1<?php 2 3/** 4 * webtrees: online genealogy 5 * Copyright (C) 2023 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\DB; 23use Fisharebest\Webtrees\Encodings\UTF8; 24use Fisharebest\Webtrees\Family; 25use Fisharebest\Webtrees\GedcomRecord; 26use Fisharebest\Webtrees\Header; 27use Fisharebest\Webtrees\I18N; 28use Fisharebest\Webtrees\Individual; 29use Fisharebest\Webtrees\Media; 30use Fisharebest\Webtrees\Registry; 31use Fisharebest\Webtrees\Site; 32use Fisharebest\Webtrees\Source; 33use Fisharebest\Webtrees\Tree; 34use Illuminate\Database\Query\Expression; 35use Illuminate\Database\Query\JoinClause; 36use Illuminate\Support\Collection; 37use League\Flysystem\FilesystemException; 38use League\Flysystem\FilesystemOperator; 39use League\Flysystem\StorageAttributes; 40 41use function array_map; 42use function array_unique; 43use function explode; 44use function fclose; 45use function fread; 46use function implode; 47use function preg_match; 48use function sort; 49 50/** 51 * Utilities for the control panel. 52 */ 53class AdminService 54{ 55 /** 56 * Count of XREFs used by two trees at the same time. 57 * 58 * @param Tree $tree1 59 * @param Tree $tree2 60 * 61 * @return int 62 */ 63 public function countCommonXrefs(Tree $tree1, Tree $tree2): int 64 { 65 $subquery1 = DB::table('individuals') 66 ->where('i_file', '=', $tree1->id()) 67 ->select(['i_id AS xref']) 68 ->union(DB::table('families') 69 ->where('f_file', '=', $tree1->id()) 70 ->select(['f_id AS xref'])) 71 ->union(DB::table('sources') 72 ->where('s_file', '=', $tree1->id()) 73 ->select(['s_id AS xref'])) 74 ->union(DB::table('media') 75 ->where('m_file', '=', $tree1->id()) 76 ->select(['m_id AS xref'])) 77 ->union(DB::table('other') 78 ->where('o_file', '=', $tree1->id()) 79 ->whereNotIn('o_type', [Header::RECORD_TYPE, 'TRLR']) 80 ->select(['o_id AS xref'])); 81 82 $subquery2 = DB::table('change') 83 ->where('gedcom_id', '=', $tree2->id()) 84 ->select(['xref AS other_xref']) 85 ->union(DB::table('individuals') 86 ->where('i_file', '=', $tree2->id()) 87 ->select(['i_id AS xref'])) 88 ->union(DB::table('families') 89 ->where('f_file', '=', $tree2->id()) 90 ->select(['f_id AS xref'])) 91 ->union(DB::table('sources') 92 ->where('s_file', '=', $tree2->id()) 93 ->select(['s_id AS xref'])) 94 ->union(DB::table('media') 95 ->where('m_file', '=', $tree2->id()) 96 ->select(['m_id AS xref'])) 97 ->union(DB::table('other') 98 ->where('o_file', '=', $tree2->id()) 99 ->whereNotIn('o_type', [Header::RECORD_TYPE, 'TRLR']) 100 ->select(['o_id AS xref'])); 101 102 return DB::table(new Expression('(' . $subquery1->toSql() . ') AS sub1')) 103 ->mergeBindings($subquery1) 104 ->joinSub($subquery2, 'sub2', 'other_xref', '=', 'xref') 105 ->count(); 106 } 107 108 /** 109 * @param Tree $tree 110 * 111 * @return array<string,array<int,array<int,GedcomRecord>>> 112 */ 113 public function duplicateRecords(Tree $tree): array 114 { 115 // We can't do any reasonable checks using MySQL. 116 // Will need to wait for a "repositories" table. 117 $repositories = []; 118 119 $sources = DB::table('sources') 120 ->where('s_file', '=', $tree->id()) 121 ->groupBy(['s_name']) 122 ->having(new Expression('COUNT(s_id)'), '>', '1') 123 ->select([new Expression(DB::groupConcat('s_id') . ' AS xrefs')]) 124 ->orderBy('xrefs') 125 ->pluck('xrefs') 126 ->map(static fn(string $xrefs): array => array_map(static fn(string $xref): Source => Registry::sourceFactory()->make($xref, $tree), explode(',', $xrefs))) 127 ->all(); 128 129 // Database agnostic way to do GROUP_CONCAT(DISTINCT x ORDER BY x) 130 $distinct_order_by = static function (string $xrefs): string { 131 $array = explode(',', $xrefs); 132 sort($array); 133 134 return implode(',', array_unique($array)); 135 }; 136 137 $individuals = DB::table('dates') 138 ->join('name', static function (JoinClause $join): void { 139 $join 140 ->on('d_file', '=', 'n_file') 141 ->on('d_gid', '=', 'n_id'); 142 }) 143 ->where('d_file', '=', $tree->id()) 144 ->whereIn('d_fact', ['BIRT', 'CHR', 'BAPM', 'DEAT', 'BURI']) 145 ->groupBy(['d_year', 'd_month', 'd_day', 'd_type', 'd_fact', 'n_type', 'n_full']) 146 ->having(new Expression('COUNT(DISTINCT d_gid)'), '>', '1') 147 ->select([new Expression(DB::groupConcat('d_gid') . ' AS xrefs')]) 148 ->orderBy('xrefs') 149 ->pluck('xrefs') 150 ->map($distinct_order_by) 151 ->unique() 152 ->map(static fn(string $xrefs): array => array_map(static fn(string $xref): Individual => Registry::individualFactory()->make($xref, $tree), explode(',', $xrefs))) 153 ->all(); 154 155 $families = DB::table('families') 156 ->where('f_file', '=', $tree->id()) 157 ->groupBy([new Expression('LEAST(f_husb, f_wife)')]) 158 ->groupBy([new Expression('GREATEST(f_husb, f_wife)')]) 159 ->having(new Expression('COUNT(f_id)'), '>', '1') 160 ->select([new Expression(DB::groupConcat('f_id') . ' AS xrefs')]) 161 ->orderBy('xrefs') 162 ->pluck('xrefs') 163 ->map(static fn(string $xrefs): array => array_map(static fn(string $xref): Family => Registry::familyFactory()->make($xref, $tree), explode(',', $xrefs))) 164 ->all(); 165 166 $media = DB::table('media_file') 167 ->where('m_file', '=', $tree->id()) 168 ->where('descriptive_title', '<>', '') 169 ->groupBy(['descriptive_title']) 170 ->having(new Expression('COUNT(DISTINCT m_id)'), '>', '1') 171 ->select([new Expression(DB::groupConcat('m_id') . ' AS xrefs')]) 172 ->orderBy('xrefs') 173 ->pluck('xrefs') 174 ->map(static fn(string $xrefs): array => array_map(static fn(string $xref): Media => Registry::mediaFactory()->make($xref, $tree), explode(',', $xrefs))) 175 ->all(); 176 177 return [ 178 I18N::translate('Repositories') => $repositories, 179 I18N::translate('Sources') => $sources, 180 I18N::translate('Individuals') => $individuals, 181 I18N::translate('Families') => $families, 182 I18N::translate('Media objects') => $media, 183 ]; 184 } 185 186 /** 187 * Every XREF used by this tree and also used by some other tree 188 * 189 * @param Tree $tree 190 * 191 * @return array<string> 192 */ 193 public function duplicateXrefs(Tree $tree): array 194 { 195 $subquery1 = DB::table('individuals') 196 ->where('i_file', '=', $tree->id()) 197 ->select(['i_id AS xref', new Expression("'INDI' AS type")]) 198 ->union(DB::table('families') 199 ->where('f_file', '=', $tree->id()) 200 ->select(['f_id AS xref', new Expression("'FAM' AS type")])) 201 ->union(DB::table('sources') 202 ->where('s_file', '=', $tree->id()) 203 ->select(['s_id AS xref', new Expression("'SOUR' AS type")])) 204 ->union(DB::table('media') 205 ->where('m_file', '=', $tree->id()) 206 ->select(['m_id AS xref', new Expression("'OBJE' AS type")])) 207 ->union(DB::table('other') 208 ->where('o_file', '=', $tree->id()) 209 ->whereNotIn('o_type', [Header::RECORD_TYPE, 'TRLR']) 210 ->select(['o_id AS xref', 'o_type AS type'])); 211 212 $subquery2 = DB::table('change') 213 ->where('gedcom_id', '<>', $tree->id()) 214 ->select(['xref AS other_xref']) 215 ->union(DB::table('individuals') 216 ->where('i_file', '<>', $tree->id()) 217 ->select(['i_id AS xref'])) 218 ->union(DB::table('families') 219 ->where('f_file', '<>', $tree->id()) 220 ->select(['f_id AS xref'])) 221 ->union(DB::table('sources') 222 ->where('s_file', '<>', $tree->id()) 223 ->select(['s_id AS xref'])) 224 ->union(DB::table('media') 225 ->where('m_file', '<>', $tree->id()) 226 ->select(['m_id AS xref'])) 227 ->union(DB::table('other') 228 ->where('o_file', '<>', $tree->id()) 229 ->whereNotIn('o_type', [Header::RECORD_TYPE, 'TRLR']) 230 ->select(['o_id AS xref'])); 231 232 return DB::query() 233 ->fromSub($subquery1, 'sub1') 234 ->joinSub($subquery2, 'sub2', 'other_xref', '=', 'xref') 235 ->pluck('type', 'xref') 236 ->all(); 237 } 238 239 /** 240 * A list of GEDCOM files in the data folder. 241 * 242 * @param FilesystemOperator $filesystem 243 * 244 * @return Collection<int,string> 245 */ 246 public function gedcomFiles(FilesystemOperator $filesystem): Collection 247 { 248 try { 249 $files = $filesystem->listContents('') 250 ->filter(static function (StorageAttributes $attributes) use ($filesystem) { 251 if (!$attributes->isFile()) { 252 return false; 253 } 254 255 $stream = $filesystem->readStream($attributes->path()); 256 257 $header = fread($stream, 10); 258 fclose($stream); 259 260 return preg_match('/^(' . UTF8::BYTE_ORDER_MARK . ')?0 HEAD/', $header) > 0; 261 }) 262 ->map(fn(StorageAttributes $attributes) => $attributes->path()) 263 ->toArray(); 264 } catch (FilesystemException) { 265 $files = []; 266 } 267 268 return Collection::make($files)->sort(); 269 } 270 271 /** 272 * Change the behaviour a little, when there are a lot of trees. 273 * 274 * @return int 275 */ 276 public function multipleTreeThreshold(): int 277 { 278 return (int) Site::getPreference('MULTIPLE_TREE_THRESHOLD'); 279 } 280} 281