1 | //===-- CPlusPlusNameParser.h -----------------------------------*- C++ -*-===// |
2 | // |
3 | // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. |
4 | // See https://llvm.org/LICENSE.txt for license information. |
5 | // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception |
6 | // |
7 | //===----------------------------------------------------------------------===// |
8 | |
9 | #ifndef LLDB_SOURCE_PLUGINS_LANGUAGE_CPLUSPLUS_CPLUSPLUSNAMEPARSER_H |
10 | #define LLDB_SOURCE_PLUGINS_LANGUAGE_CPLUSPLUS_CPLUSPLUSNAMEPARSER_H |
11 | |
12 | #include "clang/Lex/Lexer.h" |
13 | #include "llvm/ADT/SmallVector.h" |
14 | #include "llvm/ADT/StringRef.h" |
15 | |
16 | #include "lldb/Utility/ConstString.h" |
17 | #include "lldb/lldb-private.h" |
18 | #include <optional> |
19 | |
20 | namespace lldb_private { |
21 | |
22 | // Helps to validate and obtain various parts of C++ definitions. |
23 | class CPlusPlusNameParser { |
24 | public: |
25 | CPlusPlusNameParser(llvm::StringRef text) : m_text(text) { ExtractTokens(); } |
26 | |
27 | struct ParsedName { |
28 | llvm::StringRef basename; |
29 | llvm::StringRef context; |
30 | }; |
31 | |
32 | struct ParsedFunction { |
33 | ParsedName name; |
34 | llvm::StringRef arguments; |
35 | llvm::StringRef qualifiers; |
36 | llvm::StringRef return_type; |
37 | }; |
38 | |
39 | // Treats given text as a function definition and parses it. |
40 | // Function definition might or might not have a return type and this should |
41 | // change parsing result. |
42 | // Examples: |
43 | // main(int, chat const*) |
44 | // T fun(int, bool) |
45 | // std::vector<int>::push_back(int) |
46 | // int& map<int, pair<short, int>>::operator[](short) const |
47 | // int (*get_function(const chat *))() |
48 | std::optional<ParsedFunction> ParseAsFunctionDefinition(); |
49 | |
50 | // Treats given text as a potentially nested name of C++ entity (function, |
51 | // class, field) and parses it. |
52 | // Examples: |
53 | // main |
54 | // fun |
55 | // std::vector<int>::push_back |
56 | // map<int, pair<short, int>>::operator[] |
57 | // func<C>(int, C&)::nested_class::method |
58 | std::optional<ParsedName> ParseAsFullName(); |
59 | |
60 | private: |
61 | // A C++ definition to parse. |
62 | llvm::StringRef m_text; |
63 | // Tokens extracted from m_text. |
64 | llvm::SmallVector<clang::Token, 30> m_tokens; |
65 | // Index of the next token to look at from m_tokens. |
66 | size_t m_next_token_index = 0; |
67 | |
68 | // Range of tokens saved in m_next_token_index. |
69 | struct Range { |
70 | size_t begin_index = 0; |
71 | size_t end_index = 0; |
72 | |
73 | Range() = default; |
74 | Range(size_t begin, size_t end) : begin_index(begin), end_index(end) { |
75 | assert(end >= begin); |
76 | } |
77 | |
78 | size_t size() const { return end_index - begin_index; } |
79 | |
80 | bool empty() const { return size() == 0; } |
81 | }; |
82 | |
83 | struct ParsedNameRanges { |
84 | Range basename_range; |
85 | Range context_range; |
86 | }; |
87 | |
88 | // Bookmark automatically restores parsing position (m_next_token_index) |
89 | // when destructed unless it's manually removed with Remove(). |
90 | class Bookmark { |
91 | public: |
92 | Bookmark(size_t &position) |
93 | : m_position(position), m_position_value(position) {} |
94 | Bookmark(const Bookmark &) = delete; |
95 | Bookmark(Bookmark &&b) |
96 | : m_position(b.m_position), m_position_value(b.m_position_value), |
97 | m_restore(b.m_restore) { |
98 | b.Remove(); |
99 | } |
100 | Bookmark &operator=(Bookmark &&) = delete; |
101 | Bookmark &operator=(const Bookmark &) = delete; |
102 | |
103 | void Remove() { m_restore = false; } |
104 | size_t GetSavedPosition() { return m_position_value; } |
105 | ~Bookmark() { |
106 | if (m_restore) { |
107 | m_position = m_position_value; |
108 | } |
109 | } |
110 | |
111 | private: |
112 | size_t &m_position; |
113 | size_t m_position_value; |
114 | bool m_restore = true; |
115 | }; |
116 | |
117 | bool HasMoreTokens(); |
118 | void Advance(); |
119 | void TakeBack(); |
120 | bool ConsumeToken(clang::tok::TokenKind kind); |
121 | |
122 | template <typename... Ts> bool ConsumeToken(Ts... kinds); |
123 | Bookmark SetBookmark(); |
124 | size_t GetCurrentPosition(); |
125 | clang::Token &Peek(); |
126 | bool ConsumeBrackets(clang::tok::TokenKind left, clang::tok::TokenKind right); |
127 | |
128 | std::optional<ParsedFunction> ParseFunctionImpl(bool expect_return_type); |
129 | |
130 | // Parses functions returning function pointers 'string (*f(int x))(float y)' |
131 | std::optional<ParsedFunction> ParseFuncPtr(bool expect_return_type); |
132 | |
133 | // Consumes function arguments enclosed within '(' ... ')' |
134 | bool ConsumeArguments(); |
135 | |
136 | // Consumes template arguments enclosed within '<' ... '>' |
137 | bool ConsumeTemplateArgs(); |
138 | |
139 | // Consumes '(anonymous namespace)' |
140 | bool ConsumeAnonymousNamespace(); |
141 | |
142 | // Consumes '{lambda ...}' |
143 | bool ConsumeLambda(); |
144 | |
145 | // Consumes operator declaration like 'operator *' or 'operator delete []' |
146 | bool ConsumeOperator(); |
147 | |
148 | // Skips 'const' and 'volatile' |
149 | void SkipTypeQualifiers(); |
150 | |
151 | // Skips 'const', 'volatile', '&', '&&' in the end of the function. |
152 | void SkipFunctionQualifiers(); |
153 | |
154 | // Consumes built-in types like 'int' or 'unsigned long long int' |
155 | bool ConsumeBuiltinType(); |
156 | |
157 | // Consumes types defined via decltype keyword. |
158 | bool ConsumeDecltype(); |
159 | |
160 | // Skips 'const' and 'volatile' |
161 | void SkipPtrsAndRefs(); |
162 | |
163 | // Consumes things like 'const * const &' |
164 | bool ConsumePtrsAndRefs(); |
165 | |
166 | // Consumes full type name like 'Namespace::Class<int>::Method()::InnerClass' |
167 | bool ConsumeTypename(); |
168 | |
169 | /// Consumes ABI tags enclosed within '[abi:' ... ']' |
170 | /// |
171 | /// Since there is no restriction on what the ABI tag |
172 | /// string may contain, this API supports parsing a small |
173 | /// set of special characters. |
174 | /// |
175 | /// The following regex describes the set of supported characters: |
176 | /// [A-Za-z,.\s\d]+ |
177 | bool ConsumeAbiTag(); |
178 | |
179 | std::optional<ParsedNameRanges> ParseFullNameImpl(); |
180 | llvm::StringRef GetTextForRange(const Range &range); |
181 | |
182 | // Populate m_tokens by calling clang lexer on m_text. |
183 | void (); |
184 | }; |
185 | |
186 | } // namespace lldb_private |
187 | |
188 | #endif // LLDB_SOURCE_PLUGINS_LANGUAGE_CPLUSPLUS_CPLUSPLUSNAMEPARSER_H |
189 | |