Bump bundled llhttp to 9.2.1 (#113)

CVE-2024-27982

Expose leniency flags via the new `set_dangerous_leniencies` parser
method if somebody needs to opt into the old vulnerable behavior.

Fixes: #111

httptools/parser/cparser.pxd

@@ -154,3 +154,14 @@ cdef extern from "llhttp.h":
     const char* llhttp_method_name(llhttp_method_t method)
 
     void llhttp_set_error_reason(llhttp_t* parser, const char* reason);
+
+    void llhttp_set_lenient_headers(llhttp_t* parser, bint enabled);
+    void llhttp_set_lenient_chunked_length(llhttp_t* parser, bint enabled);
+    void llhttp_set_lenient_keep_alive(llhttp_t* parser, bint enabled);
+    void llhttp_set_lenient_transfer_encoding(llhttp_t* parser, bint enabled);
+    void llhttp_set_lenient_version(llhttp_t* parser, bint enabled);
+    void llhttp_set_lenient_data_after_close(llhttp_t* parser, bint enabled);
+    void llhttp_set_lenient_optional_lf_after_cr(llhttp_t* parser, bint enabled);
+    void llhttp_set_lenient_optional_cr_before_lf(llhttp_t* parser, bint enabled);
+    void llhttp_set_lenient_optional_crlf_after_chunk(llhttp_t* parser, bint enabled);
+    void llhttp_set_lenient_spaces_after_chunk_size(llhttp_t* parser, bint enabled);

httptools/parser/parser.pyx

@@ -1,6 +1,8 @@
 #cython: language_level=3
 
 from __future__ import print_function
+from typing import Optional
+
 from cpython.mem cimport PyMem_Malloc, PyMem_Free
 from cpython cimport PyObject_GetBuffer, PyBuffer_Release, PyBUF_SIMPLE, \
                      Py_buffer, PyBytes_AsString
@@ -144,6 +146,51 @@ cdef class HttpParser:
 
     ### Public API ###
 
+    def set_dangerous_leniencies(
+        self,
+        lenient_headers: Optional[bool] = None,
+        lenient_chunked_length: Optional[bool] = None,
+        lenient_keep_alive: Optional[bool] = None,
+        lenient_transfer_encoding: Optional[bool] = None,
+        lenient_version: Optional[bool] = None,
+        lenient_data_after_close: Optional[bool] = None,
+        lenient_optional_lf_after_cr: Optional[bool] = None,
+        lenient_optional_cr_before_lf: Optional[bool] = None,
+        lenient_optional_crlf_after_chunk: Optional[bool] = None,
+        lenient_spaces_after_chunk_size: Optional[bool] = None,
+    ):
+        cdef cparser.llhttp_t* parser = self._cparser
+        if lenient_headers is not None:
+            cparser.llhttp_set_lenient_headers(
+                parser, lenient_headers)
+        if lenient_chunked_length is not None:
+            cparser.llhttp_set_lenient_chunked_length(
+                parser, lenient_chunked_length)
+        if lenient_keep_alive is not None:
+            cparser.llhttp_set_lenient_keep_alive(
+                parser, lenient_keep_alive)
+        if lenient_transfer_encoding is not None:
+            cparser.llhttp_set_lenient_transfer_encoding(
+                parser, lenient_transfer_encoding)
+        if lenient_version is not None:
+            cparser.llhttp_set_lenient_version(
+                parser, lenient_version)
+        if lenient_data_after_close is not None:
+            cparser.llhttp_set_lenient_data_after_close(
+                parser, lenient_data_after_close)
+        if lenient_optional_lf_after_cr is not None:
+            cparser.llhttp_set_lenient_optional_lf_after_cr(
+                parser, lenient_optional_lf_after_cr)
+        if lenient_optional_cr_before_lf is not None:
+            cparser.llhttp_set_lenient_optional_cr_before_lf(
+                parser, lenient_optional_cr_before_lf)
+        if lenient_optional_crlf_after_chunk is not None:
+            cparser.llhttp_set_lenient_optional_crlf_after_chunk(
+                parser, lenient_optional_crlf_after_chunk)
+        if lenient_spaces_after_chunk_size is not None:
+            cparser.llhttp_set_lenient_spaces_after_chunk_size(
+                parser, lenient_spaces_after_chunk_size)
+
     def get_http_version(self):
         cdef cparser.llhttp_t* parser = self._cparser
         return '{}.{}'.format(parser.http_major, parser.http_minor)
@@ -161,7 +208,7 @@ cdef class HttpParser:
             cparser.llhttp_errno_t err
             Py_buffer *buf
             bint owning_buf = False
-            char* err_pos
+            const char* err_pos
 
         if PyMemoryView_Check(data):
             buf = PyMemoryView_GET_BUFFER(data)

tests/test_parser.py

@@ -6,6 +6,18 @@
 
 RESPONSE1_HEAD = b'''HTTP/1.1 200 OK
 Date: Mon, 23 May 2005 22:38:34 GMT
+Server: Apache/1.3.3.7 (Unix) (Red-Hat/Linux)
+Last-Modified: Wed, 08 Jan 2003 23:11:55 GMT
+ETag: "3f80f-1b6-3e1cb03b"
+Content-Type: text/html; charset=UTF-8
+Content-Length: 130
+Accept-Ranges: bytes
+Connection: close
+
+'''.replace(b'\n', b'\r\n')
+
+RESPONSE1_SPACES_IN_HEAD = b'''HTTP/1.1 200 OK
+Date: Mon, 23 May 2005 22:38:34 GMT
 Server: Apache/1.3.3.7
         (Unix) (Red-Hat/Linux)
 Last-Modified: Wed, 08 Jan 2003 23:11:55 GMT
@@ -89,7 +101,7 @@ def test_parser_response_1(self):
         self.assertEqual(len(headers), 8)
         self.assertEqual(headers.get(b'Connection'), b'close')
         self.assertEqual(headers.get(b'Content-Type'),
-                         b'text/html;  charset=UTF-8')
+                         b'text/html; charset=UTF-8')
 
         self.assertFalse(m.on_body.called)
         p.feed_data(bytearray(RESPONSE1_BODY))
@@ -109,6 +121,53 @@ def test_parser_response_1b(self):
                 'Expected HTTP/'):
             p.feed_data(b'12123123')
 
+    def test_parser_response_leninent_headers_1(self):
+        m = mock.Mock()
+
+        headers = {}
+        m.on_header.side_effect = headers.__setitem__
+
+        p = httptools.HttpResponseParser(m)
+
+        with self.assertRaisesRegex(
+            httptools.HttpParserError,
+            "whitespace after header value",
+        ):
+            p.feed_data(memoryview(RESPONSE1_SPACES_IN_HEAD))
+
+    def test_parser_response_leninent_headers_2(self):
+        m = mock.Mock()
+
+        headers = {}
+        m.on_header.side_effect = headers.__setitem__
+
+        p = httptools.HttpResponseParser(m)
+
+        p.set_dangerous_leniencies(lenient_headers=True)
+        p.feed_data(memoryview(RESPONSE1_SPACES_IN_HEAD))
+
+        self.assertEqual(p.get_http_version(), '1.1')
+        self.assertEqual(p.get_status_code(), 200)
+
+        m.on_status.assert_called_once_with(b'OK')
+
+        m.on_headers_complete.assert_called_once_with()
+        self.assertEqual(m.on_header.call_count, 8)
+        self.assertEqual(len(headers), 8)
+        self.assertEqual(headers.get(b'Connection'), b'close')
+        self.assertEqual(headers.get(b'Content-Type'),
+                         b'text/html;  charset=UTF-8')
+
+        self.assertFalse(m.on_body.called)
+        p.feed_data(bytearray(RESPONSE1_BODY))
+        m.on_body.assert_called_once_with(RESPONSE1_BODY)
+
+        m.on_message_complete.assert_called_once_with()
+
+        self.assertFalse(m.on_url.called)
+        self.assertFalse(m.on_chunk_header.called)
+        self.assertFalse(m.on_chunk_complete.called)
+
     def test_parser_response_2(self):
         with self.assertRaisesRegex(TypeError, 'a bytes-like object'):
             httptools.HttpResponseParser(None).feed_data('')