812a2a	408	bool
68e69ce2	409	_Compiler<_FwdIter, _TraitsT>::
7c812a2a TS	410	_M_match_token(_TokenT token)
	411	{
	412	if (token == _M_scanner._M_get_token())
	413	{
	414	_M_value = _M_scanner._M_get_value();
	415	_M_scanner._M_advance();
	416	return true;
	417	}
	418	return false;
	419	}
	420
68e69ce2	421	template<typename _FwdIter, typename _TraitsT>
6cb784b6	422	int
68e69ce2	423	_Compiler<_FwdIter, _TraitsT>::
6cb784b6 TS	424	_M_cur_int_value(int __radix)
6cb784b6 TS	425	{
6cb43087	426	long __v = 0;
6cb784b6	427	for (typename _StringT::size_type __i = 0;
33fbbb76 TS	428	__i < _M_value.length(); ++__i)
33fbbb76 TS	429	__v =__v * __radix + _M_traits.value(_M_value[__i], __radix);
6cb784b6 TS	430	return __v;
	431	}
	432
bb038ece	433	template<typename _TraitsT>
7d9d2185	434	bool
bb038ece	435	_BracketMatcher<_TraitsT>::operator()(_CharT __ch) const
6cb784b6	436	{
6cb784b6	437	bool __ret = false;
756aa0c3 TS	438	if (_M_traits.isctype(__ch, _M_class_set)
	439	\|\| _M_char_set.count(_M_translate(__ch))
	440	\|\| _M_equiv_set.count(_M_traits.transform_primary(&__ch, &__ch+1)))
e3509691	441	__ret = true;
6cb784b6	442	else
e280b6ff	443	{
e3509691 TS	444	_StringT __s = _M_get_str(_M_flags & regex_constants::collate
	445	? _M_translate(__ch) : __ch);
	446	for (auto& __it : _M_range_set)
	447	if (__it.first <= __s && __s <= __it.second)
e280b6ff TS	448	{
	449	__ret = true;
	450	break;
	451	}
	452	}
6cb784b6	453	if (_M_is_non_matching)
33fbbb76 TS	454	return !__ret;
	455	else
	456	return __ret;
6cb784b6 TS	457	}
	458
	459	_GLIBCXX_END_NAMESPACE_VERSION
	460	} // namespace __detail
	461	} // namespace