[BUGFIX] Fix indexing of access protected pages

Build/Test/phpstan.neon

@@ -20,6 +20,5 @@ parameters:
     - '#^Variable \$_EXTKEY might not be defined\.#'
     - '#^Method .*\\QueryBuilder::buildSuggestQuery\(\) should return .*\\SuggestQuery but returns .*\\Search\\Query\\Query\|null.#'
     - '#^Method .*\\QueryBuilder::use.*FromTypoScript\(\) should return .*\\QueryBuilder but returns .*\\AbstractQueryBuilder.#'
-    - '#^Access to protected property .*\\PageRepository::\$where_groupAccess.#'
     - '#^Unreachable statement - code above always terminates.#'
     - '#^Parameter \#1 \$queryAlternative of method .*\\DisMax::setQueryAlternative\(\) expects string, null given.#'

Classes/EventListener/PageIndexer/FrontendGroupsModifier.php

@@ -40,9 +40,13 @@ class FrontendGroupsModifier
     public function __invoke(ModifyResolvedFrontendGroupsEvent $event): void
     {
         $pageIndexerRequest = $event->getRequest()->getAttribute('solr.pageIndexingInstructions');
-        if (!$pageIndexerRequest instanceof PageIndexerRequest) {
+        if (!$pageIndexerRequest instanceof PageIndexerRequest
+            || ((int)$pageIndexerRequest->getParameter('userGroup') === 0
+                && (int)$pageIndexerRequest->getParameter('pageUserGroup') < 1)
+        ) {
             return;
         }
+
         if (!$pageIndexerRequest->isAuthenticated()) {
             $logger = GeneralUtility::makeInstance(SolrLogManager::class, self::class);
             $logger->error(
@@ -58,7 +62,7 @@ public function __invoke(ModifyResolvedFrontendGroupsEvent $event): void
                         'error' => [
                             'code' => 403,
                             'message' => 'Invalid Index Queue Request.',
-                            ],
+                        ],
                     ],
                     403
                 ),
@@ -67,6 +71,9 @@ public function __invoke(ModifyResolvedFrontendGroupsEvent $event): void
         }
 
         $groups = $this->resolveFrontendUserGroups($pageIndexerRequest);
+        if ((int)$pageIndexerRequest->getParameter('pageUserGroup') > 0) {
+            $groups[] = (int)$pageIndexerRequest->getParameter('pageUserGroup');
+        }
         $groupData = [];
         foreach ($groups as $groupUid) {
             if (in_array($groupUid, [-2, -1])) {

Classes/IndexQueue/FrontendHelper/PageIndexer.php

@@ -54,6 +54,8 @@
  */
 class PageIndexer implements FrontendHelper, SingletonInterface
 {
+    protected bool $activated = false;
+
     /**
      * This frontend helper's executed action.
      */
@@ -92,6 +94,7 @@ class PageIndexer implements FrontendHelper, SingletonInterface
     public function activate(): void
     {
         $this->logger = GeneralUtility::makeInstance(SolrLogManager::class, __CLASS__);
+        $this->activated = true;
     }
 
     /**
@@ -141,7 +144,8 @@ protected function generatePageUrl(TypoScriptFrontendController $controller): st
     public function __invoke(AfterCacheableContentIsGeneratedEvent $event): void
     {
         $this->request = $event->getRequest()->getAttribute('solr.pageIndexingInstructions');
-        if (!$this->request) {
+
+        if (!$this->request || $this->activated === false) {
             return;
         }
         $this->setupConfiguration();
@@ -188,9 +192,9 @@ public function __invoke(AfterCacheableContentIsGeneratedEvent $event): void
     }
 
     /**
-     * @internal currently only public for tests
+     * Index item
      */
-    public function index(Item $indexQueueItem, TypoScriptFrontendController $tsfe): void
+    protected function index(Item $indexQueueItem, TypoScriptFrontendController $tsfe): void
     {
         $this->solrConnection = $this->getSolrConnection($indexQueueItem, $tsfe->getLanguage(), $this->configuration->getLoggingExceptions());
 
@@ -287,7 +291,7 @@ protected function getPageDocument(TypoScriptFrontendController $tsfe, string $u
      *
      * @return bool TRUE after successfully indexing the page, FALSE on error
      */
-    public function indexPage(
+    protected function indexPage(
         Document $pageDocument,
         Item $indexQueueItem,
         TypoScriptFrontendController $tsfe,
@@ -362,11 +366,13 @@ protected function addDocumentsToSolrIndex(array $documents): bool
     }
 
     /**
-     * @internal only used for tests
+     * Initialize PageIndexer
+     *
+     * As the Solr configuration initialization might affect the request
+     * we cannot initialize the configuration directly on activation
      */
-    public function setupConfiguration(): void
+    protected function setupConfiguration(): void
     {
-        // currently needed for tests, will be separated.
         $this->logger = new SolrLogManager(__CLASS__, GeneralUtility::makeInstance(DebugWriter::class));
         $this->configuration = Util::getSolrConfiguration();
     }
@@ -381,6 +387,19 @@ protected function getEventDispatcher(): EventDispatcherInterface
      */
     public function deactivate(PageIndexerResponse $response): void
     {
-        $response->addActionResult($this->action, $this->responseData);
+        if ($this->activated) {
+            if (!isset($this->responseData['pageIndexed'])) {
+                $this->responseData['pageIndexed'] = false;
+            }
+            $response->addActionResult($this->action, $this->responseData);
+
+            $this->setupConfiguration();
+            if ($this->configuration->getLoggingExceptions()) {
+                $this->logger->error(
+                    'Unknown exception while trying to index page',
+                );
+            }
+        }
+        $this->activated = false;
     }
 }

Classes/IndexQueue/FrontendHelper/UserGroupDetector.php

@@ -124,17 +124,13 @@ public function deactivateTcaFrontendGroupEnableFields(
      * @param int $uid The page ID
      * @param bool $disableGroupAccessCheck If set, the check for group access is disabled. VERY rarely used
      * @param PageRepository $parentObject parent \TYPO3\CMS\Core\Domain\Repository\PageRepository object
-     *
-     * @noinspection PhpMissingReturnTypeInspection
      */
     public function getPage_preProcess(
         &$uid,
         &$disableGroupAccessCheck,
         PageRepository $parentObject
-    ) {
+    ): void {
         $disableGroupAccessCheck = true;
-        // @todo: Check if reset "where_groupAccess" really wanted. Most probably the core aspect 'frontend.user' must be used instead.
-        $parentObject->where_groupAccess = ''; // just to be on the safe side
     }
 
     /**
@@ -190,7 +186,7 @@ protected function findFrontendGroups(array $record, string $table): void
         if (isset($this->originalTca[$table]['ctrl']['enablecolumns']['fe_group'])) {
             $frontendGroups = $record[$this->originalTca[$table]['ctrl']['enablecolumns']['fe_group']] ?? null;
 
-            if (empty($frontendGroups)) {
+            if (empty($frontendGroups) || $frontendGroups === '-1') {
                 // default = public access
                 $frontendGroups = 0;
             } elseif ($this->request->getParameter('loggingEnabled')) {
@@ -222,7 +218,10 @@ protected function getFrontendGroups(): array
 
         // clean up: filter double groups
         $frontendGroups = array_unique($frontendGroups);
-        $frontendGroups = array_values($frontendGroups);
+        $frontendGroups = array_filter(
+            array_values($frontendGroups),
+            static fn (int $val): bool => ($val !== -1)
+        );
 
         if (empty($frontendGroups)) {
             // most likely an empty page with no content elements => public

Classes/IndexQueue/PageIndexer.php

@@ -59,16 +59,9 @@ public function index(Item $item): bool
             return false;
         }
 
-        $solrConnections = $this->getSolrConnectionsByItem($item);
-        foreach ($solrConnections as $systemLanguageUid => $solrConnection) {
+        $systemLanguageUids = array_keys($this->getSolrConnectionsByItem($item));
+        foreach ($systemLanguageUids as $systemLanguageUid) {
             $contentAccessGroups = $this->getAccessGroupsFromContent($item, $systemLanguageUid);
-
-            if (empty($contentAccessGroups)) {
-                // might be an empty page w/no content elements or some TYPO3 error / bug
-                // FIXME logging needed
-                continue;
-            }
-
             foreach ($contentAccessGroups as $userGroup) {
                 $this->indexPage($item, $systemLanguageUid, $userGroup);
             }
@@ -299,6 +292,12 @@ protected function indexPage(Item $item, ?int $language = 0, ?int $userGroup = 0
         $request->setIndexQueueItem($item);
         $request->addAction('indexPage');
         $request->setParameter('accessRootline', $accessRootline);
+        $request->setParameter('userGroup', $userGroup);
+
+        $feGroupColumn = $GLOBALS['TCA']['pages']['ctrl']['enablecolumns']['fe_group'] ?? '';
+        if (!empty($feGroupColumn)) {
+            $request->setParameter('pageUserGroup', $item->getRecord()[$feGroupColumn]);
+        }
 
         $indexRequestUrl = $this->getDataUrl($item, $language);
         $response = $request->send($indexRequestUrl);
@@ -329,7 +328,6 @@ protected function indexPage(Item $item, ?int $language = 0, ?int $userGroup = 0
 
         if (empty($indexActionResult['pageIndexed'])) {
             $message = 'Failed indexing page Index Queue item: ' . $item->getIndexQueueUid() . ' url: ' . $indexRequestUrl;
-
             throw new \RuntimeException($message, 1331837081);
         }
 

Configuration/RequestMiddlewares.php

@@ -4,7 +4,7 @@
 $requestMiddlewares = [
     'apache-solr-for-typo3/page-indexer-initialization' => [
         'target' => \ApacheSolrForTypo3\Solr\Middleware\PageIndexerInitialization::class,
-        'before' => ['typo3/cms-frontend/tsfe'],
+        'before' => ['typo3/cms-frontend/tsfe', 'typo3/cms-frontend/authentication'],
         'after' => ['typo3/cms-core/normalized-params-attribute'],
     ],
 ];

Documentation/Releases/solr-release-12-0.rst

@@ -158,6 +158,14 @@ If you've used this class or the SolrConnection directly, you have to adapt your
 
 Note: With dropping the Node implementation we also dropped the backwards compatibility that allows to define the Solr path segment "/solr" within "solr_path_read" or "solr_path_write". Be sure your configuration doesn't contain this path segment!
 
+!!! Changed visibility of ApacheSolrForTypo3\Solr\IndexQueue\FrontendHelper\PageIndexer methods
+-----------------------------------------------------------------------------------------------
+
+For testing purposes some methods of the PageIndexer were defined as public, these methods are now protected. The tests are adapted accordingly, so that there is no need to declare the methods as public.
+If you have used one of this methods, you have to adapt your code. Affected methods:
+- setupConfiguration
+- index
+- indexPage
 
 !!! Solr route enhancer disabled by default
 -------------------------------------------

Tests/Integration/Fixtures/Search/can_search.csv

@@ -1,50 +1,6 @@
-"pages",,,,,,
-,"uid","is_siteroot","doktype","title","slug","hidden"
-,1,1,1,"Hello Search Test","/",0
-"sys_template",,,,,,
-,"uid","pid","root","clear","sorting","config",
-,1,1,1,3,100,"
-page = PAGE
-page.typeNum = 0
-
-plugin.tx_solr {
-
-    enabled = 1
-
-    index {
-        fieldProcessingInstructions {
-            changed = timestampToIsoDate
-            created = timestampToIsoDate
-            endtime = timestampToUtcIsoDate
-            rootline = pageUidToHierarchy
-        }
-
-        queue {
-
-            // mapping tableName.fields.SolrFieldName => TableFieldName (+ cObj processing)
-
-            pages = 1
-            pages {
-                initialization = ApacheSolrForTypo3\Solr\IndexQueue\Initializer\Page
-
-                // allowed page types (doktype) when indexing pages
-                allowedPageTypes = 1,7
-
-                indexingPriority = 0
-
-                indexer = ApacheSolrForTypo3\Solr\IndexQueue\PageIndexer
-                indexer {
-                    // add options for the indexer here
-                }
-
-                // Only index standard pages and mount points that are not overlayed.
-                additionalWhereClause = (doktype = 1 OR (doktype=7 AND mount_pid_ol=0)) AND no_search = 0
-
-                fields {
-                    sortSubTitle_stringS = subtitle
-                }
-            }
-
-        }
-    }
-}"
+"pages",
+,"uid","pid","is_siteroot","doktype","slug","title","subtitle","crdate","tstamp"
+,2,1,1,1,"/hello_solr","Hello Search Test","the subtitle",1449151778,1449151778
+"tx_solr_indexqueue_item",
+,"uid","root","item_type","item_uid","indexing_configuration","changed","indexed","has_indexing_properties","indexing_priority","indexed","errors"
+,2,1,"pages",2,"pages",1449151778,0,0,0,0,0

Tests/Integration/Fixtures/Search/phrase_search.csv

@@ -1,64 +1,34 @@
 "pages",,,,,,,
 ,"uid","pid","is_siteroot","doktype","hidden","slug","title"
-,1,0,1,1,0,"/","World Hello Search Test"
-,2,1,0,1,0,"/improve","Hello you can improve the precision by using phrase configuration World"
-,3,1,0,1,0,"/hello-phrase-search-world","Hello phrase search World"
-,4,1,0,1,0,"/hello-test-wonderful-world","Hello Test wonderful World"
-,5,1,0,1,0,"/hello-wonderful-test-world","Hello wonderful Test World"
-,6,1,0,1,0,"/bigram-phrases","Bigraming phrases strip down the sentence to the two-word combinations."
-,7,1,0,1,0,"/trigram-phrases","Trigraming phrases strip down the sentence to triplets phrases"
-,8,1,0,1,0,"/difference-to-bi-is-tri","Difference to bi is tri, which means building of triplets phrases."
-,9,1,0,1,0,"/hello-solr-wolrd","Hello Solr Wolrd"
-,10,1,0,1,0,"/hello-test-wolrd","Hello Test Wolrd"
-,11,1,0,1,0,"/test-hello-solr-world","Test Hello Solr World"
-,12,1,0,1,0,"/test-hello-search-world","Test Hello Search World"
-,13,1,0,1,0,"/hello-the-test-search-world","Hello the Test Search World"
-,14,1,0,1,0,"/hello-world-for-phrase-searching","Hello World for phrase searching"
-,15,1,0,1,0,"/about","Solr is blazing-fast and open source enterprise search platform built on Apache Lucene"
-"sys_template",,,,,,,
-,"uid","pid","root","clear","sorting","constants","config",
-,1,1,1,3,100,"","
-page = PAGE
-page.typeNum = 0
-
-plugin.tx_solr {
-    search.query.phrase = 1
-    enabled = 1
-
-    index {
-        fieldProcessingInstructions {
-            changed = timestampToIsoDate
-            created = timestampToIsoDate
-            endtime = timestampToUtcIsoDate
-            rootline = pageUidToHierarchy
-        }
-
-        queue {
-
-            // mapping tableName.fields.SolrFieldName => TableFieldName (+ cObj processing)
-
-            pages = 1
-            pages {
-                initialization = ApacheSolrForTypo3\Solr\IndexQueue\Initializer\Page
-
-                // allowed page types (doktype) when indexing pages
-                allowedPageTypes = 1,7
-
-                indexingPriority = 0
-
-                indexer = ApacheSolrForTypo3\Solr\IndexQueue\PageIndexer
-                indexer {
-                    // add options for the indexer here
-                }
-
-                // Only index standard pages and mount points that are not overlayed.
-                additionalWhereClause = (doktype = 1 OR (doktype=7 AND mount_pid_ol=0)) AND no_search = 0
-
-                fields {
-                    sortSubTitle_stringS = subtitle
-                }
-            }
-
-        }
-    }
-}",
+,2,1,1,1,0,"/","World Hello Search Test"
+,3,1,0,1,0,"/improve","Hello you can improve the precision by using phrase configuration World"
+,4,1,0,1,0,"/hello-phrase-search-world","Hello phrase search World"
+,5,1,0,1,0,"/hello-test-wonderful-world","Hello Test wonderful World"
+,6,1,0,1,0,"/hello-wonderful-test-world","Hello wonderful Test World"
+,7,1,0,1,0,"/bigram-phrases","Bigraming phrases strip down the sentence to the two-word combinations."
+,8,1,0,1,0,"/trigram-phrases","Trigraming phrases strip down the sentence to triplets phrases"
+,9,1,0,1,0,"/difference-to-bi-is-tri","Difference to bi is tri, which means building of triplets phrases."
+,10,1,0,1,0,"/hello-solr-wolrd","Hello Solr Wolrd"
+,11,1,0,1,0,"/hello-test-wolrd","Hello Test Wolrd"
+,12,1,0,1,0,"/test-hello-solr-world","Test Hello Solr World"
+,13,1,0,1,0,"/test-hello-search-world","Test Hello Search World"
+,14,1,0,1,0,"/hello-the-test-search-world","Hello the Test Search World"
+,15,1,0,1,0,"/hello-world-for-phrase-searching","Hello World for phrase searching"
+,16,1,0,1,0,"/about","Solr is blazing-fast and open source enterprise search platform built on Apache Lucene"
+"tx_solr_indexqueue_item",
+,"uid","root","item_type","item_uid","indexing_configuration","changed","indexed","has_indexing_properties","indexing_priority","indexed","errors"
+,2,1,"pages",2,"pages",1449151778,0,0,0,0,0
+,3,1,"pages",3,"pages",1449151778,0,0,0,0,0
+,4,1,"pages",4,"pages",1449151778,0,0,0,0,0
+,5,1,"pages",5,"pages",1449151778,0,0,0,0,0
+,6,1,"pages",6,"pages",1449151778,0,0,0,0,0
+,7,1,"pages",7,"pages",1449151778,0,0,0,0,0
+,8,1,"pages",8,"pages",1449151778,0,0,0,0,0
+,9,1,"pages",9,"pages",1449151778,0,0,0,0,0
+,10,1,"pages",10,"pages",1449151778,0,0,0,0,0
+,11,1,"pages",11,"pages",1449151778,0,0,0,0,0
+,12,1,"pages",12,"pages",1449151778,0,0,0,0,0
+,13,1,"pages",13,"pages",1449151778,0,0,0,0,0
+,14,1,"pages",14,"pages",1449151778,0,0,0,0,0
+,15,1,"pages",15,"pages",1449151778,0,0,0,0,0
+,16,1,"pages",16,"pages",1449151778,0,0,0,0,0