add logging

README.md

@@ -0,0 +1,74 @@
+# Contao Meilisearch Bundle
+
+Eine schlanke Schnittstelle zwischen **Contao CMS (4.13 / 5.6 / 5.7 ready) unter PHP 8.4** und einer **selbst gehosteten Meilisearch-Instanz**.  
+Das Bundle erweitert den Contao-Suchindex um strukturierte Daten und ermöglicht eine performante, moderne Volltextsuche.
+
+---
+
+## ✨ Features
+
+- Integration von **Meilisearch** als externe Suchmaschine
+- Indexierung von:
+    - Contao-Seiten
+    - Inhaltselementen
+    - **PDF-Dateien**
+    - **Office-Dokumenten** (DOCX, XLSX, PPTX)
+- Unterstützung für:
+    - Seiten-Prioritäten
+    - Keywords
+    - Vorschaubild
+- Kompatibel mit:
+    - Contao **4.13**, **5.6** und **5.7**
+    - PHP **8.4**
+
+---
+
+## ⏱️ Scheduled Indexing (Cron setup)
+
+Das Bundle stellt eigene Commands zur Verfügung, um Dateien zu bereinigen und den Meilisearch-Index neu aufzubauen.  
+Für den produktiven Einsatz wird empfohlen, diese Commands regelmäßig per **System-Crontab** auszuführen.
+
+Das Bundle nutzt **keinen eigenen Contao-Cron**, sondern System-Cronjobs.
+
+## Verfügbare Commands
+
+### Datei-Cleanup
+
+```
+/vendor/bin/contao-console meilisearch:files:cleanup
+```
+
+### Meilisearch-Index
+
+```
+/vendor/bin/contao-console meilisearch:index
+```
+
+## Empfohlene Reihenfolge
+
+1. Datei-Cleanup  
+   `/vendor/bin/contao-console meilisearch:files:cleanup`
+
+2. Contao-Crawl (ca. 1 Minute später)  
+   `/vendor/bin/contao-console contao:crawl`
+
+3. Meilisearch-Index (ca. 15 Minuten später)  
+   `/vendor/bin/contao-console meilisearch:index`
+
+## Beispiel Crontab
+
+```
+0 5 * * *  /usr/bin/php8.4 /path/to/project/vendor/bin/contao-console meilisearch:files:cleanup
+1 5 * * *  /usr/bin/php8.4 /path/to/project/vendor/bin/contao-console contao:crawl
+15 5 * * * /usr/bin/php8.4 /path/to/project/vendor/bin/contao-console meilisearch:index
+```
+
+## Logging
+
+```
+>> var/logs/meilisearch_cron.log 2>&1
+```
+
+## Lizenz
+
+MIT

src/Command/MeilisearchFilesCleanupCommand.php

@@ -0,0 +1,106 @@
+<?php
+
+namespace MummertMedia\ContaoMeilisearchBundle\Command;
+
+use Contao\CoreBundle\Framework\ContaoFramework;
+use Contao\Database;
+use Symfony\Component\Console\Command\Command;
+use Symfony\Component\Console\Input\InputInterface;
+use Symfony\Component\Console\Input\InputOption;
+use Symfony\Component\Console\Output\OutputInterface;
+
+class MeilisearchFilesCleanupCommand extends Command
+{
+    public function __construct(
+        private readonly ContaoFramework $framework,
+    ) {
+        parent::__construct();
+    }
+
+    protected function configure(): void
+    {
+        $this
+            ->setName('meilisearch:files:cleanup')
+            ->setDescription('Remove stale indexed files (PDF, DOCX, XLSX, PPTX) from tl_search_pdf')
+            ->addOption(
+                'grace',
+                null,
+                InputOption::VALUE_OPTIONAL,
+                'Grace period in seconds (files newer than now-grace are kept)',
+                86400 // 24 Stunden
+            )
+            ->addOption(
+                'dry-run',
+                null,
+                InputOption::VALUE_NONE,
+                'Show how many entries would be removed without deleting them'
+            );
+    }
+
+    protected function execute(InputInterface $input, OutputInterface $output): int
+    {
+        $this->framework->initialize();
+
+        $this->log('Cleaner gestartet');
+
+        try {
+            $grace  = max(0, (int) $input->getOption('grace'));
+            $dryRun = (bool) $input->getOption('dry-run');
+            $cutoff = time() - $grace;
+
+            if ($dryRun) {
+                $count = Database::getInstance()
+                    ->prepare('SELECT COUNT(*) AS cnt FROM tl_search_pdf WHERE last_seen < ?')
+                    ->execute($cutoff)
+                    ->cnt;
+
+                $message = sprintf(
+                    '[DRY-RUN] %d stale file(s) would be removed (last_seen < %s)',
+                    $count,
+                    date('Y-m-d H:i:s', $cutoff)
+                );
+
+                $output->writeln('<comment>' . $message . '</comment>');
+                $this->log($message);
+
+                $this->log('Cleaner successfully stopped');
+                return Command::SUCCESS;
+            }
+
+            $affected = Database::getInstance()
+                ->prepare('DELETE FROM tl_search_pdf WHERE last_seen < ?')
+                ->execute($cutoff)
+                ->affectedRows;
+
+            $message = sprintf(
+                'Removed %d stale file(s) (last_seen < %s)',
+                $affected,
+                date('Y-m-d H:i:s', $cutoff)
+            );
+
+            $output->writeln('<info>' . $message . '</info>');
+            $this->log($message);
+
+            $this->log('Cleaner successfully stopped');
+            return Command::SUCCESS;
+
+        } catch (\Throwable $e) {
+            $this->log('Cleaner ERROR: ' . $e->getMessage());
+            $output->writeln('<error>' . $e->getMessage() . '</error>');
+
+            return Command::FAILURE;
+        }
+    }
+
+    /**
+     * Einheitliches Logging mit Zeitstempel
+     */
+    private function log(string $message): void
+    {
+        error_log(sprintf(
+            '[%s] %s',
+            date('Y-m-d H:i:s'),
+            $message
+        ));
+    }
+}

src/Command/MeilisearchIndexCommand.php

@@ -24,12 +24,34 @@ class MeilisearchIndexCommand extends Command
 
     protected function execute(InputInterface $input, OutputInterface $output): int
     {
+        $this->log('Meilisearch index gestartet');
         $output->writeln('<info>Meilisearch index started</info>');
 
+        try {
             $this->indexService->run();
 
+            $this->log('Meilisearch index successfully stopped');
             $output->writeln('<info>Meilisearch index finished</info>');
 
             return Command::SUCCESS;
+
+        } catch (\Throwable $e) {
+            $this->log('Meilisearch index ERROR: ' . $e->getMessage());
+            $output->writeln('<error>' . $e->getMessage() . '</error>');
+
+            return Command::FAILURE;
+        }
+    }
+
+    /**
+     * Einheitliches Logging mit Zeitstempel
+     */
+    private function log(string $message): void
+    {
+        error_log(sprintf(
+            '[%s] %s',
+            date('Y-m-d H:i:s'),
+            $message
+        ));
     }
 }

src/Cron/MeilisearchIndexCron.php

@@ -1,22 +0,0 @@
-<?php
-namespace MummertMedia\ContaoMeilisearchBundle\Cron;
-
-use Contao\CoreBundle\Framework\ContaoFramework;
-use MummertMedia\ContaoMeilisearchBundle\Service\MeilisearchIndexService;
-
-class MeilisearchIndexCron
-{
-    public function __construct(
-        private readonly MeilisearchIndexService $indexService,
-        private readonly ContaoFramework $framework,
-    ) {}
-
-    public function __invoke(): void
-    {
-        // Contao initialisieren (wichtig!)
-        $this->framework->initialize();
-
-        // einmal täglich indexieren
-        $this->indexService->run();
-    }
-}

src/EventListener/IndexPageListener.php

@@ -30,22 +30,6 @@ class IndexPageListener
             'set_keys'  => array_keys($set),
         ]);
 
-        /*
-         * =====================
-         * PDF: Reset genau 1× pro Crawl
-         * =====================
-         */
-        try {
-            $this->debug('PDF resetTableOnce(): call');
-            $this->pdfIndexService->resetTableOnce();
-            $this->debug('PDF resetTableOnce(): ok');
-        } catch (\Throwable $e) {
-            $this->debug('PDF resetTableOnce(): failed', [
-                'error' => $e->getMessage(),
-                'class' => $e::class,
-            ]);
-        }
-
         /*
          * =====================
          * SEITEN-METADATEN

src/Resources/config/services.yaml

@@ -3,22 +3,22 @@ services:
   Psr\Container\ContainerInterface: '@service_container'
 
   MummertMedia\ContaoMeilisearchBundle\:
-    resource: '../../{Command,Cron,EventListener,Service}'
+    resource: '../../{Command,EventListener,Service}'
     autowire: true
     autoconfigure: true
 
+  MummertMedia\ContaoMeilisearchBundle\EventListener\MeilisearchPageMarkerListener:
+    autowire: true
+    autoconfigure: false
+    tags:
+      - { name: contao.hook, hook: outputFrontendTemplate, method: onOutputFrontendTemplate }
+
   MummertMedia\ContaoMeilisearchBundle\EventListener\IndexPageListener:
     autowire: true
     autoconfigure: false
     tags:
       - { name: contao.hook, hook: indexPage, method: onIndexPage }
 
-  MummertMedia\ContaoMeilisearchBundle\Cron\MeilisearchIndexCron:
-    autowire: true
-    autoconfigure: false
-    tags:
-      - { name: contao.cron, interval: daily, method: __invoke }
-
   MummertMedia\ContaoMeilisearchBundle\Controller\FrontendModule\MeilisearchSearchController:
     autowire: true
     autoconfigure: false

src/Resources/contao/dca/tl_search_pdf.php

@@ -8,10 +8,11 @@ $GLOBALS['TL_DCA']['tl_search_pdf'] = [
         'sql' => [
             'keys' => [
                 'id'        => 'primary',
-                'checksum' => 'unique',
                 'page_id'   => 'index',
-                'url'      => 'index',
-                'type'     => 'index', // ⬅️ NEU
+                'url'       => 'unique',
+                'type'      => 'index',
+                'checksum'  => 'index',
+                'last_seen' => 'index', // ⬅️ NEU (für Cleanup-Performance)
             ],
         ],
     ],
@@ -25,10 +26,18 @@ $GLOBALS['TL_DCA']['tl_search_pdf'] = [
             'sql' => "int(10) unsigned NOT NULL default 0",
         ],
 
+        /*
+         * Zeitpunkt, wann die Datei zuletzt beim Crawl gesehen wurde
+         * → Basis für Cleanup
+         */
+        'last_seen' => [ // ⬅️ NEU
+            'sql' => "int(10) unsigned NOT NULL default 0",
+        ],
+
         /*
          * Dateityp: pdf | docx | xlsx | pptx
          */
-        'type' => [ // ⬅️ NEU
+        'type' => [
             'sql' => "varchar(16) NOT NULL default 'pdf'",
         ],
 
@@ -64,7 +73,7 @@ $GLOBALS['TL_DCA']['tl_search_pdf'] = [
 
         /*
          * Herkunftsseite (tl_page.id)
-         * → Cleanup / Referenz
+         * → optional, Debug / Referenz
          */
         'page_id' => [
             'sql' => "int(10) unsigned NOT NULL default 0",

src/Resources/public/icons/crawl_search-index_log.csv

@@ -1,3 +0,0 @@
-Time,Source,URI,"Found on URI","Found on level",Tags,Message
-"2025-12-28 10:44:12.357228","Contao\CoreBundle\Crawl\Escargot\Subscriber\SearchIndexSubscriber",https://flowerpowerabi.de/,https://flowerpowerabi.de/sitemap.xml,3,,"Forwarded to the search indexer. Was indexed successfully."
-"2025-12-28 10:44:12.413689","Contao\CoreBundle\Crawl\Escargot\Subscriber\SearchIndexSubscriber",https://flowerpowerabi.de/testseite-fuer-meilisearch,https://flowerpowerabi.de/sitemap.xml,3,,"Forwarded to the search indexer. Was indexed successfully."

src/Service/OfficeIndexService.php

@@ -12,9 +12,6 @@ class OfficeIndexService
 {
     private string $projectDir;
 
-    // pro Crawl-Durchlauf: doppelte Verarbeitung vermeiden
-    private array $seenThisCrawl = [];
-
     public function __construct(ParameterBagInterface $params)
     {
         $this->projectDir = rtrim((string) $params->get('kernel.project_dir'), '/');
@@ -25,6 +22,10 @@ class OfficeIndexService
      */
     public function handleOfficeLinks(array $officeLinks): void
     {
+        // Dedupe nur pro Aufruf (nicht "pro Crawl")
+        $seen = [];
+        $now  = time();
+
         foreach ($officeLinks as $row) {
             $url      = (string) ($row['url'] ?? '');
             $linkText = $row['linkText'] ?? null;
@@ -33,13 +34,12 @@ class OfficeIndexService
                 continue;
             }
 
-            try {
-                // innerhalb des Crawls gleiche URL nicht mehrfach parsen
+            // doppelte URLs pro Aufruf vermeiden
             $seenKey = md5($url);
-                if (isset($this->seenThisCrawl[$seenKey])) {
+            if (isset($seen[$seenKey])) {
                 continue;
             }
-                $this->seenThisCrawl[$seenKey] = true;
+            $seen[$seenKey] = true;
 
             $normalized = $this->normalizeOfficeUrl($url);
             if ($normalized === null) {
@@ -56,27 +56,37 @@ class OfficeIndexService
             $mtime    = (int) (filemtime($absolutePath) ?: 0);
             $checksum = md5($relativePath . '|' . $mtime);
 
-                $title = $linkText ?: basename($absolutePath);
+            // existiert bereits?
+            $existing = Database::getInstance()
+                ->prepare('SELECT checksum FROM tl_search_pdf WHERE url=? LIMIT 1')
+                ->execute($relativePath)
+                ->fetchAssoc();
 
+            $needsParse = !$existing || ($existing['checksum'] ?? '') !== $checksum;
+
+            // Titel-Priorität:
+            // 1) Linktext
+            // 2) Dateiname
+            $title = $linkText ?: basename($absolutePath);
+            $text  = '';
+
+            if ($needsParse) {
                 $text = $this->parseOfficeFile($absolutePath, $type);
                 if ($text === '') {
+                    // Parsing fehlgeschlagen → nichts überschreiben
                     continue;
                 }
+            }
 
             $this->upsertOffice(
                 $relativePath,
                 $title,
-                    $text,
+                $text,      // kann '' sein → SQL überschreibt dann nicht
                 $checksum,
                 $mtime,
-                    $type
+                $type,
+                $now
             );
-
-            } catch (\Throwable $e) {
-                error_log(
-                    '[ContaoMeilisearch] Office indexing failed for "' . $url . '": ' . $e->getMessage()
-                );
-            }
         }
     }
 
@@ -88,6 +98,10 @@ class OfficeIndexService
         $decoded = html_entity_decode($url);
         $parts   = parse_url($decoded);
 
+        if (!$parts) {
+            return null;
+        }
+
         // 1) files/... (ohne führenden Slash)
         if (!empty($parts['path']) && str_starts_with($parts['path'], 'files/')) {
             $ext = strtolower(pathinfo($parts['path'], PATHINFO_EXTENSION));
@@ -117,8 +131,8 @@ class OfficeIndexService
             $ext  = strtolower(pathinfo($file, PATHINFO_EXTENSION));
 
             if (
-                str_starts_with($file, 'files/')
-                && in_array($ext, ['docx', 'xlsx', 'pptx'], true)
+                str_starts_with($file, 'files/') &&
+                in_array($ext, ['docx', 'xlsx', 'pptx'], true)
             ) {
                 return ['/' . $file, $ext];
             }
@@ -148,25 +162,28 @@ class OfficeIndexService
         string $text,
         string $checksum,
         int $mtime,
-        string $type
+        string $type,
+        int $now
     ): void {
-        try {
         Database::getInstance()
             ->prepare('
                 INSERT INTO tl_search_pdf
-                        (tstamp, type, url, title, text, checksum, file_mtime)
+                    (tstamp, last_seen, type, url, title, text, checksum, file_mtime)
                 VALUES
-                        (?, ?, ?, ?, ?, ?, ?)
+                    (?, ?, ?, ?, ?, ?, ?, ?)
                 ON DUPLICATE KEY UPDATE
                     tstamp     = VALUES(tstamp),
+                    last_seen  = VALUES(last_seen),
                     type       = VALUES(type),
                     url        = VALUES(url),
                     title      = VALUES(title),
-                        text=VALUES(text),
-                        file_mtime=VALUES(file_mtime)
+                    checksum   = VALUES(checksum),
+                    file_mtime = VALUES(file_mtime),
+                    text       = IF(VALUES(text) = "" OR VALUES(text) IS NULL, text, VALUES(text))
             ')
             ->execute(
-                    time(),
+                $now,
+                $now,
                 $type,
                 $url,
                 $title,
@@ -174,11 +191,6 @@ class OfficeIndexService
                 $checksum,
                 $mtime
             );
-        } catch (\Throwable $e) {
-            error_log(
-                '[ContaoMeilisearch] Failed to write Office index entry (' . $url . '): ' . $e->getMessage()
-            );
-        }
     }
 
     private function parseOfficeFile(string $absolutePath, string $type): string
@@ -206,10 +218,7 @@ class OfficeIndexService
             }
 
             return $this->cleanText($text);
-        } catch (\Throwable $e) {
-            error_log(
-                '[ContaoMeilisearch] Failed to parse DOCX "' . $absolutePath . '": ' . $e->getMessage()
-            );
+        } catch (\Throwable) {
             return '';
         }
     }
@@ -227,10 +236,7 @@ class OfficeIndexService
             }
 
             return $this->cleanText($text);
-        } catch (\Throwable $e) {
-            error_log(
-                '[ContaoMeilisearch] Failed to parse XLSX "' . $absolutePath . '": ' . $e->getMessage()
-            );
+        } catch (\Throwable) {
             return '';
         }
     }
@@ -250,10 +256,7 @@ class OfficeIndexService
             }
 
             return $this->cleanText($text);
-        } catch (\Throwable $e) {
-            error_log(
-                '[ContaoMeilisearch] Failed to parse PPTX "' . $absolutePath . '": ' . $e->getMessage()
-            );
+        } catch (\Throwable) {
             return '';
         }
     }

src/Service/PdfIndexService.php

@@ -10,34 +10,20 @@ class PdfIndexService
 {
     private string $projectDir;
 
-    private bool $didReset = false;
-    private array $seenThisCrawl = [];
-
     public function __construct(ParameterBagInterface $params)
     {
         $this->projectDir = rtrim((string) $params->get('kernel.project_dir'), '/');
     }
 
-    /**
-     * Wird aus dem Listener beim ersten Hook-Call pro Crawl aufgerufen.
-     */
-    public function resetTableOnce(): void
-    {
-        if ($this->didReset) {
-            return;
-        }
-
-        $this->didReset = true;
-        $this->seenThisCrawl = [];
-
-        Database::getInstance()->execute('TRUNCATE tl_search_pdf');
-    }
-
     /**
      * @param array<int,array{url:string,linkText:?string}> $pdfLinks
      */
     public function handlePdfLinks(array $pdfLinks): void
     {
+        // Dedupe nur pro Aufruf (nicht "pro Crawl")
+        $seen = [];
+        $now  = time();
+
         foreach ($pdfLinks as $row) {
             $url      = (string) ($row['url'] ?? '');
             $linkText = $row['linkText'] ?? null;
@@ -46,12 +32,12 @@ class PdfIndexService
                 continue;
             }
 
-            // innerhalb eines Crawls doppelte URLs vermeiden
+            // doppelte URLs pro Aufruf vermeiden
             $seenKey = md5($url);
-            if (isset($this->seenThisCrawl[$seenKey])) {
+            if (isset($seen[$seenKey])) {
                 continue;
             }
-            $this->seenThisCrawl[$seenKey] = true;
+            $seen[$seenKey] = true;
 
             $normalizedPath = $this->normalizePdfUrl($url);
             if ($normalizedPath === null) {
@@ -66,24 +52,39 @@ class PdfIndexService
             $mtime    = (int) (filemtime($absolutePath) ?: 0);
             $checksum = md5($normalizedPath . '|' . $mtime);
 
+            // existiert bereits?
+            $existing = Database::getInstance()
+                ->prepare('SELECT checksum FROM tl_search_pdf WHERE url=? LIMIT 1')
+                ->execute($normalizedPath)
+                ->fetchAssoc();
+
+            $needsParse = !$existing || ($existing['checksum'] ?? '') !== $checksum;
+
             // Titel-Priorität:
             // 1) Linktext
             // 2) PDF-Metadaten
             // 3) Dateiname
+            $title = $linkText ?: basename($absolutePath);
+            $text  = '';
+
+            if ($needsParse) {
                 $pdfMetaTitle = $this->readPdfMetaTitle($absolutePath);
                 $title = $linkText ?: ($pdfMetaTitle ?: basename($absolutePath));
 
                 $text = $this->parsePdf($absolutePath);
                 if ($text === '') {
+                    // wenn parsing fehlschlägt, NICHT überschreiben
                     continue;
                 }
+            }
 
             $this->upsertPdf(
                 $normalizedPath,
                 $title,
-                $text,
+                $text,      // kann '' sein → wird in SQL nicht überschrieben
                 $checksum,
-                $mtime
+                $mtime,
+                $now
             );
         }
     }
@@ -93,6 +94,10 @@ class PdfIndexService
         $decoded = html_entity_decode($url);
         $parts   = parse_url($decoded);
 
+        if (!$parts) {
+            return null;
+        }
+
         // 1) files/...pdf (ohne führenden Slash)
         if (
             !empty($parts['path'])
@@ -149,23 +154,29 @@ class PdfIndexService
         string $title,
         string $text,
         string $checksum,
-        int $mtime
+        int $mtime,
+        int $now
     ): void {
         Database::getInstance()
             ->prepare('
                 INSERT INTO tl_search_pdf
-                    (tstamp, url, title, text, checksum, file_mtime)
+                    (tstamp, last_seen, type, url, title, text, checksum, file_mtime)
                 VALUES
-                    (?, ?, ?, ?, ?, ?)
+                    (?, ?, ?, ?, ?, ?, ?, ?)
                 ON DUPLICATE KEY UPDATE
                     tstamp     = VALUES(tstamp),
+                    last_seen  = VALUES(last_seen),
+                    type       = VALUES(type),
                     url        = VALUES(url),
                     title      = VALUES(title),
-                    text=VALUES(text),
-                    file_mtime=VALUES(file_mtime)
+                    checksum   = VALUES(checksum),
+                    file_mtime = VALUES(file_mtime),
+                    text       = IF(VALUES(text) = "" OR VALUES(text) IS NULL, text, VALUES(text))
             ')
             ->execute(
-                time(),
+                $now,
+                $now,
+                'pdf',
                 $url,
                 $title,
                 $text,
@@ -203,6 +214,7 @@ class PdfIndexService
                 }
             }
         } catch (\Throwable) {
+            // ignore
         }
 
         return null;