1 /**
2 * Copyright (c) 2021-2022 Huawei Device Co., Ltd.
3 * Licensed under the Apache License, Version 2.0 (the "License");
4 * you may not use this file except in compliance with the License.
5 * You may obtain a copy of the License at
6 *
7 * http://www.apache.org/licenses/LICENSE-2.0
8 *
9 * Unless required by applicable law or agreed to in writing, software
10 * distributed under the License is distributed on an "AS IS" BASIS,
11 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
12 * See the License for the specific language governing permissions and
13 * limitations under the License.
14 */
15
16 #ifndef ASSEMBLER_ASSEMBLY_PARSER_H
17 #define ASSEMBLER_ASSEMBLY_PARSER_H
18
19 #include <iostream>
20 #include <memory>
21 #include <string>
22 #include <string_view>
23
24 #include "assembly-context.h"
25 #include "assembly-emitter.h"
26 #include "assembly-field.h"
27 #include "assembly-function.h"
28 #include "assembly-ins.h"
29 #include "assembly-label.h"
30 #include "assembly-program.h"
31 #include "assembly-record.h"
32 #include "assembly-type.h"
33 #include "define.h"
34 #include "error.h"
35 #include "ide_helpers.h"
36 #include "lexer.h"
37 #include "meta.h"
38 #include "utils/expected.h"
39
40 namespace panda::pandasm {
41
42 using Instructions = std::pair<std::vector<Ins>, Error>;
43
44 using Functions = std::pair<std::unordered_map<std::string, Function>, std::unordered_map<std::string, Record>>;
45
46 class Parser {
47 public:
48 Parser() = default;
49
50 NO_MOVE_SEMANTIC(Parser);
51 NO_COPY_SEMANTIC(Parser);
52
53 ~Parser() = default;
54
55 /*
56 * The main function of parsing, which takes a vector of token vectors and a name of the source file.
57 * Returns a program or an error value: Expected<Program, Error>
58 * This function analyzes code containing several functions:
59 * - Each function used must be declared.
60 * - The correct function declaration looks like this: .function ret_type fun_name([param_type aN,]) [<metadata>]
61 * ([data] shows that this 'data' is optional).
62 * - N in function parameters must increase when number of parameters increases
63 * (Possible: a0, a1,..., aN. Impossible: a1, a10, a13).
64 * - Each function has its own label table.
65 */
66 Expected<Program, Error> Parse(TokenSet &vectors_tokens, const std::string &file_name = "");
67
68 /*
69 * The main function of parsing, which takes a string with source and a name of the source file.
70 * Returns a program or an error value: Expected<Program, Error>
71 */
72 Expected<Program, Error> Parse(const std::string &source, const std::string &file_name = "");
73
74 /*
75 * Returns a set error
76 */
ShowError()77 Error ShowError() const
78 {
79 return err_;
80 }
81
ShowWarnings()82 ErrorList ShowWarnings() const
83 {
84 return war_;
85 }
86
87 private:
88 panda::pandasm::Program program_;
89 std::unordered_map<std::string, panda::pandasm::Label> *label_table_ = nullptr;
90 Metadata *metadata_ = nullptr;
91 Context context_; /* token iterator */
92 panda::pandasm::Record *curr_record_ = nullptr;
93 panda::pandasm::LiteralArray *curr_array_ = nullptr;
94 panda::pandasm::LiteralArray::Literal *curr_array_elem_ = nullptr;
95 panda::pandasm::Function *curr_func_ = nullptr;
96 panda::pandasm::Ins *curr_ins_ = nullptr;
97 panda::pandasm::Field *curr_fld_ = nullptr;
98 size_t line_stric_ = 0;
99 panda::pandasm::Error err_;
100 panda::pandasm::ErrorList war_;
101 bool open_ = false; /* flag of being in a code section */
102 bool record_def_ = false;
103 bool array_def_ = false;
104 bool func_def_ = false;
105 static constexpr uint32_t INTRO_CONST_ARRAY_LITERALS_NUMBER = 2;
106
107 inline Error GetError(const std::string &mess = "", Error::ErrorType err = Error::ErrorType::ERR_NONE,
108 int8_t shift = 0, int token_shift = 0, const std::string &add_mess = "") const
109 {
110 return Error(mess, line_stric_, err, add_mess,
111 context_.tokens[static_cast<int>(context_.number) + token_shift - 1].bound_left + shift,
112 context_.tokens[static_cast<int>(context_.number) + token_shift - 1].bound_right,
113 context_.tokens[static_cast<int>(context_.number) + token_shift - 1].whole_line);
114 }
115
116 inline void GetWarning(const std::string &mess = "", Error::ErrorType err = Error::ErrorType::ERR_NONE,
117 int8_t shift = 0, const std::string &add_mess = "")
118 {
119 war_.emplace_back(mess, line_stric_, err, add_mess,
120 context_.tokens[context_.number - 1].bound_left + static_cast<size_t>(shift),
121 context_.tokens[context_.number - 1].bound_right,
122 context_.tokens[context_.number - 1].whole_line, Error::ErrorClass::WARNING);
123 }
124
GetCurrentPosition(bool left_bound)125 SourcePosition GetCurrentPosition(bool left_bound) const
126 {
127 if (left_bound) {
128 return SourcePosition {line_stric_, context_.tokens[context_.number - 1].bound_left};
129 }
130 return SourcePosition {line_stric_, context_.tokens[context_.number - 1].bound_right};
131 }
132
133 bool LabelValidName();
134 bool TypeValidName();
135 bool RegValidName();
136 bool ParamValidName();
137 bool FunctionValidName();
138 bool ParseFunctionName();
139 bool ParseLabel();
140 bool ParseOperation();
141 bool ParseOperands();
142 bool ParseFunctionCode();
143 bool ParseFunctionInstruction();
144 bool ParseFunctionFullSign();
145 bool UpdateFunctionName();
146 bool ParseFunctionReturn();
147 bool ParseFunctionArg();
148 bool ParseFunctionArgComma(bool &comma);
149 bool ParseFunctionArgs();
150 bool ParseType(Type *type);
151 bool PrefixedValidName();
152 bool ParseMetaListComma(bool &comma, bool eq);
153 bool MeetExpMetaList(bool eq);
154 bool BuildMetaListAttr(bool &eq, std::string &attribute_name, std::string &attribute_value);
155 bool ParseMetaList(bool flag);
156 bool ParseMetaDef();
157 bool ParseRecordFullSign();
158 bool ParseRecordFields();
159 bool ParseRecordField();
160 bool ParseRecordName();
161 bool RecordValidName();
162 bool ParseArrayFullSign();
163 bool IsConstArray();
164 bool ParseArrayName();
165 bool ArrayValidName();
166 bool ArrayElementsValidNumber();
167 bool ParseArrayElements();
168 bool ParseArrayElement();
169 bool ParseArrayElementType();
170 bool ParseArrayElementValue();
171 bool ParseArrayElementValueInteger();
172 bool ParseArrayElementValueFloat();
173 bool ParseArrayElementValueString();
174 bool ParseFieldName();
175 bool ParseFieldType();
176 std::optional<std::string> ParseStringLiteral();
177 int64_t MnemonicToBuiltinId();
178
179 bool ParseInteger(int64_t *value);
180 bool ParseFloat(double *value, bool is_64bit);
181 bool ParseOperandVreg();
182 bool ParseOperandComma();
183 bool ParseOperandInteger();
184 bool ParseOperandFloat(bool is_64bit);
185 bool ParseOperandId();
186 bool ParseOperandLabel();
187 bool ParseOperandField();
188 bool ParseOperandType(Type::VerificationType ver_type);
189 bool ParseOperandNone();
190 bool ParseOperandString();
191 bool ParseOperandLiteralArray();
192 bool ParseOperandCall();
193 bool ParseOperandSignature(std::string *sign);
194 bool ParseOperandSignatureTypesList(std::string *sign);
195 bool ParseOperandBuiltinMnemonic();
196
197 void SetFunctionInformation();
198 void SetRecordInformation();
199 void SetArrayInformation();
200 void SetOperationInformation();
201 void ParseAsCatchall(const std::vector<Token> &tokens);
202 void ParseAsLanguage(const std::vector<Token> &tokens, bool &is_lang_parsed, bool &is_first_statement);
203 void ParseAsRecord(const std::vector<Token> &tokens);
204 void ParseAsArray(const std::vector<Token> &tokens);
205 void ParseAsFunction(const std::vector<Token> &tokens);
206 void ParseAsBraceRight(const std::vector<Token> &tokens);
207 bool ParseAfterLine(bool &is_first_statement);
208 Expected<Program, Error> ParseAfterMainLoop(const std::string &file_name);
209 void ParseResetFunctionLabelsAndParams();
210 void ParseResetTables();
211 void ParseResetFunctionTable();
212 void ParseResetRecordTable();
213 void ParseResetArrayTable();
214 void ParseAsLanguageDirective();
215 Function::CatchBlock PrepareCatchBlock(bool is_catchall, size_t size, size_t catchall_tokens_num,
216 size_t catch_tokens_num);
217 void ParseAsCatchDirective();
218 void SetError();
219 void SetMetadataContextError(const Metadata::Error &err, bool has_value);
220
221 Expected<char, Error> ParseOctalEscapeSequence(std::string_view s, size_t *i);
222 Expected<char, Error> ParseHexEscapeSequence(std::string_view s, size_t *i);
223 Expected<char, Error> ParseEscapeSequence(std::string_view s, size_t *i);
224
225 template <class T>
TryEmplaceInTable(bool flag,T & item)226 auto TryEmplaceInTable(bool flag, T &item)
227 {
228 return item.try_emplace(std::string(context_.GiveToken().data(), context_.GiveToken().length()),
229 std::string(context_.GiveToken().data(), context_.GiveToken().length()), program_.lang,
230 context_.tokens[context_.number - 1].bound_left,
231 context_.tokens[context_.number - 1].bound_right,
232 context_.tokens[context_.number - 1].whole_line, flag, line_stric_);
233 }
234
235 template <class T>
AddObjectInTable(bool flag,T & item)236 bool AddObjectInTable(bool flag, T &item)
237 {
238 auto [iter, is_inserted] = TryEmplaceInTable(flag, item);
239
240 if (is_inserted) {
241 return true;
242 }
243
244 if (iter->second.file_location->is_defined && flag) {
245 return false;
246 }
247
248 if (!iter->second.file_location->is_defined && flag) {
249 iter->second.file_location->is_defined = true;
250 return true;
251 }
252
253 if (!iter->second.file_location->is_defined) {
254 iter->second.file_location->bound_left = context_.tokens[context_.number - 1].bound_left;
255 iter->second.file_location->bound_right = context_.tokens[context_.number - 1].bound_right;
256 iter->second.file_location->whole_line = context_.tokens[context_.number - 1].whole_line;
257 iter->second.file_location->line_number = line_stric_;
258 }
259
260 return true;
261 }
262 };
263
264 template <>
TryEmplaceInTable(bool flag,std::unordered_map<std::string,panda::pandasm::Label> & item)265 inline auto Parser::TryEmplaceInTable(bool flag, std::unordered_map<std::string, panda::pandasm::Label> &item)
266 {
267 return item.try_emplace(std::string(context_.GiveToken().data(), context_.GiveToken().length()),
268 std::string(context_.GiveToken().data(), context_.GiveToken().length()),
269 context_.tokens[context_.number - 1].bound_left,
270 context_.tokens[context_.number - 1].bound_right,
271 context_.tokens[context_.number - 1].whole_line, flag, line_stric_);
272 }
273
274 } // namespace panda::pandasm
275
276 #endif // ASSEMBLER_ASSEMBLY_PARSER_H
277