Index PDF and Office file contents via Apache Tika

Solr 10 dropped the ExtractingRequestHandler, so a dedicated Tika
container on the Solr VPS (Caddy route /tika/*, own basic-auth
credential; Tika itself has no auth) extracts file text during
indexing: TikaDownloadContentIndexer listens on
BeforeDocumentIsProcessedForIndexingEvent, resolves the file through
the existing DownloadFileResolver and appends the text (mime
whitelist, 30 MB cap, 100k chars, fail-soft) to the document's content
field. Extractions are cached in var/tika-cache keyed on
path+size+mtime - a full re-index of 177 downloads drops from 2:12 min
to 27 s, replacing a file re-extracts naturally. Datasheet
specifications ("625i", "genlock") are now searchable. Spec v1.15.

packages/vitec/Classes/EventListener/TikaDownloadContentIndexer.php

@@ -0,0 +1,127 @@
+<?php
+
+declare(strict_types=1);
+
+namespace Evomedien\Vitec\EventListener;
+
+use ApacheSolrForTypo3\Solr\Event\Indexing\BeforeDocumentIsProcessedForIndexingEvent;
+use Evomedien\Vitec\Service\DownloadFileResolver;
+use TYPO3\CMS\Core\Attribute\AsEventListener;
+use TYPO3\CMS\Core\Core\Environment;
+use TYPO3\CMS\Core\Utility\GeneralUtility;
+
+/**
+ * Adds the extracted file text (PDF/Office) to the Solr `content` field of
+ * download documents. Solr 10 dropped the built-in ExtractingRequestHandler,
+ * so extraction goes through the Apache Tika server on the Solr VPS
+ * (Caddy route /tika/*, credentials via TIKA_* env in additional.php).
+ *
+ * Extracted text is cached in var/tika-cache keyed on path+size+mtime -
+ * a full re-index does not re-upload 177 unchanged PDFs, replacing a file
+ * invalidates its entry naturally. Fail-soft throughout (Clause 9.5): no
+ * Tika, no file, or an extraction error leave the document as it was -
+ * title/teaser/keywords still get it indexed.
+ */
+#[AsEventListener(identifier: 'vitec/tika-download-content')]
+final class TikaDownloadContentIndexer
+{
+    /** Files above this size are not sent to Tika (bytes). */
+    private const MAX_FILE_SIZE = 31457280;
+
+    /** Extracted text is truncated to this many characters before indexing. */
+    private const MAX_TEXT_LENGTH = 100000;
+
+    private const EXTRACTABLE_MIME_TYPES = [
+        'application/pdf',
+        'application/msword',
+        'application/vnd.openxmlformats-officedocument.wordprocessingml.document',
+        'application/vnd.ms-powerpoint',
+        'application/vnd.openxmlformats-officedocument.presentationml.presentation',
+        'application/vnd.ms-excel',
+        'application/vnd.openxmlformats-officedocument.spreadsheetml.sheet',
+    ];
+
+    public function __invoke(BeforeDocumentIsProcessedForIndexingEvent $event): void
+    {
+        try {
+            $item = $event->getIndexQueueItem();
+            if ($item->getType() !== 'tx_vitec_domain_model_download') {
+                return;
+            }
+
+            $file = DownloadFileResolver::resolve($item->getRecordUid());
+            if (
+                $file === null
+                || $file['size'] <= 0
+                || $file['size'] > self::MAX_FILE_SIZE
+                || !in_array($file['mimeType'], self::EXTRACTABLE_MIME_TYPES, true)
+            ) {
+                return;
+            }
+
+            $text = $this->extract($file['path']);
+            if ($text === '') {
+                return;
+            }
+
+            $document = $event->getDocument();
+            $existing = trim((string)($document['content'] ?? ''));
+            $document->setField('content', trim($existing . ' ' . $text));
+        } catch (\Throwable) {
+            // fail soft - the download stays indexed with its metadata
+        }
+    }
+
+    private function extract(string $path): string
+    {
+        $cacheFile = $this->cacheFileFor($path);
+        if ($cacheFile !== '' && is_file($cacheFile)) {
+            return (string)file_get_contents($cacheFile);
+        }
+
+        $baseUrl = (string)getenv('TIKA_URL');
+        if ($baseUrl === '') {
+            return '';
+        }
+
+        $context = stream_context_create(['http' => [
+            'method' => 'PUT',
+            'header' => [
+                'Authorization: Basic ' . base64_encode(getenv('TIKA_USERNAME') . ':' . getenv('TIKA_PASSWORD')),
+                'Accept: text/plain',
+                'Content-Type: application/octet-stream',
+            ],
+            'content' => (string)file_get_contents($path),
+            'timeout' => 30,
+            'ignore_errors' => true,
+        ]]);
+        $raw = file_get_contents(rtrim($baseUrl, '/') . '/tika', false, $context);
+        $status = (string)($http_response_header[0] ?? '');
+        if ($raw === false || !str_contains($status, '200')) {
+            return '';
+        }
+
+        $text = mb_substr(trim((string)preg_replace('/\s+/u', ' ', $raw)), 0, self::MAX_TEXT_LENGTH);
+
+        if ($cacheFile !== '' && $text !== '') {
+            GeneralUtility::mkdir_deep(dirname($cacheFile));
+            GeneralUtility::writeFile($cacheFile, $text);
+        }
+
+        return $text;
+    }
+
+    /**
+     * Cache key covers path, size and mtime - replacing a file re-extracts.
+     */
+    private function cacheFileFor(string $path): string
+    {
+        $size = @filesize($path);
+        $mtime = @filemtime($path);
+        if ($size === false || $mtime === false) {
+            return '';
+        }
+        return Environment::getVarPath() . '/tika-cache/'
+            . sha1($path . '|' . $size . '|' . $mtime) . '.txt';
+    }
+}

packages/vitec/Classes/UserFunc/ProductFinderJsonRenderer.php

@@ -0,0 +1,150 @@
+<?php
+
+declare(strict_types=1);
+
+namespace Evomedien\Vitec\UserFunc;
+
+use Doctrine\DBAL\ParameterType;
+use TYPO3\CMS\Core\Attribute\AsAllowedCallable;
+use TYPO3\CMS\Core\Database\ConnectionPool;
+use TYPO3\CMS\Core\Database\Query\Restriction\FrontendRestrictionContainer;
+use TYPO3\CMS\Core\Service\FlexFormService;
+use TYPO3\CMS\Core\Utility\GeneralUtility;
+use TYPO3\CMS\Frontend\ContentObject\ContentObjectRenderer;
+
+/**
+ * Data source for the product filter: the product category hierarchy, and
+ * nothing else.
+ *
+ * The editor places ONE `vitec_productfinder` element above the product
+ * lists on the products page. It renders no product cards - the React front
+ * end builds the filter selects from this payload and filters the products
+ * the `vitec_productlist` elements already delivered, client side, via URL
+ * parameters (?cat=58&subcat=66). No change to the list plugin was needed.
+ *
+ * Emitted shape:
+ *
+ *   content.productfinder.categories[] = { uid, title, subcategories[] }
+ *   subcategories[]                    = { uid, title }
+ *
+ * Level 1 below the configured root becomes the main categories, their
+ * direct children the subcategories - the same tree the product records
+ * hang on, so a filter value maps straight onto `product.categories[].uid`.
+ *
+ * TYPO3 v14 hands the ContentObjectRenderer in through the setter; without
+ * it $this->cObj stays null and the element cannot read its own FlexForm.
+ */
+final class ProductFinderJsonRenderer
+{
+    private const CATEGORY_TABLE = 'sys_category';
+
+    /** Fallback when no root is configured: the category tree the products use. */
+    private const DEFAULT_ROOT_TITLE = 'Product';
+
+    protected ?ContentObjectRenderer $cObj = null;
+
+    public function setContentObjectRenderer(ContentObjectRenderer $cObj): void
+    {
+        $this->cObj = $cObj;
+    }
+
+    #[AsAllowedCallable]
+    public function render(string $content, array $conf): string
+    {
+        try {
+            $settings = $this->settings();
+
+            $root = (int)($settings['categoryRoot'] ?? 0);
+            if ($root <= 0) {
+                $root = $this->rootByTitle(self::DEFAULT_ROOT_TITLE);
+            }
+            if ($root <= 0) {
+                return (string)json_encode(['categories' => []]);
+            }
+
+            $order = (string)($settings['sorting'] ?? 'sorting');
+            $categories = [];
+            foreach ($this->childrenOf($root, $order) as $main) {
+                $categories[] = [
+                    'uid' => (int)$main['uid'],
+                    'title' => (string)$main['title'],
+                    'subcategories' => array_map(
+                        static fn(array $sub): array => [
+                            'uid' => (int)$sub['uid'],
+                            'title' => (string)$sub['title'],
+                        ],
+                        $this->childrenOf((int)$main['uid'], $order)
+                    ),
+                ];
+            }
+
+            return (string)json_encode(['categories' => $categories], JSON_UNESCAPED_SLASHES | JSON_UNESCAPED_UNICODE);
+        } catch (\Throwable) {
+            // A broken filter must never take the page payload down with it.
+            return (string)json_encode(['categories' => []]);
+        }
+    }
+
+    /**
+     * FlexForm settings of the element being rendered.
+     *
+     * @return array<string,mixed>
+     */
+    private function settings(): array
+    {
+        $flexform = (string)($this->cObj->data['pi_flexform'] ?? '');
+        if ($flexform === '') {
+            return [];
+        }
+        $parsed = GeneralUtility::makeInstance(FlexFormService::class)
+            ->convertFlexFormContentToArray($flexform);
+
+        return is_array($parsed['settings'] ?? null) ? $parsed['settings'] : [];
+    }
+
+    /**
+     * Direct children of a category, honouring the frontend restrictions
+     * (hidden, start/end time, deleted).
+     *
+     * @return array<int,array<string,mixed>>
+     */
+    private function childrenOf(int $parent, string $order): array
+    {
+        $queryBuilder = GeneralUtility::makeInstance(ConnectionPool::class)
+            ->getQueryBuilderForTable(self::CATEGORY_TABLE);
+        $queryBuilder->setRestrictions(GeneralUtility::makeInstance(FrontendRestrictionContainer::class));
+
+        $query = $queryBuilder->select('uid', 'title', 'sorting')
+            ->from(self::CATEGORY_TABLE)
+            ->where(
+                $queryBuilder->expr()->eq('parent', $queryBuilder->createNamedParameter($parent, ParameterType::INTEGER)),
+                $queryBuilder->expr()->in('sys_language_uid', [-1, 0])
+            );
+
+        match ($order) {
+            'title' => $query->orderBy('title', 'ASC'),
+            'uid' => $query->orderBy('uid', 'ASC'),
+            default => $query->orderBy('sorting', 'ASC')->addOrderBy('title', 'ASC'),
+        };
+
+        return $query->executeQuery()->fetchAllAssociative();
+    }
+
+    /** Uid of a root-level category by title, 0 when it does not exist. */
+    private function rootByTitle(string $title): int
+    {
+        $queryBuilder = GeneralUtility::makeInstance(ConnectionPool::class)
+            ->getQueryBuilderForTable(self::CATEGORY_TABLE);
+        $queryBuilder->setRestrictions(GeneralUtility::makeInstance(FrontendRestrictionContainer::class));
+
+        $uid = $queryBuilder->select('uid')->from(self::CATEGORY_TABLE)
+            ->where(
+                $queryBuilder->expr()->eq('parent', 0),
+                $queryBuilder->expr()->eq('title', $queryBuilder->createNamedParameter($title))
+            )
+            ->setMaxResults(1)
+            ->executeQuery()->fetchOne();
+
+        return is_numeric($uid) ? (int)$uid : 0;
+    }
+}

packages/vitec/Classes/UserFunc/SearchJsonRenderer.php

@@ -73,6 +73,7 @@ class SearchJsonRenderer
         'pages' => 'page',
         'tx_vitec_domain_model_product' => 'product',
         'tx_vitec_domain_model_market' => 'market',
+        'tx_vitec_domain_model_solution' => 'solution',
         'tx_vitec_domain_model_usecase' => 'story',
         'tx_news_domain_model_news' => 'news',
         'tx_vitec_domain_model_download' => 'download',
@@ -152,6 +153,15 @@ class SearchJsonRenderer
                 $activeType = '';
             }
 
+            // Secondary facet filters: value comes verbatim from facets.<name>[].value
+            $facetFilters = [];
+            foreach (['market', 'category'] as $facetParam) {
+                $facetValue = trim((string)($params[$facetParam] ?? ''));
+                if ($facetValue !== '') {
+                    $facetFilters[$facetParam] = $facetValue;
+                }
+            }
+
             if ($query === '') {
                 return (string)json_encode($this->emptyResult());
             }
@@ -177,8 +187,15 @@ class SearchJsonRenderer
                 $search,
             );
             $arguments = ['q' => $query, 'page' => $page];
+            $filterArguments = [];
             if ($activeType !== '') {
-                $arguments['filter'] = ['type:' . $typeFilterField];
+                $filterArguments[] = 'type:' . $typeFilterField;
+            }
+            foreach ($facetFilters as $facetName => $facetValue) {
+                $filterArguments[] = $facetName . ':' . $facetValue;
+            }
+            if ($filterArguments !== []) {
+                $arguments['filter'] = $filterArguments;
             }
             $searchRequest = GeneralUtility::makeInstance(SearchRequestBuilder::class, $typoScriptConfiguration)
                 ->buildForSearch($arguments, $pageId, $languageId);
@@ -255,6 +272,7 @@ class SearchJsonRenderer
                 'numFound' => $numFound,
                 'totalPages' => $resultsPerPage > 0 ? (int)ceil($numFound / $resultsPerPage) : 0,
                 'filter' => $activeType !== '' ? $activeType : null,
+                'activeFacets' => (object)$facetFilters,
                 'facets' => (object)$facets,
                 'results' => $results,
                 'suggestions' => array_values(array_unique($suggestions)),
@@ -276,6 +294,7 @@ class SearchJsonRenderer
             'numFound' => 0,
             'totalPages' => 0,
             'filter' => null,
+            'activeFacets' => new stdClass(),
             'facets' => new \stdClass(),
             'results' => [],
             'suggestions' => [],