1<?php 2 3/** 4 * webtrees: online genealogy 5 * Copyright (C) 2021 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\Gedcom; 23 24/** 25 * Utilities for manipulating GEDCOM data. 26 */ 27class GedcomService 28{ 29 // User defined tags begin with an underscore 30 private const USER_DEFINED_TAG_PREFIX = '_'; 31 32 // Some applications, such as FTM, use GEDCOM tag names instead of the tags. 33 private const TAG_NAMES = [ 34 'ABBREVIATION' => 'ABBR', 35 'ADDRESS' => 'ADDR', 36 'ADDRESS1' => 'ADR1', 37 'ADDRESS2' => 'ADR2', 38 'ADOPTION' => 'ADOP', 39 'AGENCY' => 'AGNC', 40 'ALIAS' => 'ALIA', 41 'ANCESTORS' => 'ANCE', 42 'ANCES_INTEREST' => 'ANCI', 43 'ANULMENT' => 'ANUL', 44 'ASSOCIATES' => 'ASSO', 45 'AUTHOR' => 'AUTH', 46 'BAPTISM-LDS' => 'BAPL', 47 'BAPTISM' => 'BAPM', 48 'BAR_MITZVAH' => 'BARM', 49 'BAS_MITZVAH' => 'BASM', 50 'BIRTH' => 'BIRT', 51 'BLESSING' => 'BLES', 52 'BURIAL' => 'BURI', 53 'CALL_NUMBER' => 'CALN', 54 'CASTE' => 'CAST', 55 'CAUSE' => 'CAUS', 56 'CENSUS' => 'CENS', 57 'CHANGE' => 'CHAN', 58 'CHARACTER' => 'CHAR', 59 'CHILD' => 'CHIL', 60 'CHRISTENING' => 'CHR', 61 'ADULT_CHRISTENING' => 'CHRA', 62 'CONCATENATION' => 'CONC', 63 'CONFIRMATION' => 'CONF', 64 'CONFIRMATION-LDS' => 'CONL', 65 'CONTINUED' => 'CONT', 66 'COPYRIGHT' => 'COPY', 67 'CORPORTATE' => 'CORP', 68 'CREMATION' => 'CREM', 69 'COUNTRY' => 'CTRY', 70 'DEATH' => 'DEAT', 71 'DESCENDANTS' => 'DESC', 72 'DESCENDANTS_INT' => 'DESI', 73 'DESTINATION' => 'DEST', 74 'DIVORCE' => 'DIV', 75 'DIVORCE_FILED' => 'DIVF', 76 'PHY_DESCRIPTION' => 'DSCR', 77 'EDUCATION' => 'EDUC', 78 'EMIGRATION' => 'EMIG', 79 'ENDOWMENT' => 'ENDL', 80 'ENGAGEMENT' => 'ENGA', 81 'EVENT' => 'EVEN', 82 'FAMILY' => 'FAM', 83 'FAMILY_CHILD' => 'FAMC', 84 'FAMILY_FILE' => 'FAMF', 85 'FAMILY_SPOUSE' => 'FAMS', 86 'FACIMILIE' => 'FAX', 87 'FIRST_COMMUNION' => 'FCOM', 88 'FORMAT' => 'FORM', 89 'PHONETIC' => 'FONE', 90 'GEDCOM' => 'GEDC', 91 'GIVEN_NAME' => 'GIVN', 92 'GRADUATION' => 'GRAD', 93 'HEADER' => 'HEAD', 94 'HUSBAND' => 'HUSB', 95 'IDENT_NUMBER' => 'IDNO', 96 'IMMIGRATION' => 'IMMI', 97 'INDIVIDUAL' => 'INDI', 98 'LANGUAGE' => 'LANG', 99 'LATITUDE' => 'LATI', 100 'LONGITUDE' => 'LONG', 101 'MARRIAGE_BANN' => 'MARB', 102 'MARR_CONTRACT' => 'MARC', 103 'MARR_LICENSE' => 'MARL', 104 'MARRIAGE' => 'MARR', 105 'MEDIA' => 'MEDI', 106 'NATIONALITY' => 'NATI', 107 'NATURALIZATION' => 'NATU', 108 'CHILDREN_COUNT' => 'NCHI', 109 'NICKNAME' => 'NICK', 110 'MARRIAGE_COUNT' => 'NMR', 111 'NAME_PREFIX' => 'NPFX', 112 'NAME_SUFFIX' => 'NSFX', 113 'OBJECT' => 'OBJE', 114 'OCCUPATION' => 'OCCU', 115 'ORDINANCE' => 'ORDI', 116 'ORDINATION' => 'ORDN', 117 'PEDIGREE' => 'PEDI', 118 'PHONE' => 'PHON', 119 'PLACE' => 'PLAC', 120 'POSTAL_CODE' => 'POST', 121 'PROBATE' => 'PROB', 122 'PROPERTY' => 'PROP', 123 'PUBLICATION' => 'PUBL', 124 'QUALITY_OF_DATA' => 'QUAY', 125 'REFERENCE' => 'REFN', 126 'RELATIONSHIP' => 'RELA', 127 'RELIGION' => 'RELI', 128 'REPOSITORY' => 'REPO', 129 'RESIDENCE' => 'RESI', 130 'RESTRICTION' => 'RESN', 131 'RETIREMENT' => 'RETI', 132 'REC_FILE_NUMBER' => 'RFN', 133 'REC_ID_NUMBER' => 'RIN', 134 'ROMANIZED' => 'ROMN', 135 'SEALING_CHILD' => 'SLGC', 136 'SEALING_SPOUSE' => 'SLGS', 137 'SOURCE' => 'SOUR', 138 'SURN_PREFIX' => 'SPFX', 139 'SOC_SEC_NUMBER' => 'SSN', 140 'STATE' => 'STAE', 141 'STATUS' => 'STAT', 142 'SUBMITTER' => 'SUBM', 143 'SUBMISSION' => 'SUBN', 144 'SURNAME' => 'SURN', 145 'TEMPLE' => 'TEMP', 146 'TITLE' => 'TITL', 147 'TRAILER' => 'TRLR', 148 'VERSION' => 'VERS', 149 'WEB' => 'WWW', 150 '_DEATH_OF_SPOUSE' => 'DETS', 151 '_DEGREE' => '_DEG', 152 '_MEDICAL' => '_MCL', 153 '_MILITARY_SERVICE' => '_MILT', 154 ]; 155 156 // Custom tags used by other applications, with direct synonyms 157 private const TAG_SYNONYMS = [ 158 // Convert PhpGedView tag to webtrees 159 '_PGVU' => '_WT_USER', 160 '_PGV_OBJS' => '_WT_OBJE_SORT', 161 ]; 162 163 // SEX tags 164 private const SEX_FEMALE = 'F'; 165 private const SEX_MALE = 'M'; 166 private const SEX_UNKNOWN = 'U'; 167 168 /** 169 * Convert a GEDCOM tag to a canonical form. 170 * 171 * @param string $tag 172 * 173 * @return string 174 */ 175 public function canonicalTag(string $tag): string 176 { 177 $tag = strtoupper($tag); 178 179 $tag = self::TAG_NAMES[$tag] ?? self::TAG_SYNONYMS[$tag] ?? $tag; 180 181 return $tag; 182 } 183 184 /** 185 * @param string $tag 186 * 187 * @return bool 188 */ 189 public function isUserDefinedTag(string $tag): bool 190 { 191 return substr_compare($tag, self::USER_DEFINED_TAG_PREFIX, 0, 1) === 0; 192 } 193 194 /** 195 * @param string $text 196 * 197 * @return float|null 198 */ 199 public function readLatitude(string $text): ?float 200 { 201 return $this->readDegrees($text, Gedcom::LATITUDE_NORTH, Gedcom::LATITUDE_SOUTH); 202 } 203 204 /** 205 * @param string $text 206 * 207 * @return float|null 208 */ 209 public function readLongitude(string $text): ?float 210 { 211 return $this->readDegrees($text, Gedcom::LONGITUDE_EAST, Gedcom::LONGITUDE_WEST); 212 } 213 214 /** 215 * @param string $text 216 * @param string $positive 217 * @param string $negative 218 * 219 * @return float|null 220 */ 221 private function readDegrees(string $text, string $positive, string $negative): ?float 222 { 223 $text = trim($text); 224 $hemisphere = substr($text, 0, 1); 225 $degrees = substr($text, 1); 226 227 // Match a valid GEDCOM format 228 if (is_numeric($degrees)) { 229 $hemisphere = strtoupper($hemisphere); 230 $degrees = (float) $degrees; 231 232 if ($hemisphere === $positive) { 233 return $degrees; 234 } 235 236 if ($hemisphere === $negative) { 237 return -$degrees; 238 } 239 } 240 241 // Just a number? 242 if (is_numeric($text)) { 243 return (float) $text; 244 } 245 246 // Can't match anything. 247 return null; 248 } 249 250 /** 251 * Although empty placenames are valid "Town, , Country", it is only meaningful 252 * when structured places are used (PLAC:FORM town, county, country), and 253 * structured places are discouraged. 254 * 255 * @param string $text 256 * 257 * @return array<string> 258 */ 259 public function readPlace(string $text): array 260 { 261 $text = trim($text); 262 263 return preg_split(Gedcom::PLACE_SEPARATOR_REGEX, $text); 264 } 265 266 /** 267 * @param string[] $place 268 * 269 * @return string 270 */ 271 public function writePlace(array $place): string 272 { 273 return implode(Gedcom::PLACE_SEPARATOR, $place); 274 } 275 276 /** 277 * Some applications use non-standard values for unknown. 278 * 279 * @param string $text 280 * 281 * @return string 282 */ 283 public function readSex(string $text): string 284 { 285 $text = strtoupper($text); 286 287 if ($text !== self::SEX_MALE && $text !== self::SEX_FEMALE) { 288 $text = self::SEX_UNKNOWN; 289 } 290 291 return $text; 292 } 293} 294