Fix Python parser chunked handling with multiple Transfer-Encoding values

CHANGES/8823.bugfix.rst

@@ -0,0 +1 @@
+Fixed Python parser chunked handling with multiple Transfer-Encoding values -- by :user:`Dreamsorcerer`.

aiohttp/http_parser.py

@@ -267,8 +267,10 @@ def __init__(
         self._headers_parser = HeadersParser(max_line_size, max_field_size, self.lax)
 
     @abc.abstractmethod
-    def parse_message(self, lines: List[bytes]) -> _MsgT:
-        pass
+    def parse_message(self, lines: List[bytes]) -> _MsgT: ...
+
+    @abc.abstractmethod
+    def _is_chunked_te(self, te: str) -> bool: ...
 
     def feed_eof(self) -> Optional[_MsgT]:
         if self._payload_parser is not None:
@@ -525,10 +527,8 @@ def parse_headers(
         # chunking
         te = headers.get(hdrs.TRANSFER_ENCODING)
         if te is not None:
-            if "chunked" == te.lower():
+            if self._is_chunked_te(te):
                 chunked = True
-            else:
-                raise BadHttpMessage("Request has invalid `Transfer-Encoding`")
 
             if hdrs.CONTENT_LENGTH in headers:
                 raise BadHttpMessage(
@@ -638,6 +638,12 @@ def parse_message(self, lines: List[bytes]) -> RawRequestMessage:
             url,
         )
 
+    def _is_chunked_te(self, te: str) -> bool:
+        if te.rsplit(",", maxsplit=1)[-1].strip(" \t").lower() == "chunked":
+            return True
+        # https://www.rfc-editor.org/rfc/rfc9112#section-6.3-2.4.3
+        raise BadHttpMessage("Request has invalid `Transfer-Encoding`")
+
 
 class HttpResponseParser(HttpParser[RawResponseMessage]):
     """Read response status line and headers.
@@ -723,6 +729,10 @@ def parse_message(self, lines: List[bytes]) -> RawResponseMessage:
             chunked,
         )
 
+    def _is_chunked_te(self, te: str) -> bool:
+        # https://www.rfc-editor.org/rfc/rfc9112#section-6.3-2.4.2
+        return te.rsplit(",", maxsplit=1)[-1].strip(" \t").lower() == "chunked"
+
 
 class HttpPayloadParser:
     def __init__(

tests/test_http_parser.py

@@ -92,6 +92,7 @@
         2**16,
         max_line_size=8190,
         max_field_size=8190,
+        read_until_eof=True,
     )
 
 
@@ -532,6 +533,23 @@
         parser.feed_data(text)
 
 
+async def test_request_te_last_chunked(parser: HttpRequestParser) -> None:
+    text = b"GET /test HTTP/1.1\r\nTransfer-Encoding: not, chunked\r\n\r\n1\r\nT\r\n3\r\nest\r\n0\r\n\r\n"
+    messages, upgrade, tail = parser.feed_data(text)
+    # https://www.rfc-editor.org/rfc/rfc9112#section-6.3-2.4.3
+    assert await messages[0][1].read() == b"Test"
+
+
+def test_request_te_first_chunked(parser: HttpRequestParser) -> None:
+    text = b"GET /test HTTP/1.1\r\nTransfer-Encoding: chunked, not\r\n\r\n1\r\nT\r\n3\r\nest\r\n0\r\n\r\n"
+    # https://www.rfc-editor.org/rfc/rfc9112#section-6.3-2.4.3
+    with pytest.raises(
+        http_exceptions.BadHttpMessage,
+        match="nvalid `Transfer-Encoding`",
+    ):
+        parser.feed_data(text)
+
+
 def test_conn_upgrade(parser: HttpRequestParser) -> None:
     text = (
         b"GET /test HTTP/1.1\r\n"
@@ -1191,6 +1209,27 @@
         response.feed_data(text)
 
 
+async def test_http_response_parser_notchunked(
+    response: HttpResponseParser,
+) -> None:
+    text = b"HTTP/1.1 200 OK\r\nTransfer-Encoding: notchunked\r\n\r\n1\r\nT\r\n3\r\nest\r\n0\r\n\r\n"
+    messages, upgrade, tail = response.feed_data(text)
+    response.feed_eof()
+
+    https://www.rfc-editor.org/rfc/rfc9112#section-6.3-2.4.2
+    assert await messages[0][1].read() == b"1\r\nT\r\n3\r\nest\r\n0\r\n\r\n"
+
+
+async def test_http_response_parser_last_chunked(
+    response: HttpResponseParser,
+) -> None:
+    text = b"HTTP/1.1 200 OK\r\nTransfer-Encoding: not, chunked\r\n\r\n1\r\nT\r\n3\r\nest\r\n0\r\n\r\n"
+    messages, upgrade, tail = response.feed_data(text)
+
+    # https://www.rfc-editor.org/rfc/rfc9112#section-6.3-2.4.2
+    assert await messages[0][1].read() == b"Test"
+
+
 def test_http_response_parser_bad(response: HttpResponseParser) -> None:
     with pytest.raises(http_exceptions.BadHttpMessage):
         response.feed_data(b"HTT/1\r\n\r\n")