| 1 | // Copyright (C) 2021 The Qt Company Ltd. |
| 2 | // SPDX-License-Identifier: LicenseRef-Qt-Commercial OR GPL-3.0-only |
| 3 | |
| 4 | #include "glslparsertable_p.h" |
| 5 | #include "glsllexer_p.h" |
| 6 | #include "glslparser_p.h" |
| 7 | #include "glslengine_p.h" |
| 8 | #include <cctype> |
| 9 | #include <iostream> |
| 10 | #include <cstdio> |
| 11 | |
| 12 | QT_BEGIN_NAMESPACE |
| 13 | |
| 14 | using namespace GLSL; |
| 15 | |
| 16 | Lexer::Lexer(Engine *engine, const char *source, unsigned size) |
| 17 | : _engine(engine), |
| 18 | _source(source), |
| 19 | _it(source), |
| 20 | _size(size), |
| 21 | _yychar('\n'), |
| 22 | _lineno(0), |
| 23 | _state(0), |
| 24 | _variant(Variant_Mask & ~Variant_Reserved), // everything except reserved |
| 25 | _scanKeywords(true), |
| 26 | _scanComments(false) |
| 27 | { |
| 28 | } |
| 29 | |
| 30 | Lexer::~Lexer() |
| 31 | { |
| 32 | } |
| 33 | |
| 34 | void Lexer::yyinp() |
| 35 | { |
| 36 | _yychar = (unsigned char) *_it++; |
| 37 | if (_yychar == '\n') |
| 38 | ++_lineno; |
| 39 | } |
| 40 | |
| 41 | int Lexer::yylex(Token *tk) |
| 42 | { |
| 43 | const char *pos = nullptr; |
| 44 | int line = 0; |
| 45 | _yyval.ptr = nullptr; |
| 46 | const int kind = yylex_helper(position: &pos, line: &line); |
| 47 | tk->kind = kind; |
| 48 | tk->position = pos - _source; |
| 49 | tk->length = _it - pos - 1; |
| 50 | tk->line = line; |
| 51 | tk->ptr = _yyval.ptr; |
| 52 | return kind; |
| 53 | } |
| 54 | |
| 55 | enum { |
| 56 | State_normal, |
| 57 | |
| 58 | }; |
| 59 | |
| 60 | int Lexer::yylex_helper(const char **position, int *line) |
| 61 | { |
| 62 | using namespace GLSL; |
| 63 | again: |
| 64 | while (std::isspace(_yychar)) |
| 65 | yyinp(); |
| 66 | |
| 67 | *position = _it - 1; |
| 68 | *line = _lineno; |
| 69 | |
| 70 | if (_yychar == 0) |
| 71 | return Parser::EOF_SYMBOL; |
| 72 | |
| 73 | if (_state == State_comment) { |
| 74 | while (_yychar) { |
| 75 | if (_yychar == '*') { |
| 76 | yyinp(); |
| 77 | if (_yychar == '/') { |
| 78 | yyinp(); |
| 79 | _state = State_normal; |
| 80 | break; |
| 81 | } |
| 82 | } else { |
| 83 | yyinp(); |
| 84 | } |
| 85 | } |
| 86 | return Parser::T_COMMENT; |
| 87 | } |
| 88 | |
| 89 | const int ch = _yychar; |
| 90 | yyinp(); |
| 91 | |
| 92 | switch (ch) { |
| 93 | case '#': |
| 94 | for (; _yychar; yyinp()) { |
| 95 | if (_yychar == '\n') |
| 96 | break; |
| 97 | } |
| 98 | goto again; |
| 99 | |
| 100 | // one of `!', `!=' |
| 101 | case '!': |
| 102 | if (_yychar == '=') { |
| 103 | yyinp(); |
| 104 | return Parser::T_NE_OP; |
| 105 | } |
| 106 | return Parser::T_BANG; |
| 107 | |
| 108 | // one of |
| 109 | // % |
| 110 | // %= |
| 111 | case '%': |
| 112 | if (_yychar == '=') { |
| 113 | yyinp(); |
| 114 | return Parser::T_MOD_ASSIGN; |
| 115 | } |
| 116 | return Parser::T_PERCENT; |
| 117 | |
| 118 | // one of |
| 119 | // & |
| 120 | // && |
| 121 | // &= |
| 122 | case '&': |
| 123 | if (_yychar == '=') { |
| 124 | yyinp(); |
| 125 | return Parser::T_AND_ASSIGN; |
| 126 | } else if (_yychar == '&') { |
| 127 | yyinp(); |
| 128 | return Parser::T_AND_OP; |
| 129 | } |
| 130 | return Parser::T_AMPERSAND; |
| 131 | |
| 132 | // ( |
| 133 | case '(': |
| 134 | return Parser::T_LEFT_PAREN; |
| 135 | |
| 136 | // ) |
| 137 | case ')': |
| 138 | return Parser::T_RIGHT_PAREN; |
| 139 | |
| 140 | // one of |
| 141 | // * |
| 142 | // *= |
| 143 | case '*': |
| 144 | if (_yychar == '=') { |
| 145 | yyinp(); |
| 146 | return Parser::T_MUL_ASSIGN; |
| 147 | } |
| 148 | return Parser::T_STAR; |
| 149 | |
| 150 | // one of |
| 151 | // ++ |
| 152 | // += |
| 153 | // + |
| 154 | case '+': |
| 155 | if (_yychar == '=') { |
| 156 | yyinp(); |
| 157 | return Parser::T_ADD_ASSIGN; |
| 158 | } else if (_yychar == '+') { |
| 159 | yyinp(); |
| 160 | return Parser::T_INC_OP; |
| 161 | } |
| 162 | return Parser::T_PLUS; |
| 163 | |
| 164 | // , |
| 165 | case ',': |
| 166 | return Parser::T_COMMA; |
| 167 | |
| 168 | // one of |
| 169 | // - |
| 170 | // -- |
| 171 | // -= |
| 172 | case '-': |
| 173 | if (_yychar == '=') { |
| 174 | yyinp(); |
| 175 | return Parser::T_SUB_ASSIGN; |
| 176 | } else if (_yychar == '-') { |
| 177 | yyinp(); |
| 178 | return Parser::T_DEC_OP; |
| 179 | } |
| 180 | return Parser::T_DASH; |
| 181 | |
| 182 | // one of |
| 183 | // . |
| 184 | // float constant |
| 185 | case '.': |
| 186 | if (std::isdigit(_yychar)) { |
| 187 | const char *word = _it - 2; |
| 188 | while (std::isalnum(_yychar)) { |
| 189 | yyinp(); |
| 190 | } |
| 191 | if (_engine) |
| 192 | _yyval.string = _engine->number(s: word, n: _it - word - 1); |
| 193 | return Parser::T_NUMBER; |
| 194 | } |
| 195 | return Parser::T_DOT; |
| 196 | |
| 197 | // one of |
| 198 | // / |
| 199 | // /= |
| 200 | // comment |
| 201 | case '/': |
| 202 | if (_yychar == '/') { |
| 203 | for (; _yychar; yyinp()) { |
| 204 | if (_yychar == '\n') |
| 205 | break; |
| 206 | } |
| 207 | if (_scanComments) |
| 208 | return Parser::T_COMMENT; |
| 209 | goto again; |
| 210 | } else if (_yychar == '*') { |
| 211 | yyinp(); |
| 212 | while (_yychar) { |
| 213 | if (_yychar == '*') { |
| 214 | yyinp(); |
| 215 | if (_yychar == '/') { |
| 216 | yyinp(); |
| 217 | if (_scanComments) |
| 218 | return Parser::T_COMMENT; |
| 219 | goto again; |
| 220 | } |
| 221 | } else { |
| 222 | yyinp(); |
| 223 | } |
| 224 | } |
| 225 | if (_scanComments) { |
| 226 | _state = State_comment; |
| 227 | return Parser::T_COMMENT; |
| 228 | } |
| 229 | goto again; |
| 230 | } else if (_yychar == '=') { |
| 231 | yyinp(); |
| 232 | return Parser::T_DIV_ASSIGN; |
| 233 | } |
| 234 | return Parser::T_SLASH; |
| 235 | |
| 236 | // : |
| 237 | case ':': |
| 238 | return Parser::T_COLON; |
| 239 | |
| 240 | // ; |
| 241 | case ';': |
| 242 | return Parser::T_SEMICOLON; |
| 243 | |
| 244 | // one of |
| 245 | // < |
| 246 | // <= |
| 247 | // << |
| 248 | // <<= |
| 249 | case '<': |
| 250 | if (_yychar == '=') { |
| 251 | yyinp(); |
| 252 | return Parser::T_LE_OP; |
| 253 | } else if (_yychar == '<') { |
| 254 | yyinp(); |
| 255 | if (_yychar == '=') { |
| 256 | yyinp(); |
| 257 | return Parser::T_LEFT_ASSIGN; |
| 258 | } |
| 259 | return Parser::T_LEFT_OP; |
| 260 | } |
| 261 | return Parser::T_LEFT_ANGLE; |
| 262 | |
| 263 | // one of |
| 264 | // = |
| 265 | // == |
| 266 | case '=': |
| 267 | if (_yychar == '=') { |
| 268 | yyinp(); |
| 269 | return Parser::T_EQ_OP; |
| 270 | } |
| 271 | return Parser::T_EQUAL; |
| 272 | |
| 273 | // one of |
| 274 | // > |
| 275 | // >= |
| 276 | // >>= |
| 277 | // >> |
| 278 | case '>': |
| 279 | if (_yychar == '=') { |
| 280 | yyinp(); |
| 281 | return Parser::T_GE_OP; |
| 282 | } else if (_yychar == '>') { |
| 283 | yyinp(); |
| 284 | if (_yychar == '=') { |
| 285 | yyinp(); |
| 286 | return Parser::T_RIGHT_ASSIGN; |
| 287 | } |
| 288 | return Parser::T_RIGHT_OP; |
| 289 | } |
| 290 | return Parser::T_RIGHT_ANGLE; |
| 291 | |
| 292 | // ? |
| 293 | case '?': |
| 294 | return Parser::T_QUESTION; |
| 295 | |
| 296 | // [ |
| 297 | case '[': |
| 298 | return Parser::T_LEFT_BRACKET; |
| 299 | |
| 300 | // ] |
| 301 | case ']': |
| 302 | return Parser::T_RIGHT_BRACKET; |
| 303 | |
| 304 | // one of |
| 305 | // ^ |
| 306 | // ^= |
| 307 | case '^': |
| 308 | if (_yychar == '=') { |
| 309 | yyinp(); |
| 310 | return Parser::T_XOR_ASSIGN; |
| 311 | } else if (_yychar == '^') { |
| 312 | yyinp(); |
| 313 | return Parser::T_XOR_OP; |
| 314 | } |
| 315 | return Parser::T_CARET; |
| 316 | |
| 317 | // { |
| 318 | case '{': |
| 319 | return Parser::T_LEFT_BRACE; |
| 320 | |
| 321 | // one of |
| 322 | // | |
| 323 | // |= |
| 324 | // || |
| 325 | case '|': |
| 326 | if (_yychar == '=') { |
| 327 | yyinp(); |
| 328 | return Parser::T_OR_ASSIGN; |
| 329 | } else if (_yychar == '|') { |
| 330 | yyinp(); |
| 331 | return Parser::T_OR_OP; |
| 332 | } |
| 333 | return Parser::T_VERTICAL_BAR; |
| 334 | |
| 335 | // } |
| 336 | case '}': |
| 337 | return Parser::T_RIGHT_BRACE; |
| 338 | |
| 339 | // ~ |
| 340 | case '~': |
| 341 | return Parser::T_TILDE; |
| 342 | |
| 343 | default: |
| 344 | if (std::isalpha(ch) || ch == '_') { |
| 345 | const char *word = _it - 2; |
| 346 | while (std::isalnum(_yychar) || _yychar == '_') { |
| 347 | yyinp(); |
| 348 | } |
| 349 | if (_scanKeywords) { |
| 350 | const int k = findKeyword(word, length: _it - word - 1); |
| 351 | |
| 352 | if (k != Parser::T_IDENTIFIER) |
| 353 | return k; |
| 354 | } |
| 355 | if (_engine) |
| 356 | _yyval.string = _engine->identifier(s: word, n: _it - word - 1); |
| 357 | return Parser::T_IDENTIFIER; |
| 358 | } else if (std::isdigit(ch)) { |
| 359 | const char *word = _it - 2; |
| 360 | while (std::isalnum(_yychar) || _yychar == '.') { |
| 361 | yyinp(); |
| 362 | } |
| 363 | if (_engine) |
| 364 | _yyval.string = _engine->number(s: word, n: _it - word - 1); |
| 365 | return Parser::T_NUMBER; |
| 366 | } |
| 367 | |
| 368 | } // switch |
| 369 | |
| 370 | return Parser::T_ERROR; |
| 371 | } |
| 372 | |
| 373 | int Lexer::findKeyword(const char *word, int length) const |
| 374 | { |
| 375 | int t = classify(s: word, len: length); |
| 376 | if (!(t & Variant_Mask)) |
| 377 | return t; |
| 378 | if ((_variant & t & Variant_Mask) == 0) { |
| 379 | // Return a "reserved word" token if this keyword is not permitted |
| 380 | // in the current language variant so that the syntax highlighter |
| 381 | // can warn the user about the word. |
| 382 | if (!_scanKeywords) |
| 383 | return Parser::T_RESERVED; |
| 384 | } |
| 385 | return t & ~Variant_Mask; |
| 386 | } |
| 387 | |
| 388 | void Lexer::warning(int line, const QString &message) |
| 389 | { |
| 390 | _engine->warning(line, message); |
| 391 | } |
| 392 | |
| 393 | void Lexer::error(int line, const QString &message) |
| 394 | { |
| 395 | _engine->error(line, message); |
| 396 | } |
| 397 | |
| 398 | QT_END_NAMESPACE |
| 399 | |