bpo-34589: Add -X coerce_c_locale command line option (GH-9378)

Add a new -X coerce_c_locale command line option to control C locale
coercion (PEP 538).

4 files changed, 57 insertions, 16 deletions

diff --git a/Lib/test/test_c_locale_coercion.py b/Lib/test/test_c_locale_coercion.py
index 1db293b9c3..f62208ab20 100644
--- a/Lib/test/test_c_locale_coercion.py
+++ b/Lib/test/test_c_locale_coercion.py
@@ -139,7 +139,7 @@ class EncodingDetails(_EncodingDetails):
         return data
 
     @classmethod
-    def get_child_details(cls, env_vars):
+    def get_child_details(cls, env_vars, xoption=None):
         """Retrieves fsencoding and standard stream details from a child process
 
         Returns (encoding_details, stderr_lines):
@@ -150,10 +150,11 @@ class EncodingDetails(_EncodingDetails):
         The child is run in isolated mode if the current interpreter supports
         that.
         """
-        result, py_cmd = run_python_until_end(
-            "-X", "utf8=0", "-c", cls.CHILD_PROCESS_SCRIPT,
-            **env_vars
-        )
+        args = []
+        if xoption:
+            args.extend(("-X", f"coerce_c_locale={xoption}"))
+        args.extend(("-X", "utf8=0", "-c", cls.CHILD_PROCESS_SCRIPT))
+        result, py_cmd = run_python_until_end(*args, **env_vars)
         if not result.rc == 0:
             result.fail(py_cmd)
         # All subprocess outputs in this test case should be pure ASCII
@@ -212,7 +213,8 @@ class _LocaleHandlingTestCase(unittest.TestCase):
                                       expected_fs_encoding,
                                       expected_stream_encoding,
                                       expected_warnings,
-                                      coercion_expected):
+                                      coercion_expected,
+                                      xoption=None):
         """Check the C locale handling for the given process environment
 
         Parameters:
@@ -220,7 +222,7 @@ class _LocaleHandlingTestCase(unittest.TestCase):
             expected_stream_encoding: expected encoding for standard streams
             expected_warning: stderr output to expect (if any)
         """
-        result = EncodingDetails.get_child_details(env_vars)
+        result = EncodingDetails.get_child_details(env_vars, xoption)
         encoding_details, stderr_lines = result
         expected_details = EncodingDetails.get_expected_details(
             coercion_expected,
@@ -290,6 +292,7 @@ class LocaleCoercionTests(_LocaleHandlingTestCase):
                                  coerce_c_locale,
                                  expected_warnings=None,
                                  coercion_expected=True,
+                                 use_xoption=False,
                                  **extra_vars):
         """Check the C locale handling for various configurations
 
@@ -319,8 +322,12 @@ class LocaleCoercionTests(_LocaleHandlingTestCase):
             "PYTHONCOERCECLOCALE": "",
         }
         base_var_dict.update(extra_vars)
+        xoption = None
         if coerce_c_locale is not None:
-            base_var_dict["PYTHONCOERCECLOCALE"] = coerce_c_locale
+            if use_xoption:
+                xoption = coerce_c_locale
+            else:
+                base_var_dict["PYTHONCOERCECLOCALE"] = coerce_c_locale
 
         # Check behaviour for the default locale
         with self.subTest(default_locale=True,
@@ -342,7 +349,8 @@ class LocaleCoercionTests(_LocaleHandlingTestCase):
                                                fs_encoding,
                                                stream_encoding,
                                                _expected_warnings,
-                                               _coercion_expected)
+                                               _coercion_expected,
+                                               xoption=xoption)
 
         # Check behaviour for explicitly configured locales
         for locale_to_set in EXPECTED_C_LOCALE_EQUIVALENTS:
@@ -357,7 +365,8 @@ class LocaleCoercionTests(_LocaleHandlingTestCase):
                                                        fs_encoding,
                                                        stream_encoding,
                                                        expected_warnings,
-                                                       coercion_expected)
+                                                       coercion_expected,
+                                                       xoption=xoption)
 
     def test_PYTHONCOERCECLOCALE_not_set(self):
         # This should coerce to the first available target locale by default
@@ -404,6 +413,32 @@ class LocaleCoercionTests(_LocaleHandlingTestCase):
                                       expected_warnings=[LEGACY_LOCALE_WARNING],
                                       coercion_expected=False)
 
+    def test_xoption_set_to_1(self):
+        self._check_c_locale_coercion("utf-8", "utf-8", coerce_c_locale="1",
+                                      use_xoption=True)
+
+    def test_xoption_set_to_zero(self):
+        # The setting "0" should result in the locale coercion being disabled
+        self._check_c_locale_coercion(EXPECTED_C_LOCALE_FS_ENCODING,
+                                      EXPECTED_C_LOCALE_STREAM_ENCODING,
+                                      coerce_c_locale="0",
+                                      coercion_expected=False,
+                                      use_xoption=True)
+        # Setting LC_ALL=C shouldn't make any difference to the behaviour
+        self._check_c_locale_coercion(EXPECTED_C_LOCALE_FS_ENCODING,
+                                      EXPECTED_C_LOCALE_STREAM_ENCODING,
+                                      coerce_c_locale="0",
+                                      LC_ALL="C",
+                                      coercion_expected=False,
+                                      use_xoption=True)
+
+    def test_xoption_set_to_warn(self):
+        # -X coerce_c_locale=warn enables runtime warnings for legacy locales
+        self._check_c_locale_coercion("utf-8", "utf-8",
+                                      coerce_c_locale="warn",
+                                      expected_warnings=[CLI_COERCION_WARNING],
+                                      use_xoption=True)
+
 def test_main():
     test.support.run_unittest(
         LocaleConfigurationTests,
diff --git a/Lib/test/test_cmd_line.py b/Lib/test/test_cmd_line.py
index 21511b896c..7e967b20ab 100644
--- a/Lib/test/test_cmd_line.py
+++ b/Lib/test/test_cmd_line.py
@@ -159,13 +159,16 @@ class CmdLineTest(unittest.TestCase):
         env = os.environ.copy()
         # Use C locale to get ascii for the locale encoding
         env['LC_ALL'] = 'C'
-        env['PYTHONCOERCECLOCALE'] = '0'
         code = (
             b'import locale; '
             b'print(ascii("' + undecodable + b'"), '
                 b'locale.getpreferredencoding())')
         p = subprocess.Popen(
-            [sys.executable, "-c", code],
+            [sys.executable,
+             # Disable C locale coercion and UTF-8 Mode to not use UTF-8
+             "-X", "coerce_c_locale=0",
+             "-X", "utf8=0",
+             "-c", code],
             stdout=subprocess.PIPE, stderr=subprocess.STDOUT,
             env=env)
         stdout, stderr = p.communicate()
diff --git a/Lib/test/test_sys.py b/Lib/test/test_sys.py
index b90366d814..a7f2928271 100644
--- a/Lib/test/test_sys.py
+++ b/Lib/test/test_sys.py
@@ -656,9 +656,8 @@ class SysModuleTest(unittest.TestCase):
 
     def c_locale_get_error_handler(self, locale, isolated=False, encoding=None):
         # Force the POSIX locale
-        env = os.environ.copy()
+        env = dict(os.environ)
         env["LC_ALL"] = locale
-        env["PYTHONCOERCECLOCALE"] = "0"
         code = '\n'.join((
             'import sys',
             'def dump(name):',
@@ -668,7 +667,10 @@ class SysModuleTest(unittest.TestCase):
             'dump("stdout")',
             'dump("stderr")',
         ))
-        args = [sys.executable, "-X", "utf8=0", "-c", code]
+        args = [sys.executable,
+                "-X", "utf8=0",
+                "-X", "coerce_c_locale=0",
+                "-c", code]
         if isolated:
             args.append("-I")
         if encoding is not None:
diff --git a/Lib/test/test_utf8_mode.py b/Lib/test/test_utf8_mode.py
index 7280ce77ef..c3cbb49060 100644
--- a/Lib/test/test_utf8_mode.py
+++ b/Lib/test/test_utf8_mode.py
@@ -27,6 +27,8 @@ class UTF8ModeTests(unittest.TestCase):
         return (loc in POSIX_LOCALES)
 
     def get_output(self, *args, failure=False, **kw):
+        # Always disable the C locale coercion (PEP 538)
+        args = ('-X', 'coerce_c_locale=0', *args)
         kw = dict(self.DEFAULT_ENV, **kw)
         if failure:
             out = assert_python_failure(*args, **kw)
@@ -116,7 +118,6 @@ class UTF8ModeTests(unittest.TestCase):
             # PYTHONLEGACYWINDOWSFSENCODING disables the UTF-8 mode
             # and has the priority over -X utf8 and PYTHONUTF8
             out = self.get_output('-X', 'utf8', '-c', code,
-                                  PYTHONUTF8='strict',
                                   PYTHONLEGACYWINDOWSFSENCODING='1')
             self.assertEqual(out, 'mbcs/replace')