pythongh-109989: Fix test_c_locale_coercion when PYTHONIOENCODING is set

This fixes the existing tests when PYTHONIOENCODING is
set by unsetting PYTHONIOENCODING.

Also add a test that explicitly checks what happens
when PYTHONIOENCODING is set.

Author: ronaldoussoren

Lib/test/test_c_locale_coercion.py

@@ -85,12 +85,16 @@ class EncodingDetails(_EncodingDetails):
     ])
 
     @classmethod
-    def get_expected_details(cls, coercion_expected, fs_encoding, stream_encoding, env_vars):
+    def get_expected_details(cls, coercion_expected, fs_encoding, stream_encoding, stream_errors, env_vars):
         """Returns expected child process details for a given encoding"""
         _stream = stream_encoding + ":{}"
-        # stdin and stdout should use surrogateescape either because the
-        # coercion triggered, or because the C locale was detected
-        stream_info = 2*[_stream.format("surrogateescape")]
+        if stream_errors is None:
+            # stdin and stdout should use surrogateescape either because the
+            # coercion triggered, or because the C locale was detected
+            stream_errors = "surrogateescape"
+
+        stream_info = [_stream.format(stream_errors)] * 2
+
         # stderr should always use backslashreplace
         stream_info.append(_stream.format("backslashreplace"))
         expected_lang = env_vars.get("LANG", "not set").lower()
@@ -189,6 +193,7 @@ def _check_child_encoding_details(self,
                                       env_vars,
                                       expected_fs_encoding,
                                       expected_stream_encoding,
+                                      expected_stream_errors,
                                       expected_warnings,
                                       coercion_expected):
         """Check the C locale handling for the given process environment
@@ -204,6 +209,7 @@ def _check_child_encoding_details(self,
             coercion_expected,
             expected_fs_encoding,
             expected_stream_encoding,
+            expected_stream_errors,
             env_vars
         )
         self.assertEqual(encoding_details, expected_details)
@@ -234,6 +240,8 @@ def test_external_target_locale_configuration(self):
             "LANG": "",
             "LC_CTYPE": "",
             "LC_ALL": "",
+            "PYTHONCOERCECLOCALE": "",
+            "PYTHONIOENCODING": "",
         }
         for env_var in ("LANG", "LC_CTYPE"):
             for locale_to_set in AVAILABLE_TARGETS:
@@ -250,10 +258,43 @@ def test_external_target_locale_configuration(self):
                     self._check_child_encoding_details(var_dict,
                                                        expected_fs_encoding,
                                                        expected_stream_encoding,
+                                                       expected_stream_errors=None,
                                                        expected_warnings=None,
                                                        coercion_expected=False)
 
+    def test_with_ioencoding(self):
+        # Explicitly setting a target locale should give the same behaviour as
+        # is seen when implicitly coercing to that target locale
+        self.maxDiff = None
 
+        expected_fs_encoding = "utf-8"
+        expected_stream_encoding = "utf-8"
+
+        base_var_dict = {
+            "LANG": "",
+            "LC_CTYPE": "",
+            "LC_ALL": "",
+            "PYTHONCOERCECLOCALE": "",
+            "PYTHONIOENCODING": "UTF-8",
+        }
+        for env_var in ("LANG", "LC_CTYPE"):
+            for locale_to_set in AVAILABLE_TARGETS:
+                # XXX (ncoghlan): LANG=UTF-8 doesn't appear to work as
+                #                 expected, so skip that combination for now
+                # See https://bugs.python.org/issue30672 for discussion
+                if env_var == "LANG" and locale_to_set == "UTF-8":
+                    continue
+
+                with self.subTest(env_var=env_var,
+                                  configured_locale=locale_to_set):
+                    var_dict = base_var_dict.copy()
+                    var_dict[env_var] = locale_to_set
+                    self._check_child_encoding_details(var_dict,
+                                                       expected_fs_encoding,
+                                                       expected_stream_encoding,
+                                                       expected_stream_errors="strict",
+                                                       expected_warnings=None,
+                                                       coercion_expected=False)
 
 @test.support.cpython_only
 @unittest.skipUnless(sysconfig.get_config_var("PY_COERCE_C_LOCALE"),
@@ -292,18 +333,43 @@ def _check_c_locale_coercion(self,
             "LANG": "",
             "LC_CTYPE": "",
             "LC_ALL": "",
+            "PYTHONCOERCECLOCALE": "",
+            "PYTHONIOENCODING": "",
         }
         base_var_dict.update(extra_vars)
-        for env_var in ("LANG", "LC_CTYPE"):
-            for locale_to_set in ("", "C", "POSIX", "invalid.ascii"):
-                # XXX (ncoghlan): *BSD platforms don't behave as expected in the
-                #                 POSIX locale, so we skip that for now
-                # See https://bugs.python.org/issue30672 for discussion
-                if locale_to_set == "POSIX":
-                    continue
+        if coerce_c_locale is not None:
+            base_var_dict["PYTHONCOERCECLOCALE"] = coerce_c_locale
+
+        # Check behaviour for the default locale
+        with self.subTest(default_locale=True,
+                          PYTHONCOERCECLOCALE=coerce_c_locale):
+            if EXPECT_COERCION_IN_DEFAULT_LOCALE:
+                _expected_warnings = expected_warnings
+                _coercion_expected = coercion_expected
+            else:
+                _expected_warnings = None
+                _coercion_expected = False
+            # On Android CLI_COERCION_WARNING is not printed when all the
+            # locale environment variables are undefined or empty. When
+            # this code path is run with environ['LC_ALL'] == 'C', then
+            # LEGACY_LOCALE_WARNING is printed.
+            if (support.is_android and
+                    _expected_warnings == [CLI_COERCION_WARNING]):
+                _expected_warnings = None
+            self._check_child_encoding_details(base_var_dict,
+                                               fs_encoding,
+                                               stream_encoding,
+                                               None,
+                                               _expected_warnings,
+                                               _coercion_expected)
+
+        # Check behaviour for explicitly configured locales
+        for locale_to_set in EXPECTED_C_LOCALE_EQUIVALENTS:
+            for env_var in ("LANG", "LC_CTYPE"):
                 with self.subTest(env_var=env_var,
                                   nominal_locale=locale_to_set,
-                                  PYTHONCOERCECLOCALE=coerce_c_locale):
+                                  PYTHONCOERCECLOCALE=coerce_c_locale,
+                                  PYTHONIOENCODING=""):
                     var_dict = base_var_dict.copy()
                     var_dict[env_var] = locale_to_set
                     if coerce_c_locale is not None:
@@ -312,6 +378,7 @@ def _check_c_locale_coercion(self,
                     self._check_child_encoding_details(var_dict,
                                                        fs_encoding,
                                                        stream_encoding,
+                                                       None,
                                                        expected_warnings,
                                                        coercion_expected)