smalot /
pdfparser
| Total Complexity | 180 |
| Total Lines | 1161 |
| Duplicated Lines | 0 % |
| Changes | 5 | ||
| Bugs | 3 | Features | 0 |
Complex classes like PDFObject often do a lot of different things. To break such a class down, we need to identify a cohesive component within that class. A common approach to find such a component is to look for fields/methods that share the same prefixes, or suffixes.
Once you have determined the fields that belong together, you can apply the Extract Class refactoring. If the component makes sense as a sub-class, Extract Subclass is also a candidate, and is often faster.
While breaking up the class, it is a good idea to analyze how other classes use PDFObject, and based on these observations, apply Extract Interface, too.
| 1 | <?php |
||
| 42 | class PDFObject |
||
| 43 | { |
||
| 44 | public const TYPE = 't'; |
||
| 45 | |||
| 46 | public const OPERATOR = 'o'; |
||
| 47 | |||
| 48 | public const COMMAND = 'c'; |
||
| 49 | |||
| 50 | /** |
||
| 51 | * The recursion stack. |
||
| 52 | * |
||
| 53 | * @var array |
||
| 54 | */ |
||
| 55 | public static $recursionStack = []; |
||
| 56 | |||
| 57 | /** |
||
| 58 | * @var Document|null |
||
| 59 | */ |
||
| 60 | protected $document; |
||
| 61 | |||
| 62 | /** |
||
| 63 | * @var Header |
||
| 64 | */ |
||
| 65 | protected $header; |
||
| 66 | |||
| 67 | /** |
||
| 68 | * @var string |
||
| 69 | */ |
||
| 70 | protected $content; |
||
| 71 | |||
| 72 | /** |
||
| 73 | * @var Config|null |
||
| 74 | */ |
||
| 75 | protected $config; |
||
| 76 | |||
| 77 | /** |
||
| 78 | * @var bool |
||
| 79 | */ |
||
| 80 | protected $addPositionWhitespace = false; |
||
| 81 | |||
| 82 | public function __construct( |
||
| 83 | Document $document, |
||
| 84 | ?Header $header = null, |
||
| 85 | ?string $content = null, |
||
| 86 | ?Config $config = null |
||
| 87 | ) { |
||
| 88 | $this->document = $document; |
||
| 89 | $this->header = $header ?? new Header(); |
||
| 90 | $this->content = $content; |
||
| 91 | $this->config = $config; |
||
| 92 | } |
||
| 93 | |||
| 94 | public function init() |
||
| 95 | { |
||
| 96 | } |
||
| 97 | |||
| 98 | public function getDocument(): Document |
||
| 99 | { |
||
| 100 | return $this->document; |
||
|
Bug
Best Practice
introduced
by
|
|||
| 101 | } |
||
| 102 | |||
| 103 | public function getHeader(): ?Header |
||
| 104 | { |
||
| 105 | return $this->header; |
||
| 106 | } |
||
| 107 | |||
| 108 | public function getConfig(): ?Config |
||
| 109 | { |
||
| 110 | return $this->config; |
||
| 111 | } |
||
| 112 | |||
| 113 | /** |
||
| 114 | * @return Element|PDFObject|Header |
||
| 115 | */ |
||
| 116 | public function get(string $name) |
||
| 117 | { |
||
| 118 | return $this->header->get($name); |
||
| 119 | } |
||
| 120 | |||
| 121 | public function has(string $name): bool |
||
| 122 | { |
||
| 123 | return $this->header->has($name); |
||
| 124 | } |
||
| 125 | |||
| 126 | public function getDetails(bool $deep = true): array |
||
| 127 | { |
||
| 128 | return $this->header->getDetails($deep); |
||
| 129 | } |
||
| 130 | |||
| 131 | public function getContent(): ?string |
||
| 134 | } |
||
| 135 | |||
| 136 | /** |
||
| 137 | * Creates a duplicate of the document stream with |
||
| 138 | * strings and other items replaced by $char. Formerly |
||
| 139 | * getSectionsText() used this output to more easily gather offset |
||
| 140 | * values to extract text from the *actual* document stream. |
||
| 141 | * |
||
| 142 | * @deprecated function is no longer used and will be removed in a future release |
||
| 143 | * |
||
| 144 | * @internal |
||
| 145 | */ |
||
| 146 | public function cleanContent(string $content, string $char = 'X') |
||
| 147 | { |
||
| 148 | $char = $char[0]; |
||
| 149 | $content = str_replace(['\\\\', '\\)', '\\('], $char.$char, $content); |
||
| 150 | |||
| 151 | // Remove image bloc with binary content |
||
| 152 | preg_match_all('/\s(BI\s.*?(\sID\s).*?(\sEI))\s/s', $content, $matches, \PREG_OFFSET_CAPTURE); |
||
| 153 | foreach ($matches[0] as $part) { |
||
| 154 | $content = substr_replace($content, str_repeat($char, \strlen($part[0])), $part[1], \strlen($part[0])); |
||
| 155 | } |
||
| 156 | |||
| 157 | // Clean content in square brackets [.....] |
||
| 158 | preg_match_all('/\[((\(.*?\)|[0-9\.\-\s]*)*)\]/s', $content, $matches, \PREG_OFFSET_CAPTURE); |
||
| 159 | foreach ($matches[1] as $part) { |
||
| 160 | $content = substr_replace($content, str_repeat($char, \strlen($part[0])), $part[1], \strlen($part[0])); |
||
| 161 | } |
||
| 162 | |||
| 163 | // Clean content in round brackets (.....) |
||
| 164 | preg_match_all('/\((.*?)\)/s', $content, $matches, \PREG_OFFSET_CAPTURE); |
||
| 165 | foreach ($matches[1] as $part) { |
||
| 166 | $content = substr_replace($content, str_repeat($char, \strlen($part[0])), $part[1], \strlen($part[0])); |
||
| 167 | } |
||
| 168 | |||
| 169 | // Clean structure |
||
| 170 | if ($parts = preg_split('/(<|>)/s', $content, -1, \PREG_SPLIT_NO_EMPTY | \PREG_SPLIT_DELIM_CAPTURE)) { |
||
| 171 | $content = ''; |
||
| 172 | $level = 0; |
||
| 173 | foreach ($parts as $part) { |
||
| 174 | if ('<' == $part) { |
||
| 175 | ++$level; |
||
| 176 | } |
||
| 177 | |||
| 178 | $content .= (0 == $level ? $part : str_repeat($char, \strlen($part))); |
||
| 179 | |||
| 180 | if ('>' == $part) { |
||
| 181 | --$level; |
||
| 182 | } |
||
| 183 | } |
||
| 184 | } |
||
| 185 | |||
| 186 | // Clean BDC and EMC markup |
||
| 187 | preg_match_all( |
||
| 188 | '/(\/[A-Za-z0-9\_]*\s*'.preg_quote($char).'*BDC)/s', |
||
| 189 | $content, |
||
| 190 | $matches, |
||
| 191 | \PREG_OFFSET_CAPTURE |
||
| 192 | ); |
||
| 193 | foreach ($matches[1] as $part) { |
||
| 194 | $content = substr_replace($content, str_repeat($char, \strlen($part[0])), $part[1], \strlen($part[0])); |
||
| 195 | } |
||
| 196 | |||
| 197 | preg_match_all('/\s(EMC)\s/s', $content, $matches, \PREG_OFFSET_CAPTURE); |
||
| 198 | foreach ($matches[1] as $part) { |
||
| 199 | $content = substr_replace($content, str_repeat($char, \strlen($part[0])), $part[1], \strlen($part[0])); |
||
| 200 | } |
||
| 201 | |||
| 202 | return $content; |
||
| 203 | } |
||
| 204 | |||
| 205 | /** |
||
| 206 | * Takes a string of PDF document stream text and formats |
||
| 207 | * it into a multi-line string with one PDF command on each line, |
||
| 208 | * separated by \r\n. If the given string is null, or binary data |
||
| 209 | * is detected instead of a document stream then return an empty |
||
| 210 | * string. |
||
| 211 | */ |
||
| 212 | private function formatContent(?string $content): string |
||
| 213 | { |
||
| 214 | if (null === $content) { |
||
| 215 | return ''; |
||
| 216 | } |
||
| 217 | |||
| 218 | // Outside of (String) and inline image content in PDF document |
||
| 219 | // streams, all text should conform to UTF-8. Test for binary |
||
| 220 | // content by deleting everything after the first open- |
||
| 221 | // parenthesis ( which indicates the beginning of a string, or |
||
| 222 | // the first ID command which indicates the beginning of binary |
||
| 223 | // inline image content. Then test what remains for valid |
||
| 224 | // UTF-8. If it's not UTF-8, return an empty string as this |
||
| 225 | // $content is most likely binary. Unfortunately, using |
||
| 226 | // mb_check_encoding(..., 'UTF-8') is not strict enough, so the |
||
| 227 | // following regexp, adapted from the W3, is used. See: |
||
| 228 | // https://www.w3.org/International/questions/qa-forms-utf-8.en |
||
| 229 | // We use preg_replace() instead of preg_match() to avoid "JIT |
||
| 230 | // stack limit exhausted" errors on larger files. |
||
| 231 | $utf8Filter = preg_replace('/( |
||
| 232 | [\x09\x0A\x0D\x20-\x7E] | # ASCII |
||
| 233 | [\xC2-\xDF][\x80-\xBF] | # non-overlong 2-byte |
||
| 234 | \xE0[\xA0-\xBF][\x80-\xBF] | # excluding overlongs |
||
| 235 | [\xE1-\xEC\xEE\xEF][\x80-\xBF]{2} | # straight 3-byte |
||
| 236 | \xED[\x80-\x9F][\x80-\xBF] | # excluding surrogates |
||
| 237 | \xF0[\x90-\xBF][\x80-\xBF]{2} | # planes 1-3 |
||
| 238 | [\xF1-\xF3][\x80-\xBF]{3} | # planes 4-15 |
||
| 239 | \xF4[\x80-\x8F][\x80-\xBF]{2} # plane 16 |
||
| 240 | )/xs', '', preg_replace('/(\(|ID\s).*$/s', '', $content)); |
||
| 241 | |||
| 242 | if ('' !== $utf8Filter) { |
||
| 243 | return ''; |
||
| 244 | } |
||
| 245 | |||
| 246 | // Find all inline image content and replace them so they aren't |
||
| 247 | // affected by the next steps |
||
| 248 | $pdfInlineImages = []; |
||
| 249 | $offsetBI = 0; |
||
| 250 | while (preg_match('/\sBI\s(\/.+?)\sID\s(.+?)\sEI(?=\s|$)/s', $content, $text, \PREG_OFFSET_CAPTURE, $offsetBI)) { |
||
| 251 | // Attempt to detemine if this instance of the 'BI' command |
||
| 252 | // actually occured within a (string) using the following |
||
| 253 | // steps: |
||
| 254 | |||
| 255 | // Step 1: Remove any escaped slashes and parentheses from |
||
| 256 | // the alleged image characteristics data |
||
| 257 | $para = str_replace(['\\\\', '\\(', '\\)'], '', $text[1][0]); |
||
| 258 | |||
| 259 | // Step 2: Remove all correctly ordered and balanced |
||
| 260 | // parentheses from (strings) |
||
| 261 | do { |
||
| 262 | $paraTest = $para; |
||
| 263 | $para = preg_replace('/\(([^()]*)\)/', '$1', $paraTest); |
||
| 264 | } while ($para != $paraTest); |
||
| 265 | |||
| 266 | $paraOpen = strpos($para, '('); |
||
| 267 | $paraClose = strpos($para, ')'); |
||
| 268 | |||
| 269 | // Check: If the remaining text contains a close parenthesis |
||
| 270 | // ')' AND it occurs before any open parenthesis, then we |
||
| 271 | // are almost certain to be inside a (string) |
||
| 272 | if (0 < $paraClose && (false === $paraOpen || $paraClose < $paraOpen)) { |
||
| 273 | // Bump the search offset forward and match again |
||
| 274 | $offsetBI = (int) $text[1][1]; |
||
| 275 | continue; |
||
| 276 | } |
||
| 277 | |||
| 278 | // Step 3: Double check that this is actually inline image |
||
| 279 | // data by parsing the alleged image characteristics as a |
||
| 280 | // dictionary |
||
| 281 | $dict = $this->parseDictionary('<<'.$text[1][0].'>>'); |
||
| 282 | |||
| 283 | // Check if an image Width and Height are set in the dict |
||
| 284 | if ((isset($dict['W']) || isset($dict['Width'])) |
||
| 285 | && (isset($dict['H']) || isset($dict['Height']))) { |
||
| 286 | $id = uniqid('IMAGE_', true); |
||
| 287 | $pdfInlineImages[$id] = [ |
||
| 288 | preg_replace(['/\r\n/', '/\r/', '/\n/'], ' ', $text[1][0]), |
||
| 289 | preg_replace(['/\r\n/', '/\r/', '/\n/'], '', $text[2][0]), |
||
| 290 | ]; |
||
| 291 | $content = preg_replace( |
||
| 292 | '/'.preg_quote($text[0][0], '/').'/', |
||
| 293 | '^^^'.$id.'^^^', |
||
| 294 | $content, |
||
| 295 | 1 |
||
| 296 | ); |
||
| 297 | } else { |
||
| 298 | // If there was no valid dictionary, or a height and width |
||
| 299 | // weren't specified, then we don't know what this is, so |
||
| 300 | // just leave it alone; bump the search offset forward and |
||
| 301 | // match again |
||
| 302 | $offsetBI = (int) $text[1][1]; |
||
| 303 | } |
||
| 304 | } |
||
| 305 | |||
| 306 | // Find all strings () and replace them so they aren't affected |
||
| 307 | // by the next steps |
||
| 308 | $pdfstrings = []; |
||
| 309 | $attempt = '('; |
||
| 310 | while (preg_match('/'.preg_quote($attempt, '/').'.*?\)/s', $content, $text)) { |
||
| 311 | // Remove all escaped slashes and parentheses from the target text |
||
| 312 | $para = str_replace(['\\\\', '\\(', '\\)'], '', $text[0]); |
||
| 313 | |||
| 314 | // PDF strings can contain unescaped parentheses as long as |
||
| 315 | // they're balanced, so check for balanced parentheses |
||
| 316 | $left = preg_match_all('/\(/', $para); |
||
| 317 | $right = preg_match_all('/\)/', $para); |
||
| 318 | |||
| 319 | if (')' == $para[-1] && $left == $right) { |
||
| 320 | // Replace the string with a unique placeholder |
||
| 321 | $id = uniqid('STRING_', true); |
||
| 322 | $pdfstrings[$id] = $text[0]; |
||
| 323 | $content = preg_replace( |
||
| 324 | '/'.preg_quote($text[0], '/').'/', |
||
| 325 | '@@@'.$id.'@@@', |
||
| 326 | $content, |
||
| 327 | 1 |
||
| 328 | ); |
||
| 329 | |||
| 330 | // Reset to search for the next string |
||
| 331 | $attempt = '('; |
||
| 332 | } else { |
||
| 333 | // We had unbalanced parentheses, so use the current |
||
| 334 | // match as a base to find a longer string |
||
| 335 | $attempt = $text[0]; |
||
| 336 | } |
||
| 337 | } |
||
| 338 | |||
| 339 | // Remove all carriage returns and line-feeds from the document stream |
||
| 340 | $content = str_replace(["\r", "\n"], ' ', trim($content)); |
||
| 341 | |||
| 342 | // Find all dictionary << >> commands and replace them so they |
||
| 343 | // aren't affected by the next steps |
||
| 344 | $dictstore = []; |
||
| 345 | while (preg_match('/(<<.*?>> *)(BDC|BMC|DP|MP)/s', $content, $dicttext)) { |
||
| 346 | $dictid = uniqid('DICT_', true); |
||
| 347 | $dictstore[$dictid] = $dicttext[1]; |
||
| 348 | $content = preg_replace( |
||
| 349 | '/'.preg_quote($dicttext[0], '/').'/', |
||
| 350 | ' ###'.$dictid.'###'.$dicttext[2], |
||
| 351 | $content, |
||
| 352 | 1 |
||
| 353 | ); |
||
| 354 | } |
||
| 355 | |||
| 356 | // Normalize white-space in the document stream |
||
| 357 | $content = preg_replace('/\s{2,}/', ' ', $content); |
||
| 358 | |||
| 359 | // Find all valid PDF operators and add \r\n after each; this |
||
| 360 | // ensures there is just one command on every line |
||
| 361 | // Source: https://ia801001.us.archive.org/1/items/pdf1.7/pdf_reference_1-7.pdf - Appendix A |
||
| 362 | // Source: https://archive.org/download/pdf320002008/PDF32000_2008.pdf - Annex A |
||
| 363 | // Note: PDF Reference 1.7 lists 'I' and 'rI' as valid commands, while |
||
| 364 | // PDF 32000:2008 lists them as 'i' and 'ri' respectively. Both versions |
||
| 365 | // appear here in the list for completeness. |
||
| 366 | $operators = [ |
||
| 367 | 'b*', 'b', 'BDC', 'BMC', 'B*', 'BI', 'BT', 'BX', 'B', 'cm', 'cs', 'c', 'CS', |
||
| 368 | 'd0', 'd1', 'd', 'Do', 'DP', 'EMC', 'EI', 'ET', 'EX', 'f*', 'f', 'F', 'gs', |
||
| 369 | 'g', 'G', 'h', 'i', 'ID', 'I', 'j', 'J', 'k', 'K', 'l', 'm', 'MP', 'M', 'n', |
||
| 370 | 'q', 'Q', 're', 'rg', 'ri', 'rI', 'RG', 'scn', 'sc', 'sh', 's', 'SCN', 'SC', |
||
| 371 | 'S', 'T*', 'Tc', 'Td', 'TD', 'Tf', 'TJ', 'Tj', 'TL', 'Tm', 'Tr', 'Ts', 'Tw', |
||
| 372 | 'Tz', 'v', 'w', 'W*', 'W', 'y', '\'', '"', |
||
| 373 | ]; |
||
| 374 | foreach ($operators as $operator) { |
||
| 375 | $content = preg_replace( |
||
| 376 | '/(?<!\w|\/)'.preg_quote($operator, '/').'(?![\w10\*])/', |
||
| 377 | $operator."\r\n", |
||
| 378 | $content |
||
| 379 | ); |
||
| 380 | } |
||
| 381 | |||
| 382 | // Restore the original content of the dictionary << >> commands |
||
| 383 | $dictstore = array_reverse($dictstore, true); |
||
| 384 | foreach ($dictstore as $id => $dict) { |
||
| 385 | $content = str_replace('###'.$id.'###', $dict, $content); |
||
| 386 | } |
||
| 387 | |||
| 388 | // Restore the original string content |
||
| 389 | $pdfstrings = array_reverse($pdfstrings, true); |
||
| 390 | foreach ($pdfstrings as $id => $text) { |
||
| 391 | // Strings may contain escaped newlines, or literal newlines |
||
| 392 | // and we should clean these up before replacing the string |
||
| 393 | // back into the content stream; this ensures no strings are |
||
| 394 | // split between two lines (every command must be on one line) |
||
| 395 | $text = str_replace( |
||
| 396 | ["\\\r\n", "\\\r", "\\\n", "\r", "\n"], |
||
| 397 | ['', '', '', '\r', '\n'], |
||
| 398 | $text |
||
| 399 | ); |
||
| 400 | |||
| 401 | $content = str_replace('@@@'.$id.'@@@', $text, $content); |
||
| 402 | } |
||
| 403 | |||
| 404 | // Restore the original content of any inline images |
||
| 405 | $pdfInlineImages = array_reverse($pdfInlineImages, true); |
||
| 406 | foreach ($pdfInlineImages as $id => $image) { |
||
| 407 | $content = str_replace( |
||
| 408 | '^^^'.$id.'^^^', |
||
| 409 | "\r\nBI\r\n".$image[0]." ID\r\n".$image[1]." EI\r\n", |
||
| 410 | $content |
||
| 411 | ); |
||
| 412 | } |
||
| 413 | |||
| 414 | $content = trim(preg_replace(['/(\r\n){2,}/', '/\r\n +/'], "\r\n", $content)); |
||
| 415 | |||
| 416 | return $content; |
||
| 417 | } |
||
| 418 | |||
| 419 | /** |
||
| 420 | * getSectionsText() now takes an entire, unformatted |
||
| 421 | * document stream as a string, cleans it, then filters out |
||
| 422 | * commands that aren't needed for text positioning/extraction. It |
||
| 423 | * returns an array of unprocessed PDF commands, one command per |
||
| 424 | * element. |
||
| 425 | * |
||
| 426 | * @internal |
||
| 427 | */ |
||
| 428 | public function getSectionsText(?string $content): array |
||
| 429 | { |
||
| 430 | $sections = []; |
||
| 431 | |||
| 432 | // A cleaned stream has one command on every line, so split the |
||
| 433 | // cleaned stream content on \r\n into an array |
||
| 434 | $textCleaned = preg_split( |
||
| 435 | '/(\r\n|\n|\r)/', |
||
| 436 | $this->formatContent($content), |
||
| 437 | -1, |
||
| 438 | \PREG_SPLIT_NO_EMPTY |
||
| 439 | ); |
||
| 440 | |||
| 441 | $inTextBlock = false; |
||
| 442 | foreach ($textCleaned as $line) { |
||
| 443 | $line = trim($line); |
||
| 444 | |||
| 445 | // Skip empty lines |
||
| 446 | if ('' === $line) { |
||
| 447 | continue; |
||
| 448 | } |
||
| 449 | |||
| 450 | // If a 'BT' is encountered, set the $inTextBlock flag |
||
| 451 | if (preg_match('/BT$/', $line)) { |
||
| 452 | $inTextBlock = true; |
||
| 453 | $sections[] = $line; |
||
| 454 | |||
| 455 | // If an 'ET' is encountered, unset the $inTextBlock flag |
||
| 456 | } elseif ('ET' == $line) { |
||
| 457 | $inTextBlock = false; |
||
| 458 | $sections[] = $line; |
||
| 459 | } elseif ($inTextBlock) { |
||
| 460 | // If we are inside a BT ... ET text block, save all lines |
||
| 461 | $sections[] = trim($line); |
||
| 462 | } else { |
||
| 463 | // Otherwise, if we are outside of a text block, only |
||
| 464 | // save specific, necessary lines. Care should be taken |
||
| 465 | // to ensure a command being checked for *only* matches |
||
| 466 | // that command. For instance, a simple search for 'c' |
||
| 467 | // may also match the 'sc' command. See the command |
||
| 468 | // list in the formatContent() method above. |
||
| 469 | // Add more commands to save here as you find them in |
||
| 470 | // weird PDFs! |
||
| 471 | if ('q' == $line[-1] || 'Q' == $line[-1]) { |
||
| 472 | // Save and restore graphics state commands |
||
| 473 | $sections[] = $line; |
||
| 474 | } elseif (preg_match('/(?<!\w)B[DM]C$/', $line)) { |
||
| 475 | // Begin marked content sequence |
||
| 476 | $sections[] = $line; |
||
| 477 | } elseif (preg_match('/(?<!\w)[DM]P$/', $line)) { |
||
| 478 | // Marked content point |
||
| 479 | $sections[] = $line; |
||
| 480 | } elseif (preg_match('/(?<!\w)EMC$/', $line)) { |
||
| 481 | // End marked content sequence |
||
| 482 | $sections[] = $line; |
||
| 483 | } elseif (preg_match('/(?<!\w)cm$/', $line)) { |
||
| 484 | // Graphics position change commands |
||
| 485 | $sections[] = $line; |
||
| 486 | } elseif (preg_match('/(?<!\w)Tf$/', $line)) { |
||
| 487 | // Font change commands |
||
| 488 | $sections[] = $line; |
||
| 489 | } elseif (preg_match('/(?<!\w)Do$/', $line)) { |
||
| 490 | // Invoke named XObject command |
||
| 491 | $sections[] = $line; |
||
| 492 | } |
||
| 493 | } |
||
| 494 | } |
||
| 495 | |||
| 496 | return $sections; |
||
| 497 | } |
||
| 498 | |||
| 499 | private function getDefaultFont(?Page $page = null): Font |
||
| 500 | { |
||
| 501 | $fonts = []; |
||
| 502 | if (null !== $page) { |
||
| 503 | $fonts = $page->getFonts(); |
||
| 504 | } |
||
| 505 | |||
| 506 | $firstFont = $this->document->getFirstFont(); |
||
| 507 | if (null !== $firstFont) { |
||
| 508 | $fonts[] = $firstFont; |
||
| 509 | } |
||
| 510 | |||
| 511 | if (\count($fonts) > 0) { |
||
| 512 | return reset($fonts); |
||
| 513 | } |
||
| 514 | |||
| 515 | return new Font($this->document, null, null, $this->config); |
||
| 516 | } |
||
| 517 | |||
| 518 | /** |
||
| 519 | * Decode a '[]TJ' command and attempt to use alternate |
||
| 520 | * fonts if the current font results in output that contains |
||
| 521 | * Unicode control characters. |
||
| 522 | * |
||
| 523 | * @internal |
||
| 524 | * |
||
| 525 | * @param array<int,array<string,string|bool>> $command |
||
| 526 | */ |
||
| 527 | private function getTJUsingFontFallback(Font $font, array $command, ?Page $page = null, float $fontFactor = 4): string |
||
| 555 | } |
||
| 556 | |||
| 557 | /** |
||
| 558 | * Expects a string that is a full PDF dictionary object, |
||
| 559 | * including the outer enclosing << >> angle brackets |
||
| 560 | * |
||
| 561 | * @internal |
||
| 562 | * |
||
| 563 | * @throws InvalidDictionaryObjectException |
||
| 564 | */ |
||
| 565 | public function parseDictionary(string $dictionary): array |
||
| 566 | { |
||
| 567 | // Normalize whitespace |
||
| 568 | $dictionary = preg_replace(['/\r/', '/\n/', '/\s{2,}/'], ' ', trim($dictionary)); |
||
| 569 | |||
| 570 | if ('<<' != substr($dictionary, 0, 2)) { |
||
| 571 | throw new InvalidDictionaryObjectException('Not a valid dictionary object.'); |
||
| 572 | } |
||
| 573 | |||
| 574 | $parsed = []; |
||
| 575 | $stack = []; |
||
| 576 | $currentName = ''; |
||
| 577 | $arrayTypeNumeric = false; |
||
| 578 | |||
| 579 | // Remove outer layer of dictionary, and split on tokens |
||
| 580 | $split = preg_split( |
||
| 581 | '/(<<|>>|\[|\]|\/[^\s\/\[\]\(\)<>]*)/', |
||
| 582 | trim(preg_replace('/^<<|>>$/', '', $dictionary)), |
||
| 583 | -1, |
||
| 584 | \PREG_SPLIT_NO_EMPTY | \PREG_SPLIT_DELIM_CAPTURE |
||
| 585 | ); |
||
| 586 | |||
| 587 | foreach ($split as $token) { |
||
| 588 | $token = trim($token); |
||
| 589 | switch ($token) { |
||
| 590 | case '': |
||
| 591 | break; |
||
| 592 | |||
| 593 | // Open numeric array |
||
| 594 | case '[': |
||
| 595 | $parsed[$currentName] = []; |
||
| 596 | $arrayTypeNumeric = true; |
||
| 597 | |||
| 598 | // Move up one level in the stack |
||
| 599 | $stack[\count($stack)] = &$parsed; |
||
| 600 | $parsed = &$parsed[$currentName]; |
||
| 601 | $currentName = ''; |
||
| 602 | break; |
||
| 603 | |||
| 604 | // Open hashed array |
||
| 605 | case '<<': |
||
| 606 | $parsed[$currentName] = []; |
||
| 607 | $arrayTypeNumeric = false; |
||
| 608 | |||
| 609 | // Move up one level in the stack |
||
| 610 | $stack[\count($stack)] = &$parsed; |
||
| 611 | $parsed = &$parsed[$currentName]; |
||
| 612 | $currentName = ''; |
||
| 613 | break; |
||
| 614 | |||
| 615 | // Close numeric array |
||
| 616 | case ']': |
||
| 617 | // Revert string type arrays back to a single element |
||
| 618 | if (\is_array($parsed) && 1 == \count($parsed) |
||
| 619 | && isset($parsed[0]) && \is_string($parsed[0]) |
||
| 620 | && '' !== $parsed[0] && '/' != $parsed[0][0]) { |
||
| 621 | $parsed = '['.$parsed[0].']'; |
||
| 622 | } |
||
| 623 | // Close hashed array |
||
| 624 | // no break |
||
| 625 | case '>>': |
||
| 626 | $arrayTypeNumeric = false; |
||
| 627 | |||
| 628 | // Move down one level in the stack |
||
| 629 | $parsed = &$stack[\count($stack) - 1]; |
||
| 630 | unset($stack[\count($stack) - 1]); |
||
| 631 | break; |
||
| 632 | |||
| 633 | default: |
||
| 634 | // If value begins with a slash, then this is a name |
||
| 635 | // Add it to the appropriate array |
||
| 636 | if ('/' == substr($token, 0, 1)) { |
||
| 637 | $currentName = substr($token, 1); |
||
| 638 | if (true == $arrayTypeNumeric) { |
||
| 639 | $parsed[] = $currentName; |
||
| 640 | $currentName = ''; |
||
| 641 | } |
||
| 642 | } elseif ('' != $currentName) { |
||
| 643 | if (false == $arrayTypeNumeric) { |
||
| 644 | $parsed[$currentName] = $token; |
||
| 645 | } |
||
| 646 | $currentName = ''; |
||
| 647 | } elseif ('' == $currentName) { |
||
| 648 | $parsed[] = $token; |
||
| 649 | } |
||
| 650 | } |
||
| 651 | } |
||
| 652 | |||
| 653 | return $parsed; |
||
| 654 | } |
||
| 655 | |||
| 656 | /** |
||
| 657 | * Returns the text content of a PDF as a string. Attempts to add |
||
| 658 | * whitespace for spacing and line-breaks where appropriate. |
||
| 659 | * |
||
| 660 | * getText() leverages getTextArray() to get the content |
||
| 661 | * of the document, setting the addPositionWhitespace flag to true |
||
| 662 | * so whitespace is inserted in a logical way for reading by |
||
| 663 | * humans. |
||
| 664 | */ |
||
| 665 | public function getText(?Page $page = null): string |
||
| 672 | } |
||
| 673 | |||
| 674 | /** |
||
| 675 | * Returns the text content of a PDF as an array of strings. No |
||
| 676 | * extra whitespace is inserted besides what is actually encoded in |
||
| 677 | * the PDF text. |
||
| 678 | * |
||
| 679 | * @throws \Exception |
||
| 680 | */ |
||
| 681 | public function getTextArray(?Page $page = null): array |
||
| 682 | { |
||
| 683 | $result = []; |
||
| 684 | $text = []; |
||
| 685 | |||
| 686 | $marked_stack = []; |
||
| 687 | $last_written_position = false; |
||
| 688 | |||
| 689 | $sections = $this->getSectionsText($this->content); |
||
| 690 | $current_font = $this->getDefaultFont($page); |
||
| 691 | $current_font_size = 1; |
||
| 692 | $current_text_leading = 0; |
||
| 693 | |||
| 694 | $current_position = ['x' => false, 'y' => false]; |
||
| 695 | $current_position_tm = [ |
||
| 696 | 'a' => 1, 'b' => 0, 'c' => 0, |
||
| 697 | 'i' => 0, 'j' => 1, 'k' => 0, |
||
| 698 | 'x' => 0, 'y' => 0, 'z' => 1, |
||
| 699 | ]; |
||
| 700 | $current_position_td = ['x' => 0, 'y' => 0]; |
||
| 701 | $current_position_cm = [ |
||
| 702 | 'a' => 1, 'b' => 0, 'c' => 0, |
||
| 703 | 'i' => 0, 'j' => 1, 'k' => 0, |
||
| 704 | 'x' => 0, 'y' => 0, 'z' => 1, |
||
| 705 | ]; |
||
| 706 | |||
| 707 | $clipped_font = []; |
||
| 708 | $clipped_position_cm = []; |
||
| 709 | |||
| 710 | self::$recursionStack[] = $this->getUniqueId(); |
||
| 711 | |||
| 712 | foreach ($sections as $section) { |
||
| 713 | $commands = $this->getCommandsText($section); |
||
| 714 | foreach ($commands as $command) { |
||
| 715 | switch ($command[self::OPERATOR]) { |
||
| 716 | // Begin text object |
||
| 717 | case 'BT': |
||
| 718 | // Reset text positioning matrices |
||
| 719 | $current_position_tm = [ |
||
| 720 | 'a' => 1, 'b' => 0, 'c' => 0, |
||
| 721 | 'i' => 0, 'j' => 1, 'k' => 0, |
||
| 722 | 'x' => 0, 'y' => 0, 'z' => 1, |
||
| 723 | ]; |
||
| 724 | $current_position_td = ['x' => 0, 'y' => 0]; |
||
| 725 | $current_text_leading = 0; |
||
| 726 | break; |
||
| 727 | |||
| 728 | // Begin marked content sequence with property list |
||
| 729 | case 'BDC': |
||
| 730 | if (preg_match('/(<<.*>>)$/', $command[self::COMMAND], $match)) { |
||
| 731 | $dict = $this->parseDictionary($match[1]); |
||
| 732 | |||
| 733 | // Check for ActualText block |
||
| 734 | if (isset($dict['ActualText']) && \is_string($dict['ActualText']) && '' !== $dict['ActualText']) { |
||
| 735 | if ('[' == $dict['ActualText'][0]) { |
||
| 736 | // Simulate a 'TJ' command on the stack |
||
| 737 | $marked_stack[] = [ |
||
| 738 | 'ActualText' => $this->getCommandsText($dict['ActualText'].'TJ')[0], |
||
| 739 | ]; |
||
| 740 | } elseif ('<' == $dict['ActualText'][0] || '(' == $dict['ActualText'][0]) { |
||
| 741 | // Simulate a 'Tj' command on the stack |
||
| 742 | $marked_stack[] = [ |
||
| 743 | 'ActualText' => $this->getCommandsText($dict['ActualText'].'Tj')[0], |
||
| 744 | ]; |
||
| 745 | } |
||
| 746 | } |
||
| 747 | } |
||
| 748 | break; |
||
| 749 | |||
| 750 | // Begin marked content sequence |
||
| 751 | case 'BMC': |
||
| 752 | if ('ReversedChars' == $command[self::COMMAND]) { |
||
| 753 | // Upon encountering a ReversedChars command, |
||
| 754 | // add the characters we've built up so far to |
||
| 755 | // the result array |
||
| 756 | $result = array_merge($result, $text); |
||
| 757 | |||
| 758 | // Start a fresh $text array that will contain |
||
| 759 | // reversed characters |
||
| 760 | $text = []; |
||
| 761 | |||
| 762 | // Add the reversed text flag to the stack |
||
| 763 | $marked_stack[] = ['ReversedChars' => true]; |
||
| 764 | } |
||
| 765 | break; |
||
| 766 | |||
| 767 | // set graphics position matrix |
||
| 768 | case 'cm': |
||
| 769 | $args = preg_split('/\s+/s', $command[self::COMMAND]); |
||
| 770 | $current_position_cm = [ |
||
| 771 | 'a' => (float) $args[0], 'b' => (float) $args[1], 'c' => 0, |
||
| 772 | 'i' => (float) $args[2], 'j' => (float) $args[3], 'k' => 0, |
||
| 773 | 'x' => (float) $args[4], 'y' => (float) $args[5], 'z' => 1, |
||
| 774 | ]; |
||
| 775 | break; |
||
| 776 | |||
| 777 | case 'Do': |
||
| 778 | if (is_null($page)) { |
||
| 779 | break; |
||
| 780 | } |
||
| 781 | |||
| 782 | $args = preg_split('/\s/s', $command[self::COMMAND]); |
||
| 783 | $id = trim(array_pop($args), '/ '); |
||
| 784 | $xobject = $page->getXObject($id); |
||
| 785 | |||
| 786 | // Check we got a PDFObject back. |
||
| 787 | if (!$xobject instanceof self) { |
||
| 788 | break; |
||
| 789 | } |
||
| 790 | |||
| 791 | // If the PDFObject is an image, do nothing, as images aren't text. |
||
| 792 | if ($xobject instanceof Image) { |
||
| 793 | break; |
||
| 794 | } |
||
| 795 | |||
| 796 | // Check this is not a circular reference. |
||
| 797 | if (!\in_array($xobject->getUniqueId(), self::$recursionStack, true)) { |
||
| 798 | $text[] = $xobject->getText($page); |
||
| 799 | } |
||
| 800 | break; |
||
| 801 | |||
| 802 | // Marked content point with (DP) & without (MP) property list |
||
| 803 | case 'DP': |
||
| 804 | case 'MP': |
||
| 805 | break; |
||
| 806 | |||
| 807 | // End text object |
||
| 808 | case 'ET': |
||
| 809 | break; |
||
| 810 | |||
| 811 | // Store current selected font and graphics matrix |
||
| 812 | case 'q': |
||
| 813 | $clipped_font[] = [$current_font, $current_font_size]; |
||
| 814 | $clipped_position_cm[] = $current_position_cm; |
||
| 815 | break; |
||
| 816 | |||
| 817 | // Restore previous selected font and graphics matrix |
||
| 818 | case 'Q': |
||
| 819 | list($current_font, $current_font_size) = array_pop($clipped_font); |
||
| 820 | $current_position_cm = array_pop($clipped_position_cm); |
||
| 821 | break; |
||
| 822 | |||
| 823 | // End marked content sequence |
||
| 824 | case 'EMC': |
||
| 825 | $data = false; |
||
| 826 | if (\count($marked_stack)) { |
||
| 827 | $marked = array_pop($marked_stack); |
||
| 828 | $action = key($marked); |
||
| 829 | $data = $marked[$action]; |
||
| 830 | |||
| 831 | switch ($action) { |
||
| 832 | // If we are in ReversedChars mode... |
||
| 833 | case 'ReversedChars': |
||
| 834 | // Reverse the characters we've built up so far |
||
| 835 | foreach ($text as $key => $t) { |
||
| 836 | $text[$key] = implode('', array_reverse( |
||
| 837 | mb_str_split($t, 1, mb_internal_encoding()) |
||
| 838 | )); |
||
| 839 | } |
||
| 840 | |||
| 841 | // Add these characters to the result array |
||
| 842 | $result = array_merge($result, $text); |
||
| 843 | |||
| 844 | // Start a fresh $text array that will contain |
||
| 845 | // non-reversed characters |
||
| 846 | $text = []; |
||
| 847 | break; |
||
| 848 | |||
| 849 | case 'ActualText': |
||
| 850 | // Use the content of the ActualText as a command |
||
| 851 | $command = $data; |
||
| 852 | break; |
||
| 853 | } |
||
| 854 | } |
||
| 855 | |||
| 856 | // If this EMC command has been transformed into a 'Tj' |
||
| 857 | // or 'TJ' command because of being ActualText, then bypass |
||
| 858 | // the break to proceed to the writing section below. |
||
| 859 | if ('Tj' != $command[self::OPERATOR] && 'TJ' != $command[self::OPERATOR]) { |
||
| 860 | break; |
||
| 861 | } |
||
| 862 | |||
| 863 | // no break |
||
| 864 | case "'": |
||
| 865 | case '"': |
||
| 866 | if ("'" == $command[self::OPERATOR] || '"' == $command[self::OPERATOR]) { |
||
| 867 | // Move to next line and write text |
||
| 868 | $current_position['x'] = 0; |
||
| 869 | $current_position_td['x'] = 0; |
||
| 870 | $current_position_td['y'] += $current_text_leading; |
||
| 871 | } |
||
| 872 | // no break |
||
| 873 | case 'Tj': |
||
| 874 | $command[self::COMMAND] = [$command]; |
||
| 875 | // no break |
||
| 876 | case 'TJ': |
||
| 877 | // Check the marked content stack for flags |
||
| 878 | $actual_text = false; |
||
| 879 | $reverse_text = false; |
||
| 880 | foreach ($marked_stack as $marked) { |
||
| 881 | if (isset($marked['ActualText'])) { |
||
| 882 | $actual_text = true; |
||
| 883 | } |
||
| 884 | if (isset($marked['ReversedChars'])) { |
||
| 885 | $reverse_text = true; |
||
| 886 | } |
||
| 887 | } |
||
| 888 | |||
| 889 | // Account for text position ONLY just before we write text |
||
| 890 | if (false === $actual_text && \is_array($last_written_position)) { |
||
| 891 | // If $last_written_position is an array, that |
||
| 892 | // means we have stored text position coordinates |
||
| 893 | // for placing an ActualText |
||
| 894 | $currentX = $last_written_position[0]; |
||
| 895 | $currentY = $last_written_position[1]; |
||
| 896 | $last_written_position = false; |
||
| 897 | } else { |
||
| 898 | $currentX = $current_position_cm['x'] + $current_position_tm['x'] + $current_position_td['x']; |
||
| 899 | $currentY = $current_position_cm['y'] + $current_position_tm['y'] + $current_position_td['y']; |
||
| 900 | } |
||
| 901 | $whiteSpace = ''; |
||
| 902 | |||
| 903 | $factorX = -$current_font_size * $current_position_tm['a'] - $current_font_size * $current_position_tm['i']; |
||
| 904 | $factorY = $current_font_size * $current_position_tm['b'] + $current_font_size * $current_position_tm['j']; |
||
| 905 | |||
| 906 | if (true === $this->addPositionWhitespace && false !== $current_position['x']) { |
||
| 907 | $curY = $currentY - $current_position['y']; |
||
| 908 | if (abs($curY) >= abs($factorY) / 4) { |
||
| 909 | $whiteSpace = "\n"; |
||
| 910 | } else { |
||
| 911 | if (true === $reverse_text) { |
||
| 912 | $curX = $current_position['x'] - $currentX; |
||
| 913 | } else { |
||
| 914 | $curX = $currentX - $current_position['x']; |
||
| 915 | } |
||
| 916 | |||
| 917 | // In abs($factorX * 7) below, the 7 is chosen arbitrarily |
||
| 918 | // as the number of apparent "spaces" in a document we |
||
| 919 | // would need before considering them a "tab". In the |
||
| 920 | // future, we might offer this value to users as a config |
||
| 921 | // option. |
||
| 922 | if ($curX >= abs($factorX * 7)) { |
||
| 923 | $whiteSpace = "\t"; |
||
| 924 | } elseif ($curX >= abs($factorX * 2)) { |
||
| 925 | $whiteSpace = ' '; |
||
| 926 | } |
||
| 927 | } |
||
| 928 | } |
||
| 929 | |||
| 930 | $newtext = $this->getTJUsingFontFallback( |
||
| 931 | $current_font, |
||
| 932 | $command[self::COMMAND], |
||
| 933 | $page, |
||
| 934 | $factorX |
||
| 935 | ); |
||
| 936 | |||
| 937 | // If there is no ActualText pending then write |
||
| 938 | if (false === $actual_text) { |
||
| 939 | $newtext = str_replace(["\r", "\n"], '', $newtext); |
||
| 940 | if (false !== $reverse_text) { |
||
| 941 | // If we are in ReversedChars mode, add the whitespace last |
||
| 942 | $text[] = preg_replace('/ $/', ' ', $newtext.$whiteSpace); |
||
| 943 | } else { |
||
| 944 | // Otherwise add the whitespace first |
||
| 945 | if (' ' === $whiteSpace && isset($text[\count($text) - 1])) { |
||
| 946 | $text[\count($text) - 1] = preg_replace('/ $/', '', $text[\count($text) - 1]); |
||
| 947 | } |
||
| 948 | $text[] = preg_replace('/^[ \t]{2}/', ' ', $whiteSpace.$newtext); |
||
| 949 | } |
||
| 950 | |||
| 951 | // Record the position of this inserted text for comparison |
||
| 952 | // with the next text block. |
||
| 953 | // Provide a 'fudge' factor guess on how wide this text block |
||
| 954 | // is based on the number of characters. This helps limit the |
||
| 955 | // number of tabs inserted, but isn't perfect. |
||
| 956 | $factor = $factorX / 2; |
||
| 957 | $current_position = [ |
||
| 958 | 'x' => $currentX - mb_strlen($newtext) * $factor, |
||
| 959 | 'y' => $currentY, |
||
| 960 | ]; |
||
| 961 | } elseif (false === $last_written_position) { |
||
| 962 | // If there is an ActualText in the pipeline |
||
| 963 | // store the position this undisplayed text |
||
| 964 | // *would* have been written to, so the |
||
| 965 | // ActualText is displayed in the right spot |
||
| 966 | $last_written_position = [$currentX, $currentY]; |
||
| 967 | $current_position['x'] = $currentX; |
||
| 968 | } |
||
| 969 | break; |
||
| 970 | |||
| 971 | // move to start of next line |
||
| 972 | case 'T*': |
||
| 973 | $current_position['x'] = 0; |
||
| 974 | $current_position_td['x'] = 0; |
||
| 975 | $current_position_td['y'] += $current_text_leading; |
||
| 976 | break; |
||
| 977 | |||
| 978 | // set character spacing |
||
| 979 | case 'Tc': |
||
| 980 | break; |
||
| 981 | |||
| 982 | // move text current point and set leading |
||
| 983 | case 'Td': |
||
| 984 | case 'TD': |
||
| 985 | // move text current point |
||
| 986 | $args = preg_split('/\s+/s', $command[self::COMMAND]); |
||
| 987 | $y = (float) array_pop($args); |
||
| 988 | $x = (float) array_pop($args); |
||
| 989 | |||
| 990 | if ('TD' == $command[self::OPERATOR]) { |
||
| 991 | $current_text_leading = -$y * $current_position_tm['b'] - $y * $current_position_tm['j']; |
||
| 992 | } |
||
| 993 | |||
| 994 | $current_position_td = [ |
||
| 995 | 'x' => $current_position_td['x'] + $x * $current_position_tm['a'] + $x * $current_position_tm['i'], |
||
| 996 | 'y' => $current_position_td['y'] + $y * $current_position_tm['b'] + $y * $current_position_tm['j'], |
||
| 997 | ]; |
||
| 998 | break; |
||
| 999 | |||
| 1000 | case 'Tf': |
||
| 1001 | $args = preg_split('/\s/s', $command[self::COMMAND]); |
||
| 1002 | $size = (float) array_pop($args); |
||
| 1003 | $id = trim(array_pop($args), '/'); |
||
| 1004 | if (null !== $page) { |
||
| 1005 | $new_font = $page->getFont($id); |
||
| 1006 | // If an invalid font ID is given, do not update the font. |
||
| 1007 | // This should theoretically never happen, as the PDF spec states for the Tf operator: |
||
| 1008 | // "The specified font value shall match a resource name in the Font entry of the default resource dictionary" |
||
| 1009 | // (https://www.adobe.com/content/dam/acom/en/devnet/pdf/pdfs/PDF32000_2008.pdf, page 435) |
||
| 1010 | // But we want to make sure that malformed PDFs do not simply crash. |
||
| 1011 | if (null !== $new_font) { |
||
| 1012 | $current_font = $new_font; |
||
| 1013 | $current_font_size = $size; |
||
| 1014 | } |
||
| 1015 | } |
||
| 1016 | break; |
||
| 1017 | |||
| 1018 | // set leading |
||
| 1019 | case 'TL': |
||
| 1020 | $y = (float) $command[self::COMMAND]; |
||
| 1021 | $current_text_leading = -$y * $current_position_tm['b'] + -$y * $current_position_tm['j']; |
||
| 1022 | break; |
||
| 1023 | |||
| 1024 | // set text position matrix |
||
| 1025 | case 'Tm': |
||
| 1026 | $args = preg_split('/\s+/s', $command[self::COMMAND]); |
||
| 1027 | $current_position_tm = [ |
||
| 1028 | 'a' => (float) $args[0], 'b' => (float) $args[1], 'c' => 0, |
||
| 1029 | 'i' => (float) $args[2], 'j' => (float) $args[3], 'k' => 0, |
||
| 1030 | 'x' => (float) $args[4], 'y' => (float) $args[5], 'z' => 1, |
||
| 1031 | ]; |
||
| 1032 | break; |
||
| 1033 | |||
| 1034 | // set text rendering mode |
||
| 1035 | case 'Ts': |
||
| 1036 | break; |
||
| 1037 | |||
| 1038 | // set super/subscripting text rise |
||
| 1039 | case 'Ts': |
||
| 1040 | break; |
||
| 1041 | |||
| 1042 | // set word spacing |
||
| 1043 | case 'Tw': |
||
| 1044 | break; |
||
| 1045 | |||
| 1046 | // set horizontal scaling |
||
| 1047 | case 'Tz': |
||
| 1048 | break; |
||
| 1049 | |||
| 1050 | default: |
||
| 1051 | } |
||
| 1052 | } |
||
| 1053 | } |
||
| 1054 | |||
| 1055 | $result = array_merge($result, $text); |
||
| 1056 | |||
| 1057 | return $result; |
||
| 1058 | } |
||
| 1059 | |||
| 1060 | /** |
||
| 1061 | * getCommandsText() expects the content of $text_part to be an |
||
| 1062 | * already formatted, single-line command from a document stream. |
||
| 1063 | * The companion function getSectionsText() returns a document |
||
| 1064 | * stream as an array of single commands for just this purpose. |
||
| 1065 | * Because of this, the argument $offset is no longer used, and |
||
| 1066 | * may be removed in a future PdfParser release. |
||
| 1067 | * |
||
| 1068 | * A better name for this function would be getCommandText() |
||
| 1069 | * since it now always works on just one command. |
||
| 1070 | */ |
||
| 1071 | public function getCommandsText(string $text_part, int &$offset = 0): array |
||
| 1072 | { |
||
| 1073 | $commands = $matches = []; |
||
| 1074 | |||
| 1075 | preg_match('/^(([\/\[\(<])?.*)(?<!\w)([a-z01\'\"*]+)$/i', $text_part, $matches); |
||
| 1076 | |||
| 1077 | // If no valid command is detected, return an empty array |
||
| 1078 | if (!isset($matches[1]) || !isset($matches[2]) || !isset($matches[3])) { |
||
| 1079 | return []; |
||
| 1080 | } |
||
| 1081 | |||
| 1082 | $type = $matches[2]; |
||
| 1083 | $operator = $matches[3]; |
||
| 1084 | $command = trim($matches[1]); |
||
| 1085 | |||
| 1086 | if ('TJ' == $operator) { |
||
| 1087 | $subcommand = []; |
||
| 1088 | $command = trim($command, '[]'); |
||
| 1089 | do { |
||
| 1090 | $oldCommand = $command; |
||
| 1091 | |||
| 1092 | // Search for parentheses string () format |
||
| 1093 | if (preg_match('/^ *\((.*?)(?<![^\\\\]\\\\)\) *(-?[\d.]+)?/', $command, $tjmatch)) { |
||
| 1094 | $subcommand[] = [ |
||
| 1095 | self::TYPE => '(', |
||
| 1096 | self::OPERATOR => 'TJ', |
||
| 1097 | self::COMMAND => $tjmatch[1], |
||
| 1098 | ]; |
||
| 1099 | if (isset($tjmatch[2]) && trim($tjmatch[2])) { |
||
| 1100 | $subcommand[] = [ |
||
| 1101 | self::TYPE => 'n', |
||
| 1102 | self::OPERATOR => '', |
||
| 1103 | self::COMMAND => $tjmatch[2], |
||
| 1104 | ]; |
||
| 1105 | } |
||
| 1106 | $command = substr($command, \strlen($tjmatch[0])); |
||
| 1107 | } |
||
| 1108 | |||
| 1109 | // Search for hexadecimal <> format |
||
| 1110 | if (preg_match('/^ *<([0-9a-f\s]*)> *(-?[\d.]+)?/i', $command, $tjmatch)) { |
||
| 1111 | $tjmatch[1] = preg_replace('/\s/', '', $tjmatch[1]); |
||
| 1112 | $subcommand[] = [ |
||
| 1113 | self::TYPE => '<', |
||
| 1114 | self::OPERATOR => 'TJ', |
||
| 1115 | self::COMMAND => $tjmatch[1], |
||
| 1116 | ]; |
||
| 1117 | if (isset($tjmatch[2]) && trim($tjmatch[2])) { |
||
| 1118 | $subcommand[] = [ |
||
| 1119 | self::TYPE => 'n', |
||
| 1120 | self::OPERATOR => '', |
||
| 1121 | self::COMMAND => $tjmatch[2], |
||
| 1122 | ]; |
||
| 1123 | } |
||
| 1124 | $command = substr($command, \strlen($tjmatch[0])); |
||
| 1125 | } |
||
| 1126 | } while ($command != $oldCommand); |
||
| 1127 | |||
| 1128 | $command = $subcommand; |
||
| 1129 | } elseif ('Tj' == $operator || "'" == $operator || '"' == $operator) { |
||
| 1130 | // Depending on the string type, trim the data of the |
||
| 1131 | // appropriate delimiters |
||
| 1132 | if ('(' == $type) { |
||
| 1133 | // Don't use trim() here since a () string may end with |
||
| 1134 | // a balanced or escaped right parentheses, and trim() |
||
| 1135 | // will delete both. Both strings below are valid: |
||
| 1136 | // eg. (String()) |
||
| 1137 | // eg. (String\)) |
||
| 1138 | $command = preg_replace('/^\(|\)$/', '', $command); |
||
| 1139 | } elseif ('<' == $type) { |
||
| 1140 | $command = trim($command, '<>'); |
||
| 1141 | } |
||
| 1142 | } elseif ('/' == $type) { |
||
| 1143 | $command = substr($command, 1); |
||
| 1144 | } |
||
| 1145 | |||
| 1146 | $commands[] = [ |
||
| 1147 | self::TYPE => $type, |
||
| 1148 | self::OPERATOR => $operator, |
||
| 1149 | self::COMMAND => $command, |
||
| 1150 | ]; |
||
| 1151 | |||
| 1152 | return $commands; |
||
| 1153 | } |
||
| 1154 | |||
| 1155 | public static function factory( |
||
| 1156 | Document $document, |
||
| 1157 | Header $header, |
||
| 1158 | ?string $content, |
||
| 1159 | ?Config $config = null |
||
| 1160 | ): self { |
||
| 1161 | switch ($header->get('Type')->getContent()) { |
||
| 1162 | case 'XObject': |
||
| 1163 | switch ($header->get('Subtype')->getContent()) { |
||
| 1164 | case 'Image': |
||
| 1165 | return new Image($document, $header, $config->getRetainImageContent() ? $content : null, $config); |
||
| 1166 | |||
| 1167 | case 'Form': |
||
| 1168 | return new Form($document, $header, $content, $config); |
||
| 1169 | } |
||
| 1170 | |||
| 1171 | return new self($document, $header, $content, $config); |
||
| 1172 | |||
| 1173 | case 'Pages': |
||
| 1174 | return new Pages($document, $header, $content, $config); |
||
| 1175 | |||
| 1176 | case 'Page': |
||
| 1177 | return new Page($document, $header, $content, $config); |
||
| 1178 | |||
| 1179 | case 'Encoding': |
||
| 1180 | return new Encoding($document, $header, $content, $config); |
||
| 1181 | |||
| 1182 | case 'Font': |
||
| 1183 | $subtype = $header->get('Subtype')->getContent(); |
||
| 1184 | $classname = '\Smalot\PdfParser\Font\Font'.$subtype; |
||
| 1185 | |||
| 1186 | if (class_exists($classname)) { |
||
| 1187 | return new $classname($document, $header, $content, $config); |
||
| 1188 | } |
||
| 1189 | |||
| 1190 | return new Font($document, $header, $content, $config); |
||
| 1191 | |||
| 1192 | default: |
||
| 1193 | return new self($document, $header, $content, $config); |
||
| 1194 | } |
||
| 1195 | } |
||
| 1196 | |||
| 1197 | /** |
||
| 1198 | * Returns unique id identifying the object. |
||
| 1199 | */ |
||
| 1200 | protected function getUniqueId(): string |
||
| 1203 | } |
||
| 1204 | } |
||
| 1205 |
Loading history...
