1<?php 2 3 4use ComboStrap\DatabasePageRow; 5use ComboStrap\DokuwikiId; 6use ComboStrap\ExceptionBadArgument; 7use ComboStrap\ExceptionBadSyntax; 8use ComboStrap\ExceptionCompile; 9use ComboStrap\ExceptionSqliteNotAvailable; 10use ComboStrap\ExecutionContext; 11use ComboStrap\FileSystems; 12use ComboStrap\HttpResponse; 13use ComboStrap\HttpResponseStatus; 14use ComboStrap\Identity; 15use ComboStrap\LogUtility; 16use ComboStrap\MarkupPath; 17use ComboStrap\Meta\Field\AliasType; 18use ComboStrap\Mime; 19use ComboStrap\PageId; 20use ComboStrap\PageRules; 21use ComboStrap\PageUrlPath; 22use ComboStrap\PageUrlType; 23use ComboStrap\RouterBestEndPage; 24use ComboStrap\Site; 25use ComboStrap\SiteConfig; 26use ComboStrap\Sqlite; 27use ComboStrap\Web\Url; 28use ComboStrap\Web\UrlEndpoint; 29use ComboStrap\Web\UrlRewrite; 30use ComboStrap\WikiPath; 31 32require_once(__DIR__ . '/../vendor/autoload.php'); 33 34/** 35 * Class action_plugin_combo_url 36 * 37 * The actual URL manager 38 * 39 * 40 */ 41class action_plugin_combo_router extends DokuWiki_Action_Plugin 42{ 43 44 /** 45 * @deprecated 46 */ 47 const URL_MANAGER_ENABLE_CONF = "enableUrlManager"; 48 const ROUTER_ENABLE_CONF = "enableRouter"; 49 50 // The redirect type 51 const REDIRECT_TRANSPARENT_METHOD = 'transparent'; // was (Id) 52 // For permanent, see https://developers.google.com/search/docs/advanced/crawling/301-redirects 53 const REDIRECT_PERMANENT_METHOD = 'permanent'; // was `Http` (301) 54 const REDIRECT_NOTFOUND_METHOD = "notfound"; // 404 (See other) (when best page name is calculated) 55 56 public const PERMANENT_REDIRECT_CANONICAL = "permanent:redirect"; 57 58 // Where the target id value comes from 59 const TARGET_ORIGIN_WELL_KNOWN = 'well-known'; 60 const TARGET_ORIGIN_PAGE_RULES = 'pageRules'; 61 /** 62 * Named Permalink (canonical) 63 */ 64 const TARGET_ORIGIN_CANONICAL = 'canonical'; 65 const TARGET_ORIGIN_ALIAS = 'alias'; 66 /** 67 * Identifier Permalink (full page id) 68 */ 69 const TARGET_ORIGIN_PERMALINK = "permalink"; 70 /** 71 * Extended Permalink (abbreviated page id at the end) 72 */ 73 const TARGET_ORIGIN_PERMALINK_EXTENDED = "extendedPermalink"; 74 const TARGET_ORIGIN_START_PAGE = 'startPage'; 75 const TARGET_ORIGIN_BEST_PAGE_NAME = 'bestPageName'; 76 const TARGET_ORIGIN_BEST_NAMESPACE = 'bestNamespace'; 77 const TARGET_ORIGIN_SEARCH_ENGINE = 'searchEngine'; 78 const TARGET_ORIGIN_BEST_END_PAGE_NAME = 'bestEndPageName'; 79 const TARGET_ORIGIN_SHADOW_BANNED = "shadowBanned"; 80 81 82 // The constant parameters 83 const GO_TO_SEARCH_ENGINE = 'GoToSearchEngine'; 84 const GO_TO_BEST_NAMESPACE = 'GoToBestNamespace'; 85 const GO_TO_BEST_PAGE_NAME = 'GoToBestPageName'; 86 const GO_TO_BEST_END_PAGE_NAME = 'GoToBestEndPageName'; 87 const GO_TO_NS_START_PAGE = 'GoToNsStartPage'; 88 const GO_TO_EDIT_MODE = 'GoToEditMode'; 89 const NOTHING = 'Nothing'; 90 91 /** @var string - a name used in log and other places */ 92 const NAME = 'Url Manager'; 93 const CANONICAL = 'router'; 94 const PAGE_404 = "<html lang=\"en\"><body></body></html>"; 95 const REFRESH_HEADER_NAME = "Refresh"; 96 const REFRESH_HEADER_PREFIX = self::REFRESH_HEADER_NAME . ': 0;url='; 97 const LOCATION_HEADER_PREFIX = HttpResponse::LOCATION_HEADER_NAME . ": "; 98 public const URL_MANAGER_NAME = "Router"; 99 100 101 /** 102 * @var PageRules 103 */ 104 private $pageRules; 105 106 107 function __construct() 108 { 109 // enable direct access to language strings 110 // ie $this->lang 111 $this->setupLocale(); 112 113 } 114 115 /** 116 * @param string $refreshHeader 117 * @return false|string 118 */ 119 public static function getUrlFromRefresh(string $refreshHeader) 120 { 121 return substr($refreshHeader, strlen(action_plugin_combo_router::REFRESH_HEADER_PREFIX)); 122 } 123 124 public static function getUrlFromLocation($refreshHeader) 125 { 126 return substr($refreshHeader, strlen(action_plugin_combo_router::LOCATION_HEADER_PREFIX)); 127 } 128 129 /** 130 * @return string|null 131 * 132 * Return the original id from the request 133 * ie `howto:how-to-get-started-with-combostrap-m3i8vga8` 134 * if `/howto/how-to-get-started-with-combostrap-m3i8vga8` 135 * 136 * Unfortunately, DOKUWIKI_STARTED is not the first event 137 * The id may have been changed by 138 * {@link action_plugin_combo_lang::load_lang()} 139 * function, that's why we have this function 140 * to get the original requested id 141 */ 142 private static function getOriginalIdFromRequest(): ?string 143 { 144 $originalId = $_GET["id"] ?? null; 145 if ($originalId === null) { 146 return null; 147 } 148 // We get a `/` as first character 149 // because we return an id, we need to delete it 150 $originalId = substr($originalId,1); 151 // transform / to : 152 return str_replace("/", WikiPath::NAMESPACE_SEPARATOR_DOUBLE_POINT, $originalId); 153 } 154 155 /** 156 * Determine if the request should be banned based on the id 157 * 158 * @param string $id 159 * @return bool 160 * 161 * See also {@link https://perishablepress.com/7g-firewall/#features} 162 * for blocking rules on http request data such as: 163 * * query_string 164 * * user_agent, 165 * * remote host 166 */ 167 public static function isShadowBanned(string $id): bool 168 { 169 /** 170 * ie 171 * wp-json:api:flutter_woo:config_file 172 * wp-content:plugins:wpdiscuz:themes:default:style-rtl.css 173 * wp-admin 174 * 2020:wp-includes:wlwmanifest.xml 175 * wp-content:start 176 * wp-admin:css:start 177 * sito:wp-includes:wlwmanifest.xml 178 * site:wp-includes:wlwmanifest.xml 179 * cms:wp-includes:wlwmanifest.xml 180 * test:wp-includes:wlwmanifest.xml 181 * media:wp-includes:wlwmanifest.xml 182 * wp2:wp-includes:wlwmanifest.xml 183 * 2019:wp-includes:wlwmanifest.xml 184 * shop:wp-includes:wlwmanifest.xml 185 * wp1:wp-includes:wlwmanifest.xml 186 * news:wp-includes:wlwmanifest.xml 187 * 2018:wp-includes:wlwmanifest.xml 188 */ 189 if (strpos($id, 'wp-') !== false) { 190 return true; 191 } 192 193 /** 194 * db:oracle:long_or_1_utl_inaddr.get_host_address_chr_33_chr_126_chr_33_chr_65_chr_66_chr_67_chr_49_chr_52_chr_53_chr_90_chr_81_chr_54_chr_50_chr_68_chr_87_chr_81_chr_65_chr_70_chr_80_chr_79_chr_73_chr_89_chr_67_chr_70_chr_68_chr_33_chr_126_chr_33 195 * db:oracle:999999.9:union:all:select_null:from_dual 196 * db:oracle:999999.9:union:all:select_null:from_dual_and_0_0 197 */ 198 if (preg_match('/_chr_|_0_0/', $id) === 1) { 199 return true; 200 } 201 202 203 /** 204 * ie 205 * git:objects: 206 * git:refs:heads:stable 207 * git:logs:refs:heads:main 208 * git:logs:refs:heads:stable 209 * git:hooks:pre-push.sample 210 * git:hooks:pre-receive.sample 211 */ 212 if (strpos($id, "git:") === 0) { 213 return true; 214 } 215 216 return false; 217 218 } 219 220 /** 221 * @param string $id 222 * @return bool 223 * well-known:traffic-advice = https://github.com/buettner/private-prefetch-proxy/blob/main/traffic-advice.md 224 * .well-known/security.txt, id=well-known:security.txt = https://securitytxt.org/ 225 * well-known:dnt-policy.txt 226 */ 227 public static function isWellKnownFile(string $id): bool 228 { 229 return strpos($id, "well-known") === 0; 230 } 231 232 233 function register(Doku_Event_Handler $controller) 234 { 235 236 if (SiteConfig::getConfValue(self::ROUTER_ENABLE_CONF, 1)) { 237 238 /** 239 * This will call the function {@link action_plugin_combo_router::_router()} 240 * The event is not DOKUWIKI_STARTED because this is not the first one 241 * 242 * https://www.dokuwiki.org/devel:event:init_lang_load 243 */ 244 $controller->register_hook('DOKUWIKI_STARTED', 245 'BEFORE', 246 $this, 247 'router', 248 array()); 249 250 /** 251 * This is the real first call of Dokuwiki 252 * Unfortunately, it does not create the environment 253 * We just ban to spare server resources 254 * 255 * https://www.dokuwiki.org/devel:event:init_lang_load 256 */ 257 $controller->register_hook('INIT_LANG_LOAD', 'BEFORE', $this, 'ban', array()); 258 259 } 260 261 262 } 263 264 /** 265 * 266 * We have created a spacial ban function that is 267 * called before the first function 268 * {@link action_plugin_combo_metalang::load_lang()} 269 * to spare CPU. 270 * 271 * @param $event 272 * @throws Exception 273 */ 274 function ban(&$event) 275 { 276 277 $id = self::getOriginalIdFromRequest(); 278 if ($id === null) { 279 return; 280 } 281 $page = MarkupPath::createMarkupFromId($id); 282 if (!FileSystems::exists($page)) { 283 // Well known 284 if (self::isWellKnownFile($id)) { 285 $this->logRedirection($id, "", self::TARGET_ORIGIN_WELL_KNOWN, self::REDIRECT_NOTFOUND_METHOD); 286 ExecutionContext::getActualOrCreateFromEnv() 287 ->response() 288 ->setStatus(HttpResponseStatus::NOT_FOUND) 289 ->end(); 290 return; 291 } 292 293 // Shadow banned 294 if (self::isShadowBanned($id)) { 295 $webSiteHomePage = Site::getIndexPageName(); 296 $this->executeTransparentRedirect($webSiteHomePage, self::TARGET_ORIGIN_SHADOW_BANNED); 297 } 298 } 299 } 300 301 /** 302 * @param $event Doku_Event 303 * @param $param 304 * @return void 305 * @throws Exception 306 */ 307 function router(&$event, $param) 308 { 309 310 /** 311 * Just the {@link ExecutionContext::SHOW_ACTION} 312 * may be redirected 313 */ 314 $executionContext = ExecutionContext::getActualOrCreateFromEnv(); 315 if ($executionContext->getExecutingAction() !== ExecutionContext::SHOW_ACTION) { 316 return; 317 } 318 319 $urlRewrite = Site::getUrlRewrite(); 320 if ($urlRewrite == UrlRewrite::VALUE_DOKU_REWRITE) { 321 UrlRewrite::sendErrorMessage(); 322 return; 323 } 324 325 global $ID; 326 327 /** 328 * Without SQLite, this module does not work further 329 */ 330 try { 331 Sqlite::createOrGetSqlite(); 332 } catch (ExceptionSqliteNotAvailable $e) { 333 return; 334 } 335 336 $this->pageRules = new PageRules(); 337 338 339 /** 340 * Unfortunately, DOKUWIKI_STARTED is not the first event 341 * The id may have been changed by 342 * {@link action_plugin_combo_lang::load_lang()} 343 * function, that's why we check against the {@link $_REQUEST} 344 * and not the global ID 345 */ 346 $originalId = self::getOriginalIdFromRequest(); 347 348 /** 349 * Page is an existing id ? 350 */ 351 $requestedMarkupPath = MarkupPath::createMarkupFromId($ID); 352 if (FileSystems::exists($requestedMarkupPath)) { 353 354 /** 355 * If this is not the root home page 356 * and if the canonical id is the not the same (the id has changed) 357 * and if this is not a historical page (revision) 358 * redirect 359 */ 360 if ( 361 $originalId !== $requestedMarkupPath->getUrlId() // The id may have been changed 362 && $ID != Site::getIndexPageName() 363 && !isset($_REQUEST["rev"]) 364 ) { 365 /** 366 * TODO: When saving for the first time, the page is not stored in the database 367 * but that's not the case actually 368 */ 369 $databasePageRow = $requestedMarkupPath->getDatabasePage(); 370 if ($databasePageRow->exists()) { 371 /** 372 * A move may leave the database in a bad state, 373 * unfortunately (ie page is not in index, unable to update, ...) 374 * We test therefore if the database page id exists 375 */ 376 $targetPageId = $databasePageRow->getFromRow("id"); 377 $targetPath = WikiPath::createMarkupPathFromId($targetPageId); 378 if (FileSystems::exists($targetPath)) { 379 $this->executePermanentRedirect( 380 $requestedMarkupPath->getCanonicalUrl()->toAbsoluteUrlString(), 381 self::TARGET_ORIGIN_PERMALINK_EXTENDED 382 ); 383 } 384 } 385 } 386 return; 387 } 388 389 390 $identifier = $ID; 391 392 393 /** 394 * Page Id in the url 395 */ 396 $shortPageId = PageUrlPath::getShortEncodedPageIdFromUrlId($requestedMarkupPath->getPathObject()->getLastNameWithoutExtension()); 397 if ($shortPageId != null) { 398 $pageId = PageUrlPath::decodePageId($shortPageId); 399 } else { 400 /** 401 * Permalink with id 402 */ 403 $pageId = PageUrlPath::decodePageId($identifier); 404 } 405 if ($pageId !== null) { 406 407 if ($requestedMarkupPath->getParent() === null) { 408 $page = DatabasePageRow::createFromPageId($pageId)->getMarkupPath(); 409 if ($page !== null && $page->exists()) { 410 $this->executePermanentRedirect( 411 $page->getCanonicalUrl()->toAbsoluteUrlString(), 412 self::TARGET_ORIGIN_PERMALINK 413 ); 414 return; 415 } 416 } 417 418 /** 419 * Page Id Abbr ? 420 * {@link PageUrlType::CONF_CANONICAL_URL_TYPE} 421 */ 422 $page = DatabasePageRow::createFromPageIdAbbr($pageId)->getMarkupPath(); 423 if ($page === null) { 424 // or the length of the abbr has changed 425 $canonicalDatabasePage = new DatabasePageRow(); 426 $row = $canonicalDatabasePage->getDatabaseRowFromAttribute("substr(" . PageId::PROPERTY_NAME . ", 1, " . strlen($pageId) . ")", $pageId); 427 if ($row !== null) { 428 $canonicalDatabasePage->setRow($row); 429 $page = $canonicalDatabasePage->getMarkupPath(); 430 } 431 } 432 if ($page !== null && $page->exists()) { 433 /** 434 * If the url canonical id has changed, we show it 435 * to the writer by performing a permanent redirect 436 */ 437 if ($identifier != $page->getUrlId()) { 438 // Google asks for a redirect 439 // https://developers.google.com/search/docs/advanced/crawling/301-redirects 440 // People access your site through several different URLs. 441 // If, for example, your home page can be reached in multiple ways 442 // (for instance, http://example.com/home, http://home.example.com, or http://www.example.com), 443 // it's a good idea to pick one of those URLs as your preferred (canonical) destination, 444 // and use redirects to send traffic from the other URLs to your preferred URL. 445 $this->executePermanentRedirect( 446 $page->getCanonicalUrl()->toAbsoluteUrlString(), 447 self::TARGET_ORIGIN_PERMALINK_EXTENDED 448 ); 449 return; 450 } 451 452 $this->executeTransparentRedirect($page->getWikiId(), self::TARGET_ORIGIN_PERMALINK_EXTENDED); 453 return; 454 455 } 456 // permanent url not yet in the database 457 // Other permanent such as permanent canonical ? 458 // We let the process go with the new identifier 459 460 } 461 462 // Global variable needed in the process 463 global $conf; 464 465 /** 466 * Identifier is a Canonical ? 467 */ 468 $canonicalDatabasePage = DatabasePageRow::createFromCanonical($identifier); 469 $canonicalPage = $canonicalDatabasePage->getMarkupPath(); 470 if ($canonicalPage !== null && $canonicalPage->exists()) { 471 /** 472 * Does the canonical url is canonical name based 473 * ie {@link PageUrlType::CONF_VALUE_CANONICAL_PATH} 474 */ 475 if ($canonicalPage->getUrlId() === $identifier) { 476 $res = $this->executeTransparentRedirect( 477 $canonicalPage->getWikiId(), 478 self::TARGET_ORIGIN_CANONICAL 479 ); 480 } else { 481 $res = $this->executePermanentRedirect( 482 $canonicalPage->getWikiId(), // not the url because, it allows to add url query redirection property 483 self::TARGET_ORIGIN_CANONICAL 484 ); 485 } 486 if ($res) { 487 return; 488 } 489 } 490 491 /** 492 * Identifier is an alias 493 */ 494 $aliasRequestedPage = DatabasePageRow::createFromAlias($identifier)->getMarkupPath(); 495 if ( 496 $aliasRequestedPage !== null 497 && $aliasRequestedPage->exists() 498 // The build alias is the file system metadata alias 499 // it may be null if the replication in the database was not successful 500 && $aliasRequestedPage->getBuildAlias() !== null 501 ) { 502 $buildAlias = $aliasRequestedPage->getBuildAlias(); 503 switch ($buildAlias->getType()) { 504 case AliasType::REDIRECT: 505 $res = $this->executePermanentRedirect($aliasRequestedPage->getCanonicalUrl()->toAbsoluteUrlString(), self::TARGET_ORIGIN_ALIAS); 506 if ($res) { 507 return; 508 } 509 break; 510 case AliasType::SYNONYM: 511 $res = $this->executeTransparentRedirect($aliasRequestedPage->getWikiId(), self::TARGET_ORIGIN_ALIAS); 512 if ($res) { 513 return; 514 } 515 break; 516 default: 517 LogUtility::msg("The alias type ({$buildAlias->getType()}) is unknown. A permanent redirect was performed for the alias $identifier"); 518 $res = $this->executePermanentRedirect($aliasRequestedPage->getCanonicalUrl()->toAbsoluteUrlString(), self::TARGET_ORIGIN_ALIAS); 519 if ($res) { 520 return; 521 } 522 break; 523 } 524 } 525 526 527 // If there is a redirection defined in the page rules 528 $result = $this->processingPageRules(); 529 if ($result) { 530 // A redirection has occurred 531 // finish the process 532 return; 533 } 534 535 /** 536 * 537 * There was no redirection found, redirect to edit mode if writer 538 * 539 */ 540 if (Identity::isWriter() && $this->getConf(self::GO_TO_EDIT_MODE) == 1) { 541 542 $this->gotToEditMode($event); 543 // Stop here 544 return; 545 546 } 547 548 /** 549 * We are still a reader, the redirection does not exist the user is not allowed to edit the page (public of other) 550 */ 551 if ($this->getConf('ActionReaderFirst') == self::NOTHING) { 552 return; 553 } 554 555 // We are reader and their is no redirection set, we apply the algorithm 556 $readerAlgorithms = array(); 557 $readerAlgorithms[0] = $this->getConf('ActionReaderFirst'); 558 $readerAlgorithms[1] = $this->getConf('ActionReaderSecond'); 559 $readerAlgorithms[2] = $this->getConf('ActionReaderThird'); 560 561 while ( 562 ($algorithm = array_shift($readerAlgorithms)) != null 563 ) { 564 565 switch ($algorithm) { 566 567 case self::NOTHING: 568 return; 569 570 case self::GO_TO_BEST_END_PAGE_NAME: 571 572 /** 573 * @var MarkupPath $bestEndPage 574 */ 575 list($bestEndPage, $method) = RouterBestEndPage::process($requestedMarkupPath); 576 if ($bestEndPage != null && $bestEndPage->getWikiId() !== $requestedMarkupPath->getWikiId()) { 577 $res = false; 578 switch ($method) { 579 case self::REDIRECT_PERMANENT_METHOD: 580 $res = $this->executePermanentRedirect($bestEndPage->getWikiId(), self::TARGET_ORIGIN_BEST_END_PAGE_NAME); 581 break; 582 case self::REDIRECT_NOTFOUND_METHOD: 583 $res = $this->performNotFoundRedirect($bestEndPage->getWikiId(), self::TARGET_ORIGIN_BEST_END_PAGE_NAME); 584 break; 585 default: 586 LogUtility::msg("This redirection method ($method) was not expected for the redirection algorithm ($algorithm)"); 587 } 588 if ($res) { 589 // Redirection has succeeded 590 return; 591 } 592 } 593 break; 594 595 case self::GO_TO_NS_START_PAGE: 596 597 // Start page with the conf['start'] parameter 598 $startPage = getNS($identifier) . ':' . $conf['start']; 599 if (page_exists($startPage)) { 600 $res = $this->performNotFoundRedirect($startPage, self::TARGET_ORIGIN_START_PAGE); 601 if ($res) { 602 return; 603 } 604 } 605 606 // Start page with the same name than the namespace 607 $startPage = getNS($identifier) . ':' . curNS($identifier); 608 if (page_exists($startPage)) { 609 $res = $this->performNotFoundRedirect($startPage, self::TARGET_ORIGIN_START_PAGE); 610 if ($res) { 611 return; 612 } 613 } 614 break; 615 616 case self::GO_TO_BEST_PAGE_NAME: 617 618 $bestPageId = null; 619 620 $bestPage = $this->getBestPage($identifier); 621 $bestPageId = $bestPage['id']; 622 $scorePageName = $bestPage['score']; 623 624 // Get Score from a Namespace 625 $bestNamespace = $this->scoreBestNamespace($identifier); 626 $bestNamespaceId = $bestNamespace['namespace']; 627 $namespaceScore = $bestNamespace['score']; 628 629 // Compare the two score 630 if ($scorePageName > 0 or $namespaceScore > 0) { 631 if ($scorePageName > $namespaceScore) { 632 $this->performNotFoundRedirect($bestPageId, self::TARGET_ORIGIN_BEST_PAGE_NAME); 633 } else { 634 $this->performNotFoundRedirect($bestNamespaceId, self::TARGET_ORIGIN_BEST_PAGE_NAME); 635 } 636 return; 637 } 638 break; 639 640 case self::GO_TO_BEST_NAMESPACE: 641 642 $scoreNamespace = $this->scoreBestNamespace($identifier); 643 $bestNamespaceId = $scoreNamespace['namespace']; 644 $score = $scoreNamespace['score']; 645 646 if ($score > 0) { 647 $this->performNotFoundRedirect($bestNamespaceId, self::TARGET_ORIGIN_BEST_NAMESPACE); 648 return; 649 } 650 break; 651 652 case self::GO_TO_SEARCH_ENGINE: 653 654 $this->redirectToSearchEngine(); 655 656 return; 657 658 // End Switch Action 659 } 660 661 // End While Action 662 } 663 664 665 } 666 667 668 /** 669 * getBestNamespace 670 * Return a list with 'BestNamespaceId Score' 671 * @param $id 672 * @return array 673 */ 674 private 675 function scoreBestNamespace($id) 676 { 677 678 global $conf; 679 680 // Parameters 681 $pageNameSpace = getNS($id); 682 683 // If the page has an existing namespace start page take it, other search other namespace 684 $startPageNameSpace = $pageNameSpace . ":"; 685 $dateAt = ''; 686 // $startPageNameSpace will get a full path (ie with start or the namespace 687 resolve_pageid($pageNameSpace, $startPageNameSpace, $exists, $dateAt, true); 688 if (page_exists($startPageNameSpace)) { 689 $nameSpaces = array($startPageNameSpace); 690 } else { 691 $nameSpaces = ft_pageLookup($conf['start']); 692 } 693 694 // Parameters and search the best namespace 695 $pathNames = explode(':', $pageNameSpace); 696 $bestNbWordFound = 0; 697 $bestNamespaceId = ''; 698 foreach ($nameSpaces as $nameSpace) { 699 700 $nbWordFound = 0; 701 foreach ($pathNames as $pathName) { 702 if (strlen($pathName) > 2) { 703 $nbWordFound = $nbWordFound + substr_count($nameSpace, $pathName); 704 } 705 } 706 if ($nbWordFound > $bestNbWordFound) { 707 // Take only the smallest namespace 708 if (strlen($nameSpace) < strlen($bestNamespaceId) or $nbWordFound > $bestNbWordFound) { 709 $bestNbWordFound = $nbWordFound; 710 $bestNamespaceId = $nameSpace; 711 } 712 } 713 } 714 715 $startPageFactor = $this->getConf('WeightFactorForStartPage'); 716 $nameSpaceFactor = $this->getConf('WeightFactorForSameNamespace'); 717 if ($bestNbWordFound > 0) { 718 $bestNamespaceScore = $bestNbWordFound * $nameSpaceFactor + $startPageFactor; 719 } else { 720 $bestNamespaceScore = 0; 721 } 722 723 724 return array( 725 'namespace' => $bestNamespaceId, 726 'score' => $bestNamespaceScore 727 ); 728 729 } 730 731 /** 732 * @param $event 733 */ 734 private 735 function gotToEditMode(&$event) 736 { 737 global $ACT; 738 $ACT = 'edit'; 739 740 } 741 742 743 /** 744 * Redirect to an internal page ie: 745 * * on the same domain 746 * * no HTTP redirect 747 * * id rewrite 748 * @param string $targetPageId - target page id 749 * @param string $targetOriginId - the source of the target (redirect) 750 * @return bool - return true if the user has the permission and that the redirect was done 751 * @throws Exception 752 */ 753 private 754 function executeTransparentRedirect(string $targetPageId, string $targetOriginId): bool 755 { 756 /** 757 * Because we set the ID globally for the ID redirect 758 * we make sure that this is not a {@link MarkupPath} 759 * object otherwise we got an error in the {@link \ComboStrap\AnalyticsMenuItem} 760 * because the constructor takes it {@link \dokuwiki\Menu\Item\AbstractItem} 761 */ 762 if (is_object($targetPageId)) { 763 $class = get_class($targetPageId); 764 LogUtility::msg("The parameters targetPageId ($targetPageId) is an object of the class ($class) and it should be a page id"); 765 } 766 767 if (is_object($targetOriginId)) { 768 $class = get_class($targetOriginId); 769 LogUtility::msg("The parameters targetOriginId ($targetOriginId) is an object of the class ($class) and it should be a page id"); 770 } 771 772 // If the user does not have the right to see the target page 773 // don't do anything 774 if (!(Identity::isReader($targetPageId))) { 775 return false; 776 } 777 778 // Change the id 779 global $ID; 780 global $INFO; 781 $sourceId = $ID; 782 $ID = $targetPageId; 783 if (isset($_REQUEST["id"])) { 784 $_REQUEST["id"] = $targetPageId; 785 } 786 if (isset($_GET["id"])) { 787 $_GET["id"] = $targetPageId; 788 } 789 790 /** 791 * Refresh the $INFO data 792 * 793 * the info attributes are used elsewhere 794 * 'id': for the sidebar 795 * 'exist' : for the meta robot = noindex,follow, see {@link tpl_metaheaders()} 796 * 'rev' : for the edit button to be sure that the page is still the same 797 */ 798 $INFO = pageinfo(); 799 800 /** 801 * Not compatible with 802 * https://www.dokuwiki.org/config:send404 is enabled 803 * 804 * This check happens before that dokuwiki is started 805 * and send an header in doku.php 806 * 807 * We send a warning 808 */ 809 global $conf; 810 if ($conf['send404'] == true) { 811 LogUtility::msg("The <a href=\"https://www.dokuwiki.org/config:send404\">dokuwiki send404 configuration</a> is on and should be disabled when using the url manager", LogUtility::LVL_MSG_ERROR, self::CANONICAL); 812 } 813 814 // Redirection 815 $this->logRedirection($sourceId, $targetPageId, $targetOriginId, self::REDIRECT_TRANSPARENT_METHOD); 816 817 return true; 818 819 } 820 821 private function executePermanentRedirect(string $targetIdOrUrl, $targetOrigin): bool 822 { 823 return $this->executeHttpRedirect($targetIdOrUrl, $targetOrigin, self::REDIRECT_PERMANENT_METHOD); 824 } 825 826 /** 827 * The general HTTP Redirect method to an internal page 828 * where the redirection method decide which type of redirection 829 * @param string $targetIdOrUrl - a dokuwiki id or an url 830 * @param string $targetOrigin - the origin of the target (the algorithm used to get the target origin) 831 * @param string $method - the redirection method 832 */ 833 private 834 function executeHttpRedirect(string $targetIdOrUrl, string $targetOrigin, string $method): bool 835 { 836 837 global $ID; 838 839 840 // Log the redirections 841 $this->logRedirection($ID, $targetIdOrUrl, $targetOrigin, $method); 842 843 844 // An http external url ? 845 try { 846 $isHttpUrl = Url::createFromString($targetIdOrUrl)->isHttpUrl(); 847 } catch (ExceptionBadSyntax|ExceptionBadArgument $e) { 848 $isHttpUrl = false; 849 } 850 851 // If there is a bug in the isValid function for an internal url 852 // We get a loop. 853 // The Url becomes the id, the id is unknown and we do a redirect again 854 // 855 // We check then if the target starts with the base url 856 // if this is the case, it's valid 857 if (!$isHttpUrl && strpos($targetIdOrUrl, DOKU_URL) === 0) { 858 $isHttpUrl = true; 859 } 860 if ($isHttpUrl) { 861 862 // defend against HTTP Response Splitting 863 // https://owasp.org/www-community/attacks/HTTP_Response_Splitting 864 $targetUrl = stripctl($targetIdOrUrl); 865 866 } else { 867 868 869 // Explode the page ID and the anchor (#) 870 $link = explode('#', $targetIdOrUrl, 2); 871 872 $url = UrlEndpoint::createDokuUrl(); 873 874 $urlParams = []; 875 // if this is search engine redirect 876 if ($targetOrigin == self::TARGET_ORIGIN_SEARCH_ENGINE) { 877 $replacementPart = array(':', '_', '-'); 878 $query = str_replace($replacementPart, ' ', $ID); 879 $url->setQueryParameter(ExecutionContext::DO_ATTRIBUTE, ExecutionContext::SEARCH_ACTION); 880 $url->setQueryParameter("q", $query); 881 } 882 883 /** 884 * Doing a permanent redirect with a added query string 885 * create a new page url on the search engine 886 * 887 * ie 888 * http://host/page 889 * is not the same 890 * than 891 * http://host/page?whatever 892 * 893 * We can't pass query string otherwise, we get 894 * the SEO warning / error 895 * `Alternative page with proper canonical tag` 896 * 897 * Use HTTP X header for debug 898 */ 899 if ($method !== self::REDIRECT_PERMANENT_METHOD) { 900 $url->setQueryParameter(action_plugin_combo_routermessage::ORIGIN_PAGE, $ID); 901 $url->setQueryParameter(action_plugin_combo_routermessage::ORIGIN_TYPE, $targetOrigin); 902 } 903 904 $id = $link[0]; 905 $url->setQueryParameter(DokuwikiId::DOKUWIKI_ID_ATTRIBUTE, $id); 906 if (array_key_exists(1, $link)) { 907 $url->setFragment($link[1]); 908 } 909 $targetUrl = $url->toAbsoluteUrlString(); 910 911 } 912 913 /** 914 * The dokuwiki function {@link send_redirect()} 915 * set the `Location header` and in php, the header function 916 * in this case change the status code to 302 Arghhhh. 917 * The code below is adapted from this function {@link send_redirect()} 918 */ 919 global $MSG; // are there any undisplayed messages? keep them in session for display 920 if (isset($MSG) && count($MSG) && !defined('NOSESSION')) { 921 //reopen session, store data and close session again 922 @session_start(); 923 $_SESSION[DOKU_COOKIE]['msg'] = $MSG; 924 } 925 session_write_close(); // always close the session 926 927 switch ($method) { 928 929 case self::REDIRECT_PERMANENT_METHOD: 930 ExecutionContext::getActualOrCreateFromEnv() 931 ->response() 932 ->setStatus(HttpResponseStatus::PERMANENT_REDIRECT) 933 ->addHeader(self::LOCATION_HEADER_PREFIX . $targetUrl) 934 ->end(); 935 return true; 936 937 case self::REDIRECT_NOTFOUND_METHOD: 938 939 940 // Empty 404 body to not get the standard 404 page of the browser 941 // but a blank page to avoid a sort of FOUC. 942 // ie the user see a page briefly 943 ExecutionContext::getActualOrCreateFromEnv() 944 ->response() 945 ->setStatus(HttpResponseStatus::NOT_FOUND) 946 ->addHeader(self::REFRESH_HEADER_PREFIX . $targetUrl) 947 ->setBody(self::PAGE_404, Mime::getHtml()) 948 ->end(); 949 return true; 950 951 default: 952 LogUtility::msg("The method ($method) is not an http redirection"); 953 return false; 954 } 955 956 957 } 958 959 /** 960 * @param $id 961 * @return array 962 */ 963 private 964 function getBestPage($id): array 965 { 966 967 // The return parameters 968 $bestPageId = null; 969 $scorePageName = null; 970 971 // Get Score from a page 972 $pageName = noNS($id); 973 $pagesWithSameName = ft_pageLookup($pageName); 974 if (count($pagesWithSameName) > 0) { 975 976 // Search same namespace in the page found than in the Id page asked. 977 $bestNbWordFound = 0; 978 979 980 $wordsInPageSourceId = explode(':', $id); 981 foreach ($pagesWithSameName as $targetPageId => $title) { 982 983 // Nb of word found in the target page id 984 // that are in the source page id 985 $nbWordFound = 0; 986 foreach ($wordsInPageSourceId as $word) { 987 $nbWordFound = $nbWordFound + substr_count($targetPageId, $word); 988 } 989 990 if ($bestPageId == null) { 991 992 $bestNbWordFound = $nbWordFound; 993 $bestPageId = $targetPageId; 994 995 } else { 996 997 if ($nbWordFound >= $bestNbWordFound && strlen($bestPageId) > strlen($targetPageId)) { 998 999 $bestNbWordFound = $nbWordFound; 1000 $bestPageId = $targetPageId; 1001 1002 } 1003 1004 } 1005 1006 } 1007 $scorePageName = $this->getConf('WeightFactorForSamePageName') + ($bestNbWordFound - 1) * $this->getConf('WeightFactorForSameNamespace'); 1008 return array( 1009 'id' => $bestPageId, 1010 'score' => $scorePageName); 1011 } 1012 return array( 1013 'id' => $bestPageId, 1014 'score' => $scorePageName 1015 ); 1016 1017 } 1018 1019 1020 /** 1021 * Redirect to the search engine 1022 */ 1023 private 1024 function redirectToSearchEngine() 1025 { 1026 1027 global $ID; 1028 $this->performNotFoundRedirect($ID, self::TARGET_ORIGIN_SEARCH_ENGINE); 1029 1030 } 1031 1032 1033 /** 1034 * 1035 * * For a conf file, it will update the Redirection Action Data as Referrer, Count Of Redirection, Redirection Date 1036 * * For a SQlite database, it will add a row into the log 1037 * 1038 * @param string $sourcePageId 1039 * @param $targetPageId 1040 * @param $algorithmic 1041 * @param $method - http or rewrite 1042 */ 1043 function logRedirection(string $sourcePageId, $targetPageId, $algorithmic, $method) 1044 { 1045 1046 $row = array( 1047 "TIMESTAMP" => date("c"), 1048 "SOURCE" => $sourcePageId, 1049 "TARGET" => $targetPageId, 1050 "REFERRER" => $_SERVER['HTTP_REFERER'] ?? null, 1051 "TYPE" => $algorithmic, 1052 "METHOD" => $method 1053 ); 1054 $request = Sqlite::createOrGetBackendSqlite() 1055 ->createRequest() 1056 ->setTableRow('redirections_log', $row); 1057 try { 1058 $request 1059 ->execute(); 1060 } catch (ExceptionCompile $e) { 1061 LogUtility::msg("Redirection Log Insert Error. {$e->getMessage()}"); 1062 } finally { 1063 $request->close(); 1064 } 1065 1066 1067 } 1068 1069 /** 1070 * This function check if there is a redirection declared 1071 * in the redirection table 1072 * @return bool - true if a rewrite or redirection occurs 1073 * @throws Exception 1074 */ 1075 private function processingPageRules(): bool 1076 { 1077 global $ID; 1078 1079 $calculatedTarget = null; 1080 $ruleMatcher = null; // Used in a warning message if the target page does not exist 1081 // Known redirection in the table 1082 // Get the page from redirection data 1083 $rules = $this->pageRules->getRules(); 1084 foreach ($rules as $rule) { 1085 1086 $ruleMatcher = strtolower($rule[PageRules::MATCHER_NAME]); 1087 $ruleTarget = $rule[PageRules::TARGET_NAME]; 1088 1089 // Glob to Rexgexp 1090 $regexpPattern = '/' . str_replace("*", "(.*)", $ruleMatcher) . '/i'; 1091 1092 // Match ? 1093 // https://www.php.net/manual/en/function.preg-match.php 1094 $pregMatchResult = @preg_match($regexpPattern, $ID, $matches); 1095 if ($pregMatchResult === false) { 1096 // The `if` to take into account this problem 1097 // PHP Warning: preg_match(): Unknown modifier 'd' in /opt/www/datacadamia.com/lib/plugins/combo/action/router.php on line 972 1098 LogUtility::log2file("processing Page Rules An error occurred with the pattern ($regexpPattern)", LogUtility::LVL_MSG_WARNING); 1099 return false; 1100 } 1101 if ($pregMatchResult) { 1102 $calculatedTarget = $ruleTarget; 1103 foreach ($matches as $key => $match) { 1104 if ($key == 0) { 1105 continue; 1106 } else { 1107 $calculatedTarget = str_replace('$' . $key, $match, $calculatedTarget); 1108 } 1109 } 1110 break; 1111 } 1112 } 1113 1114 if ($calculatedTarget == null) { 1115 return false; 1116 } 1117 1118 // If this is an external redirect (other domain) 1119 try { 1120 $isHttpUrl = Url::createFromString($calculatedTarget)->isHttpUrl(); 1121 } catch (ExceptionBadSyntax $e) { 1122 $isHttpUrl = false; 1123 } 1124 if ($isHttpUrl) { 1125 $this->executeHttpRedirect($calculatedTarget, self::TARGET_ORIGIN_PAGE_RULES, self::REDIRECT_PERMANENT_METHOD); 1126 return true; 1127 } 1128 1129 // If the page exist 1130 if (page_exists($calculatedTarget)) { 1131 1132 // This is DokuWiki Id and should always be lowercase 1133 // The page rule may have change that 1134 $calculatedTarget = strtolower($calculatedTarget); 1135 $res = $this->executeHttpRedirect($calculatedTarget, self::TARGET_ORIGIN_PAGE_RULES, self::REDIRECT_PERMANENT_METHOD); 1136 if ($res) { 1137 return true; 1138 } else { 1139 return false; 1140 } 1141 1142 } else { 1143 1144 LogUtility::msg("The calculated target page ($calculatedTarget) (for the non-existing page `$ID` with the matcher `$ruleMatcher`) does not exist", LogUtility::LVL_MSG_ERROR); 1145 return false; 1146 1147 } 1148 1149 } 1150 1151 private function performNotFoundRedirect(string $targetId, string $origin): bool 1152 { 1153 return $this->executeHttpRedirect($targetId, $origin, self::REDIRECT_NOTFOUND_METHOD); 1154 } 1155 1156 1157} 1158