Merge branch 'main' into fix-compression-block-decompress-error-handling

Author: greateggsgreg

Lib/locale.py

@@ -570,10 +570,6 @@ def _getdefaultlocale(envvars=('LC_ALL', 'LC_CTYPE', 'LANG', 'LANGUAGE')):
     except (ImportError, AttributeError):
         pass
     else:
-        # make sure the code/encoding values are valid
-        if sys.platform == "win32" and code and code[:2] == "0x":
-            # map windows language identifier to language name
-            code = windows_locale.get(int(code, 0))
         # ...add other platform-specific processing here, if
         # necessary...
         return code, encoding

Lib/test/test_py_compile.py

@@ -239,6 +239,14 @@ def test_quiet(self):
             with self.assertRaises(py_compile.PyCompileError):
                 py_compile.compile(bad_coding, self.pyc_path, doraise=True, quiet=1)
 
+    def test_utf7_decoded_cr_compiles(self):
+        with open(self.source_path, 'wb') as file:
+            file.write(b"#coding=U7+AA0''\n")
+
+        pyc_path = py_compile.compile(self.source_path, self.pyc_path, doraise=True)
+        self.assertEqual(pyc_path, self.pyc_path)
+        self.assertTrue(os.path.exists(self.pyc_path))
+
 
 class PyCompileTestsWithSourceEpoch(PyCompileTestsBase,
                                     unittest.TestCase,

Lib/test/test_sqlite3/test_dbapi.py

@@ -1379,6 +1379,11 @@ def test_blob_get_slice(self):
     def test_blob_get_empty_slice(self):
         self.assertEqual(self.blob[5:5], b"")
 
+    def test_blob_get_empty_slice_oob_indices(self):
+        self.cx.execute("insert into test(b) values (?)", (b"abc",))
+        with self.cx.blobopen("test", "b", 2) as blob:
+            self.assertEqual(blob[5:-5], b"")
+
     def test_blob_get_slice_negative_index(self):
         self.assertEqual(self.blob[5:-5], self.data[5:-5])
 

Misc/NEWS.d/next/Core_and_Builtins/2026-02-26-21-36-00.gh-issue-145234.w0mQ9n.rst

@@ -0,0 +1,5 @@
+Fixed a ``SystemError`` in the parser when an encoding cookie (for example,
+UTF-7) decodes to carriage returns (``\r``). Newlines are now normalized after
+decoding in the string tokenizer.
+
+Patch by Pablo Galindo.

Misc/NEWS.d/next/Library/2025-12-16-13-34-48.gh-issue-142787.wNitJX.rst

@@ -0,0 +1,2 @@
+Fix assertion failure in :mod:`sqlite3` blob subscript when slicing with
+indices that result in an empty slice.

Misc/NEWS.d/next/Library/2026-02-21-17-34-53.gh-issue-123853.6RUwWh.rst

@@ -0,0 +1 @@
+Removed Windows 95 compatibility for :func:`locale.getdefaultlocale`.

Modules/_localemodule.c

@@ -555,19 +555,7 @@ _locale__getdefaultlocale_impl(PyObject *module)
             return Py_BuildValue("ss", locale, encoding);
     }
 
-    /* If we end up here, this windows version didn't know about
-       ISO639/ISO3166 names (it's probably Windows 95).  Return the
-       Windows language identifier instead (a hexadecimal number) */
-
-    locale[0] = '0';
-    locale[1] = 'x';
-    if (GetLocaleInfoA(LOCALE_USER_DEFAULT, LOCALE_IDEFAULTLANGUAGE,
-                      locale+2, sizeof(locale)-2)) {
-        return Py_BuildValue("ss", locale, encoding);
-    }
-
     /* cannot determine the language code (very unlikely) */
-    Py_INCREF(Py_None);
     return Py_BuildValue("Os", Py_None, encoding);
 }
 #endif

Modules/_sqlite/blob.c

@@ -439,6 +439,10 @@ subscript_slice(pysqlite_Blob *self, PyObject *item)
         return NULL;
     }
 
+    if (len == 0) {
+        return PyBytes_FromStringAndSize(NULL, 0);
+    }
+
     if (step == 1) {
         return read_multiple(self, len, start);
     }

Parser/tokenizer/string_tokenizer.c

@@ -108,6 +108,19 @@ decode_str(const char *input, int single, struct tok_state *tok, int preserve_cr
     else if (!_PyTokenizer_ensure_utf8(str, tok, 1)) {
         return _PyTokenizer_error_ret(tok);
     }
+    if (utf8 != NULL) {
+        char *translated = _PyTokenizer_translate_newlines(
+            str, single, preserve_crlf, tok);
+        if (translated == NULL) {
+            Py_DECREF(utf8);
+            return _PyTokenizer_error_ret(tok);
+        }
+        PyMem_Free(tok->input);
+        tok->input = translated;
+        str = translated;
+        Py_CLEAR(utf8);
+    }
+    tok->str = str;
     assert(tok->decoding_buffer == NULL);
     tok->decoding_buffer = utf8; /* CAUTION */
     return str;