Fix hexadecimal decoding (issue smalot#715)

krzyc · May 27, 2024 · a604278 · a604278
1 parent 4b86c66
commit a604278
Show file tree

Hide file tree

Showing 5 changed files with 22 additions and 5 deletions.
diff --git a/samples/bugs/Issue715.pdf b/samples/bugs/Issue715.pdf
diff --git a/src/Smalot/PdfParser/Element/ElementHexa.php b/src/Smalot/PdfParser/Element/ElementHexa.php
@@ -47,12 +47,11 @@ public static function parse(string $content, ?Document $document = null, int &$
         if (preg_match('/^\s*\<(?P<name>[A-F0-9]+)\>/is', $content, $match)) {
             $name = $match['name'];
             $offset += strpos($content, '<'.$name) + \strlen($name) + 2; // 1 for '>'
-            // repackage string as standard
-            $name = '('.self::decode($name).')';
-            $element = ElementDate::parse($name, $document);
+            $name = self::decode($name);
+            $element = ElementDate::parse('('.$name.')', $document);
 
             if (!$element) {
-                $element = ElementString::parse($name, $document);
+                $element = new self($name);
             }
 
             return $element;

diff --git a/src/Smalot/PdfParser/Parser.php b/src/Smalot/PdfParser/Parser.php
@@ -296,7 +296,7 @@ protected function parseHeaderElement(?string $type, $value, ?Document $document
                 return ElementString::parse('('.$value.')', $document);
 
             case '<':
-                return $this->parseHeaderElement('(', ElementHexa::decode($value), $document);
+                return ElementHexa::parse('<'.$value.'>', $document);
 
             case '/':
                 return ElementName::parse('/'.$value, $document);

diff --git a/tests/PHPUnit/Integration/Element/ElementHexaTest.php b/tests/PHPUnit/Integration/Element/ElementHexaTest.php
@@ -130,5 +130,9 @@ public function testParse(): void
 007000610  073007100750061002c0020> ');
 
         $this->assertEquals('pasqua, primavera, resurrezione, festa cristiana, gesù, uova di cioccolata, coniglietti, pulcini, pasquale, campane, dina rebucci, uova di pasqua, ', $element);
+
+        $testString = '()\\';
+        $element = ElementHexa::parse('<'.bin2hex($testString).'>', null, $offset);
+        $this->assertEquals($testString, (string) $element);
     }
 }
diff --git a/tests/PHPUnit/Integration/ParserTest.php b/tests/PHPUnit/Integration/ParserTest.php
@@ -438,6 +438,20 @@ public function testIgnoreEncryption(): void
 
         // without the configuration option set, an exception would be thrown.
     }
+
+    /**
+     * Tests special chars encoded as hex.
+     */
+    public function testSpecialCharsEncodedAsHex(): void
+    {
+        $filename = $this->rootDir.'/samples/bugs/Issue715.pdf';
+
+        $this->fixture = new Parser();
+        $document = $this->fixture->parseFile($filename);
+        $result = (string) ($document->getObjectsByType('Sig')['4_0'] ?? null)?->getHeader()->get('Contents');
+
+        $this->assertEquals('()\\', $result);
+    }
 }
 
 class ParserSub extends Parser