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 'FORM': 146 // Consistent commas 147 $data = preg_replace('/ *, */', ', ', $data); 148 break; 149 case 'HEAD': 150 // HEAD records don't have an XREF or DATA 151 if ($level === '0') { 152 $xref = ''; 153 $data = ''; 154 } 155 break; 156 case 'NAME': 157 // Tidy up non-printing characters 158 $data = preg_replace('/ +/', ' ', trim($data)); 159 break; 160 case 'PEDI': 161 // PEDI values are lower case 162 $data = strtolower($data); 163 break; 164 case 'PLAC': 165 // Consistent commas 166 $data = preg_replace('/ *[,，،] */u', ', ', $data); 167 // The Master Genealogist stores LAT/LONG data in the PLAC field, e.g. Pennsylvania, USA, 395945N0751013W 168 if (preg_match('/(.*), (\d\d)(\d\d)(\d\d)([NS])(\d\d\d)(\d\d)(\d\d)([EW])$/', $data, $match)) { 169 $data = 170 $match[1] . "\n" . 171 ($level + 1) . " MAP\n" . 172 ($level + 2) . ' LATI ' . ($match[5] . round($match[2] + ($match[3] / 60) + ($match[4] / 3600), 4)) . "\n" . 173 ($level + 2) . ' LONG ' . ($match[9] . round($match[6] + ($match[7] / 60) + ($match[8] / 3600), 4)); 174 } 175 break; 176 case 'RESN': 177 // RESN values are lower case (confidential, privacy, locked, none) 178 $data = strtolower($data); 179 if ($data === 'invisible') { 180 $data = 'confidential'; // From old versions of Legacy. 181 } 182 break; 183 case 'SEX': 184 $data = strtoupper($data); 185 break; 186 case 'TRLR': 187 // TRLR records don't have an XREF or DATA 188 if ($level === '0') { 189 $xref = ''; 190 $data = ''; 191 } 192 break; 193 } 194 // Suppress "Y", for facts/events with a DATE or PLAC 195 if ($data === 'y') { 196 $data = 'Y'; 197 } 198 if ($level === '1' && $data === 'Y') { 199 for ($i = $n + 1; $i < $num_matches - 1 && $matches[$i][1] !== '1'; ++$i) { 200 if ($matches[$i][3] === 'DATE' || $matches[$i][3] === 'PLAC') { 201 $data = ''; 202 break; 203 } 204 } 205 } 206 // Reassemble components back into a single line 207 switch ($tag) { 208 default: 209 // Remove tabs and multiple/leading/trailing spaces 210 $data = strtr($data, ["\t" => ' ']); 211 $data = trim($data, ' '); 212 while (str_contains($data, ' ')) { 213 $data = strtr($data, [' ' => ' ']); 214 } 215 $newrec .= ($newrec ? "\n" : '') . $level . ' ' . ($level === '0' && $xref ? $xref . ' ' : '') . $tag . ($data === '' && $tag !== 'NOTE' ? '' : ' ' . $data); 216 break; 217 case 'NOTE': 218 case 'TEXT': 219 case 'DATA': 220 case 'CONT': 221 $newrec .= ($newrec ? "\n" : '') . $level . ' ' . ($level === '0' && $xref ? $xref . ' ' : '') . $tag . ($data === '' && $tag !== 'NOTE' ? '' : ' ' . $data); 222 break; 223 case 'FILE': 224 // Strip off the user-defined path prefix 225 $GEDCOM_MEDIA_PATH = $tree->getPreference('GEDCOM_MEDIA_PATH'); 226 if ($GEDCOM_MEDIA_PATH !== '' && str_starts_with($data, $GEDCOM_MEDIA_PATH)) { 227 $data = substr($data, strlen($GEDCOM_MEDIA_PATH)); 228 } 229 // convert backslashes in filenames to forward slashes 230 $data = preg_replace("/\\\\/", '/', $data); 231 232 $newrec .= ($newrec ? "\n" : '') . $level . ' ' . ($level === '0' && $xref ? $xref . ' ' : '') . $tag . ($data === '' && $tag !== 'NOTE' ? '' : ' ' . $data); 233 break; 234 case 'CONC': 235 // Merge CONC lines, to simplify access later on. 236 $newrec .= ($tree->getPreference('WORD_WRAPPED_NOTES') ? ' ' : '') . $data; 237 break; 238 } 239 } 240 241 return $newrec; 242 } 243 244 /** 245 * import record into database 246 * this function will parse the given gedcom record and add it to the database 247 * 248 * @param string $gedrec the raw gedcom record to parse 249 * @param Tree $tree import the record into this tree 250 * @param bool $update whether this is an updated record that has been accepted 251 * 252 * @return void 253 * @throws GedcomErrorException 254 */ 255 public function importRecord(string $gedrec, Tree $tree, bool $update): void 256 { 257 $tree_id = $tree->id(); 258 259 // Escaped @ signs (only if importing from file) 260 if (!$update) { 261 $gedrec = str_replace('@@', '@', $gedrec); 262 } 263 264 // Standardise gedcom format 265 $gedrec = $this->reformatRecord($gedrec, $tree); 266 267 // import different types of records 268 if (preg_match('/^0 @(' . Gedcom::REGEX_XREF . ')@ (' . Gedcom::REGEX_TAG . ')/', $gedrec, $match)) { 269 [, $xref, $type] = $match; 270 } elseif (str_starts_with($gedrec, '0 HEAD')) { 271 $type = 'HEAD'; 272 $xref = 'HEAD'; // For records without an XREF, use the type as a pseudo XREF. 273 } elseif (str_starts_with($gedrec, '0 TRLR')) { 274 $tree->setPreference('imported', '1'); 275 $type = 'TRLR'; 276 $xref = 'TRLR'; // For records without an XREF, use the type as a pseudo XREF. 277 } elseif (str_starts_with($gedrec, '0 _PLAC_DEFN')) { 278 $this->importLegacyPlacDefn($gedrec); 279 280 return; 281 } elseif (str_starts_with($gedrec, '0 _PLAC ')) { 282 $this->importTNGPlac($gedrec); 283 284 return; 285 } else { 286 foreach (Gedcom::CUSTOM_RECORDS_WITHOUT_XREFS as $record_type) { 287 if (preg_match('/^0 ' . $record_type . '\b/', $gedrec) === 1) { 288 return; 289 } 290 } 291 292 throw new GedcomErrorException($gedrec); 293 } 294 295 // Add a _UID 296 if ($tree->getPreference('GENERATE_UIDS') === '1' && !str_contains($gedrec, "\n1 _UID ")) { 297 $element = Registry::elementFactory()->make($type . ':_UID'); 298 if (!$element instanceof UnknownElement) { 299 $gedrec .= "\n1 _UID " . $element->default($tree); 300 } 301 } 302 303 // If the user has downloaded their GEDCOM data (containing media objects) and edited it 304 // using an application which does not support (and deletes) media objects, then add them 305 // back in. 306 if ($tree->getPreference('keep_media')) { 307 $old_linked_media = DB::table('link') 308 ->where('l_from', '=', $xref) 309 ->where('l_file', '=', $tree_id) 310 ->where('l_type', '=', 'OBJE') 311 ->pluck('l_to'); 312 313 // Delete these links - so that we do not insert them again in updateLinks() 314 DB::table('link') 315 ->where('l_from', '=', $xref) 316 ->where('l_file', '=', $tree_id) 317 ->where('l_type', '=', 'OBJE') 318 ->delete(); 319 320 foreach ($old_linked_media as $media_id) { 321 $gedrec .= "\n1 OBJE @" . $media_id . '@'; 322 } 323 } 324 325 // Convert inline media into media objects 326 $gedrec = $this->convertInlineMedia($tree, $gedrec); 327 328 switch ($type) { 329 case Individual::RECORD_TYPE: 330 $record = Registry::individualFactory()->new($xref, $gedrec, null, $tree); 331 332 if (preg_match('/\n1 RIN (.+)/', $gedrec, $match)) { 333 $rin = $match[1]; 334 } else { 335 $rin = $xref; 336 } 337 338 // The database can only store MFU, and many of the stats queries assume this. 339 $sex = $record->sex(); 340 $sex = $sex === 'M' || $sex === 'F' ? $sex : 'U'; 341 342 DB::table('individuals')->insert([ 343 'i_id' => $xref, 344 'i_file' => $tree_id, 345 'i_rin' => $rin, 346 'i_sex' => $sex, 347 'i_gedcom' => $gedrec, 348 ]); 349 350 // Update the cross-reference/index tables. 351 $this->updatePlaces($xref, $tree, $gedrec); 352 $this->updateDates($xref, $tree_id, $gedrec); 353 $this->updateNames($xref, $tree_id, $record); 354 break; 355 356 case Family::RECORD_TYPE: 357 if (preg_match('/\n1 HUSB @(' . Gedcom::REGEX_XREF . ')@/', $gedrec, $match)) { 358 $husb = $match[1]; 359 } else { 360 $husb = ''; 361 } 362 if (preg_match('/\n1 WIFE @(' . Gedcom::REGEX_XREF . ')@/', $gedrec, $match)) { 363 $wife = $match[1]; 364 } else { 365 $wife = ''; 366 } 367 $nchi = preg_match_all('/\n1 CHIL @(' . Gedcom::REGEX_XREF . ')@/', $gedrec, $match); 368 if (preg_match('/\n1 NCHI (\d+)/', $gedrec, $match)) { 369 $nchi = max($nchi, $match[1]); 370 } 371 372 DB::table('families')->insert([ 373 'f_id' => $xref, 374 'f_file' => $tree_id, 375 'f_husb' => $husb, 376 'f_wife' => $wife, 377 'f_gedcom' => $gedrec, 378 'f_numchil' => $nchi, 379 ]); 380 381 // Update the cross-reference/index tables. 382 $this->updatePlaces($xref, $tree, $gedrec); 383 $this->updateDates($xref, $tree_id, $gedrec); 384 break; 385 386 case Source::RECORD_TYPE: 387 if (preg_match('/\n1 TITL (.+)/', $gedrec, $match)) { 388 $name = $match[1]; 389 } elseif (preg_match('/\n1 ABBR (.+)/', $gedrec, $match)) { 390 $name = $match[1]; 391 } else { 392 $name = $xref; 393 } 394 395 DB::table('sources')->insert([ 396 's_id' => $xref, 397 's_file' => $tree_id, 398 's_name' => mb_substr($name, 0, 255), 399 's_gedcom' => $gedrec, 400 ]); 401 break; 402 403 case Repository::RECORD_TYPE: 404 case Note::RECORD_TYPE: 405 case Submission::RECORD_TYPE: 406 case Submitter::RECORD_TYPE: 407 case Location::RECORD_TYPE: 408 DB::table('other')->insert([ 409 'o_id' => $xref, 410 'o_file' => $tree_id, 411 'o_type' => $type, 412 'o_gedcom' => $gedrec, 413 ]); 414 break; 415 416 case Header::RECORD_TYPE: 417 // Force HEAD records to have a creation date. 418 if (!str_contains($gedrec, "\n1 DATE ")) { 419 $today = strtoupper(date('d M Y')); 420 $gedrec .= "\n1 DATE " . $today; 421 } 422 423 DB::table('other')->insert([ 424 'o_id' => $xref, 425 'o_file' => $tree_id, 426 'o_type' => Header::RECORD_TYPE, 427 'o_gedcom' => $gedrec, 428 ]); 429 break; 430 431 432 case Media::RECORD_TYPE: 433 $record = Registry::mediaFactory()->new($xref, $gedrec, null, $tree); 434 435 DB::table('media')->insert([ 436 'm_id' => $xref, 437 'm_file' => $tree_id, 438 'm_gedcom' => $gedrec, 439 ]); 440 441 foreach ($record->mediaFiles() as $media_file) { 442 DB::table('media_file')->insert([ 443 'm_id' => $xref, 444 'm_file' => $tree_id, 445 'multimedia_file_refn' => mb_substr($media_file->filename(), 0, 248), 446 'multimedia_format' => mb_substr($media_file->format(), 0, 4), 447 'source_media_type' => mb_substr($media_file->type(), 0, 15), 448 'descriptive_title' => mb_substr($media_file->title(), 0, 248), 449 ]); 450 } 451 break; 452 453 default: // Custom record types. 454 DB::table('other')->insert([ 455 'o_id' => $xref, 456 'o_file' => $tree_id, 457 'o_type' => mb_substr($type, 0, 15), 458 'o_gedcom' => $gedrec, 459 ]); 460 break; 461 } 462 463 // Update the cross-reference/index tables. 464 $this->updateLinks($xref, $tree_id, $gedrec); 465 } 466 467 /** 468 * Legacy Family Tree software generates _PLAC_DEFN records containing LAT/LONG values 469 * 470 * @param string $gedcom 471 */ 472 private function importLegacyPlacDefn(string $gedcom): void 473 { 474 $gedcom_service = new GedcomService(); 475 476 if (preg_match('/\n1 PLAC (.+)/', $gedcom, $match)) { 477 $place_name = $match[1]; 478 } else { 479 return; 480 } 481 482 if (preg_match('/\n3 LATI ([NS].+)/', $gedcom, $match)) { 483 $latitude = $gedcom_service->readLatitude($match[1]); 484 } else { 485 return; 486 } 487 488 if (preg_match('/\n3 LONG ([EW].+)/', $gedcom, $match)) { 489 $longitude = $gedcom_service->readLongitude($match[1]); 490 } else { 491 return; 492 } 493 494 $location = new PlaceLocation($place_name); 495 496 if ($location->latitude() === null && $location->longitude() === null) { 497 DB::table('place_location') 498 ->where('id', '=', $location->id()) 499 ->update([ 500 'latitude' => $latitude, 501 'longitude' => $longitude, 502 ]); 503 } 504 } 505 506 /** 507 * Legacy Family Tree software generates _PLAC records containing LAT/LONG values 508 * 509 * @param string $gedcom 510 */ 511 private function importTNGPlac(string $gedcom): void 512 { 513 if (preg_match('/^0 _PLAC (.+)/', $gedcom, $match)) { 514 $place_name = $match[1]; 515 } else { 516 return; 517 } 518 519 if (preg_match('/\n2 LATI (.+)/', $gedcom, $match)) { 520 $latitude = (float) $match[1]; 521 } else { 522 return; 523 } 524 525 if (preg_match('/\n2 LONG (.+)/', $gedcom, $match)) { 526 $longitude = (float) $match[1]; 527 } else { 528 return; 529 } 530 531 $location = new PlaceLocation($place_name); 532 533 if ($location->latitude() === null && $location->longitude() === null) { 534 DB::table('place_location') 535 ->where('id', '=', $location->id()) 536 ->update([ 537 'latitude' => $latitude, 538 'longitude' => $longitude, 539 ]); 540 } 541 } 542 543 /** 544 * Extract all level 2 places from the given record and insert them into the places table 545 * 546 * @param string $xref 547 * @param Tree $tree 548 * @param string $gedrec 549 * 550 * @return void 551 */ 552 public function updatePlaces(string $xref, Tree $tree, string $gedrec): void 553 { 554 // Insert all new rows together 555 $rows = []; 556 557 preg_match_all('/\n2 PLAC (.+)/', $gedrec, $matches); 558 559 $places = array_unique($matches[1]); 560 561 foreach ($places as $place_name) { 562 $place = new Place($place_name, $tree); 563 564 // Calling Place::id() will create the entry in the database, if it doesn't already exist. 565 while ($place->id() !== 0) { 566 $rows[] = [ 567 'pl_p_id' => $place->id(), 568 'pl_gid' => $xref, 569 'pl_file' => $tree->id(), 570 ]; 571 572 $place = $place->parent(); 573 } 574 } 575 576 // array_unique doesn't work with arrays of arrays 577 $rows = array_intersect_key($rows, array_unique(array_map('serialize', $rows))); 578 579 // PDO has a limit of 65535 placeholders, and each row requires 3 placeholders. 580 foreach (array_chunk($rows, 20000) as $chunk) { 581 DB::table('placelinks')->insert($chunk); 582 } 583 } 584 585 /** 586 * Extract all the dates from the given record and insert them into the database. 587 * 588 * @param string $xref 589 * @param int $ged_id 590 * @param string $gedrec 591 * 592 * @return void 593 */ 594 private function updateDates(string $xref, int $ged_id, string $gedrec): void 595 { 596 // Insert all new rows together 597 $rows = []; 598 599 preg_match_all("/\n1 (\w+).*(?:\n[2-9].*)*\n2 DATE (.+)(?:\n[2-9].*)*/", $gedrec, $matches, PREG_SET_ORDER); 600 601 foreach ($matches as $match) { 602 $fact = $match[1]; 603 $date = new Date($match[2]); 604 $rows[] = [ 605 'd_day' => $date->minimumDate()->day, 606 'd_month' => $date->minimumDate()->format('%O'), 607 'd_mon' => $date->minimumDate()->month, 608 'd_year' => $date->minimumDate()->year, 609 'd_julianday1' => $date->minimumDate()->minimumJulianDay(), 610 'd_julianday2' => $date->minimumDate()->maximumJulianDay(), 611 'd_fact' => $fact, 612 'd_gid' => $xref, 613 'd_file' => $ged_id, 614 'd_type' => $date->minimumDate()->format('%@'), 615 ]; 616 617 $rows[] = [ 618 'd_day' => $date->maximumDate()->day, 619 'd_month' => $date->maximumDate()->format('%O'), 620 'd_mon' => $date->maximumDate()->month, 621 'd_year' => $date->maximumDate()->year, 622 'd_julianday1' => $date->maximumDate()->minimumJulianDay(), 623 'd_julianday2' => $date->maximumDate()->maximumJulianDay(), 624 'd_fact' => $fact, 625 'd_gid' => $xref, 626 'd_file' => $ged_id, 627 'd_type' => $date->minimumDate()->format('%@'), 628 ]; 629 } 630 631 // array_unique doesn't work with arrays of arrays 632 $rows = array_intersect_key($rows, array_unique(array_map('serialize', $rows))); 633 634 DB::table('dates')->insert($rows); 635 } 636 637 /** 638 * Extract all the links from the given record and insert them into the database 639 * 640 * @param string $xref 641 * @param int $ged_id 642 * @param string $gedrec 643 * 644 * @return void 645 */ 646 private function updateLinks(string $xref, int $ged_id, string $gedrec): void 647 { 648 // Insert all new rows together 649 $rows = []; 650 651 preg_match_all('/\n\d+ (' . Gedcom::REGEX_TAG . ') @(' . Gedcom::REGEX_XREF . ')@/', $gedrec, $matches, PREG_SET_ORDER); 652 653 foreach ($matches as $match) { 654 // Some applications (e.g. GenoPro) create links longer than 15 characters. 655 $link = mb_substr($match[1], 0, 15); 656 657 // Take care of "duplicates" that differ on case/collation, e.g. "SOUR @S1@" and "SOUR @s1@" 658 $rows[$link . strtoupper($match[2])] = [ 659 'l_from' => $xref, 660 'l_to' => $match[2], 661 'l_type' => $link, 662 'l_file' => $ged_id, 663 ]; 664 } 665 666 DB::table('link')->insert($rows); 667 } 668 669 /** 670 * Extract all the names from the given record and insert them into the database. 671 * 672 * @param string $xref 673 * @param int $ged_id 674 * @param Individual $record 675 * 676 * @return void 677 */ 678 private function updateNames(string $xref, int $ged_id, Individual $record): void 679 { 680 // Insert all new rows together 681 $rows = []; 682 683 foreach ($record->getAllNames() as $n => $name) { 684 if ($name['givn'] === Individual::PRAENOMEN_NESCIO) { 685 $soundex_givn_std = null; 686 $soundex_givn_dm = null; 687 } else { 688 $soundex_givn_std = Soundex::russell($name['givn']); 689 $soundex_givn_dm = Soundex::daitchMokotoff($name['givn']); 690 } 691 692 if ($name['surn'] === Individual::NOMEN_NESCIO) { 693 $soundex_surn_std = null; 694 $soundex_surn_dm = null; 695 } else { 696 $soundex_surn_std = Soundex::russell($name['surname']); 697 $soundex_surn_dm = Soundex::daitchMokotoff($name['surname']); 698 } 699 700 $rows[] = [ 701 'n_file' => $ged_id, 702 'n_id' => $xref, 703 'n_num' => $n, 704 'n_type' => $name['type'], 705 'n_sort' => mb_substr($name['sort'], 0, 255), 706 'n_full' => mb_substr($name['fullNN'], 0, 255), 707 'n_surname' => mb_substr($name['surname'], 0, 255), 708 'n_surn' => mb_substr($name['surn'], 0, 255), 709 'n_givn' => mb_substr($name['givn'], 0, 255), 710 'n_soundex_givn_std' => $soundex_givn_std, 711 'n_soundex_surn_std' => $soundex_surn_std, 712 'n_soundex_givn_dm' => $soundex_givn_dm, 713 'n_soundex_surn_dm' => $soundex_surn_dm, 714 ]; 715 } 716 717 DB::table('name')->insert($rows); 718 } 719 720 /** 721 * Extract inline media data, and convert to media objects. 722 * 723 * @param Tree $tree 724 * @param string $gedcom 725 * 726 * @return string 727 */ 728 private function convertInlineMedia(Tree $tree, string $gedcom): string 729 { 730 while (preg_match('/\n1 OBJE(?:\n[2-9].+)+/', $gedcom, $match)) { 731 $xref = $this->createMediaObject($match[0], $tree); 732 $gedcom = strtr($gedcom, [$match[0] => "\n1 OBJE @" . $xref . '@']); 733 } 734 while (preg_match('/\n2 OBJE(?:\n[3-9].+)+/', $gedcom, $match)) { 735 $xref = $this->createMediaObject($match[0], $tree); 736 $gedcom = strtr($gedcom, [$match[0] => "\n2 OBJE @" . $xref . '@']); 737 } 738 while (preg_match('/\n3 OBJE(?:\n[4-9].+)+/', $gedcom, $match)) { 739 $xref = $this->createMediaObject($match[0], $tree); 740 $gedcom = strtr($gedcom, [$match[0] => "\n3 OBJE @" . $xref . '@']); 741 } 742 743 return $gedcom; 744 } 745 746 /** 747 * Create a new media object, from inline media data. 748 * 749 * GEDCOM 5.5.1 specifies: +1 FILE / +2 FORM / +3 MEDI / +1 TITL 750 * GEDCOM 5.5 specifies: +1 FILE / +1 FORM / +1 TITL 751 * GEDCOM 5.5.1 says that GEDCOM 5.5 specifies: +1 FILE / +1 FORM / +2 MEDI 752 * 753 * Legacy generates: +1 FORM / +1 FILE / +1 TITL / +1 _SCBK / +1 _PRIM / +1 _TYPE / +1 NOTE 754 * RootsMagic generates: +1 FILE / +1 FORM / +1 TITL 755 * 756 * @param string $gedcom 757 * @param Tree $tree 758 * 759 * @return string 760 */ 761 private function createMediaObject(string $gedcom, Tree $tree): string 762 { 763 preg_match('/\n\d FILE (.+)/', $gedcom, $match); 764 $file = $match[1] ?? ''; 765 766 preg_match('/\n\d TITL (.+)/', $gedcom, $match); 767 $title = $match[1] ?? ''; 768 769 preg_match('/\n\d FORM (.+)/', $gedcom, $match); 770 $format = $match[1] ?? ''; 771 772 preg_match('/\n\d MEDI (.+)/', $gedcom, $match); 773 $media = $match[1] ?? ''; 774 775 preg_match('/\n\d _SCBK (.+)/', $gedcom, $match); 776 $scrapbook = $match[1] ?? ''; 777 778 preg_match('/\n\d _PRIM (.+)/', $gedcom, $match); 779 $primary = $match[1] ?? ''; 780 781 preg_match('/\n\d _TYPE (.+)/', $gedcom, $match); 782 if ($media === '') { 783 // Legacy uses _TYPE instead of MEDI 784 $media = $match[1] ?? ''; 785 $type = ''; 786 } else { 787 $type = $match[1] ?? ''; 788 } 789 790 preg_match_all('/\n\d NOTE (.+(?:\n\d CONT.*)*)/', $gedcom, $matches); 791 $notes = $matches[1] ?? []; 792 793 // Have we already created a media object with the same title/filename? 794 $xref = DB::table('media_file') 795 ->where('m_file', '=', $tree->id()) 796 ->where('descriptive_title', '=', mb_substr($title, 0, 248)) 797 ->where('multimedia_file_refn', '=', mb_substr($file, 0, 248)) 798 ->value('m_id'); 799 800 if ($xref === null) { 801 $xref = Registry::xrefFactory()->make(Media::RECORD_TYPE); 802 803 // convert to a media-object 804 $gedcom = '0 @' . $xref . "@ OBJE\n1 FILE " . $file; 805 806 if ($format !== '') { 807 $gedcom .= "\n2 FORM " . $format; 808 809 if ($media !== '') { 810 $gedcom .= "\n3 TYPE " . $media; 811 } 812 } 813 814 if ($title !== '') { 815 $gedcom .= "\n3 TITL " . $title; 816 } 817 818 if ($scrapbook !== '') { 819 $gedcom .= "\n1 _SCBK " . $scrapbook; 820 } 821 822 if ($primary !== '') { 823 $gedcom .= "\n1 _PRIM " . $primary; 824 } 825 826 if ($type !== '') { 827 $gedcom .= "\n1 _TYPE " . $type; 828 } 829 830 foreach ($notes as $note) { 831 $gedcom .= "\n1 NOTE " . strtr($note, ["\n3" => "\n2", "\n4" => "\n2", "\n5" => "\n2"]); 832 } 833 834 DB::table('media')->insert([ 835 'm_id' => $xref, 836 'm_file' => $tree->id(), 837 'm_gedcom' => $gedcom, 838 ]); 839 840 DB::table('media_file')->insert([ 841 'm_id' => $xref, 842 'm_file' => $tree->id(), 843 'multimedia_file_refn' => mb_substr($file, 0, 248), 844 'multimedia_format' => mb_substr($format, 0, 4), 845 'source_media_type' => mb_substr($media, 0, 15), 846 'descriptive_title' => mb_substr($title, 0, 248), 847 ]); 848 } 849 850 return $xref; 851 } 852 853 /** 854 * update a record in the database 855 * 856 * @param string $gedrec 857 * @param Tree $tree 858 * @param bool $delete 859 * 860 * @return void 861 * @throws GedcomErrorException 862 */ 863 public function updateRecord(string $gedrec, Tree $tree, bool $delete): void 864 { 865 if (preg_match('/^0 @(' . Gedcom::REGEX_XREF . ')@ (' . Gedcom::REGEX_TAG . ')/', $gedrec, $match)) { 866 [, $gid, $type] = $match; 867 } elseif (preg_match('/^0 (HEAD)(?:\n|$)/', $gedrec, $match)) { 868 // The HEAD record has no XREF. Any others? 869 $gid = $match[1]; 870 $type = $match[1]; 871 } else { 872 throw new GedcomErrorException($gedrec); 873 } 874 875 // Place links 876 DB::table('placelinks') 877 ->where('pl_gid', '=', $gid) 878 ->where('pl_file', '=', $tree->id()) 879 ->delete(); 880 881 // Orphaned places. If we're deleting "Westminster, London, England", 882 // then we may also need to delete "London, England" and "England". 883 do { 884 $affected = DB::table('places') 885 ->leftJoin('placelinks', function (JoinClause $join): void { 886 $join 887 ->on('p_id', '=', 'pl_p_id') 888 ->on('p_file', '=', 'pl_file'); 889 }) 890 ->whereNull('pl_p_id') 891 ->delete(); 892 } while ($affected > 0); 893 894 DB::table('dates') 895 ->where('d_gid', '=', $gid) 896 ->where('d_file', '=', $tree->id()) 897 ->delete(); 898 899 DB::table('name') 900 ->where('n_id', '=', $gid) 901 ->where('n_file', '=', $tree->id()) 902 ->delete(); 903 904 DB::table('link') 905 ->where('l_from', '=', $gid) 906 ->where('l_file', '=', $tree->id()) 907 ->delete(); 908 909 switch ($type) { 910 case Individual::RECORD_TYPE: 911 DB::table('individuals') 912 ->where('i_id', '=', $gid) 913 ->where('i_file', '=', $tree->id()) 914 ->delete(); 915 break; 916 917 case Family::RECORD_TYPE: 918 DB::table('families') 919 ->where('f_id', '=', $gid) 920 ->where('f_file', '=', $tree->id()) 921 ->delete(); 922 break; 923 924 case Source::RECORD_TYPE: 925 DB::table('sources') 926 ->where('s_id', '=', $gid) 927 ->where('s_file', '=', $tree->id()) 928 ->delete(); 929 break; 930 931 case Media::RECORD_TYPE: 932 DB::table('media_file') 933 ->where('m_id', '=', $gid) 934 ->where('m_file', '=', $tree->id()) 935 ->delete(); 936 937 DB::table('media') 938 ->where('m_id', '=', $gid) 939 ->where('m_file', '=', $tree->id()) 940 ->delete(); 941 break; 942 943 default: 944 DB::table('other') 945 ->where('o_id', '=', $gid) 946 ->where('o_file', '=', $tree->id()) 947 ->delete(); 948 break; 949 } 950 951 if (!$delete) { 952 $this->importRecord($gedrec, $tree, true); 953 } 954 } 955} 956