Code Coverage |
||||||||||
Lines |
Functions and Methods |
Classes and Traits |
||||||||
| Total | |
91.04% |
325 / 357 |
|
59.26% |
16 / 27 |
CRAP | |
0.00% |
0 / 1 |
| Tokenizer | |
91.04% |
325 / 357 |
|
59.26% |
16 / 27 |
217.27 | |
0.00% |
0 / 1 |
| __construct | |
100.00% |
3 / 3 |
|
100.00% |
1 / 1 |
3 | |||
| tokenize | |
100.00% |
3 / 3 |
|
100.00% |
1 / 1 |
2 | |||
| nextToken | |
100.00% |
5 / 5 |
|
100.00% |
1 / 1 |
4 | |||
| step | |
93.38% |
127 / 136 |
|
0.00% |
0 / 1 |
50.72 | |||
| consumeComments | |
91.67% |
11 / 12 |
|
0.00% |
0 / 1 |
7.03 | |||
| consumeString | |
68.18% |
15 / 22 |
|
0.00% |
0 / 1 |
10.06 | |||
| consumeIdentLikeToken | |
100.00% |
13 / 13 |
|
100.00% |
1 / 1 |
8 | |||
| consumeUrlToken | |
96.55% |
28 / 29 |
|
0.00% |
0 / 1 |
15 | |||
| consumeRemnantsOfBadUrl | |
80.00% |
8 / 10 |
|
0.00% |
0 / 1 |
7.39 | |||
| consumeIdentSequence | |
100.00% |
13 / 13 |
|
100.00% |
1 / 1 |
5 | |||
| consumeNumericToken | |
100.00% |
8 / 8 |
|
100.00% |
1 / 1 |
3 | |||
| consumeNumber | |
93.94% |
31 / 33 |
|
0.00% |
0 / 1 |
16.06 | |||
| consumeEscape | |
80.95% |
17 / 21 |
|
0.00% |
0 / 1 |
14.17 | |||
| isValidEscape | |
100.00% |
3 / 3 |
|
100.00% |
1 / 1 |
2 | |||
| wouldStartIdentSequence | |
90.91% |
10 / 11 |
|
0.00% |
0 / 1 |
8.05 | |||
| wouldStartNumber | |
75.00% |
9 / 12 |
|
0.00% |
0 / 1 |
7.77 | |||
| preprocess | |
100.00% |
2 / 2 |
|
100.00% |
1 / 1 |
1 | |||
| peek | |
100.00% |
2 / 2 |
|
100.00% |
1 / 1 |
2 | |||
| advance | |
100.00% |
1 / 1 |
|
100.00% |
1 / 1 |
1 | |||
| emit | |
100.00% |
1 / 1 |
|
100.00% |
1 / 1 |
1 | |||
| isWhitespace | |
100.00% |
1 / 1 |
|
100.00% |
1 / 1 |
3 | |||
| isDigit | |
100.00% |
1 / 1 |
|
100.00% |
1 / 1 |
2 | |||
| isHexDigit | |
100.00% |
1 / 1 |
|
100.00% |
1 / 1 |
5 | |||
| isLetter | |
100.00% |
1 / 1 |
|
100.00% |
1 / 1 |
4 | |||
| isIdentStartCodePoint | |
80.00% |
4 / 5 |
|
0.00% |
0 / 1 |
4.13 | |||
| isIdentCodePoint | |
100.00% |
1 / 1 |
|
100.00% |
1 / 1 |
3 | |||
| isNonPrintable | |
85.71% |
6 / 7 |
|
0.00% |
0 / 1 |
7.14 | |||
| 1 | <?php |
| 2 | |
| 3 | declare(strict_types=1); |
| 4 | |
| 5 | namespace Phpdftk\Css; |
| 6 | |
| 7 | use Phpdftk\Css\Token\AtKeywordToken; |
| 8 | use Phpdftk\Css\Token\BadStringToken; |
| 9 | use Phpdftk\Css\Token\BadUrlToken; |
| 10 | use Phpdftk\Css\Token\CdcToken; |
| 11 | use Phpdftk\Css\Token\CdoToken; |
| 12 | use Phpdftk\Css\Token\ColonToken; |
| 13 | use Phpdftk\Css\Token\CommaToken; |
| 14 | use Phpdftk\Css\Token\DelimToken; |
| 15 | use Phpdftk\Css\Token\DimensionToken; |
| 16 | use Phpdftk\Css\Token\EofToken; |
| 17 | use Phpdftk\Css\Token\FunctionToken; |
| 18 | use Phpdftk\Css\Token\HashToken; |
| 19 | use Phpdftk\Css\Token\HashTokenType; |
| 20 | use Phpdftk\Css\Token\IdentToken; |
| 21 | use Phpdftk\Css\Token\LeftBraceToken; |
| 22 | use Phpdftk\Css\Token\LeftBracketToken; |
| 23 | use Phpdftk\Css\Token\LeftParenToken; |
| 24 | use Phpdftk\Css\Token\NumberToken; |
| 25 | use Phpdftk\Css\Token\NumberTokenType; |
| 26 | use Phpdftk\Css\Token\PercentageToken; |
| 27 | use Phpdftk\Css\Token\RightBraceToken; |
| 28 | use Phpdftk\Css\Token\RightBracketToken; |
| 29 | use Phpdftk\Css\Token\RightParenToken; |
| 30 | use Phpdftk\Css\Token\SemicolonToken; |
| 31 | use Phpdftk\Css\Token\StringToken; |
| 32 | use Phpdftk\Css\Token\Token; |
| 33 | use Phpdftk\Css\Token\UrlToken; |
| 34 | use Phpdftk\Css\Token\WhitespaceToken; |
| 35 | |
| 36 | /** |
| 37 | * CSS Syntax Module 3 tokenizer (§4). Walks the preprocessed input |
| 38 | * character-by-character, dispatching by the next character into one of the |
| 39 | * "consume X" sub-procedures. |
| 40 | * |
| 41 | * Preprocessing per §3.3 is applied at construction: CR / CRLF / FF → LF; |
| 42 | * NULL → U+FFFD (handled inline during consume). Comments are stripped |
| 43 | * (not emitted as tokens). |
| 44 | * |
| 45 | * Both `tokenize()` (returns full array) and `nextToken()` (streaming) are |
| 46 | * provided, mirroring the html tokenizer's contract so downstream parsers |
| 47 | * can drive either way. |
| 48 | */ |
| 49 | final class Tokenizer |
| 50 | { |
| 51 | /** @var list<string> input as an array of UTF-8 single-codepoint strings */ |
| 52 | private array $chars; |
| 53 | private int $length; |
| 54 | private int $pos = 0; |
| 55 | /** @var list<Token> */ |
| 56 | private array $emitted = []; |
| 57 | private int $emittedCursor = 0; |
| 58 | private bool $done = false; |
| 59 | |
| 60 | public function __construct(string $input) |
| 61 | { |
| 62 | $normalised = $this->preprocess($input); |
| 63 | $this->chars = $normalised === '' ? [] : (mb_str_split($normalised, 1, 'UTF-8') ?: []); |
| 64 | $this->length = count($this->chars); |
| 65 | } |
| 66 | |
| 67 | /** @return list<Token> */ |
| 68 | public function tokenize(): array |
| 69 | { |
| 70 | while (!$this->done) { |
| 71 | $this->step(); |
| 72 | } |
| 73 | return $this->emitted; |
| 74 | } |
| 75 | |
| 76 | public function nextToken(): ?Token |
| 77 | { |
| 78 | while ($this->emittedCursor >= count($this->emitted) && !$this->done) { |
| 79 | $this->step(); |
| 80 | } |
| 81 | if ($this->emittedCursor < count($this->emitted)) { |
| 82 | return $this->emitted[$this->emittedCursor++]; |
| 83 | } |
| 84 | return null; |
| 85 | } |
| 86 | |
| 87 | private function step(): void |
| 88 | { |
| 89 | // Comments are stripped here so the dispatch below doesn't see them. |
| 90 | $this->consumeComments(); |
| 91 | $c = $this->peek(0); |
| 92 | if ($c === null) { |
| 93 | $this->emit(new EofToken()); |
| 94 | $this->done = true; |
| 95 | return; |
| 96 | } |
| 97 | if (self::isWhitespace($c)) { |
| 98 | while (($next = $this->peek(0)) !== null && self::isWhitespace($next)) { |
| 99 | $this->advance(); |
| 100 | } |
| 101 | $this->emit(new WhitespaceToken()); |
| 102 | return; |
| 103 | } |
| 104 | if ($c === '"' || $c === "'") { |
| 105 | $this->advance(); |
| 106 | $this->emit($this->consumeString($c)); |
| 107 | return; |
| 108 | } |
| 109 | if ($c === '#') { |
| 110 | // Hash if followed by ident-code-point or escape; else Delim('#'). |
| 111 | $next = $this->peek(1); |
| 112 | if ($next !== null && (self::isIdentCodePoint($next) || $this->isValidEscape(1))) { |
| 113 | $this->advance(); // consume '#' |
| 114 | $type = $this->wouldStartIdentSequence(0) ? HashTokenType::Id : HashTokenType::Unrestricted; |
| 115 | $name = $this->consumeIdentSequence(); |
| 116 | $this->emit(new HashToken($name, $type)); |
| 117 | return; |
| 118 | } |
| 119 | $this->advance(); |
| 120 | $this->emit(new DelimToken('#')); |
| 121 | return; |
| 122 | } |
| 123 | if ($c === '(') { |
| 124 | $this->advance(); |
| 125 | $this->emit(new LeftParenToken()); |
| 126 | return; |
| 127 | } |
| 128 | if ($c === ')') { |
| 129 | $this->advance(); |
| 130 | $this->emit(new RightParenToken()); |
| 131 | return; |
| 132 | } |
| 133 | if ($c === '+' || $c === '.') { |
| 134 | if ($this->wouldStartNumber(0)) { |
| 135 | $this->emit($this->consumeNumericToken()); |
| 136 | return; |
| 137 | } |
| 138 | $this->advance(); |
| 139 | $this->emit(new DelimToken($c)); |
| 140 | return; |
| 141 | } |
| 142 | if ($c === ',') { |
| 143 | $this->advance(); |
| 144 | $this->emit(new CommaToken()); |
| 145 | return; |
| 146 | } |
| 147 | if ($c === '-') { |
| 148 | if ($this->wouldStartNumber(0)) { |
| 149 | $this->emit($this->consumeNumericToken()); |
| 150 | return; |
| 151 | } |
| 152 | if ($this->peek(1) === '-' && $this->peek(2) === '>') { |
| 153 | $this->advance(); |
| 154 | $this->advance(); |
| 155 | $this->advance(); |
| 156 | $this->emit(new CdcToken()); |
| 157 | return; |
| 158 | } |
| 159 | if ($this->wouldStartIdentSequence(0)) { |
| 160 | $this->emit($this->consumeIdentLikeToken()); |
| 161 | return; |
| 162 | } |
| 163 | $this->advance(); |
| 164 | $this->emit(new DelimToken('-')); |
| 165 | return; |
| 166 | } |
| 167 | if ($c === ':') { |
| 168 | $this->advance(); |
| 169 | $this->emit(new ColonToken()); |
| 170 | return; |
| 171 | } |
| 172 | if ($c === ';') { |
| 173 | $this->advance(); |
| 174 | $this->emit(new SemicolonToken()); |
| 175 | return; |
| 176 | } |
| 177 | if ($c === '<') { |
| 178 | if ($this->peek(1) === '!' && $this->peek(2) === '-' && $this->peek(3) === '-') { |
| 179 | $this->advance(); |
| 180 | $this->advance(); |
| 181 | $this->advance(); |
| 182 | $this->advance(); |
| 183 | $this->emit(new CdoToken()); |
| 184 | return; |
| 185 | } |
| 186 | // XHTML CDATA section opener: `<![CDATA[`. CSS Syntax doesn't |
| 187 | // tokenise these but browsers ignore the brackets when parsing |
| 188 | // an XHTML stylesheet, so consume + emit nothing — the inner |
| 189 | // body tokenises as normal CSS. |
| 190 | if ( |
| 191 | $this->peek(1) === '!' |
| 192 | && $this->peek(2) === '[' |
| 193 | && $this->peek(3) === 'C' |
| 194 | && $this->peek(4) === 'D' |
| 195 | && $this->peek(5) === 'A' |
| 196 | && $this->peek(6) === 'T' |
| 197 | && $this->peek(7) === 'A' |
| 198 | && $this->peek(8) === '[' |
| 199 | ) { |
| 200 | for ($i = 0; $i < 9; $i++) { |
| 201 | $this->advance(); |
| 202 | } |
| 203 | return; |
| 204 | } |
| 205 | $this->advance(); |
| 206 | $this->emit(new DelimToken('<')); |
| 207 | return; |
| 208 | } |
| 209 | if ($c === '@') { |
| 210 | if ($this->wouldStartIdentSequence(1)) { |
| 211 | $this->advance(); // consume '@' |
| 212 | $name = $this->consumeIdentSequence(); |
| 213 | $this->emit(new AtKeywordToken($name)); |
| 214 | return; |
| 215 | } |
| 216 | $this->advance(); |
| 217 | $this->emit(new DelimToken('@')); |
| 218 | return; |
| 219 | } |
| 220 | if ($c === '[') { |
| 221 | $this->advance(); |
| 222 | $this->emit(new LeftBracketToken()); |
| 223 | return; |
| 224 | } |
| 225 | if ($c === ']') { |
| 226 | // XHTML CDATA section closer: `]]>`. Mirror the opener — eat |
| 227 | // the three characters silently so the stylesheet body parses. |
| 228 | if ($this->peek(1) === ']' && $this->peek(2) === '>') { |
| 229 | $this->advance(); |
| 230 | $this->advance(); |
| 231 | $this->advance(); |
| 232 | return; |
| 233 | } |
| 234 | $this->advance(); |
| 235 | $this->emit(new RightBracketToken()); |
| 236 | return; |
| 237 | } |
| 238 | if ($c === '\\') { |
| 239 | if ($this->isValidEscape(0)) { |
| 240 | $this->emit($this->consumeIdentLikeToken()); |
| 241 | return; |
| 242 | } |
| 243 | $this->advance(); |
| 244 | $this->emit(new DelimToken('\\')); |
| 245 | return; |
| 246 | } |
| 247 | if ($c === '{') { |
| 248 | $this->advance(); |
| 249 | $this->emit(new LeftBraceToken()); |
| 250 | return; |
| 251 | } |
| 252 | if ($c === '}') { |
| 253 | $this->advance(); |
| 254 | $this->emit(new RightBraceToken()); |
| 255 | return; |
| 256 | } |
| 257 | if (self::isDigit($c)) { |
| 258 | $this->emit($this->consumeNumericToken()); |
| 259 | return; |
| 260 | } |
| 261 | if (self::isIdentStartCodePoint($c)) { |
| 262 | $this->emit($this->consumeIdentLikeToken()); |
| 263 | return; |
| 264 | } |
| 265 | $this->advance(); |
| 266 | $this->emit(new DelimToken($c)); |
| 267 | } |
| 268 | |
| 269 | // ============================================================ |
| 270 | // Sub-procedures |
| 271 | // ============================================================ |
| 272 | |
| 273 | private function consumeComments(): void |
| 274 | { |
| 275 | while ($this->peek(0) === '/' && $this->peek(1) === '*') { |
| 276 | $this->advance(); |
| 277 | $this->advance(); |
| 278 | while (true) { |
| 279 | $c = $this->peek(0); |
| 280 | if ($c === null) { |
| 281 | return; // EOF inside comment is a parse error per spec; we just stop. |
| 282 | } |
| 283 | if ($c === '*' && $this->peek(1) === '/') { |
| 284 | $this->advance(); |
| 285 | $this->advance(); |
| 286 | break; |
| 287 | } |
| 288 | $this->advance(); |
| 289 | } |
| 290 | } |
| 291 | } |
| 292 | |
| 293 | private function consumeString(string $terminator): Token |
| 294 | { |
| 295 | $buf = ''; |
| 296 | while (true) { |
| 297 | $c = $this->peek(0); |
| 298 | if ($c === null) { |
| 299 | return new StringToken($buf); // EOF — parse error per spec; return what we have. |
| 300 | } |
| 301 | if ($c === $terminator) { |
| 302 | $this->advance(); |
| 303 | return new StringToken($buf); |
| 304 | } |
| 305 | if ($c === "\n") { |
| 306 | return new BadStringToken(); |
| 307 | } |
| 308 | if ($c === '\\') { |
| 309 | if ($this->peek(1) === null) { |
| 310 | $this->advance(); |
| 311 | continue; |
| 312 | } |
| 313 | if ($this->peek(1) === "\n") { |
| 314 | $this->advance(); |
| 315 | $this->advance(); |
| 316 | continue; |
| 317 | } |
| 318 | $buf .= $this->consumeEscape(); |
| 319 | continue; |
| 320 | } |
| 321 | $buf .= $c; |
| 322 | $this->advance(); |
| 323 | } |
| 324 | } |
| 325 | |
| 326 | private function consumeIdentLikeToken(): Token |
| 327 | { |
| 328 | $name = $this->consumeIdentSequence(); |
| 329 | if (strcasecmp($name, 'url') === 0 && $this->peek(0) === '(') { |
| 330 | $this->advance(); // consume '(' |
| 331 | // Skip leading whitespace. |
| 332 | while (($next = $this->peek(0)) !== null && self::isWhitespace($next)) { |
| 333 | $this->advance(); |
| 334 | } |
| 335 | // If quote follows, it's a function-call url() with a string arg. |
| 336 | $n = $this->peek(0); |
| 337 | if ($n === '"' || $n === "'") { |
| 338 | return new FunctionToken($name); |
| 339 | } |
| 340 | return $this->consumeUrlToken(); |
| 341 | } |
| 342 | if ($this->peek(0) === '(') { |
| 343 | $this->advance(); |
| 344 | return new FunctionToken($name); |
| 345 | } |
| 346 | return new IdentToken($name); |
| 347 | } |
| 348 | |
| 349 | private function consumeUrlToken(): Token |
| 350 | { |
| 351 | $buf = ''; |
| 352 | while (true) { |
| 353 | $c = $this->peek(0); |
| 354 | if ($c === null) { |
| 355 | return new UrlToken($buf); |
| 356 | } |
| 357 | if ($c === ')') { |
| 358 | $this->advance(); |
| 359 | return new UrlToken($buf); |
| 360 | } |
| 361 | if (self::isWhitespace($c)) { |
| 362 | while (($next = $this->peek(0)) !== null && self::isWhitespace($next)) { |
| 363 | $this->advance(); |
| 364 | } |
| 365 | if ($this->peek(0) === ')') { |
| 366 | $this->advance(); |
| 367 | return new UrlToken($buf); |
| 368 | } |
| 369 | if ($this->peek(0) === null) { |
| 370 | return new UrlToken($buf); |
| 371 | } |
| 372 | $this->consumeRemnantsOfBadUrl(); |
| 373 | return new BadUrlToken(); |
| 374 | } |
| 375 | if ($c === '"' || $c === "'" || $c === '(' || self::isNonPrintable($c)) { |
| 376 | $this->consumeRemnantsOfBadUrl(); |
| 377 | return new BadUrlToken(); |
| 378 | } |
| 379 | if ($c === '\\') { |
| 380 | if ($this->isValidEscape(0)) { |
| 381 | $buf .= $this->consumeEscape(); |
| 382 | continue; |
| 383 | } |
| 384 | $this->consumeRemnantsOfBadUrl(); |
| 385 | return new BadUrlToken(); |
| 386 | } |
| 387 | $buf .= $c; |
| 388 | $this->advance(); |
| 389 | } |
| 390 | } |
| 391 | |
| 392 | private function consumeRemnantsOfBadUrl(): void |
| 393 | { |
| 394 | while (true) { |
| 395 | $c = $this->peek(0); |
| 396 | if ($c === null || $c === ')') { |
| 397 | if ($c === ')') { |
| 398 | $this->advance(); |
| 399 | } |
| 400 | return; |
| 401 | } |
| 402 | if ($c === '\\' && $this->isValidEscape(0)) { |
| 403 | $this->consumeEscape(); |
| 404 | continue; |
| 405 | } |
| 406 | $this->advance(); |
| 407 | } |
| 408 | } |
| 409 | |
| 410 | private function consumeIdentSequence(): string |
| 411 | { |
| 412 | $out = ''; |
| 413 | while (true) { |
| 414 | $c = $this->peek(0); |
| 415 | if ($c === null) { |
| 416 | return $out; |
| 417 | } |
| 418 | if (self::isIdentCodePoint($c)) { |
| 419 | $out .= $c; |
| 420 | $this->advance(); |
| 421 | continue; |
| 422 | } |
| 423 | if ($this->isValidEscape(0)) { |
| 424 | $out .= $this->consumeEscape(); |
| 425 | continue; |
| 426 | } |
| 427 | return $out; |
| 428 | } |
| 429 | } |
| 430 | |
| 431 | private function consumeNumericToken(): Token |
| 432 | { |
| 433 | [$value, $type] = $this->consumeNumber(); |
| 434 | if ($this->wouldStartIdentSequence(0)) { |
| 435 | $unit = $this->consumeIdentSequence(); |
| 436 | return new DimensionToken($value, $unit, $type); |
| 437 | } |
| 438 | if ($this->peek(0) === '%') { |
| 439 | $this->advance(); |
| 440 | return new PercentageToken($value); |
| 441 | } |
| 442 | return new NumberToken($value, $type); |
| 443 | } |
| 444 | |
| 445 | /** @return array{0: float, 1: NumberTokenType} */ |
| 446 | private function consumeNumber(): array |
| 447 | { |
| 448 | $type = NumberTokenType::Integer; |
| 449 | $buf = ''; |
| 450 | $c = $this->peek(0); |
| 451 | if ($c === '+' || $c === '-') { |
| 452 | $buf .= $c; |
| 453 | $this->advance(); |
| 454 | } |
| 455 | while (self::isDigit($this->peek(0) ?? '')) { |
| 456 | $buf .= $this->peek(0); |
| 457 | $this->advance(); |
| 458 | } |
| 459 | if ($this->peek(0) === '.' && self::isDigit($this->peek(1) ?? '')) { |
| 460 | $buf .= $this->peek(0) . $this->peek(1); |
| 461 | $this->advance(); |
| 462 | $this->advance(); |
| 463 | $type = NumberTokenType::Number; |
| 464 | while (self::isDigit($this->peek(0) ?? '')) { |
| 465 | $buf .= $this->peek(0); |
| 466 | $this->advance(); |
| 467 | } |
| 468 | } |
| 469 | $c = $this->peek(0); |
| 470 | $next = $this->peek(1); |
| 471 | $next2 = $this->peek(2); |
| 472 | if (($c === 'e' || $c === 'E') |
| 473 | && (self::isDigit($next ?? '') |
| 474 | || (($next === '+' || $next === '-') && self::isDigit($next2 ?? ''))) |
| 475 | ) { |
| 476 | $buf .= $this->peek(0); |
| 477 | $this->advance(); |
| 478 | if ($this->peek(0) === '+' || $this->peek(0) === '-') { |
| 479 | $buf .= $this->peek(0); |
| 480 | $this->advance(); |
| 481 | } |
| 482 | $type = NumberTokenType::Number; |
| 483 | while (self::isDigit($this->peek(0) ?? '')) { |
| 484 | $buf .= $this->peek(0); |
| 485 | $this->advance(); |
| 486 | } |
| 487 | } |
| 488 | return [(float) $buf, $type]; |
| 489 | } |
| 490 | |
| 491 | private function consumeEscape(): string |
| 492 | { |
| 493 | // Caller has positioned us on '\\'; advance past it. |
| 494 | $this->advance(); |
| 495 | $c = $this->peek(0); |
| 496 | if ($c === null) { |
| 497 | return "\u{FFFD}"; |
| 498 | } |
| 499 | if (self::isHexDigit($c)) { |
| 500 | $hex = ''; |
| 501 | for ($i = 0; $i < 6; $i++) { |
| 502 | $n = $this->peek(0); |
| 503 | if ($n === null || !self::isHexDigit($n)) { |
| 504 | break; |
| 505 | } |
| 506 | $hex .= $n; |
| 507 | $this->advance(); |
| 508 | } |
| 509 | $next = $this->peek(0); |
| 510 | if ($next !== null && self::isWhitespace($next)) { |
| 511 | $this->advance(); |
| 512 | } |
| 513 | $cp = (int) hexdec($hex); |
| 514 | if ($cp === 0 || $cp > 0x10FFFF || ($cp >= 0xD800 && $cp <= 0xDFFF)) { |
| 515 | return "\u{FFFD}"; |
| 516 | } |
| 517 | return mb_chr($cp, 'UTF-8') ?: "\u{FFFD}"; |
| 518 | } |
| 519 | $this->advance(); |
| 520 | return $c; |
| 521 | } |
| 522 | |
| 523 | // ============================================================ |
| 524 | // Lookahead helpers (CSS Syntax 3 §4.3.8 / §4.3.9) |
| 525 | // ============================================================ |
| 526 | |
| 527 | private function isValidEscape(int $offset): bool |
| 528 | { |
| 529 | if ($this->peek($offset) !== '\\') { |
| 530 | return false; |
| 531 | } |
| 532 | return $this->peek($offset + 1) !== "\n"; |
| 533 | } |
| 534 | |
| 535 | private function wouldStartIdentSequence(int $offset): bool |
| 536 | { |
| 537 | $c1 = $this->peek($offset); |
| 538 | if ($c1 === '-') { |
| 539 | $c2 = $this->peek($offset + 1); |
| 540 | if ($c2 !== null && (self::isIdentStartCodePoint($c2) || $c2 === '-')) { |
| 541 | return true; |
| 542 | } |
| 543 | return $this->isValidEscape($offset + 1); |
| 544 | } |
| 545 | if ($c1 !== null && self::isIdentStartCodePoint($c1)) { |
| 546 | return true; |
| 547 | } |
| 548 | if ($c1 === '\\') { |
| 549 | return $this->isValidEscape($offset); |
| 550 | } |
| 551 | return false; |
| 552 | } |
| 553 | |
| 554 | private function wouldStartNumber(int $offset): bool |
| 555 | { |
| 556 | $c1 = $this->peek($offset); |
| 557 | if ($c1 === '+' || $c1 === '-') { |
| 558 | $c2 = $this->peek($offset + 1); |
| 559 | if (self::isDigit($c2 ?? '')) { |
| 560 | return true; |
| 561 | } |
| 562 | if ($c2 === '.') { |
| 563 | $c3 = $this->peek($offset + 2); |
| 564 | return self::isDigit($c3 ?? ''); |
| 565 | } |
| 566 | return false; |
| 567 | } |
| 568 | if ($c1 === '.') { |
| 569 | return self::isDigit($this->peek($offset + 1) ?? ''); |
| 570 | } |
| 571 | return $c1 !== null && self::isDigit($c1); |
| 572 | } |
| 573 | |
| 574 | // ============================================================ |
| 575 | // I/O helpers |
| 576 | // ============================================================ |
| 577 | |
| 578 | private function preprocess(string $input): string |
| 579 | { |
| 580 | $input = str_replace(["\r\n", "\r", "\f"], "\n", $input); |
| 581 | return str_replace("\0", "\u{FFFD}", $input); |
| 582 | } |
| 583 | |
| 584 | private function peek(int $offset): ?string |
| 585 | { |
| 586 | $i = $this->pos + $offset; |
| 587 | return $i < $this->length ? $this->chars[$i] : null; |
| 588 | } |
| 589 | |
| 590 | private function advance(): void |
| 591 | { |
| 592 | $this->pos++; |
| 593 | } |
| 594 | |
| 595 | private function emit(Token $t): void |
| 596 | { |
| 597 | $this->emitted[] = $t; |
| 598 | } |
| 599 | |
| 600 | // ============================================================ |
| 601 | // Character classification |
| 602 | // ============================================================ |
| 603 | |
| 604 | private static function isWhitespace(string $c): bool |
| 605 | { |
| 606 | return $c === ' ' || $c === "\t" || $c === "\n"; |
| 607 | } |
| 608 | |
| 609 | private static function isDigit(string $c): bool |
| 610 | { |
| 611 | return $c >= '0' && $c <= '9'; |
| 612 | } |
| 613 | |
| 614 | private static function isHexDigit(string $c): bool |
| 615 | { |
| 616 | return self::isDigit($c) || ($c >= 'A' && $c <= 'F') || ($c >= 'a' && $c <= 'f'); |
| 617 | } |
| 618 | |
| 619 | private static function isLetter(string $c): bool |
| 620 | { |
| 621 | return ($c >= 'a' && $c <= 'z') || ($c >= 'A' && $c <= 'Z'); |
| 622 | } |
| 623 | |
| 624 | private static function isIdentStartCodePoint(string $c): bool |
| 625 | { |
| 626 | if ($c === '') { |
| 627 | return false; |
| 628 | } |
| 629 | if (self::isLetter($c) || $c === '_') { |
| 630 | return true; |
| 631 | } |
| 632 | return mb_ord($c, 'UTF-8') >= 0x80; |
| 633 | } |
| 634 | |
| 635 | private static function isIdentCodePoint(string $c): bool |
| 636 | { |
| 637 | return self::isIdentStartCodePoint($c) || self::isDigit($c) || $c === '-'; |
| 638 | } |
| 639 | |
| 640 | private static function isNonPrintable(string $c): bool |
| 641 | { |
| 642 | $cp = mb_ord($c, 'UTF-8'); |
| 643 | if ($cp === false) { |
| 644 | return false; |
| 645 | } |
| 646 | return ($cp >= 0x00 && $cp <= 0x08) |
| 647 | || $cp === 0x0B |
| 648 | || ($cp >= 0x0E && $cp <= 0x1F) |
| 649 | || $cp === 0x7F; |
| 650 | } |
| 651 | } |