Add cleanup command for stale indexed files

2026-01-05 10:21:37 +01:00
parent 0e20a813af
commit b684267541
1 changed files with 77 additions and 0 deletions
@@ -0,0 +1,77 @@
+<?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
+{
+    protected static $defaultName = 'meilisearch:files:cleanup';
+
+    public function __construct(
+        private readonly ContaoFramework $framework,
+    ) {
+        parent::__construct();
+    }
+
+    protected function configure(): void
+    {
+        $this
+            ->setDescription('Removes 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)',
+                3600 // 1 Stunde Default
+            )
+            ->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();
+
+        $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;
+
+            $output->writeln(sprintf(
+                '[DRY-RUN] %d stale file(s) would be removed (last_seen < %s)',
+                $count,
+                date('Y-m-d H:i:s', $cutoff)
+            ));
+
+            return Command::SUCCESS;
+        }
+
+        $affected = Database::getInstance()
+            ->prepare('DELETE FROM tl_search_pdf WHERE last_seen < ?')
+            ->execute($cutoff)
+            ->affectedRows;
+
+        $output->writeln(sprintf(
+            'Removed %d stale file(s) from tl_search_pdf (last_seen < %s)',
+            $affected,
+            date('Y-m-d H:i:s', $cutoff)
+        ));
+
+        return Command::SUCCESS;
+    }
+}