From 2c84d7875f35d3d1d0fbc170271227cc95752fa5 Mon Sep 17 00:00:00 2001 From: "Gregory P. Smith" <68491+gpshead@users.noreply.github.com> Date: Sat, 17 Jan 2026 10:23:57 -0800 Subject: [PATCH] [3.11] gh-143916: Reject control characters in wsgiref.headers.Headers (GH-143917) (GH-143973) gh-143916: Reject control characters in wsgiref.headers.Headers (GH-143917) * Add 'test.support' fixture for C0 control characters * gh-143916: Reject control characters in wsgiref.headers.Headers (cherry picked from commit f7fceed79ca1bceae8dbe5ba5bc8928564da7211) (cherry picked from commit 22e4d55285cee52bc4dbe061324e5f30bd4dee58) Co-authored-by: Gregory P. Smith <68491+gpshead@users.noreply.github.com> Co-authored-by: Seth Michael Larson --- Lib/test/support/__init__.py | 7 ++ Lib/test/test_wsgiref.py | 18 +++++ Lib/wsgiref/headers.py | 34 ++++++---- Misc/NEWS.d/next/Security/2026-01-16-11-07-36.gh-issue-143916.dpWeOD.rst | 2 4 files changed, 47 insertions(+), 14 deletions(-) create mode 100644 Misc/NEWS.d/next/Security/2026-01-16-11-07-36.gh-issue-143916.dpWeOD.rst Index: Python-3.11.14/Lib/test/support/__init__.py =================================================================== --- Python-3.11.14.orig/Lib/test/support/__init__.py 2026-02-11 23:22:45.373477280 +0100 +++ Python-3.11.14/Lib/test/support/__init__.py 2026-02-11 23:23:25.641652256 +0100 @@ -2294,3 +2294,10 @@ fails_with_expat_2_6_0 = (unittest.expectedFailure if is_expat_2_6_0 else lambda test: test) + + +def control_characters_c0() -> list[str]: + """Returns a list of C0 control characters as strings. + C0 control characters defined as the byte range 0x00-0x1F, and 0x7F. + """ + return [chr(c) for c in range(0x00, 0x20)] + ["\x7F"] Index: Python-3.11.14/Lib/test/test_wsgiref.py =================================================================== --- Python-3.11.14.orig/Lib/test/test_wsgiref.py 2026-02-11 23:22:38.512011986 +0100 +++ Python-3.11.14/Lib/test/test_wsgiref.py 2026-02-11 23:24:19.545119499 +0100 @@ -1,6 +1,6 @@ from unittest import mock from test import support -from test.support import socket_helper +from test.support import socket_helper, control_characters_c0 from test.test_httpservers import NoLogRequestHandler from unittest import TestCase from wsgiref.util import setup_testing_defaults @@ -503,6 +503,22 @@ '\r\n' ) + def testRaisesControlCharacters(self): + for c0 in control_characters_c0(): + with self.subTest(c0): + headers = Headers() + self.assertRaises(ValueError, headers.__setitem__, f"key{c0}", "val") + self.assertRaises(ValueError, headers.add_header, f"key{c0}", "val", param="param") + # HTAB (\x09) is allowed in values, not names. + if c0 == "\t": + headers["key"] = f"val{c0}" + headers.add_header("key", f"val{c0}") + headers.setdefault(f"key", f"val{c0}") + else: + self.assertRaises(ValueError, headers.__setitem__, "key", f"val{c0}") + self.assertRaises(ValueError, headers.add_header, "key", f"val{c0}", param="param") + self.assertRaises(ValueError, headers.add_header, "key", "val", param=f"param{c0}") + class ErrorHandler(BaseCGIHandler): """Simple handler subclass for testing BaseHandler""" Index: Python-3.11.14/Lib/wsgiref/headers.py =================================================================== --- Python-3.11.14.orig/Lib/wsgiref/headers.py 2026-02-11 23:22:38.927685306 +0100 +++ Python-3.11.14/Lib/wsgiref/headers.py 2026-02-11 23:24:19.545709612 +0100 @@ -9,6 +9,11 @@ # existence of which force quoting of the parameter value. import re tspecials = re.compile(r'[ \(\)<>@,;:\\"/\[\]\?=]') +# Disallowed characters for headers and values. +# HTAB (\x09) is allowed in header values, but +# not in header names. (RFC 9110 Section 5.5) +_name_disallowed_re = re.compile(r'[\x00-\x1F\x7F]') +_value_disallowed_re = re.compile(r'[\x00-\x08\x0A-\x1F\x7F]') def _formatparam(param, value=None, quote=1): """Convenience function to format and return a key=value pair. @@ -35,12 +40,15 @@ self._headers = headers if __debug__: for k, v in headers: - self._convert_string_type(k) - self._convert_string_type(v) + self._convert_string_type(k, name=True) + self._convert_string_type(v, name=False) - def _convert_string_type(self, value): + def _convert_string_type(self, value, *, name): """Convert/check value type.""" if type(value) is str: + regex = (_name_disallowed_re if name else _value_disallowed_re) + if regex.search(value): + raise ValueError("Control characters not allowed in headers") return value raise AssertionError("Header names/values must be" " of type str (got {0})".format(repr(value))) @@ -53,14 +61,14 @@ """Set the value of a header.""" del self[name] self._headers.append( - (self._convert_string_type(name), self._convert_string_type(val))) + (self._convert_string_type(name, name=True), self._convert_string_type(val, name=False))) def __delitem__(self,name): """Delete all occurrences of a header, if present. Does *not* raise an exception if the header is missing. """ - name = self._convert_string_type(name.lower()) + name = self._convert_string_type(name.lower(), name=True) self._headers[:] = [kv for kv in self._headers if kv[0].lower() != name] def __getitem__(self,name): @@ -87,13 +95,13 @@ fields deleted and re-inserted are always appended to the header list. If no fields exist with the given name, returns an empty list. """ - name = self._convert_string_type(name.lower()) + name = self._convert_string_type(name.lower(), name=True) return [kv[1] for kv in self._headers if kv[0].lower()==name] def get(self,name,default=None): """Get the first header value for 'name', or return 'default'""" - name = self._convert_string_type(name.lower()) + name = self._convert_string_type(name.lower(), name=True) for k,v in self._headers: if k.lower()==name: return v @@ -148,8 +156,8 @@ and value 'value'.""" result = self.get(name) if result is None: - self._headers.append((self._convert_string_type(name), - self._convert_string_type(value))) + self._headers.append((self._convert_string_type(name, name=True), + self._convert_string_type(value, name=False))) return value else: return result @@ -172,13 +180,13 @@ """ parts = [] if _value is not None: - _value = self._convert_string_type(_value) + _value = self._convert_string_type(_value, name=False) parts.append(_value) for k, v in _params.items(): - k = self._convert_string_type(k) + k = self._convert_string_type(k, name=True) if v is None: parts.append(k.replace('_', '-')) else: - v = self._convert_string_type(v) + v = self._convert_string_type(v, name=False) parts.append(_formatparam(k.replace('_', '-'), v)) - self._headers.append((self._convert_string_type(_name), "; ".join(parts))) + self._headers.append((self._convert_string_type(_name, name=True), "; ".join(parts))) Index: Python-3.11.14/Misc/NEWS.d/next/Security/2026-01-16-11-07-36.gh-issue-143916.dpWeOD.rst =================================================================== --- /dev/null 1970-01-01 00:00:00.000000000 +0000 +++ Python-3.11.14/Misc/NEWS.d/next/Security/2026-01-16-11-07-36.gh-issue-143916.dpWeOD.rst 2026-02-11 23:22:49.891193395 +0100 @@ -0,0 +1,2 @@ +Reject C0 control characters within wsgiref.headers.Headers fields, values, +and parameters.