Inventory/src/Command/MigrateDocumentsToFilesystemCommand.php

<?php

declare(strict_types=1);

namespace App\Command;

use App\Repository\DocumentRepository;
use App\Service\DocumentStorageService;
use Doctrine\ORM\EntityManagerInterface;
use Symfony\Component\Console\Attribute\AsCommand;
use Symfony\Component\Console\Command\Command;
use Symfony\Component\Console\Input\InputInterface;
use Symfony\Component\Console\Input\InputOption;
use Symfony\Component\Console\Output\OutputInterface;
use Symfony\Component\Console\Style\SymfonyStyle;
use Throwable;

use function count;
use function strlen;

#[AsCommand(
    name: 'app:migrate-documents-to-filesystem',
    description: 'Migrate document storage from Base64 in DB to filesystem',
)]
class MigrateDocumentsToFilesystemCommand extends Command
{
    public function __construct(
        private readonly DocumentRepository $documentRepository,
        private readonly EntityManagerInterface $em,
        private readonly DocumentStorageService $storageService,
    ) {
        parent::__construct();
    }

    protected function configure(): void
    {
        $this
            ->addOption('dry-run', null, InputOption::VALUE_NONE, 'Show what would be migrated without making changes')
            ->addOption('batch-size', null, InputOption::VALUE_REQUIRED, 'Number of documents to process before flushing', '50')
            ->addOption('limit', null, InputOption::VALUE_REQUIRED, 'Max documents to migrate (for testing)', '0')
        ;
    }

    protected function execute(InputInterface $input, OutputInterface $output): int
    {
        $io        = new SymfonyStyle($input, $output);
        $dryRun    = $input->getOption('dry-run');
        $batchSize = (int) $input->getOption('batch-size');
        $limit     = (int) $input->getOption('limit');

        $io->title('Document Storage Migration: Base64 → Filesystem');

        // Verify storage directory is writable
        $storageDir = $this->storageService->getStorageDir();
        if (!$dryRun) {
            if (!is_dir($storageDir)) {
                mkdir($storageDir, 0o775, true);
            }
            if (!is_writable($storageDir)) {
                $io->error("Storage directory is not writable: {$storageDir}");

                return Command::FAILURE;
            }
            $io->text("Storage directory: {$storageDir}");
        }

        // Step 1: fetch only IDs of Base64 documents (no heavy path column loaded)
        $conn       = $this->em->getConnection();
        $ids        = $conn->fetchFirstColumn("SELECT id FROM documents WHERE path LIKE 'data:%'");
        $total      = count($ids);
        $migrated   = 0;
        $skipped    = 0;
        $errors     = 0;
        $totalBytes = 0;

        $io->text(sprintf('Found %d documents with Base64 data to migrate', $total));

        if (0 === $total) {
            $io->success('Nothing to migrate — all documents are already file-based.');

            return Command::SUCCESS;
        }

        // Step 2: process one document at a time to avoid memory exhaustion
        foreach ($ids as $index => $docId) {
            if ($limit > 0 && $migrated >= $limit) {
                $io->text("Reached limit of {$limit} documents.");

                break;
            }

            // Fetch single row with raw SQL to keep memory flat
            $row = $conn->fetchAssociative(
                'SELECT id, name, filename, path, mimetype, size FROM documents WHERE id = ?',
                [$docId]
            );

            if (!$row) {
                ++$skipped;

                continue;
            }

            $path = $row['path'];
            if (!$this->storageService->isBase64DataUri($path)) {
                ++$skipped;

                continue;
            }

            $docName  = $row['name'] ?: $row['filename'];
            $filename = $row['filename'] ?: $row['name'];
            $mimeType = $row['mimetype'] ?? 'application/octet-stream';

            // Extract binary content from data URI
            $parts   = explode(',', $path, 2);
            $base64  = $parts[1] ?? '';
            $content = base64_decode($base64, true);

            // Free the raw row immediately
            unset($row, $path, $base64, $parts);

            if (false === $content || '' === $content) {
                $io->warning(sprintf('[%d/%d] Cannot decode: %s (id: %s)', $index + 1, $total, $docName, $docId));
                ++$errors;

                continue;
            }

            $fileSize  = strlen($content);
            $extension = $this->storageService->extensionFromFilename(
                $filename ?: ('file.'.$this->storageService->extensionFromMimeType($mimeType))
            );

            if ($dryRun) {
                $io->text(sprintf(
                    '  [DRY-RUN] Would migrate: %s (%s)',
                    $docName,
                    $this->formatBytes($fileSize)
                ));
                ++$migrated;
                $totalBytes += $fileSize;
                unset($content);

                continue;
            }

            try {
                $relativePath = $this->storageService->store($content, $docId, $extension);
                unset($content);

                // Update DB directly — avoid loading entity with huge path
                $conn->executeStatement(
                    'UPDATE documents SET path = ?, size = ? WHERE id = ?',
                    [$relativePath, $fileSize, $docId]
                );

                ++$migrated;
                $totalBytes += $fileSize;

                $io->text(sprintf(
                    '  [OK] %s → %s (%s)',
                    $docName,
                    $relativePath,
                    $this->formatBytes($fileSize)
                ));
            } catch (Throwable $e) {
                unset($content);
                $io->error(sprintf(
                    '  [FAIL] %s: %s',
                    $docName,
                    $e->getMessage()
                ));
                ++$errors;

                continue;
            }

            if (0 === $migrated % $batchSize) {
                $io->text(sprintf('  ... %d migrated so far', $migrated));
            }
        }

        $io->newLine();
        $io->table(
            ['Metric', 'Count'],
            [
                ['Total documents', (string) $total],
                ['Migrated', (string) $migrated],
                ['Skipped (already file-based)', (string) $skipped],
                ['Errors', (string) $errors],
                ['Total bytes written', $this->formatBytes($totalBytes)],
            ]
        );

        if ($dryRun) {
            $io->info('Dry run completed. No changes were made.');
        } elseif ($errors > 0) {
            $io->warning(sprintf('Migration completed with %d errors.', $errors));
        } else {
            $io->success('Migration completed successfully.');
        }

        return $errors > 0 ? Command::FAILURE : Command::SUCCESS;
    }

    private function formatBytes(int $bytes): string
    {
        $units = ['B', 'KB', 'MB', 'GB'];
        $i     = 0;
        while ($bytes >= 1024 && $i < count($units) - 1) {
            $bytes /= 1024;
            ++$i;
        }

        return round($bytes, 2).' '.$units[$i];
    }
}