1<?php 2 3/** 4 * webtrees: online genealogy 5 * Copyright (C) 2022 webtrees development team 6 * This program is free software: you can redistribute it and/or modify 7 * it under the terms of the GNU General Public License as published by 8 * the Free Software Foundation, either version 3 of the License, or 9 * (at your option) any later version. 10 * This program is distributed in the hope that it will be useful, 11 * but WITHOUT ANY WARRANTY; without even the implied warranty of 12 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the 13 * GNU General Public License for more details. 14 * You should have received a copy of the GNU General Public License 15 * along with this program. If not, see <https://www.gnu.org/licenses/>. 16 */ 17 18declare(strict_types=1); 19 20namespace Fisharebest\Webtrees\Services; 21 22use Fisharebest\Webtrees\Date; 23use Fisharebest\Webtrees\Elements\UnknownElement; 24use Fisharebest\Webtrees\Exceptions\GedcomErrorException; 25use Fisharebest\Webtrees\Family; 26use Fisharebest\Webtrees\Gedcom; 27use Fisharebest\Webtrees\Header; 28use Fisharebest\Webtrees\Individual; 29use Fisharebest\Webtrees\Location; 30use Fisharebest\Webtrees\Media; 31use Fisharebest\Webtrees\Note; 32use Fisharebest\Webtrees\Place; 33use Fisharebest\Webtrees\PlaceLocation; 34use Fisharebest\Webtrees\Registry; 35use Fisharebest\Webtrees\Repository; 36use Fisharebest\Webtrees\Soundex; 37use Fisharebest\Webtrees\Source; 38use Fisharebest\Webtrees\Submission; 39use Fisharebest\Webtrees\Submitter; 40use Fisharebest\Webtrees\Tree; 41use Illuminate\Database\Capsule\Manager as DB; 42use Illuminate\Database\Query\JoinClause; 43 44use function app; 45use function array_chunk; 46use function array_intersect_key; 47use function array_map; 48use function array_unique; 49use function assert; 50use function date; 51use function explode; 52use function max; 53use function mb_substr; 54use function preg_match; 55use function preg_match_all; 56use function preg_replace; 57use function round; 58use function str_contains; 59use function str_replace; 60use function str_starts_with; 61use function strlen; 62use function strtolower; 63use function strtoupper; 64use function strtr; 65use function substr; 66use function trim; 67 68use const PREG_SET_ORDER; 69 70/** 71 * Class GedcomImportService - import GEDCOM data 72 */ 73class GedcomImportService 74{ 75 /** 76 * Tidy up a gedcom record on import, so that we can access it consistently/efficiently. 77 * 78 * @param string $rec 79 * @param Tree $tree 80 * 81 * @return string 82 */ 83 private function reformatRecord(string $rec, Tree $tree): string 84 { 85 $gedcom_service = app(GedcomService::class); 86 assert($gedcom_service instanceof GedcomService); 87 88 // Strip out mac/msdos line endings 89 $rec = preg_replace("/[\r\n]+/", "\n", $rec); 90 91 // Extract lines from the record; lines consist of: level + optional xref + tag + optional data 92 $num_matches = preg_match_all('/^[ \t]*(\d+)[ \t]*(@[^@]*@)?[ \t]*(\w+)[ \t]?(.*)$/m', $rec, $matches, PREG_SET_ORDER); 93 94 // Process the record line-by-line 95 $newrec = ''; 96 foreach ($matches as $n => $match) { 97 [, $level, $xref, $tag, $data] = $match; 98 99 $tag = $gedcom_service->canonicalTag($tag); 100 101 switch ($tag) { 102 case 'DATE': 103 // Preserve text from INT dates 104 if (str_contains($data, '(')) { 105 [$date, $text] = explode('(', $data, 2); 106 $text = ' (' . $text; 107 } else { 108 $date = $data; 109 $text = ''; 110 } 111 // Capitals 112 $date = strtoupper($date); 113 // Temporarily add leading/trailing spaces, to allow efficient matching below 114 $date = ' ' . $date . ' '; 115 // Ensure space digits and letters 116 $date = preg_replace('/([A-Z])(\d)/', '$1 $2', $date); 117 $date = preg_replace('/(\d)([A-Z])/', '$1 $2', $date); 118 // Ensure space before/after calendar escapes 119 $date = preg_replace('/@#[^@]+@/', ' $0 ', $date); 120 // "BET." => "BET" 121 $date = preg_replace('/(\w\w)\./', '$1', $date); 122 // "CIR" => "ABT" 123 $date = str_replace(' CIR ', ' ABT ', $date); 124 $date = str_replace(' APX ', ' ABT ', $date); 125 // B.C. => BC (temporarily, to allow easier handling of ".") 126 $date = str_replace(' B.C. ', ' BC ', $date); 127 // TMG uses "EITHER X OR Y" 128 $date = preg_replace('/^ EITHER (.+) OR (.+)/', ' BET $1 AND $2', $date); 129 // "BET X - Y " => "BET X AND Y" 130 $date = preg_replace('/^(.* BET .+) - (.+)/', '$1 AND $2', $date); 131 $date = preg_replace('/^(.* FROM .+) - (.+)/', '$1 TO $2', $date); 132 // "@#ESC@ FROM X TO Y" => "FROM @#ESC@ X TO @#ESC@ Y" 133 $date = preg_replace('/^ +(@#[^@]+@) +FROM +(.+) +TO +(.+)/', ' FROM $1 $2 TO $1 $3', $date); 134 $date = preg_replace('/^ +(@#[^@]+@) +BET +(.+) +AND +(.+)/', ' BET $1 $2 AND $1 $3', $date); 135 // "@#ESC@ AFT X" => "AFT @#ESC@ X" 136 $date = preg_replace('/^ +(@#[^@]+@) +(FROM|BET|TO|AND|BEF|AFT|CAL|EST|INT|ABT) +(.+)/', ' $2 $1 $3', $date); 137 // Ignore any remaining punctuation, e.g. "14-MAY, 1900" => "14 MAY 1900" 138 // (don't change "/" - it is used in NS/OS dates) 139 $date = preg_replace('/[.,:;-]/', ' ', $date); 140 // BC => B.C. 141 $date = str_replace(' BC ', ' B.C. ', $date); 142 // Append the "INT" text 143 $data = $date . $text; 144 break; 145 case 'HEAD': 146 case 'TRLR': 147 // HEAD and TRLR records do not have an XREF or DATA 148 if ($level === '0') { 149 $xref = ''; 150 $data = ''; 151 } 152 break; 153 case 'NAME': 154 // Tidy up non-printing characters 155 $data = preg_replace('/ +/', ' ', trim($data)); 156 break; 157 case 'PLAC': 158 // Consistent commas 159 $data = preg_replace('/ *[,，،] */u', ', ', $data); 160 // The Master Genealogist stores LAT/LONG data in the PLAC field, e.g. Pennsylvania, USA, 395945N0751013W 161 if (preg_match('/(.*), (\d\d)(\d\d)(\d\d)([NS])(\d\d\d)(\d\d)(\d\d)([EW])$/', $data, $match)) { 162 $data = 163 $match[1] . "\n" . 164 ($level + 1) . " MAP\n" . 165 ($level + 2) . ' LATI ' . ($match[5] . round($match[2] + ($match[3] / 60) + ($match[4] / 3600), 4)) . "\n" . 166 ($level + 2) . ' LONG ' . ($match[9] . round($match[6] + ($match[7] / 60) + ($match[8] / 3600), 4)); 167 } 168 break; 169 case 'SEX': 170 $data = strtoupper($data); 171 break; 172 } 173 // Suppress "Y", for facts/events with a DATE or PLAC 174 if ($data === 'y') { 175 $data = 'Y'; 176 } 177 if ($level === '1' && $data === 'Y') { 178 for ($i = $n + 1; $i < $num_matches - 1 && $matches[$i][1] !== '1'; ++$i) { 179 if ($matches[$i][3] === 'DATE' || $matches[$i][3] === 'PLAC') { 180 $data = ''; 181 break; 182 } 183 } 184 } 185 // Reassemble components back into a single line 186 switch ($tag) { 187 default: 188 // Remove tabs and multiple/leading/trailing spaces 189 $data = strtr($data, ["\t" => ' ']); 190 $data = trim($data, ' '); 191 while (str_contains($data, ' ')) { 192 $data = strtr($data, [' ' => ' ']); 193 } 194 $newrec .= ($newrec ? "\n" : '') . $level . ' ' . ($level === '0' && $xref ? $xref . ' ' : '') . $tag . ($data === '' && $tag !== 'NOTE' ? '' : ' ' . $data); 195 break; 196 case 'NOTE': 197 case 'TEXT': 198 case 'DATA': 199 case 'CONT': 200 $newrec .= ($newrec ? "\n" : '') . $level . ' ' . ($level === '0' && $xref ? $xref . ' ' : '') . $tag . ($data === '' && $tag !== 'NOTE' ? '' : ' ' . $data); 201 break; 202 case 'FILE': 203 // Strip off the user-defined path prefix 204 $GEDCOM_MEDIA_PATH = $tree->getPreference('GEDCOM_MEDIA_PATH'); 205 if ($GEDCOM_MEDIA_PATH !== '' && str_starts_with($data, $GEDCOM_MEDIA_PATH)) { 206 $data = substr($data, strlen($GEDCOM_MEDIA_PATH)); 207 } 208 // convert backslashes in filenames to forward slashes 209 $data = preg_replace("/\\\\/", '/', $data); 210 211 $newrec .= ($newrec ? "\n" : '') . $level . ' ' . ($level === '0' && $xref ? $xref . ' ' : '') . $tag . ($data === '' && $tag !== 'NOTE' ? '' : ' ' . $data); 212 break; 213 case 'CONC': 214 // Merge CONC lines, to simplify access later on. 215 $newrec .= ($tree->getPreference('WORD_WRAPPED_NOTES') ? ' ' : '') . $data; 216 break; 217 } 218 } 219 220 return $newrec; 221 } 222 223 /** 224 * import record into database 225 * this function will parse the given gedcom record and add it to the database 226 * 227 * @param string $gedrec the raw gedcom record to parse 228 * @param Tree $tree import the record into this tree 229 * @param bool $update whether this is an updated record that has been accepted 230 * 231 * @return void 232 * @throws GedcomErrorException 233 */ 234 public function importRecord(string $gedrec, Tree $tree, bool $update): void 235 { 236 $tree_id = $tree->id(); 237 238 // Escaped @ signs (only if importing from file) 239 if (!$update) { 240 $gedrec = str_replace('@@', '@', $gedrec); 241 } 242 243 // Standardise gedcom format 244 $gedrec = $this->reformatRecord($gedrec, $tree); 245 246 // import different types of records 247 if (preg_match('/^0 @(' . Gedcom::REGEX_XREF . ')@ (' . Gedcom::REGEX_TAG . ')/', $gedrec, $match)) { 248 [, $xref, $type] = $match; 249 } elseif (str_starts_with($gedrec, '0 HEAD')) { 250 $type = 'HEAD'; 251 $xref = 'HEAD'; // For records without an XREF, use the type as a pseudo XREF. 252 } elseif (str_starts_with($gedrec, '0 TRLR')) { 253 $tree->setPreference('imported', '1'); 254 $type = 'TRLR'; 255 $xref = 'TRLR'; // For records without an XREF, use the type as a pseudo XREF. 256 } elseif (str_starts_with($gedrec, '0 _PLAC_DEFN')) { 257 $this->importLegacyPlacDefn($gedrec); 258 259 return; 260 } elseif (str_starts_with($gedrec, '0 _PLAC ')) { 261 $this->importTNGPlac($gedrec); 262 263 return; 264 } else { 265 foreach (Gedcom::CUSTOM_RECORDS_WITHOUT_XREFS as $record_type) { 266 if (preg_match('/^0 ' . $record_type . '\b/', $gedrec) === 1) { 267 return; 268 } 269 } 270 271 throw new GedcomErrorException($gedrec); 272 } 273 274 // Add a _UID 275 if ($tree->getPreference('GENERATE_UIDS') === '1' && !str_contains($gedrec, "\n1 _UID ")) { 276 $element = Registry::elementFactory()->make($type . ':_UID'); 277 if (!$element instanceof UnknownElement) { 278 $gedrec .= "\n1 _UID " . $element->default($tree); 279 } 280 } 281 282 // If the user has downloaded their GEDCOM data (containing media objects) and edited it 283 // using an application which does not support (and deletes) media objects, then add them 284 // back in. 285 if ($tree->getPreference('keep_media')) { 286 $old_linked_media = DB::table('link') 287 ->where('l_from', '=', $xref) 288 ->where('l_file', '=', $tree_id) 289 ->where('l_type', '=', 'OBJE') 290 ->pluck('l_to'); 291 292 // Delete these links - so that we do not insert them again in updateLinks() 293 DB::table('link') 294 ->where('l_from', '=', $xref) 295 ->where('l_file', '=', $tree_id) 296 ->where('l_type', '=', 'OBJE') 297 ->delete(); 298 299 foreach ($old_linked_media as $media_id) { 300 $gedrec .= "\n1 OBJE @" . $media_id . '@'; 301 } 302 } 303 304 // Convert inline media into media objects 305 $gedrec = $this->convertInlineMedia($tree, $gedrec); 306 307 switch ($type) { 308 case Individual::RECORD_TYPE: 309 $record = Registry::individualFactory()->new($xref, $gedrec, null, $tree); 310 311 if (preg_match('/\n1 RIN (.+)/', $gedrec, $match)) { 312 $rin = $match[1]; 313 } else { 314 $rin = $xref; 315 } 316 317 // The database can only store MFU, and many of the stats queries assume this. 318 $sex = $record->sex(); 319 $sex = $sex === 'M' || $sex === 'F' ? $sex : 'U'; 320 321 DB::table('individuals')->insert([ 322 'i_id' => $xref, 323 'i_file' => $tree_id, 324 'i_rin' => $rin, 325 'i_sex' => $sex, 326 'i_gedcom' => $gedrec, 327 ]); 328 329 // Update the cross-reference/index tables. 330 $this->updatePlaces($xref, $tree, $gedrec); 331 $this->updateDates($xref, $tree_id, $gedrec); 332 $this->updateNames($xref, $tree_id, $record); 333 break; 334 335 case Family::RECORD_TYPE: 336 if (preg_match('/\n1 HUSB @(' . Gedcom::REGEX_XREF . ')@/', $gedrec, $match)) { 337 $husb = $match[1]; 338 } else { 339 $husb = ''; 340 } 341 if (preg_match('/\n1 WIFE @(' . Gedcom::REGEX_XREF . ')@/', $gedrec, $match)) { 342 $wife = $match[1]; 343 } else { 344 $wife = ''; 345 } 346 $nchi = preg_match_all('/\n1 CHIL @(' . Gedcom::REGEX_XREF . ')@/', $gedrec, $match); 347 if (preg_match('/\n1 NCHI (\d+)/', $gedrec, $match)) { 348 $nchi = max($nchi, $match[1]); 349 } 350 351 DB::table('families')->insert([ 352 'f_id' => $xref, 353 'f_file' => $tree_id, 354 'f_husb' => $husb, 355 'f_wife' => $wife, 356 'f_gedcom' => $gedrec, 357 'f_numchil' => $nchi, 358 ]); 359 360 // Update the cross-reference/index tables. 361 $this->updatePlaces($xref, $tree, $gedrec); 362 $this->updateDates($xref, $tree_id, $gedrec); 363 break; 364 365 case Source::RECORD_TYPE: 366 if (preg_match('/\n1 TITL (.+)/', $gedrec, $match)) { 367 $name = $match[1]; 368 } elseif (preg_match('/\n1 ABBR (.+)/', $gedrec, $match)) { 369 $name = $match[1]; 370 } else { 371 $name = $xref; 372 } 373 374 DB::table('sources')->insert([ 375 's_id' => $xref, 376 's_file' => $tree_id, 377 's_name' => mb_substr($name, 0, 255), 378 's_gedcom' => $gedrec, 379 ]); 380 break; 381 382 case Repository::RECORD_TYPE: 383 case Note::RECORD_TYPE: 384 case Submission::RECORD_TYPE: 385 case Submitter::RECORD_TYPE: 386 case Location::RECORD_TYPE: 387 DB::table('other')->insert([ 388 'o_id' => $xref, 389 'o_file' => $tree_id, 390 'o_type' => $type, 391 'o_gedcom' => $gedrec, 392 ]); 393 break; 394 395 case Header::RECORD_TYPE: 396 // Force HEAD records to have a creation date. 397 if (!str_contains($gedrec, "\n1 DATE ")) { 398 $today = strtoupper(date('d M Y')); 399 $gedrec .= "\n1 DATE " . $today; 400 } 401 402 DB::table('other')->insert([ 403 'o_id' => $xref, 404 'o_file' => $tree_id, 405 'o_type' => Header::RECORD_TYPE, 406 'o_gedcom' => $gedrec, 407 ]); 408 break; 409 410 411 case Media::RECORD_TYPE: 412 $record = Registry::mediaFactory()->new($xref, $gedrec, null, $tree); 413 414 DB::table('media')->insert([ 415 'm_id' => $xref, 416 'm_file' => $tree_id, 417 'm_gedcom' => $gedrec, 418 ]); 419 420 foreach ($record->mediaFiles() as $media_file) { 421 DB::table('media_file')->insert([ 422 'm_id' => $xref, 423 'm_file' => $tree_id, 424 'multimedia_file_refn' => mb_substr($media_file->filename(), 0, 248), 425 'multimedia_format' => mb_substr($media_file->format(), 0, 4), 426 'source_media_type' => mb_substr($media_file->type(), 0, 15), 427 'descriptive_title' => mb_substr($media_file->title(), 0, 248), 428 ]); 429 } 430 break; 431 432 default: // Custom record types. 433 DB::table('other')->insert([ 434 'o_id' => $xref, 435 'o_file' => $tree_id, 436 'o_type' => mb_substr($type, 0, 15), 437 'o_gedcom' => $gedrec, 438 ]); 439 break; 440 } 441 442 // Update the cross-reference/index tables. 443 $this->updateLinks($xref, $tree_id, $gedrec); 444 } 445 446 /** 447 * Legacy Family Tree software generates _PLAC_DEFN records containing LAT/LONG values 448 * 449 * @param string $gedcom 450 */ 451 private function importLegacyPlacDefn(string $gedcom): void 452 { 453 $gedcom_service = new GedcomService(); 454 455 if (preg_match('/\n1 PLAC (.+)/', $gedcom, $match)) { 456 $place_name = $match[1]; 457 } else { 458 return; 459 } 460 461 if (preg_match('/\n3 LATI ([NS].+)/', $gedcom, $match)) { 462 $latitude = $gedcom_service->readLatitude($match[1]); 463 } else { 464 return; 465 } 466 467 if (preg_match('/\n3 LONG ([EW].+)/', $gedcom, $match)) { 468 $longitude = $gedcom_service->readLongitude($match[1]); 469 } else { 470 return; 471 } 472 473 $location = new PlaceLocation($place_name); 474 475 if ($location->latitude() === null && $location->longitude() === null) { 476 DB::table('place_location') 477 ->where('id', '=', $location->id()) 478 ->update([ 479 'latitude' => $latitude, 480 'longitude' => $longitude, 481 ]); 482 } 483 } 484 485 /** 486 * Legacy Family Tree software generates _PLAC records containing LAT/LONG values 487 * 488 * @param string $gedcom 489 */ 490 private function importTNGPlac(string $gedcom): void 491 { 492 if (preg_match('/^0 _PLAC (.+)/', $gedcom, $match)) { 493 $place_name = $match[1]; 494 } else { 495 return; 496 } 497 498 if (preg_match('/\n2 LATI (.+)/', $gedcom, $match)) { 499 $latitude = (float) $match[1]; 500 } else { 501 return; 502 } 503 504 if (preg_match('/\n2 LONG (.+)/', $gedcom, $match)) { 505 $longitude = (float) $match[1]; 506 } else { 507 return; 508 } 509 510 $location = new PlaceLocation($place_name); 511 512 if ($location->latitude() === null && $location->longitude() === null) { 513 DB::table('place_location') 514 ->where('id', '=', $location->id()) 515 ->update([ 516 'latitude' => $latitude, 517 'longitude' => $longitude, 518 ]); 519 } 520 } 521 522 /** 523 * Extract all level 2 places from the given record and insert them into the places table 524 * 525 * @param string $xref 526 * @param Tree $tree 527 * @param string $gedrec 528 * 529 * @return void 530 */ 531 public function updatePlaces(string $xref, Tree $tree, string $gedrec): void 532 { 533 // Insert all new rows together 534 $rows = []; 535 536 preg_match_all('/\n2 PLAC (.+)/', $gedrec, $matches); 537 538 $places = array_unique($matches[1]); 539 540 foreach ($places as $place_name) { 541 $place = new Place($place_name, $tree); 542 543 // Calling Place::id() will create the entry in the database, if it doesn't already exist. 544 while ($place->id() !== 0) { 545 $rows[] = [ 546 'pl_p_id' => $place->id(), 547 'pl_gid' => $xref, 548 'pl_file' => $tree->id(), 549 ]; 550 551 $place = $place->parent(); 552 } 553 } 554 555 // array_unique doesn't work with arrays of arrays 556 $rows = array_intersect_key($rows, array_unique(array_map('serialize', $rows))); 557 558 // PDO has a limit of 65535 placeholders, and each row requires 3 placeholders. 559 foreach (array_chunk($rows, 20000) as $chunk) { 560 DB::table('placelinks')->insert($chunk); 561 } 562 } 563 564 /** 565 * Extract all the dates from the given record and insert them into the database. 566 * 567 * @param string $xref 568 * @param int $ged_id 569 * @param string $gedrec 570 * 571 * @return void 572 */ 573 private function updateDates(string $xref, int $ged_id, string $gedrec): void 574 { 575 // Insert all new rows together 576 $rows = []; 577 578 preg_match_all("/\n1 (\w+).*(?:\n[2-9].*)*\n2 DATE (.+)(?:\n[2-9].*)*/", $gedrec, $matches, PREG_SET_ORDER); 579 580 foreach ($matches as $match) { 581 $fact = $match[1]; 582 $date = new Date($match[2]); 583 $rows[] = [ 584 'd_day' => $date->minimumDate()->day, 585 'd_month' => $date->minimumDate()->format('%O'), 586 'd_mon' => $date->minimumDate()->month, 587 'd_year' => $date->minimumDate()->year, 588 'd_julianday1' => $date->minimumDate()->minimumJulianDay(), 589 'd_julianday2' => $date->minimumDate()->maximumJulianDay(), 590 'd_fact' => $fact, 591 'd_gid' => $xref, 592 'd_file' => $ged_id, 593 'd_type' => $date->minimumDate()->format('%@'), 594 ]; 595 596 $rows[] = [ 597 'd_day' => $date->maximumDate()->day, 598 'd_month' => $date->maximumDate()->format('%O'), 599 'd_mon' => $date->maximumDate()->month, 600 'd_year' => $date->maximumDate()->year, 601 'd_julianday1' => $date->maximumDate()->minimumJulianDay(), 602 'd_julianday2' => $date->maximumDate()->maximumJulianDay(), 603 'd_fact' => $fact, 604 'd_gid' => $xref, 605 'd_file' => $ged_id, 606 'd_type' => $date->minimumDate()->format('%@'), 607 ]; 608 } 609 610 // array_unique doesn't work with arrays of arrays 611 $rows = array_intersect_key($rows, array_unique(array_map('serialize', $rows))); 612 613 DB::table('dates')->insert($rows); 614 } 615 616 /** 617 * Extract all the links from the given record and insert them into the database 618 * 619 * @param string $xref 620 * @param int $ged_id 621 * @param string $gedrec 622 * 623 * @return void 624 */ 625 private function updateLinks(string $xref, int $ged_id, string $gedrec): void 626 { 627 // Insert all new rows together 628 $rows = []; 629 630 preg_match_all('/\n\d+ (' . Gedcom::REGEX_TAG . ') @(' . Gedcom::REGEX_XREF . ')@/', $gedrec, $matches, PREG_SET_ORDER); 631 632 foreach ($matches as $match) { 633 // Some applications (e.g. GenoPro) create links longer than 15 characters. 634 $link = mb_substr($match[1], 0, 15); 635 636 // Take care of "duplicates" that differ on case/collation, e.g. "SOUR @S1@" and "SOUR @s1@" 637 $rows[$link . strtoupper($match[2])] = [ 638 'l_from' => $xref, 639 'l_to' => $match[2], 640 'l_type' => $link, 641 'l_file' => $ged_id, 642 ]; 643 } 644 645 DB::table('link')->insert($rows); 646 } 647 648 /** 649 * Extract all the names from the given record and insert them into the database. 650 * 651 * @param string $xref 652 * @param int $ged_id 653 * @param Individual $record 654 * 655 * @return void 656 */ 657 private function updateNames(string $xref, int $ged_id, Individual $record): void 658 { 659 // Insert all new rows together 660 $rows = []; 661 662 foreach ($record->getAllNames() as $n => $name) { 663 if ($name['givn'] === Individual::PRAENOMEN_NESCIO) { 664 $soundex_givn_std = null; 665 $soundex_givn_dm = null; 666 } else { 667 $soundex_givn_std = Soundex::russell($name['givn']); 668 $soundex_givn_dm = Soundex::daitchMokotoff($name['givn']); 669 } 670 671 if ($name['surn'] === Individual::NOMEN_NESCIO) { 672 $soundex_surn_std = null; 673 $soundex_surn_dm = null; 674 } else { 675 $soundex_surn_std = Soundex::russell($name['surname']); 676 $soundex_surn_dm = Soundex::daitchMokotoff($name['surname']); 677 } 678 679 $rows[] = [ 680 'n_file' => $ged_id, 681 'n_id' => $xref, 682 'n_num' => $n, 683 'n_type' => $name['type'], 684 'n_sort' => mb_substr($name['sort'], 0, 255), 685 'n_full' => mb_substr($name['fullNN'], 0, 255), 686 'n_surname' => mb_substr($name['surname'], 0, 255), 687 'n_surn' => mb_substr($name['surn'], 0, 255), 688 'n_givn' => mb_substr($name['givn'], 0, 255), 689 'n_soundex_givn_std' => $soundex_givn_std, 690 'n_soundex_surn_std' => $soundex_surn_std, 691 'n_soundex_givn_dm' => $soundex_givn_dm, 692 'n_soundex_surn_dm' => $soundex_surn_dm, 693 ]; 694 } 695 696 DB::table('name')->insert($rows); 697 } 698 699 /** 700 * Extract inline media data, and convert to media objects. 701 * 702 * @param Tree $tree 703 * @param string $gedcom 704 * 705 * @return string 706 */ 707 private function convertInlineMedia(Tree $tree, string $gedcom): string 708 { 709 while (preg_match('/\n1 OBJE(?:\n[2-9].+)+/', $gedcom, $match)) { 710 $xref = $this->createMediaObject($match[0], $tree); 711 $gedcom = strtr($gedcom, [$match[0] => "\n1 OBJE @" . $xref . '@']); 712 } 713 while (preg_match('/\n2 OBJE(?:\n[3-9].+)+/', $gedcom, $match)) { 714 $xref = $this->createMediaObject($match[0], $tree); 715 $gedcom = strtr($gedcom, [$match[0] => "\n2 OBJE @" . $xref . '@']); 716 } 717 while (preg_match('/\n3 OBJE(?:\n[4-9].+)+/', $gedcom, $match)) { 718 $xref = $this->createMediaObject($match[0], $tree); 719 $gedcom = strtr($gedcom, [$match[0] => "\n3 OBJE @" . $xref . '@']); 720 } 721 722 return $gedcom; 723 } 724 725 /** 726 * Create a new media object, from inline media data. 727 * 728 * GEDCOM 5.5.1 specifies: +1 FILE / +2 FORM / +3 MEDI / +1 TITL 729 * GEDCOM 5.5 specifies: +1 FILE / +1 FORM / +1 TITL 730 * GEDCOM 5.5.1 says that GEDCOM 5.5 specifies: +1 FILE / +1 FORM / +2 MEDI 731 * 732 * Legacy generates: +1 FORM / +1 FILE / +1 TITL / +1 _SCBK / +1 _PRIM / +1 _TYPE / +1 NOTE 733 * RootsMagic generates: +1 FILE / +1 FORM / +1 TITL 734 * 735 * @param string $gedcom 736 * @param Tree $tree 737 * 738 * @return string 739 */ 740 private function createMediaObject(string $gedcom, Tree $tree): string 741 { 742 preg_match('/\n\d FILE (.+)/', $gedcom, $match); 743 $file = $match[1] ?? ''; 744 745 preg_match('/\n\d TITL (.+)/', $gedcom, $match); 746 $title = $match[1] ?? ''; 747 748 preg_match('/\n\d FORM (.+)/', $gedcom, $match); 749 $format = $match[1] ?? ''; 750 751 preg_match('/\n\d MEDI (.+)/', $gedcom, $match); 752 $media = $match[1] ?? ''; 753 754 preg_match('/\n\d _SCBK (.+)/', $gedcom, $match); 755 $scrapbook = $match[1] ?? ''; 756 757 preg_match('/\n\d _PRIM (.+)/', $gedcom, $match); 758 $primary = $match[1] ?? ''; 759 760 preg_match('/\n\d _TYPE (.+)/', $gedcom, $match); 761 if ($media === '') { 762 // Legacy uses _TYPE instead of MEDI 763 $media = $match[1] ?? ''; 764 $type = ''; 765 } else { 766 $type = $match[1] ?? ''; 767 } 768 769 preg_match_all('/\n\d NOTE (.+(?:\n\d CONT.*)*)/', $gedcom, $matches); 770 $notes = $matches[1] ?? []; 771 772 // Have we already created a media object with the same title/filename? 773 $xref = DB::table('media_file') 774 ->where('m_file', '=', $tree->id()) 775 ->where('descriptive_title', '=', mb_substr($title, 0, 248)) 776 ->where('multimedia_file_refn', '=', mb_substr($file, 0, 248)) 777 ->value('m_id'); 778 779 if ($xref === null) { 780 $xref = Registry::xrefFactory()->make(Media::RECORD_TYPE); 781 782 // convert to a media-object 783 $gedcom = '0 @' . $xref . "@ OBJE\n1 FILE " . $file; 784 785 if ($format !== '') { 786 $gedcom .= "\n2 FORM " . $format; 787 788 if ($media !== '') { 789 $gedcom .= "\n3 TYPE " . $media; 790 } 791 } 792 793 if ($title !== '') { 794 $gedcom .= "\n3 TITL " . $title; 795 } 796 797 if ($scrapbook !== '') { 798 $gedcom .= "\n1 _SCBK " . $scrapbook; 799 } 800 801 if ($primary !== '') { 802 $gedcom .= "\n1 _PRIM " . $primary; 803 } 804 805 if ($type !== '') { 806 $gedcom .= "\n1 _TYPE " . $type; 807 } 808 809 foreach ($notes as $note) { 810 $gedcom .= "\n1 NOTE " . strtr($note, ["\n3" => "\n2", "\n4" => "\n2", "\n5" => "\n2"]); 811 } 812 813 DB::table('media')->insert([ 814 'm_id' => $xref, 815 'm_file' => $tree->id(), 816 'm_gedcom' => $gedcom, 817 ]); 818 819 DB::table('media_file')->insert([ 820 'm_id' => $xref, 821 'm_file' => $tree->id(), 822 'multimedia_file_refn' => mb_substr($file, 0, 248), 823 'multimedia_format' => mb_substr($format, 0, 4), 824 'source_media_type' => mb_substr($media, 0, 15), 825 'descriptive_title' => mb_substr($title, 0, 248), 826 ]); 827 } 828 829 return $xref; 830 } 831 832 /** 833 * update a record in the database 834 * 835 * @param string $gedrec 836 * @param Tree $tree 837 * @param bool $delete 838 * 839 * @return void 840 * @throws GedcomErrorException 841 */ 842 public function updateRecord(string $gedrec, Tree $tree, bool $delete): void 843 { 844 if (preg_match('/^0 @(' . Gedcom::REGEX_XREF . ')@ (' . Gedcom::REGEX_TAG . ')/', $gedrec, $match)) { 845 [, $gid, $type] = $match; 846 } elseif (preg_match('/^0 (HEAD)(?:\n|$)/', $gedrec, $match)) { 847 // The HEAD record has no XREF. Any others? 848 $gid = $match[1]; 849 $type = $match[1]; 850 } else { 851 throw new GedcomErrorException($gedrec); 852 } 853 854 // Place links 855 DB::table('placelinks') 856 ->where('pl_gid', '=', $gid) 857 ->where('pl_file', '=', $tree->id()) 858 ->delete(); 859 860 // Orphaned places. If we're deleting "Westminster, London, England", 861 // then we may also need to delete "London, England" and "England". 862 do { 863 $affected = DB::table('places') 864 ->leftJoin('placelinks', function (JoinClause $join): void { 865 $join 866 ->on('p_id', '=', 'pl_p_id') 867 ->on('p_file', '=', 'pl_file'); 868 }) 869 ->whereNull('pl_p_id') 870 ->delete(); 871 } while ($affected > 0); 872 873 DB::table('dates') 874 ->where('d_gid', '=', $gid) 875 ->where('d_file', '=', $tree->id()) 876 ->delete(); 877 878 DB::table('name') 879 ->where('n_id', '=', $gid) 880 ->where('n_file', '=', $tree->id()) 881 ->delete(); 882 883 DB::table('link') 884 ->where('l_from', '=', $gid) 885 ->where('l_file', '=', $tree->id()) 886 ->delete(); 887 888 switch ($type) { 889 case Individual::RECORD_TYPE: 890 DB::table('individuals') 891 ->where('i_id', '=', $gid) 892 ->where('i_file', '=', $tree->id()) 893 ->delete(); 894 break; 895 896 case Family::RECORD_TYPE: 897 DB::table('families') 898 ->where('f_id', '=', $gid) 899 ->where('f_file', '=', $tree->id()) 900 ->delete(); 901 break; 902 903 case Source::RECORD_TYPE: 904 DB::table('sources') 905 ->where('s_id', '=', $gid) 906 ->where('s_file', '=', $tree->id()) 907 ->delete(); 908 break; 909 910 case Media::RECORD_TYPE: 911 DB::table('media_file') 912 ->where('m_id', '=', $gid) 913 ->where('m_file', '=', $tree->id()) 914 ->delete(); 915 916 DB::table('media') 917 ->where('m_id', '=', $gid) 918 ->where('m_file', '=', $tree->id()) 919 ->delete(); 920 break; 921 922 default: 923 DB::table('other') 924 ->where('o_id', '=', $gid) 925 ->where('o_file', '=', $tree->id()) 926 ->delete(); 927 break; 928 } 929 930 if (!$delete) { 931 $this->importRecord($gedrec, $tree, true); 932 } 933 } 934} 935