1 /* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
2 * Use of this file is governed by the BSD 3-clause license that
3 * can be found in the LICENSE.txt file in the project root.
4 */
5
6 #include "atn/ATN.h"
7 #include "atn/ATNDeserializer.h"
8 #include "Vocabulary.h"
9
10 #include "misc/InterpreterDataReader.h"
11
12 using namespace antlr4::dfa;
13 using namespace antlr4::atn;
14 using namespace antlr4::misc;
15
InterpreterData(std::vector<std::string> const & literalNames,std::vector<std::string> const & symbolicNames)16 InterpreterData::InterpreterData(std::vector<std::string> const& literalNames, std::vector<std::string> const& symbolicNames)
17 : vocabulary(literalNames, symbolicNames) {
18 }
19
parseFile(std::string const & fileName)20 InterpreterData InterpreterDataReader::parseFile(std::string const& fileName) {
21 // The structure of the data file is very simple. Everything is line based with empty lines
22 // separating the different parts. For lexers the layout is:
23 // token literal names:
24 // ...
25 //
26 // token symbolic names:
27 // ...
28 //
29 // rule names:
30 // ...
31 //
32 // channel names:
33 // ...
34 //
35 // mode names:
36 // ...
37 //
38 // atn:
39 // <a single line with comma separated int values> enclosed in a pair of squared brackets.
40 //
41 // Data for a parser does not contain channel and mode names.
42
43 std::ifstream input(fileName);
44 if (!input.good())
45 return {};
46
47 std::vector<std::string> literalNames;
48 std::vector<std::string> symbolicNames;
49
50 std::string line;
51
52 std::getline(input, line, '\n');
53 assert(line == "token literal names:");
54 while (true) {
55 std::getline(input, line, '\n');
56 if (line.empty())
57 break;
58
59 literalNames.push_back(line == "null" ? "" : line);
60 };
61
62 std::getline(input, line, '\n');
63 assert(line == "token symbolic names:");
64 while (true) {
65 std::getline(input, line, '\n');
66 if (line.empty())
67 break;
68
69 symbolicNames.push_back(line == "null" ? "" : line);
70 };
71 InterpreterData result(literalNames, symbolicNames);
72
73 std::getline(input, line, '\n');
74 assert(line == "rule names:");
75 while (true) {
76 std::getline(input, line, '\n');
77 if (line.empty())
78 break;
79
80 result.ruleNames.push_back(line);
81 };
82
83 std::getline(input, line, '\n');
84 if (line == "channel names:") {
85 while (true) {
86 std::getline(input, line, '\n');
87 if (line.empty())
88 break;
89
90 result.channels.push_back(line);
91 };
92
93 std::getline(input, line, '\n');
94 assert(line == "mode names:");
95 while (true) {
96 std::getline(input, line, '\n');
97 if (line.empty())
98 break;
99
100 result.modes.push_back(line);
101 };
102 }
103
104 std::vector<uint16_t> serializedATN;
105
106 std::getline(input, line, '\n');
107 assert(line == "atn:");
108 std::getline(input, line, '\n');
109 std::stringstream tokenizer(line);
110 std::string value;
111 while (tokenizer.good()) {
112 std::getline(tokenizer, value, ',');
113 unsigned long number;
114 if (value[0] == '[')
115 number = std::strtoul(&value[1], nullptr, 10);
116 else
117 number = std::strtoul(value.c_str(), nullptr, 10);
118 serializedATN.push_back(static_cast<uint16_t>(number));
119 }
120
121 ATNDeserializer deserializer;
122 result.atn = deserializer.deserialize(serializedATN);
123 return result;
124 }
125