parse broken boundaries

opcodesio · Aug 25, 2023 · 7668596 · 7668596
1 parent 94d5da9
commit 7668596
Show file tree

Hide file tree

Showing 2 changed files with 46 additions and 1 deletion.
diff --git a/src/Message.php b/src/Message.php
@@ -224,7 +224,7 @@ protected function parse()
             }
 
             if (preg_match('/^(?<key>[A-Za-z\-0-9]+): (?<value>.*)$/', $line, $matches)) {
-                if (strtolower($matches['key']) === 'content-type' && !isset($this->boundary)) {
+                if (strtolower($matches['key']) === 'content-type' && !isset($this->boundary) && !str_contains($matches['value'], 'multipart/mixed')) {
                     // this might be a single-part message. Let's start collecting the body.
                     $collectingBody = true;
                     $currentBody = '';
@@ -249,6 +249,14 @@ protected function parse()
                 continue;
             }
 
+            if (preg_match("~^--(?<boundary>[0-9A-Za-z'()+_,-./:=?]{0,68}[0-9A-Za-z'()+_,-./=?])~", $line, $matches)) {
+                $this->boundary = trim($matches['boundary']);
+                $collectingBody = true;
+                $currentBody = '';
+                $currentBodyHeaders = [];
+                continue;
+            }
+
             // The line is not part of the email message. Let's remove it altogether.
             $this->message = ltrim(substr($this->message, strlen($line)));
         }

diff --git a/tests/Unit/MessageTest.php b/tests/Unit/MessageTest.php
@@ -296,3 +296,40 @@
 EOF)
         ->and($message->getPArts()[1]->getContent())->toBe('This is a test string');
 });
+
+it('still parses with a broken boundary', function () {
+    $messageString = <<<EOF
+From: [email protected]
+To: [email protected]
+Subject: This is an email with common headers
+Date: Thu, 24 Aug 2023 21:15:01 PST
+MIME-Version: 1.0
+Content-Type: multipart/mixed; boundary¨cQXEYh
+
+--a8cQXEYh
+Content-Type: text/html; charset="utf-8"
+
+<html>
+<head>
+<title>This is an HTML email</title>
+</head>
+<body>
+<h1>This is the HTML version of the email</h1>
+</body>
+</html>--a8cQXEYh
+Content-Type: text/plain; name=test.txt
+Content-Transfer-Encoding: base64
+Content-Disposition: attachment; name=test.txt;
+ filename="test.txt"; name="test.txt"
+
+
+--a8cQXEYh--
+EOF;
+    $messageString = str_replace("\n", "\r\n", $messageString);
+
+    $message = Message::fromString($messageString);
+
+    expect($message->getParts())->toHaveCount(2)
+        ->and($message->getParts()[1]->isAttachment())->toBe(true)
+        ->and($message->getParts()[1]->getContent())->toBeEmpty();
+});