Closing round bracket encoded in hexadecimal format breaks parsing

[krzyc]

• PHP Version: 8.2.10
• PDFParser Version: master (4b86c66)

Description:

Closing round bracket encoded in hexadecimal format breaks parsing - string is truncated.
String is truncated here:

pdfparser/src/Smalot/PdfParser/Element/ElementString.php

Lines 62 to 74 in 4b86c66

	// Find next ')' not escaped.
	$cur_start_text = $start_search_end = 0;
	while (false !== ($cur_start_pos = strpos($name, ')', $start_search_end))) {
	$cur_extract = substr($name, $cur_start_text, $cur_start_pos - $cur_start_text);
	preg_match('/(?P<escape>[\\\]*)$/s', $cur_extract, $match);
	if (!(\strlen($match['escape']) % 2)) {
	break;
	}
	$start_search_end = $cur_start_pos + 1;
	}
	// Extract string.
	$name = substr($name, 0, (int) $cur_start_pos);

Because this is my first contact with pdfparser I probably have no competence to provide a safe patch.

Test

public function testHexadecimalEncodedBracket(): void
{
    $document = new Document();

    $testString = '()';
    $content = '<< /Contents <'.bin2hex($testString).'> >>';
    $header = Header::parse($content, $document);
    $this->assertEquals($testString, (string) $header->get('Contents'));
}

Expected output & actual output

Test should pass, but returns:
Failed asserting that two strings are equal.
--- Expected
+++ Actual
@@ @@
-'()'
+'('

[k00ni]

@krzyc this looks decent enough to me to have a deeper look. Can you create a pull request with your changes + the test and we will discuss there how to proceed?

[krzyc]

@k00ni I have provided PR and I am awaiting suggestions. It works for my case (extracting binary Contents from Sig object). All tests are passing.
Edited And there are more possible problems with round square parsing, I have added another test cases, which I think should pass.