Mirror
/
cpython
mirror of https://github.com/python/cpython
4
1
Fork 0
Code Issues Releases Wiki Activity

Issue #21118: Add unit test for invalid character replacement (code point higher than U+10ffff)

This commit is contained in:
Victor Stinner 2014-04-05 11:56:37 +02:00
parent 89a76abf20
commit 4ff33af257
2 changed files with 18 additions and 10 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
Lib/test/test_unicode.py
View File

@ -280,6 +280,14 @@ class UnicodeTest(string_tests.CommonTest,
self.assertEqual("[\xe9]".translate(str.maketrans({'\xe9': None})), self.assertEqual("[\xe9]".translate(str.maketrans({'\xe9': None})),
"[]") "[]")
# invalid Unicode characters
invalid_char = 0x10ffff+1
for before in "a\xe9\u20ac\U0010ffff":
mapping = str.maketrans({before: invalid_char})
text = "[%s]" % before
self.assertRaises(ValueError, text.translate, mapping)
# errors
self.assertRaises(TypeError, self.type2test.maketrans) self.assertRaises(TypeError, self.type2test.maketrans)
self.assertRaises(ValueError, self.type2test.maketrans, 'abc', 'defg') self.assertRaises(ValueError, self.type2test.maketrans, 'abc', 'defg')
self.assertRaises(TypeError, self.type2test.maketrans, 2, 'def') self.assertRaises(TypeError, self.type2test.maketrans, 2, 'def')

20
Objects/unicodeobject.c
View File

@ -8473,10 +8473,10 @@ charmaptranslate_lookup(Py_UCS4 c, PyObject *mapping, PyObject **result)
} }
else if (PyLong_Check(x)) { else if (PyLong_Check(x)) {
long value = PyLong_AS_LONG(x); long value = PyLong_AS_LONG(x);
long max = PyUnicode_GetMax(); if (value < 0 || value > MAX_UNICODE) {
if (value < 0 || value > max) { PyErr_Format(PyExc_ValueError,
PyErr_Format(PyExc_TypeError, "character mapping must be in range(0x%x)",
"character mapping must be in range(0x%x)", max+1); MAX_UNICODE+1);
Py_DECREF(x); Py_DECREF(x);
return -1; return -1;
} }
@ -8522,7 +8522,9 @@ charmaptranslate_output(Py_UCS4 ch, PyObject *mapping,
} }
if (PyLong_Check(item)) { if (PyLong_Check(item)) {
Py_UCS4 ch = (Py_UCS4)PyLong_AS_LONG(item); long ch = (Py_UCS4)PyLong_AS_LONG(item);
/* PyLong_AS_LONG() cannot fail, charmaptranslate_lookup() already
used it */
if (_PyUnicodeWriter_WriteCharInline(writer, ch) < 0) { if (_PyUnicodeWriter_WriteCharInline(writer, ch) < 0) {
Py_DECREF(item); Py_DECREF(item);
return -1; return -1;
@ -8570,11 +8572,9 @@ unicode_fast_translate_lookup(PyObject *mapping, Py_UCS1 ch,
if (PyLong_Check(item)) { if (PyLong_Check(item)) {
long replace = (Py_UCS4)PyLong_AS_LONG(item); long replace = (Py_UCS4)PyLong_AS_LONG(item);
if (replace == -1) { /* PyLong_AS_LONG() cannot fail, charmaptranslate_lookup() already
Py_DECREF(item); used it */
return -1; if (127 < replace) {
}
if (replace < 0 || 127 < replace) {
/* invalid character or character outside ASCII: /* invalid character or character outside ASCII:
skip the fast translate */ skip the fast translate */
goto exit; goto exit;