Merge pull request #39 from nanjekyejoannah/warn_unicode_parsing

Add state tracking for unicode
softdevteam · Apr 9, 2024 · 71af9e4 · 71af9e4
2 parents b03f43c + 9312ad1
commit 71af9e4
Show file tree

Hide file tree

Showing 2 changed files with 19 additions and 7 deletions.
diff --git a/Lib/test/test_unicode.py b/Lib/test/test_unicode.py
@@ -743,13 +743,13 @@ def __str__(self):
             self.assertRaises(
                 TypeError,
                 unicode,
-                u'decoding unicode is not supported',
+                u"'decode()' is not supported on Unicode in 3.x: convert the string to bytes.",
                 'utf-8',
                 'strict'
             )
 
         self.assertEqual(
-            unicode('strings are decoded to unicode', 'utf-8', 'strict'),
+            unicode("'decode()' is not supported on Unicode in 3.x: convert the string to bytes.", 'utf-8', 'strict'),
             u'strings are decoded to unicode'
         )
 

diff --git a/Objects/unicodeobject.c b/Objects/unicodeobject.c
@@ -1210,9 +1210,15 @@ PyObject *PyUnicode_FromEncodedObject(register PyObject *obj,
     }
 #else
     if (PyUnicode_Check(obj)) {
-        PyErr_SetString(PyExc_TypeError,
-                        "decoding Unicode is not supported");
-        return NULL;
+        if (PyString_CheckExact(self)) {
+            self->ob_bstate = BSTATE_BYTE;
+        }
+
+        if ((self->ob_bstate == BSTATE_BYTE) &&
+            PyErr_WarnPy3k(
+                "'decode()' is not supported on Unicode in 3.x: convert the string to bytes.", 1) < 0) {
+            return NULL;
+        }
     }
 #endif
 
@@ -1304,8 +1310,14 @@ PyObject *PyUnicode_AsDecodedObject(PyObject *unicode,
         goto onError;
     }
 
-    if (PyErr_WarnPy3k("decoding Unicode is not supported in 3.x", 1) < 0)
-        goto onError;
+    if (PyString_CheckExact(self)) {
+        self->ob_bstate = BSTATE_BYTE;
+    }
+
+    if ((self->ob_bstate == BSTATE_BYTE) &&
+        PyErr_WarnPy3k("'decode()' is not supported on Unicode in 3.x: convert the string to bytes.", 1) < 0) {
+        return NULL;
+    }
 
     if (encoding == NULL)
         encoding = PyUnicode_GetDefaultEncoding();