From a5118305d3b3764054fc2790465fce6d700c0d19 Mon Sep 17 00:00:00 2001
From: Matthieu <mtholot19@gmail.com>
Date: Sun, 25 Jan 2026 19:02:26 +0100
Subject: [PATCH] feat : automatic PDF compression on upload

- Add PdfCompressorService for lossless compression with qpdf
- Add DocumentPdfCompressorListener for automatic compression on persist/update
- Add app:compress-pdf command for batch compression of existing PDFs

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
---
 src/Command/CompressPdfCommand.php            | 175 ++++++++++++++++++
 .../DocumentPdfCompressorListener.php         |  54 ++++++
 src/Service/PdfCompressorService.php          |  73 ++++++++
 3 files changed, 302 insertions(+)
 create mode 100644 src/Command/CompressPdfCommand.php
 create mode 100644 src/EventListener/DocumentPdfCompressorListener.php
 create mode 100644 src/Service/PdfCompressorService.php

diff --git a/src/Command/CompressPdfCommand.php b/src/Command/CompressPdfCommand.php
new file mode 100644
index 0000000..ee3232c
--- /dev/null
+++ b/src/Command/CompressPdfCommand.php
@@ -0,0 +1,175 @@
+<?php
+
+declare(strict_types=1);
+
+namespace App\Command;
+
+use App\Repository\DocumentRepository;
+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;
+
+#[AsCommand(
+    name: 'app:compress-pdf',
+    description: 'Compress all PDF documents stored in database without quality loss',
+)]
+class CompressPdfCommand extends Command
+{
+    public function __construct(
+        private readonly DocumentRepository $documentRepository,
+        private readonly EntityManagerInterface $em,
+    ) {
+        parent::__construct();
+    }
+
+    protected function configure(): void
+    {
+        $this
+            ->addOption('dry-run', null, InputOption::VALUE_NONE, 'Show what would be compressed without actually doing it')
+        ;
+    }
+
+    protected function execute(InputInterface $input, OutputInterface $output): int
+    {
+        $io     = new SymfonyStyle($input, $output);
+        $dryRun = $input->getOption('dry-run');
+
+        // Check if qpdf is installed
+        exec('which qpdf', $qpdfPath, $returnCode);
+        if (0 !== $returnCode) {
+            $io->error('qpdf is not installed. Run: sudo apt install qpdf');
+
+            return Command::FAILURE;
+        }
+
+        $documents = $this->documentRepository->findBy(['mimeType' => 'application/pdf']);
+
+        if (empty($documents)) {
+            $io->info('No PDF documents found.');
+
+            return Command::SUCCESS;
+        }
+
+        $io->title('PDF Compression');
+        $io->text(sprintf('Found %d PDF documents', count($documents)));
+
+        $totalSaved = 0;
+        $compressed = 0;
+
+        foreach ($documents as $document) {
+            $base64Data = $document->getPath();
+
+            // Remove data URI prefix if present
+            if (str_contains($base64Data, ',')) {
+                $base64Data = explode(',', $base64Data, 2)[1];
+            }
+
+            $pdfContent = base64_decode($base64Data, true);
+            if (false === $pdfContent) {
+                $io->warning(sprintf('Failed to decode document: %s', $document->getName()));
+
+                continue;
+            }
+
+            $originalSize = strlen($pdfContent);
+
+            if ($dryRun) {
+                $io->text(sprintf(
+                    '  [DRY-RUN] Would compress: %s (%s)',
+                    $document->getName(),
+                    $this->formatBytes($originalSize)
+                ));
+
+                continue;
+            }
+
+            // Create temp files
+            $tempInput  = tempnam(sys_get_temp_dir(), 'pdf_in_');
+            $tempOutput = tempnam(sys_get_temp_dir(), 'pdf_out_');
+
+            file_put_contents($tempInput, $pdfContent);
+
+            // Compress with qpdf (lossless)
+            $command = sprintf(
+                'qpdf --linearize --object-streams=generate %s %s 2>&1',
+                escapeshellarg($tempInput),
+                escapeshellarg($tempOutput)
+            );
+
+            exec($command, $cmdOutput, $returnCode);
+
+            if (0 !== $returnCode || !file_exists($tempOutput)) {
+                $io->warning(sprintf('Failed to compress: %s', $document->getName()));
+                @unlink($tempInput);
+                @unlink($tempOutput);
+
+                continue;
+            }
+
+            $compressedContent = file_get_contents($tempOutput);
+            $compressedSize    = strlen($compressedContent);
+
+            // Only update if we actually saved space
+            if ($compressedSize < $originalSize) {
+                $saved = $originalSize - $compressedSize;
+                $totalSaved += $saved;
+                ++$compressed;
+
+                // Rebuild base64 with data URI prefix
+                $newBase64 = 'data:application/pdf;base64,'.base64_encode($compressedContent);
+                $document->setPath($newBase64);
+                $document->setSize($compressedSize);
+
+                $io->text(sprintf(
+                    '  ✓ %s: %s → %s (-%s, -%.1f%%)',
+                    $document->getName(),
+                    $this->formatBytes($originalSize),
+                    $this->formatBytes($compressedSize),
+                    $this->formatBytes($saved),
+                    ($saved / $originalSize) * 100
+                ));
+            } else {
+                $io->text(sprintf(
+                    '  - %s: Already optimal (%s)',
+                    $document->getName(),
+                    $this->formatBytes($originalSize)
+                ));
+            }
+
+            @unlink($tempInput);
+            @unlink($tempOutput);
+        }
+
+        if (!$dryRun && $compressed > 0) {
+            $this->em->flush();
+            $io->success(sprintf(
+                'Compressed %d/%d PDFs. Total space saved: %s',
+                $compressed,
+                count($documents),
+                $this->formatBytes($totalSaved)
+            ));
+        } elseif ($dryRun) {
+            $io->info('Dry run completed. No changes made.');
+        } else {
+            $io->info('No PDFs needed compression.');
+        }
+
+        return 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];
+    }
+}
diff --git a/src/EventListener/DocumentPdfCompressorListener.php b/src/EventListener/DocumentPdfCompressorListener.php
new file mode 100644
index 0000000..85e00b9
--- /dev/null
+++ b/src/EventListener/DocumentPdfCompressorListener.php
@@ -0,0 +1,54 @@
+<?php
+
+declare(strict_types=1);
+
+namespace App\EventListener;
+
+use App\Entity\Document;
+use App\Service\PdfCompressorService;
+use Doctrine\Bundle\DoctrineBundle\Attribute\AsEntityListener;
+use Doctrine\ORM\Events;
+use Psr\Log\LoggerInterface;
+
+#[AsEntityListener(event: Events::prePersist, method: 'prePersist', entity: Document::class)]
+#[AsEntityListener(event: Events::preUpdate, method: 'preUpdate', entity: Document::class)]
+class DocumentPdfCompressorListener
+{
+    public function __construct(
+        private readonly PdfCompressorService $pdfCompressor,
+        private readonly ?LoggerInterface $logger = null,
+    ) {}
+
+    public function prePersist(Document $document): void
+    {
+        $this->compressIfPdf($document);
+    }
+
+    public function preUpdate(Document $document): void
+    {
+        $this->compressIfPdf($document);
+    }
+
+    private function compressIfPdf(Document $document): void
+    {
+        if ('application/pdf' !== $document->getMimeType()) {
+            return;
+        }
+
+        $result = $this->pdfCompressor->compressBase64Pdf($document->getPath());
+
+        if (null === $result) {
+            return;
+        }
+
+        $document->setPath($result['path']);
+        $document->setSize($result['size']);
+
+        $this->logger?->info('PDF compressed', [
+            'document'       => $document->getName(),
+            'originalSize'   => $result['originalSize'],
+            'compressedSize' => $result['size'],
+            'saved'          => $result['saved'],
+        ]);
+    }
+}
diff --git a/src/Service/PdfCompressorService.php b/src/Service/PdfCompressorService.php
new file mode 100644
index 0000000..9ebe502
--- /dev/null
+++ b/src/Service/PdfCompressorService.php
@@ -0,0 +1,73 @@
+<?php
+
+declare(strict_types=1);
+
+namespace App\Service;
+
+class PdfCompressorService
+{
+    public function compressBase64Pdf(string $base64Data): ?array
+    {
+        // Check if qpdf is available
+        exec('which qpdf', $qpdfPath, $returnCode);
+        if (0 !== $returnCode) {
+            return null;
+        }
+
+        // Remove data URI prefix if present
+        $originalBase64 = $base64Data;
+        if (str_contains($base64Data, ',')) {
+            $base64Data = explode(',', $base64Data, 2)[1];
+        }
+
+        $pdfContent = base64_decode($base64Data, true);
+        if (false === $pdfContent) {
+            return null;
+        }
+
+        $originalSize = strlen($pdfContent);
+
+        // Create temp files
+        $tempInput  = tempnam(sys_get_temp_dir(), 'pdf_in_');
+        $tempOutput = tempnam(sys_get_temp_dir(), 'pdf_out_');
+
+        file_put_contents($tempInput, $pdfContent);
+
+        // Compress with qpdf (lossless)
+        $command = sprintf(
+            'qpdf --linearize --object-streams=generate %s %s 2>&1',
+            escapeshellarg($tempInput),
+            escapeshellarg($tempOutput)
+        );
+
+        exec($command, $cmdOutput, $returnCode);
+
+        if (0 !== $returnCode || !file_exists($tempOutput)) {
+            @unlink($tempInput);
+            @unlink($tempOutput);
+
+            return null;
+        }
+
+        $compressedContent = file_get_contents($tempOutput);
+        $compressedSize    = strlen($compressedContent);
+
+        @unlink($tempInput);
+        @unlink($tempOutput);
+
+        // Only return compressed version if it's smaller
+        if ($compressedSize >= $originalSize) {
+            return null;
+        }
+
+        // Rebuild with data URI prefix
+        $newBase64 = 'data:application/pdf;base64,'.base64_encode($compressedContent);
+
+        return [
+            'path'         => $newBase64,
+            'size'         => $compressedSize,
+            'originalSize' => $originalSize,
+            'saved'        => $originalSize - $compressedSize,
+        ];
+    }
+}