Add PDF signature zone parsing functionality

This update introduces new services into the ChillDocStoreBundle for signature zone parsing within PDFs. The PDFSignatureZoneParser service identifies signature zones within PDF content while the additional classes, PDFPage and PDFSignatureZone, help define these zones and pages. Corresponding tests have also been
This commit is contained in:
Julien Fastré 2024-06-19 12:17:25 +02:00
parent 5bc542a567
commit a9f0059743
Signed by: julienfastre
GPG Key ID: BDE2190974723FCB
5 changed files with 195 additions and 0 deletions

View File

@ -0,0 +1,28 @@
<?php
declare(strict_types=1);
/*
* Chill is a software for social workers
*
* For the full copyright and license information, please view
* the LICENSE file that was distributed with this source code.
*/
namespace Chill\DocStoreBundle\Service\Signature;
final readonly class PDFPage
{
public function __construct(
public int $index,
public float $width,
public float $height,
) {}
public function equals(self $page): bool
{
return $page->index === $this->index
&& round($page->width, 2) === round($this->width, 2)
&& round($page->height, 2) === round($this->height, 2);
}
}

View File

@ -0,0 +1,33 @@
<?php
declare(strict_types=1);
/*
* Chill is a software for social workers
*
* For the full copyright and license information, please view
* the LICENSE file that was distributed with this source code.
*/
namespace Chill\DocStoreBundle\Service\Signature;
final readonly class PDFSignatureZone
{
public function __construct(
public float $x,
public float $y,
public float $height,
public float $width,
public PDFPage $PDFPage,
) {}
public function equals(self $other): bool
{
return
$this->x == $other->x
&& $this->y == $other->y
&& $this->height == $other->height
&& $this->width == $other->width
&& $this->PDFPage->equals($other->PDFPage);
}
}

View File

@ -0,0 +1,57 @@
<?php
declare(strict_types=1);
/*
* Chill is a software for social workers
*
* For the full copyright and license information, please view
* the LICENSE file that was distributed with this source code.
*/
namespace Chill\DocStoreBundle\Service\Signature;
use Smalot\PdfParser\Parser;
class PDFSignatureZoneParser
{
public const ZONE_SIGNATURE_START = 'signature_zone';
private Parser $parser;
public function __construct(
public float $defaultHeight = 180.0,
public float $defaultWidth = 180.0,
) {
$this->parser = new Parser();
}
/**
* @return list<PDFSignatureZone>
*/
public function findSignatureZones(string $fileContent): array
{
$pdf = $this->parser->parseContent($fileContent);
$zones = [];
$defaults = $pdf->getObjectsByType('Pages');
$defaultPage = reset($defaults);
$defaultPageDetails = $defaultPage->getDetails();
foreach ($pdf->getPages() as $index => $page) {
$pdfPage = new PDFPage(
$index,
$page['MediaBox'][2] ?? $defaultPageDetails['MediaBox'][2],
$page['MediaBox'][3] ?? $defaultPageDetails['MediaBox'][3],
);
foreach ($page->getDataTm() as $dataTm) {
if (str_starts_with($dataTm[1], self::ZONE_SIGNATURE_START)) {
$zones[] = new PDFSignatureZone($dataTm[0][4], $dataTm[0][5], $this->defaultHeight, $this->defaultWidth, $pdfPage);
}
}
}
return $zones;
}
}

View File

@ -0,0 +1,77 @@
<?php
declare(strict_types=1);
/*
* Chill is a software for social workers
*
* For the full copyright and license information, please view
* the LICENSE file that was distributed with this source code.
*/
namespace Tests\Service\Signature;
use Chill\DocStoreBundle\Service\Signature\PDFPage;
use PHPUnit\Framework\TestCase;
use Chill\DocStoreBundle\Service\Signature\PDFSignatureZone;
use Chill\DocStoreBundle\Service\Signature\PDFSignatureZoneParser;
/**
* @internal
*
* @coversNothing
*/
class PDFSignatureZoneParserTest extends TestCase
{
private static PDFSignatureZoneParser $parser;
public static function setUpBeforeClass(): void
{
self::$parser = new PDFSignatureZoneParser();
}
/**
* @dataProvider provideFiles
*
* @param list<PDFSignatureZone> $expected
*/
public function testFindSignatureZones(string $filePath, array $expected): void
{
$content = file_get_contents($filePath);
if (false === $content) {
throw new \LogicException("Unable to read file {$filePath}");
}
$actual = self::$parser->findSignatureZones($content);
self::assertEquals(count($expected), count($actual));
foreach ($actual as $index => $signatureZone) {
self::assertObjectEquals($expected[$index], $signatureZone);
}
}
public static function provideFiles(): iterable
{
yield [
__DIR__.'/data/signature_2_signature_page_1.pdf',
[
new PDFSignatureZone(
127.7,
95.289,
180.0,
180.0,
$page = new PDFPage(0, 595.30393, 841.8897)
),
new PDFSignatureZone(
269.5,
95.289,
180.0,
180.0,
$page,
),
],
];
}
}