regex-repeat-matcher.cpp
Go to the documentation of this file.
1 /* -*- Mode:C++; c-file-style:"gnu"; indent-tabs-mode:nil; -*- */
2 /*
3  * Copyright (c) 2013-2021 Regents of the University of California.
4  *
5  * This file is part of ndn-cxx library (NDN C++ library with eXperimental eXtensions).
6  *
7  * ndn-cxx library is free software: you can redistribute it and/or modify it under the
8  * terms of the GNU Lesser General Public License as published by the Free Software
9  * Foundation, either version 3 of the License, or (at your option) any later version.
10  *
11  * ndn-cxx library is distributed in the hope that it will be useful, but WITHOUT ANY
12  * WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A
13  * PARTICULAR PURPOSE. See the GNU Lesser General Public License for more details.
14  *
15  * You should have received copies of the GNU General Public License and GNU Lesser
16  * General Public License along with ndn-cxx, e.g., in COPYING.md file. If not, see
17  * <http://www.gnu.org/licenses/>.
18  *
19  * See AUTHORS.md for complete list of ndn-cxx authors and contributors.
20  *
21  * @author Yingdi Yu <http://irl.cs.ucla.edu/~yingdi/>
22  */
23 
27 
28 #include <regex>
29 
30 namespace ndn {
31 
33  shared_ptr<RegexBackrefManager> backrefManager,
34  size_t indicator)
35  : RegexMatcher(expr, EXPR_REPEAT_PATTERN, std::move(backrefManager))
36  , m_indicator(indicator)
37 {
38  compile();
39 }
40 
41 void
42 RegexRepeatMatcher::compile()
43 {
44  if ('(' == m_expr[0]) {
45  auto matcher = make_shared<RegexBackrefMatcher>(m_expr.substr(0, m_indicator),
47  m_backrefManager->pushRef(matcher);
48  matcher->compile();
49  m_matchers.push_back(std::move(matcher));
50  }
51  else {
52  m_matchers.push_back(make_shared<RegexComponentSetMatcher>(m_expr.substr(0, m_indicator),
54  }
55 
56  parseRepetition();
57 }
58 
59 void
60 RegexRepeatMatcher::parseRepetition()
61 {
62  constexpr size_t MAX_REPETITIONS = std::numeric_limits<size_t>::max();
63  size_t exprSize = m_expr.size();
64 
65  if (exprSize == m_indicator) {
66  m_repeatMin = 1;
67  m_repeatMax = 1;
68  }
69  else if (exprSize == m_indicator + 1) {
70  switch (m_expr[m_indicator]) {
71  case '?':
72  m_repeatMin = 0;
73  m_repeatMax = 1;
74  break;
75  case '+':
76  m_repeatMin = 1;
77  m_repeatMax = MAX_REPETITIONS;
78  break;
79  case '*':
80  m_repeatMin = 0;
81  m_repeatMax = MAX_REPETITIONS;
82  break;
83  default:
84  NDN_THROW(Error("Unrecognized quantifier '"s + m_expr[m_indicator] + "' in regex: " + m_expr));
85  }
86  }
87  else {
88  std::string repeatStruct = m_expr.substr(m_indicator, exprSize - m_indicator);
89  size_t rsSize = repeatStruct.size();
90 
91  try {
92  if (std::regex_match(repeatStruct, std::regex("\\{[0-9]+,[0-9]+\\}"))) {
93  size_t separator = repeatStruct.find_first_of(',', 0);
94  m_repeatMin = std::stoul(repeatStruct.substr(1, separator - 1));
95  m_repeatMax = std::stoul(repeatStruct.substr(separator + 1, rsSize - separator - 2));
96  if (m_repeatMin > m_repeatMax) {
97  NDN_THROW(Error("Invalid number of repetitions '" + repeatStruct + "' in regex: " + m_expr));
98  }
99  }
100  else if (std::regex_match(repeatStruct, std::regex("\\{,[0-9]+\\}"))) {
101  size_t separator = repeatStruct.find_first_of(',', 0);
102  m_repeatMin = 0;
103  m_repeatMax = std::stoul(repeatStruct.substr(separator + 1, rsSize - separator - 2));
104  }
105  else if (std::regex_match(repeatStruct, std::regex("\\{[0-9]+,\\}"))) {
106  size_t separator = repeatStruct.find_first_of(',', 0);
107  m_repeatMin = std::stoul(repeatStruct.substr(1, separator));
108  m_repeatMax = MAX_REPETITIONS;
109  }
110  else if (std::regex_match(repeatStruct, std::regex("\\{[0-9]+\\}"))) {
111  m_repeatMin = std::stoul(repeatStruct.substr(1, rsSize - 1));
112  m_repeatMax = m_repeatMin;
113  }
114  else {
115  NDN_THROW(Error("Invalid quantifier '" + repeatStruct + "' in regex: " + m_expr));
116  }
117  }
118  // std::stoul can throw invalid_argument or out_of_range, both are derived from logic_error
119  catch (const std::logic_error&) {
120  NDN_THROW_NESTED(Error("Invalid number of repetitions '" + repeatStruct + "' in regex: " + m_expr));
121  }
122  }
123 }
124 
125 bool
126 RegexRepeatMatcher::match(const Name& name, size_t offset, size_t len)
127 {
128  m_matchResult.clear();
129 
130  if (m_repeatMin == 0 && len == 0) {
131  return true;
132  }
133 
134  if (recursiveMatch(0, name, offset, len)) {
135  for (size_t i = offset; i < offset + len; i++) {
136  m_matchResult.push_back(name.get(i));
137  }
138  return true;
139  }
140 
141  return false;
142 }
143 
144 bool
145 RegexRepeatMatcher::recursiveMatch(size_t repeat, const Name& name, size_t offset, size_t len)
146 {
147  if (0 < len && repeat >= m_repeatMax) {
148  return false;
149  }
150 
151  if (0 == len && repeat < m_repeatMin) {
152  return false;
153  }
154 
155  if (0 == len && repeat >= m_repeatMin) {
156  return true;
157  }
158 
159  const auto& matcher = m_matchers[0];
160  ssize_t tried = static_cast<ssize_t>(len);
161  while (tried >= 0) {
162  if (matcher->match(name, offset, tried) &&
163  recursiveMatch(repeat + 1, name, offset + tried, len - tried)) {
164  return true;
165  }
166  tried--;
167  }
168 
169  return false;
170 }
171 
172 } // namespace ndn
Represents an absolute name.
Definition: name.hpp:44
const Component & get(ssize_t i) const noexcept
Returns an immutable reference to the component at the specified index.
Definition: name.hpp:167
std::vector< name::Component > m_matchResult
shared_ptr< RegexBackrefManager > m_backrefManager
const std::string m_expr
std::vector< shared_ptr< RegexMatcher > > m_matchers
bool match(const Name &name, size_t offset, size_t len) override
RegexRepeatMatcher(const std::string &expr, shared_ptr< RegexBackrefManager > backrefManager, size_t indicator)
#define NDN_THROW_NESTED(e)
Definition: exception.hpp:71
#define NDN_THROW(e)
Definition: exception.hpp:61
Definition: data.cpp:25