libstdc++
regex_executor.h
Go to the documentation of this file.
1// class template regex -*- C++ -*-
2
3// Copyright (C) 2013-2022 Free Software Foundation, Inc.
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_executor.h
27 * This is an internal header file, included by other library headers.
28 * Do not attempt to use it directly. @headername{regex}
29 */
30
31// FIXME convert comments to doxygen format.
32
33namespace std _GLIBCXX_VISIBILITY(default)
34{
35_GLIBCXX_BEGIN_NAMESPACE_VERSION
36
37namespace __detail
38{
39 /**
40 * @addtogroup regex-detail
41 * @{
42 */
43
44 /**
45 * @brief Takes a regex and an input string and does the matching.
46 *
47 * The %_Executor class has two modes: DFS mode and BFS mode, controlled
48 * by the template parameter %__dfs_mode.
49 */
50 template<typename _BiIter, typename _Alloc, typename _TraitsT,
51 bool __dfs_mode>
53 {
55 using __dfs = true_type;
56 using __bfs = false_type;
57
58 enum class _Match_mode : unsigned char { _Exact, _Prefix };
59
60 public:
61 typedef typename iterator_traits<_BiIter>::value_type _CharT;
63 typedef _GLIBCXX_STD_C::vector<sub_match<_BiIter>, _Alloc> _ResultsVec;
65 typedef typename _TraitsT::char_class_type _ClassT;
66 typedef _NFA<_TraitsT> _NFAT;
67
68 public:
69 _Executor(_BiIter __begin,
70 _BiIter __end,
71 _ResultsVec& __results,
72 const _RegexT& __re,
73 _FlagT __flags)
74 : _M_begin(__begin),
75 _M_end(__end),
76 _M_re(__re),
77 _M_nfa(*__re._M_automaton),
78 _M_results(__results),
79 _M_rep_count(_M_nfa.size()),
80 _M_states(_M_nfa._M_start(), _M_nfa.size()),
81 _M_flags(__flags)
82 {
83 using namespace regex_constants;
84 if (__flags & match_prev_avail) // ignore not_bol and not_bow
85 _M_flags &= ~(match_not_bol | match_not_bow);
86 }
87
88 // Set matched when string exactly matches the pattern.
89 bool
90 _M_match()
91 {
92 _M_current = _M_begin;
93 return _M_main(_Match_mode::_Exact);
94 }
95
96 // Set matched when some prefix of the string matches the pattern.
97 bool
98 _M_search_from_first()
99 {
100 _M_current = _M_begin;
101 return _M_main(_Match_mode::_Prefix);
102 }
103
104 bool
105 _M_search();
106
107 private:
108 void
109 _M_rep_once_more(_Match_mode __match_mode, _StateIdT);
110
111 void
112 _M_handle_repeat(_Match_mode, _StateIdT);
113
114 void
115 _M_handle_subexpr_begin(_Match_mode, _StateIdT);
116
117 void
118 _M_handle_subexpr_end(_Match_mode, _StateIdT);
119
120 void
121 _M_handle_line_begin_assertion(_Match_mode, _StateIdT);
122
123 void
124 _M_handle_line_end_assertion(_Match_mode, _StateIdT);
125
126 void
127 _M_handle_word_boundary(_Match_mode, _StateIdT);
128
129 void
130 _M_handle_subexpr_lookahead(_Match_mode, _StateIdT);
131
132 void
133 _M_handle_match(_Match_mode, _StateIdT);
134
135 void
136 _M_handle_backref(_Match_mode, _StateIdT);
137
138 void
139 _M_handle_accept(_Match_mode, _StateIdT);
140
141 void
142 _M_handle_alternative(_Match_mode, _StateIdT);
143
144 void
145 _M_dfs(_Match_mode __match_mode, _StateIdT __start);
146
147 bool
148 _M_main(_Match_mode __match_mode)
149 { return _M_main_dispatch(__match_mode, __search_mode{}); }
150
151 bool
152 _M_main_dispatch(_Match_mode __match_mode, __dfs);
153
154 bool
155 _M_main_dispatch(_Match_mode __match_mode, __bfs);
156
157 bool
158 _M_is_word(_CharT __ch) const
159 {
160 static const _CharT __s[2] = { 'w' };
161 return _M_re._M_automaton->_M_traits.isctype
162 (__ch, _M_re._M_automaton->_M_traits.lookup_classname(__s, __s+1));
163 }
164
165 bool
166 _M_at_begin() const
167 {
168 if (_M_current == _M_begin)
169 {
170 // match_not_bol means ^ does not match [_M_begin,_M_begin)
171 if (_M_flags & regex_constants::match_not_bol)
172 return false;
173 // match_prev_avail means _M_begin is not the start of the input.
175 {
176 // For ECMAScript multiline matches, check if the previous
177 // character is a line terminator.
178 if (_M_match_multiline())
179 return _M_is_line_terminator(*std::prev(_M_current));
180 else
181 return false;
182 }
183 else // ^ matches at _M_begin
184 return true;
185 }
186 else if (_M_match_multiline())
187 return _M_is_line_terminator(*std::prev(_M_current));
188 else
189 return false;
190 }
191
192 bool
193 _M_at_end() const
194 {
195 if (_M_current == _M_end)
196 return !(_M_flags & regex_constants::match_not_eol);
197 else if (_M_match_multiline())
198 return _M_is_line_terminator(*_M_current);
199 else
200 return false;
201 }
202
203 bool
204 _M_word_boundary() const;
205
206 bool
207 _M_lookahead(_StateIdT __next);
208
209 bool
210 _M_is_line_terminator(_CharT __c) const
211 {
212 const auto& __traits = _M_re._M_automaton->_M_traits;
213 const auto& __ct = use_facet<ctype<_CharT>>(__traits.getloc());
214 const char __n{ __ct.narrow(__c, ' ') };
215 if (__n == '\n')
216 return true;
217 if (_M_re._M_automaton->_M_options() & regex_constants::ECMAScript)
218 {
219 if (__n == '\r')
220 return true;
221 // FIXME: U+2028 (line separator) and U+2029 (paragraph separator)
222 }
223 return false;
224 }
225
226 bool
227 _M_match_multiline() const noexcept
228 {
229 constexpr auto __m
231 return (_M_re._M_automaton->_M_options() & __m) == __m;
232 }
233
234 // Holds additional information used in BFS-mode.
235 template<typename _SearchMode, typename _ResultsVec>
236 struct _State_info;
237
238 template<typename _ResultsVec>
239 struct _State_info<__bfs, _ResultsVec>
240 {
241 explicit
242 _State_info(_StateIdT __start, size_t __n)
243 : _M_visited_states(new bool[__n]()), _M_start(__start)
244 { }
245
246 ~_State_info() { delete[] _M_visited_states; }
247
248 _State_info(const _State_info&) = delete;
249 _State_info& operator=(const _State_info&) = delete;
250
251 bool _M_visited(_StateIdT __i)
252 {
253 if (_M_visited_states[__i])
254 return true;
255 _M_visited_states[__i] = true;
256 return false;
257 }
258
259 void _M_queue(_StateIdT __i, const _ResultsVec& __res)
260 { _M_match_queue.emplace_back(__i, __res); }
261
262 // Dummy implementations for BFS mode.
263 _BiIter* _M_get_sol_pos() { return nullptr; }
264
265 // Saves states that need to be considered for the next character.
266 _GLIBCXX_STD_C::vector<pair<_StateIdT, _ResultsVec>> _M_match_queue;
267 // Indicates which states are already visited.
268 bool* _M_visited_states;
269 // To record current solution.
270 _StateIdT _M_start;
271 };
272
273 template<typename _ResultsVec>
274 struct _State_info<__dfs, _ResultsVec>
275 {
276 explicit
277 _State_info(_StateIdT __start, size_t) : _M_start(__start)
278 { }
279
280 // Dummy implementations for DFS mode.
281 bool _M_visited(_StateIdT) const { return false; }
282 void _M_queue(_StateIdT, const _ResultsVec&) { }
283
284 _BiIter* _M_get_sol_pos() { return &_M_sol_pos; }
285
286 // To record current solution.
287 _StateIdT _M_start;
288 _BiIter _M_sol_pos;
289 };
290
291 public:
292 _ResultsVec _M_cur_results;
293 _BiIter _M_current;
294 _BiIter _M_begin;
295 const _BiIter _M_end;
296 const _RegexT& _M_re;
297 const _NFAT& _M_nfa;
298 _ResultsVec& _M_results;
299 _GLIBCXX_STD_C::vector<pair<_BiIter, int>> _M_rep_count;
300 _State_info<__search_mode, _ResultsVec> _M_states;
301 _FlagT _M_flags;
302 // Do we have a solution so far?
303 bool _M_has_sol;
304 };
305
306 ///@} regex-detail
307} // namespace __detail
308_GLIBCXX_END_NAMESPACE_VERSION
309} // namespace std
310
integral_constant< bool, true > true_type
The type used as a compile-time boolean with true value.
Definition: type_traits:82
integral_constant< bool, false > false_type
The type used as a compile-time boolean with false value.
Definition: type_traits:85
ISO C++ entities toplevel namespace is std.
constexpr match_flag_type match_not_bol
constexpr syntax_option_type ECMAScript
constexpr syntax_option_type __multiline
Extension: Equivalent to regex_constants::multiline for C++11 and C++14.
constexpr match_flag_type match_not_eol
match_flag_type
This is a bitmask type indicating regex matching rules.
constexpr match_flag_type match_prev_avail
integral_constant
Definition: type_traits:63
Traits class for iterators.
Takes a regex and an input string and does the matching.