summaryrefslogtreecommitdiffhomepage
path: root/lexer.cpp
blob: 9ef29b17e75390c67e2d751b623378d5dbf678fd (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
#include "lexer.h"

#include <algorithm>

using namespace Lex;

size_t Lexer::newState(std::string state_type)
{
 m_state_types.push_back(state_type);
 return states++;
}

void Lexer::addTransition(size_t state0, size_t state1, char c)
{
 auto it{transitions.find(state0)};

 if (it == transitions.end()) { // new entry is a list with 1 entry
  transitions[state0] = {{state1, c}};
 } else { // extend list entry
  it->second.push_back({state1, c});
 }
}

void Lexer::removeTransition(size_t state0, size_t state1, char c)
{
 auto it{transitions.find(state0)};

 if (it != transitions.end()) {
  std::pair<size_t, char> reference{state1, c};
  it->second.erase(std::remove(it->second.begin(), it->second.end(), reference),
                   it->second.end());
 } else {
  throw std::runtime_error("Transition not found: "s + std::to_string(state0) + "->"s + std::to_string(state1));
 }
}

std::vector<std::pair<size_t, char>> Lexer::getSuccessorsViaEmpty(size_t state)
{
 std::vector<std::pair<size_t, char>> result;

 auto it{transitions.find(state)};

 if (it != transitions.end()) {
  for (auto &i: it->second) {
   if (i.second == '\0') { // add empty transitions
    result.push_back(i);
    // and more, recursively
    auto successors { getSuccessorsViaEmpty(i.first) };
    result.insert(result.end(), successors.begin(), successors.end());
   }
  }
 }

 return result;
}

std::vector<std::pair<size_t, char>> Lexer::getSuccessors(size_t state)
{
 auto it{transitions.find(state)};

 if (it != transitions.end()) {
   return it->second;
 }

 return {};
}

// Helper function
void Lexer::addPathOrTransition(size_t state0, size_t state1, std::string symbol, std::string type)
{
 if (isTerminal(m_bnf, symbol)) { // end recursion with new transition
  if (symbol.size() != 1)
   throw std::runtime_error("Bad sized terminal symbol: "s + symbol);
  addTransition(state0, state1, symbol[0]);
 } else { // recurse via non-terminal symbol
  addPath(state0, state1, symbol, type);
 }
}

// Helper function: add one rule
void Lexer::addRule(const std::vector<std::string>& list, size_t list_index_from, size_t list_index_to, size_t state0, size_t state1, const std::string& rule_symbol, std::string type)
{
 size_t previousState{state0};
 
 // add intermediate states with transitions
 for (size_t i = list_index_from; i < list_index_to - 1; i++) {
  std::string symbol{list[i]};
  if (symbol == rule_symbol)
   throw std::runtime_error("Recursion found but not allowed");

  size_t state{newState(type)};
  addPathOrTransition(previousState, state, symbol, type);
  previousState = state;
 }
 if (list.back() == rule_symbol)
  throw std::runtime_error("Recursion found but not allowed");

 // last transition
 addPathOrTransition(previousState, state1, list.back(), type);
}

// Add paths between state0 and state1, including new states and transitions
void Lexer::addPath(size_t state0, size_t state1, std::string s, std::string type)
{
 if (type == "" && s != "" && s != m_top)
  type = s;

 // state0 -> [paths] -> state01 -> state1
 //                      ^     v
 //                 [recursion paths]
 size_t state01{newState(type)};
 auto it {m_bnf.find(s)};

 if (it == m_bnf.end())
  throw std::runtime_error("Path ("s + std::to_string(state0) + ", "s + std::to_string(state1) + ") not possible."s);

 for (auto& list: it->second) { // for every path between state0 and state1
  size_t list_size{list.size()};
  if (list_size < 1)
   throw std::runtime_error("List too small in rule "s + s);

  if (list[0] == s) { // recursion rule
   addRule(list, 1, list_size, state01, state01, s, type);
  } else { // non-recursion rule
   addRule(list, 0, list_size, state0, state01, s, type);
  }
 }
 addTransition(state01, state1, 0); // empty transition to end
}

void Lexer::replaceEmptyTransitions()
{
 // iterate over all transitions
 for (auto& [state0, list]: transitions) {
  std::vector<std::pair<size_t, char>> list_extension; // which elements must be added to list
  for (auto& [state1, currentChar]: list) {
   if (currentChar != '\0') { // for every non-empty transition
    // add extension via following empty transitions
    auto successors { getSuccessorsViaEmpty(state1) };
    for (auto &[state, dummyChar] : successors) {
     list_extension.emplace_back(state, currentChar);
    }
   }
  }
  list.insert(list.end(), list_extension.begin(), list_extension.end());
 }
}

void Lexer::removeEmptyTransitions()
{
 for (auto& [state0, list]: transitions) {
  list.erase(std::remove_if(list.begin(), list.end(), [](const std::pair<size_t, char>& pair){return pair.second == '\0';}),
             list.end());
 }
}

Lexer::Lexer(const BNF& bnf, const std::string& top): m_bnf(bnf), m_top(top), m_startState(newState()), m_endState(newState())
{
 addPath(m_startState, m_endState, m_top, "");

 replaceEmptyTransitions();
 removeEmptyTransitions();
}

Token Lexer::getToken(const std::string& s, Location& location)
{
 Location oldLocation{location}; // start of token

 std::vector<size_t> states{m_startState}; // can be in multiple states at once
 std::vector<size_t> newStates;

 Location found;
 std::string state_type;

 // match as much as possible
 while (location.pos < s.size() && states.size() > 0) {
  char currentChar{s[location.pos]};
  std::cout << "DEBUG: Char: " << currentChar << std::endl;

  for (const auto& state: states) {
   std::vector<std::pair<size_t, char>> successors{getSuccessors(state)};
   for (const auto& [nextState, c]: successors) {
    if (c == currentChar) {
     if (nextState == m_endState) { // save intermediate result upon match
      found = location;
      found.advance();
      state_type = m_state_types[state];
     } else {
      newStates.push_back(nextState);
     }
    }
   }
  }
  states = newStates;
  newStates.clear();
  location.advance(currentChar == '\n');
 }

 std::string value {s.substr(oldLocation.pos, found.pos - oldLocation.pos)};

 if (found.pos > 0)
  std::cout << "DEBUG: Matched " << found.pos - oldLocation.pos << " chars: " << value << "|" << state_type << std::endl;
 else
  throw std::runtime_error("Bad Token at "s + oldLocation.toString());

 location = found; // reset to end of match

 return {state_type, value, oldLocation};
}

void Lexer::skipWhitespace(const std::string& s, Location& location)
{
 while (location.pos < s.size() && std::isspace(s[location.pos])) {
  location.advance(s[location.pos] == '\n');
 }
}

std::vector<Token> Lexer::Lex(const std::string& s)
{
 std::vector<Token> result;

 Location location;
 skipWhitespace(s, location);
 while (location.pos < s.size()) {
  result.emplace_back(getToken(s, location));
  skipWhitespace(s, location);
 }

 return result;
}