1<?php 2 3/** 4 * webtrees: online genealogy 5 * Copyright (C) 2019 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 <http://www.gnu.org/licenses/>. 16 */ 17 18declare(strict_types=1); 19 20namespace Fisharebest\Webtrees\Services; 21 22/** 23 * Utilities for manipulating GEDCOM data. 24 */ 25class GedcomService 26{ 27 // Gedcom allows 255 characters (not bytes), including the EOL character. 28 private const EOL = "\r\n"; 29 private const EOL_REGEX = '\r|\r\n|\n|\n\r'; 30 private const LINE_LENGTH = 255 - 2; 31 32 33 // User defined tags begin with an underscore 34 private const USER_DEFINED_TAG_PREFIX = '_'; 35 36 // Some applications, such as FTM, use GEDCOM tag names instead of the tags. 37 private const TAG_NAMES = [ 38 'ABBREVIATION' => 'ABBR', 39 'ADDRESS' => 'ADDR', 40 'ADDRESS1' => 'ADR1', 41 'ADDRESS2' => 'ADR2', 42 'ADOPTION' => 'ADOP', 43 'AGENCY' => 'AGNC', 44 'ALIAS' => 'ALIA', 45 'ANCESTORS' => 'ANCE', 46 'ANCES_INTEREST' => 'ANCI', 47 'ANULMENT' => 'ANUL', 48 'ASSOCIATES' => 'ASSO', 49 'AUTHOR' => 'AUTH', 50 'BAPTISM-LDS' => 'BAPL', 51 'BAPTISM' => 'BAPM', 52 'BAR_MITZVAH' => 'BARM', 53 'BAS_MITZVAH' => 'BASM', 54 'BIRTH' => 'BIRT', 55 'BLESSING' => 'BLES', 56 'BURIAL' => 'BURI', 57 'CALL_NUMBER' => 'CALN', 58 'CASTE' => 'CAST', 59 'CAUSE' => 'CAUS', 60 'CENSUS' => 'CENS', 61 'CHANGE' => 'CHAN', 62 'CHARACTER' => 'CHAR', 63 'CHILD' => 'CHIL', 64 'CHRISTENING' => 'CHR', 65 'ADULT_CHRISTENING' => 'CHRA', 66 'CONCATENATION' => 'CONC', 67 'CONFIRMATION' => 'CONF', 68 'CONFIRMATION-LDS' => 'CONL', 69 'CONTINUED' => 'CONT', 70 'COPYRIGHT' => 'COPY', 71 'CORPORTATE' => 'CORP', 72 'CREMATION' => 'CREM', 73 'COUNTRY' => 'CTRY', 74 'DEATH' => 'DEAT', 75 'DESCENDANTS' => 'DESC', 76 'DESCENDANTS_INT' => 'DESI', 77 'DESTINATION' => 'DEST', 78 'DIVORCE' => 'DIV', 79 'DIVORCE_FILED' => 'DIVF', 80 'PHY_DESCRIPTION' => 'DSCR', 81 'EDUCATION' => 'EDUC', 82 'EMAIL' => 'EMAI', 83 'EMIGRATION' => 'EMIG', 84 'ENDOWMENT' => 'ENDL', 85 'ENGAGEMENT' => 'ENGA', 86 'EVENT' => 'EVEN', 87 'FAMILY' => 'FAM', 88 'FAMILY_CHILD' => 'FAMC', 89 'FAMILY_FILE' => 'FAMF', 90 'FAMILY_SPOUSE' => 'FAMS', 91 'FACIMILIE' => 'FAX', 92 'FIRST_COMMUNION' => 'FCOM', 93 'FORMAT' => 'FORM', 94 'PHONETIC' => 'FONE', 95 'GEDCOM' => 'GEDC', 96 'GIVEN_NAME' => 'GIVN', 97 'GRADUATION' => 'GRAD', 98 'HEADER' => 'HEAD', 99 'HUSBAND' => 'HUSB', 100 'IDENT_NUMBER' => 'IDNO', 101 'IMMIGRATION' => 'IMMI', 102 'INDIVIDUAL' => 'INDI', 103 'LANGUAGE' => 'LANG', 104 'LATITUDE' => 'LATI', 105 'LONGITUDE' => 'LONG', 106 'MARRIAGE_BANN' => 'MARB', 107 'MARR_CONTRACT' => 'MARC', 108 'MARR_LICENSE' => 'MARL', 109 'MARRIAGE' => 'MARR', 110 'MEDIA' => 'MEDI', 111 'NATIONALITY' => 'NATI', 112 'NATURALIZATION' => 'NATU', 113 'CHILDREN_COUNT' => 'NCHI', 114 'NICKNAME' => 'NICK', 115 'MARRIAGE_COUNT' => 'NMR', 116 'NAME_PREFIX' => 'NPFX', 117 'NAME_SUFFIX' => 'NSFX', 118 'OBJECT' => 'OBJE', 119 'OCCUPATION' => 'OCCU', 120 'ORDINANCE' => 'ORDI', 121 'ORDINATION' => 'ORDN', 122 'PEDIGREE' => 'PEDI', 123 'PHONE' => 'PHON', 124 'PLACE' => 'PLAC', 125 'POSTAL_CODE' => 'POST', 126 'PROBATE' => 'PROB', 127 'PROPERTY' => 'PROP', 128 'PUBLICATION' => 'PUBL', 129 'QUALITY_OF_DATA' => 'QUAY', 130 'REFERENCE' => 'REFN', 131 'RELATIONSHIP' => 'RELA', 132 'RELIGION' => 'RELI', 133 'REPOSITORY' => 'REPO', 134 'RESIDENCE' => 'RESI', 135 'RESTRICTION' => 'RESN', 136 'RETIREMENT' => 'RETI', 137 'REC_FILE_NUMBER' => 'RFN', 138 'REC_ID_NUMBER' => 'RIN', 139 'ROMANIZED' => 'ROMN', 140 'SEALING_CHILD' => 'SLGC', 141 'SEALING_SPOUSE' => 'SLGS', 142 'SOURCE' => 'SOUR', 143 'SURN_PREFIX' => 'SPFX', 144 'SOC_SEC_NUMBER' => 'SSN', 145 'STATE' => 'STAE', 146 'STATUS' => 'STAT', 147 'SUBMITTER' => 'SUBM', 148 'SUBMISSION' => 'SUBN', 149 'SURNAME' => 'SURN', 150 'TEMPLE' => 'TEMP', 151 'TITLE' => 'TITL', 152 'TRAILER' => 'TRLR', 153 'VERSION' => 'VERS', 154 'WEB' => 'WWW', 155 '_DEATH_OF_SPOUSE' => 'DETS', 156 '_DEGREE' => '_DEG', 157 '_MEDICAL' => '_MCL', 158 '_MILITARY_SERVICE' => '_MILT', 159 ]; 160 161 // Custom tags used by other applications, with direct synonyms 162 private const TAG_SYNONYMS = [ 163 // Convert PhpGedView tag to webtrees 164 '_PGVU' => '_WT_USER', 165 '_PGV_OBJS' => '_WT_OBJE_SORT', 166 ]; 167 168 // LATI and LONG tags 169 private const DEGREE_FORMAT = ' % .5f%s'; 170 private const LATITUDE_NORTH = 'N'; 171 private const LATITUDE_SOUTH = 'S'; 172 private const LONGITUDE_EAST = 'E'; 173 private const LONGITUDE_WEST = 'W'; 174 175 // PLAC tags 176 private const PLACE_SEPARATOR = ', '; 177 private const PLACE_SEPARATOR_REGEX = ' *, *'; 178 179 // SEX tags 180 private const SEX_FEMALE = 'F'; 181 private const SEX_MALE = 'M'; 182 private const SEX_UNKNOWN = 'U'; 183 184 /** 185 * Convert a GEDCOM tag to a canonical form. 186 * 187 * @param string $tag 188 * 189 * @return string 190 */ 191 public function canonicalTag(string $tag): string 192 { 193 $tag = strtoupper($tag); 194 195 $tag = self::TAG_NAMES[$tag] ?? self::TAG_SYNONYMS[$tag] ?? $tag; 196 197 return $tag; 198 } 199 200 /** 201 * @param string $tag 202 * 203 * @return bool 204 */ 205 public function isUserDefinedTag(string $tag): bool 206 { 207 return substr_compare($tag, self::USER_DEFINED_TAG_PREFIX, 0, 1) === 0; 208 } 209 210 /** 211 * @param string $text 212 * 213 * @return float 214 */ 215 public function readLatitude(string $text): float 216 { 217 return $this->readDegrees($text, self::LATITUDE_NORTH, self::LATITUDE_SOUTH); 218 } 219 220 /** 221 * @param string $text 222 * 223 * @return float 224 */ 225 public function readLongitude(string $text): float 226 { 227 return $this->readDegrees($text, self::LONGITUDE_EAST, self::LONGITUDE_WEST); 228 } 229 230 /** 231 * @param string $text 232 * @param string $positive 233 * @param string $negative 234 * 235 * @return float 236 */ 237 private function readDegrees(string $text, string $positive, string $negative): float 238 { 239 $text = trim($text); 240 $hemisphere = substr($text, 0, 1); 241 $degrees = substr($text, 1); 242 243 // Match a valid GEDCOM format 244 if (is_numeric($degrees)) { 245 $hemisphere = strtoupper($hemisphere); 246 $degrees = (float) $degrees; 247 248 if ($hemisphere === $positive) { 249 return $degrees; 250 } 251 252 if ($hemisphere === $negative) { 253 return -$degrees; 254 } 255 } 256 257 // Just a number? 258 if (is_numeric($text)) { 259 return (float) $text; 260 } 261 262 // Can't match anything. 263 return 0.0; 264 } 265 266 /** 267 * @param float $latitude 268 * 269 * @return string 270 */ 271 public function writeLatitude(float $latitude): string 272 { 273 return $this->writeDegrees($latitude, self::LATITUDE_NORTH, self::LATITUDE_SOUTH); 274 } 275 276 /** 277 * @param float $longitude 278 * 279 * @return string 280 */ 281 public function writeLongitude(float $longitude): string 282 { 283 return $this->writeDegrees($longitude, self::LONGITUDE_EAST, self::LONGITUDE_WEST); 284 } 285 286 /** 287 * @param float $degrees 288 * @param string $positive 289 * @param string $negative 290 * 291 * @return string 292 */ 293 private function writeDegrees(float $degrees, string $positive, string $negative): string 294 { 295 if ($degrees < 0.0) { 296 return sprintf(self::DEGREE_FORMAT, $degrees, $negative); 297 } 298 299 return sprintf(self::DEGREE_FORMAT, $degrees, $positive); 300 } 301 302 /** 303 * Although empty placenames are valid "Town, , Country", it is only meaningful 304 * when structured places are used (PLAC:FORM town, county, country), and 305 * structured places are discouraged. 306 * 307 * @param string $text 308 * 309 * @return string[] 310 */ 311 public function readPlace(string $text): array 312 { 313 $text = trim($text); 314 315 return preg_split(self::PLACE_SEPARATOR_REGEX, $text, PREG_SPLIT_NO_EMPTY); 316 } 317 318 /** 319 * @param string[] $place 320 * 321 * @return string 322 */ 323 public function writePlace(array $place): string 324 { 325 return implode(self::PLACE_SEPARATOR, $place); 326 } 327 328 /** 329 * Some applications use non-standard values for unknown. 330 * 331 * @param string $text 332 * 333 * @return string 334 */ 335 public function readSex(string $text): string 336 { 337 $text = strtoupper($text); 338 339 if ($text !== self::SEX_MALE && $text !== self::SEX_FEMALE) { 340 $text = self::SEX_UNKNOWN; 341 } 342 343 return $text; 344 } 345} 346