Merge pull request #3451 from dave-shawley/allow-tab-in-header

Allow horizontal tabs in header values

Author: bdarnell

tornado/test/web_test.py

@@ -706,6 +706,26 @@ def get(self):
                 raise
 
 
+class SetHeaderHandler(RequestHandler):
+    def get(self):
+        # tests the validity of web.RequestHandler._VALID_HEADER_CHARS
+        illegal_chars = [chr(o) for o in range(0, 0x20)]
+        illegal_chars.append(chr(0x7f))
+        illegal_chars.remove('\t')
+        for char in illegal_chars:
+            try:
+                self.set_header("X-Foo", "foo" + char + "bar")
+                raise Exception("Didn't get expected exception")
+            except ValueError as e:
+                if "Unsafe header value" not in str(e):
+                    raise
+
+        # an empty header value is valid as well
+        self.set_header("X-Foo", "")
+
+        self.finish(b"ok")
+
+
 class GetArgumentHandler(RequestHandler):
     def prepare(self):
         if self.get_argument("source", None) == "query":
@@ -790,6 +810,7 @@ def get_handlers(self):
             url("/header_injection", HeaderInjectionHandler),
             url("/get_argument", GetArgumentHandler),
             url("/get_arguments", GetArgumentsHandler),
+            url("/set_header", SetHeaderHandler),
         ]
         return urls
 
@@ -938,6 +959,10 @@ def test_header_injection(self):
         response = self.fetch("/header_injection")
         self.assertEqual(response.body, b"ok")
 
+    def test_set_header(self):
+        response = self.fetch("/set_header")
+        self.assertEqual(response.body, b"ok")
+
     def test_get_argument(self):
         response = self.fetch("/get_argument?foo=bar")
         self.assertEqual(response.body, b"bar")

tornado/web.py

@@ -399,7 +399,8 @@ def clear_header(self, name: str) -> None:
         if name in self._headers:
             del self._headers[name]
 
-    _INVALID_HEADER_CHAR_RE = re.compile(r"[\x00-\x1f]")
+    # https://www.rfc-editor.org/rfc/rfc9110#name-field-values
+    _VALID_HEADER_CHARS = re.compile(r"[\x09\x20-\x7e\x80-\xff]*")
 
     def _convert_header_value(self, value: _HeaderTypes) -> str:
         # Convert the input value to a str. This type check is a bit
@@ -421,7 +422,7 @@ def _convert_header_value(self, value: _HeaderTypes) -> str:
             raise TypeError("Unsupported header value %r" % value)
         # If \n is allowed into the header, it is possible to inject
         # additional headers or split the request.
-        if RequestHandler._INVALID_HEADER_CHAR_RE.search(retval):
+        if RequestHandler._VALID_HEADER_CHARS.fullmatch(retval) is None:
             raise ValueError("Unsafe header value %r", retval)
         return retval