pythongh-117557: Improve error messages when a string, bytes or bytearray of length 1 are expected (pythonGH-117631)

Author: serhiy-storchaka

Lib/test/clinic.test.c

@@ -2,7 +2,7 @@
 import sys
 import unittest
 from ctypes import (CDLL, Structure, Array, CFUNCTYPE,
-                    byref, POINTER, pointer, ArgumentError,
+                    byref, POINTER, pointer, ArgumentError, sizeof,
                     c_char, c_wchar, c_byte, c_char_p, c_wchar_p,
                     c_short, c_int, c_long, c_longlong, c_void_p,
                     c_float, c_double, c_longdouble)
@@ -72,7 +72,8 @@ def callback(*args):
 
         self.assertEqual(str(cm.exception),
                          "argument 1: TypeError: one character bytes, "
-                         "bytearray or integer expected")
+                         "bytearray, or an integer in range(256) expected, "
+                         "not bytes of length 3")
 
     def test_wchar_parm(self):
         f = dll._testfunc_i_bhilfd
@@ -84,14 +85,27 @@ def test_wchar_parm(self):
         with self.assertRaises(ArgumentError) as cm:
             f(1, 2, 3, 4, 5.0, 6.0)
         self.assertEqual(str(cm.exception),
-                         "argument 2: TypeError: unicode string expected "
-                         "instead of int instance")
+                         "argument 2: TypeError: a unicode character expected, "
+                         "not instance of int")
 
         with self.assertRaises(ArgumentError) as cm:
             f(1, "abc", 3, 4, 5.0, 6.0)
         self.assertEqual(str(cm.exception),
-                         "argument 2: TypeError: one character unicode string "
-                         "expected")
+                         "argument 2: TypeError: a unicode character expected, "
+                         "not a string of length 3")
+
+        with self.assertRaises(ArgumentError) as cm:
+            f(1, "", 3, 4, 5.0, 6.0)
+        self.assertEqual(str(cm.exception),
+                         "argument 2: TypeError: a unicode character expected, "
+                         "not a string of length 0")
+
+        if sizeof(c_wchar) < 4:
+            with self.assertRaises(ArgumentError) as cm:
+                f(1, "\U0001f40d", 3, 4, 5.0, 6.0)
+            self.assertEqual(str(cm.exception),
+                             "argument 2: TypeError: the string '\\U0001f40d' "
+                             "cannot be converted to a single wchar_t character")
 
     def test_c_char_p_parm(self):
         """Test the error message when converting an incompatible type to c_char_p."""

Lib/test/test_ctypes/test_functions.py

@@ -3,7 +3,7 @@
 from ctypes import (CDLL, PyDLL, ArgumentError,
                     Structure, Array, Union,
                     _Pointer, _SimpleCData, _CFuncPtr,
-                    POINTER, pointer, byref,
+                    POINTER, pointer, byref, sizeof,
                     c_void_p, c_char_p, c_wchar_p, py_object,
                     c_bool,
                     c_char, c_wchar,
@@ -87,19 +87,33 @@ def test_c_char(self):
         with self.assertRaises(TypeError) as cm:
             c_char.from_param(b"abc")
         self.assertEqual(str(cm.exception),
-                         "one character bytes, bytearray or integer expected")
+                         "one character bytes, bytearray, or an integer "
+                         "in range(256) expected, not bytes of length 3")
 
     def test_c_wchar(self):
         with self.assertRaises(TypeError) as cm:
             c_wchar.from_param("abc")
         self.assertEqual(str(cm.exception),
-                         "one character unicode string expected")
+                         "a unicode character expected, not a string of length 3")
 
+        with self.assertRaises(TypeError) as cm:
+            c_wchar.from_param("")
+        self.assertEqual(str(cm.exception),
+                         "a unicode character expected, not a string of length 0")
 
         with self.assertRaises(TypeError) as cm:
             c_wchar.from_param(123)
         self.assertEqual(str(cm.exception),
-                         "unicode string expected instead of int instance")
+                         "a unicode character expected, not instance of int")
+
+        if sizeof(c_wchar) < 4:
+            with self.assertRaises(TypeError) as cm:
+                c_wchar.from_param('\U0001f40d')
+            self.assertEqual(str(cm.exception),
+                             "the string '\\U0001f40d' cannot be converted to "
+                             "a single wchar_t character")
+
+
 
     def test_int_pointers(self):
         LPINT = POINTER(c_int)

Lib/test/test_ctypes/test_parameters.py

@@ -1578,7 +1578,7 @@ def __int__(self):
         self.assertRaisesRegex(TypeError, '%u format: a real number is required, not complex', operator.mod, '%u', 3j)
         self.assertRaisesRegex(TypeError, '%i format: a real number is required, not complex', operator.mod, '%i', 2j)
         self.assertRaisesRegex(TypeError, '%d format: a real number is required, not complex', operator.mod, '%d', 1j)
-        self.assertRaisesRegex(TypeError, '%c requires int or char', operator.mod, '%c', pi)
+        self.assertRaisesRegex(TypeError, r'%c requires an int or a unicode character, not .*\.PseudoFloat', operator.mod, '%c', pi)
 
         class RaisingNumber:
             def __int__(self):

Lib/test/test_str.py

@@ -0,0 +1,2 @@
+Improve error messages when a string, bytes or bytearray object of length 1
+is expected.

Misc/NEWS.d/next/Core and Builtins/2024-04-08-13-49-02.gh-issue-117558.9lSEpR.rst

@@ -1100,25 +1100,45 @@ O_set(void *ptr, PyObject *value, Py_ssize_t size)
 static PyObject *
 c_set(void *ptr, PyObject *value, Py_ssize_t size)
 {
-    if (PyBytes_Check(value) && PyBytes_GET_SIZE(value) == 1) {
+    if (PyBytes_Check(value)) {
+        if (PyBytes_GET_SIZE(value) != 1) {
+            PyErr_Format(PyExc_TypeError,
+                        "one character bytes, bytearray, or an integer "
+                        "in range(256) expected, not bytes of length %zd",
+                        PyBytes_GET_SIZE(value));
+            return NULL;
+        }
         *(char *)ptr = PyBytes_AS_STRING(value)[0];
         _RET(value);
     }
-    if (PyByteArray_Check(value) && PyByteArray_GET_SIZE(value) == 1) {
+    if (PyByteArray_Check(value)) {
+        if (PyByteArray_GET_SIZE(value) != 1) {
+            PyErr_Format(PyExc_TypeError,
+                        "one character bytes, bytearray, or an integer "
+                        "in range(256) expected, not bytearray of length %zd",
+                        PyByteArray_GET_SIZE(value));
+            return NULL;
+        }
         *(char *)ptr = PyByteArray_AS_STRING(value)[0];
         _RET(value);
     }
-    if (PyLong_Check(value))
-    {
-        long longval = PyLong_AsLong(value);
-        if (longval < 0 || longval >= 256)
-            goto error;
+    if (PyLong_Check(value)) {
+        int overflow;
+        long longval = PyLong_AsLongAndOverflow(value, &overflow);
+        if (longval == -1 && PyErr_Occurred()) {
+            return NULL;
+        }
+        if (overflow || longval < 0 || longval >= 256) {
+            PyErr_SetString(PyExc_TypeError, "integer not in range(256)");
+            return NULL;
+        }
         *(char *)ptr = (char)longval;
         _RET(value);
     }
-  error:
     PyErr_Format(PyExc_TypeError,
-                 "one character bytes, bytearray or integer expected");
+                 "one character bytes, bytearray, or an integer "
+                 "in range(256) expected, not %T",
+                 value);
     return NULL;
 }
 
@@ -1137,22 +1157,27 @@ u_set(void *ptr, PyObject *value, Py_ssize_t size)
     wchar_t chars[2];
     if (!PyUnicode_Check(value)) {
         PyErr_Format(PyExc_TypeError,
-                        "unicode string expected instead of %s instance",
-                        Py_TYPE(value)->tp_name);
+                     "a unicode character expected, not instance of %T",
+                     value);
         return NULL;
-    } else
-        Py_INCREF(value);
+    }
 
     len = PyUnicode_AsWideChar(value, chars, 2);
     if (len != 1) {
-        Py_DECREF(value);
-        PyErr_SetString(PyExc_TypeError,
-                        "one character unicode string expected");
+        if (PyUnicode_GET_LENGTH(value) != 1) {
+            PyErr_Format(PyExc_TypeError,
+                         "a unicode character expected, not a string of length %zd",
+                         PyUnicode_GET_LENGTH(value));
+        }
+        else {
+            PyErr_Format(PyExc_TypeError,
+                         "the string %A cannot be converted to a single wchar_t character",
+                         value);
+        }
         return NULL;
     }
 
     *(wchar_t *)ptr = chars[0];
-    Py_DECREF(value);
 
     _RET(value);
 }

Modules/_ctypes/cfield.c

@@ -233,13 +233,20 @@ static int
 PyCurses_ConvertToChtype(PyCursesWindowObject *win, PyObject *obj, chtype *ch)
 {
     long value;
-    if(PyBytes_Check(obj) && PyBytes_Size(obj) == 1) {
+    if (PyBytes_Check(obj)) {
+        if (PyBytes_GET_SIZE(obj) != 1) {
+            PyErr_Format(PyExc_TypeError,
+                         "expect int or bytes or str of length 1, "
+                         "got a bytes of length %zd",
+                         PyBytes_GET_SIZE(obj));
+            return 0;
+        }
         value = (unsigned char)PyBytes_AsString(obj)[0];
     }
     else if (PyUnicode_Check(obj)) {
-        if (PyUnicode_GetLength(obj) != 1) {
+        if (PyUnicode_GET_LENGTH(obj) != 1) {
             PyErr_Format(PyExc_TypeError,
-                         "expect bytes or str of length 1, or int, "
+                         "expect int or bytes or str of length 1, "
                          "got a str of length %zi",
                          PyUnicode_GET_LENGTH(obj));
             return 0;
@@ -272,7 +279,7 @@ PyCurses_ConvertToChtype(PyCursesWindowObject *win, PyObject *obj, chtype *ch)
     }
     else {
         PyErr_Format(PyExc_TypeError,
-                     "expect bytes or str of length 1, or int, got %s",
+                     "expect int or bytes or str of length 1, got %s",
                      Py_TYPE(obj)->tp_name);
         return 0;
     }
@@ -315,7 +322,7 @@ PyCurses_ConvertToCchar_t(PyCursesWindowObject *win, PyObject *obj,
 #ifdef HAVE_NCURSESW
         if (PyUnicode_AsWideChar(obj, buffer, 2) != 1) {
             PyErr_Format(PyExc_TypeError,
-                         "expect bytes or str of length 1, or int, "
+                         "expect int or bytes or str of length 1, "
                          "got a str of length %zi",
                          PyUnicode_GET_LENGTH(obj));
             return 0;
@@ -326,7 +333,14 @@ PyCurses_ConvertToCchar_t(PyCursesWindowObject *win, PyObject *obj,
         return PyCurses_ConvertToChtype(win, obj, ch);
 #endif
     }
-    else if(PyBytes_Check(obj) && PyBytes_Size(obj) == 1) {
+    else if (PyBytes_Check(obj)) {
+        if (PyBytes_GET_SIZE(obj) != 1) {
+            PyErr_Format(PyExc_TypeError,
+                         "expect int or bytes or str of length 1, "
+                         "got a bytes of length %zd",
+                         PyBytes_GET_SIZE(obj));
+            return 0;
+        }
         value = (unsigned char)PyBytes_AsString(obj)[0];
     }
     else if (PyLong_CheckExact(obj)) {
@@ -340,7 +354,7 @@ PyCurses_ConvertToCchar_t(PyCursesWindowObject *win, PyObject *obj,
     }
     else {
         PyErr_Format(PyExc_TypeError,
-                     "expect bytes or str of length 1, or int, got %s",
+                     "expect int or bytes or str of length 1, got %s",
                      Py_TYPE(obj)->tp_name);
         return 0;
     }
@@ -4443,7 +4457,7 @@ PyCurses_ConvertToWchar_t(PyObject *obj,
         wchar_t buffer[2];
         if (PyUnicode_AsWideChar(obj, buffer, 2) != 1) {
             PyErr_Format(PyExc_TypeError,
-                         "expect str of length 1 or int, "
+                         "expect int or str of length 1, "
                          "got a str of length %zi",
                          PyUnicode_GET_LENGTH(obj));
             return 0;
@@ -4470,7 +4484,7 @@ PyCurses_ConvertToWchar_t(PyObject *obj,
     }
     else {
         PyErr_Format(PyExc_TypeError,
-                     "expect str of length 1 or int, got %s",
+                     "expect int or str of length 1, got %s",
                      Py_TYPE(obj)->tp_name);
         return 0;
     }

Modules/_cursesmodule.c

@@ -260,20 +260,32 @@ u_getitem(arrayobject *ap, Py_ssize_t i)
 static int
 u_setitem(arrayobject *ap, Py_ssize_t i, PyObject *v)
 {
-    PyObject *u;
-    if (!PyArg_Parse(v, "U;array item must be unicode character", &u)) {
+    if (!PyUnicode_Check(v)) {
+        PyErr_Format(PyExc_TypeError,
+                     "array item must be a unicode character, not %T",
+                     v);
         return -1;
     }
 
-    Py_ssize_t len = PyUnicode_AsWideChar(u, NULL, 0);
+    Py_ssize_t len = PyUnicode_AsWideChar(v, NULL, 0);
     if (len != 2) {
-        PyErr_SetString(PyExc_TypeError,
-                        "array item must be unicode character");
+        if (PyUnicode_GET_LENGTH(v) != 1) {
+            PyErr_Format(PyExc_TypeError,
+                         "array item must be a unicode character, "
+                         "not a string of length %zd",
+                         PyUnicode_GET_LENGTH(v));
+        }
+        else {
+            PyErr_Format(PyExc_TypeError,
+                         "string %A cannot be converted to "
+                         "a single wchar_t character",
+                         v);
+        }
         return -1;
     }
 
     wchar_t w;
-    len = PyUnicode_AsWideChar(u, &w, 1);
+    len = PyUnicode_AsWideChar(v, &w, 1);
     assert(len == 1);
 
     if (i >= 0) {
@@ -291,19 +303,23 @@ w_getitem(arrayobject *ap, Py_ssize_t i)
 static int
 w_setitem(arrayobject *ap, Py_ssize_t i, PyObject *v)
 {
-    PyObject *u;
-    if (!PyArg_Parse(v, "U;array item must be unicode character", &u)) {
+    if (!PyUnicode_Check(v)) {
+        PyErr_Format(PyExc_TypeError,
+                     "array item must be a unicode character, not %T",
+                     v);
         return -1;
     }
 
-    if (PyUnicode_GetLength(u) != 1) {
-        PyErr_SetString(PyExc_TypeError,
-                        "array item must be unicode character");
+    if (PyUnicode_GET_LENGTH(v) != 1) {
+        PyErr_Format(PyExc_TypeError,
+                     "array item must be a unicode character, "
+                     "not a string of length %zd",
+                     PyUnicode_GET_LENGTH(v));
         return -1;
     }
 
     if (i >= 0) {
-        ((Py_UCS4 *)ap->ob_item)[i] = PyUnicode_READ_CHAR(u, 0);
+        ((Py_UCS4 *)ap->ob_item)[i] = PyUnicode_READ_CHAR(v, 0);
     }
     return 0;
 }