diff --git a/Lib/test/test_complex.py b/Lib/test/test_complex.py index 2d883c5348e..21c6eaed605 100644 --- a/Lib/test/test_complex.py +++ b/Lib/test/test_complex.py @@ -345,6 +345,9 @@ class ComplexTest(unittest.TestCase): self.assertEqual(type(complex("1"*500)), complex) # check whitespace processing self.assertEqual(complex('\N{EM SPACE}(\N{EN SPACE}1+1j ) '), 1+1j) + # Invalid unicode string + # See bpo-34087 + self.assertRaises(ValueError, complex, '\u3053\u3093\u306b\u3061\u306f') class EvilExc(Exception): pass diff --git a/Lib/test/test_float.py b/Lib/test/test_float.py index 17174dd295d..06ea90c207f 100644 --- a/Lib/test/test_float.py +++ b/Lib/test/test_float.py @@ -60,6 +60,9 @@ class GeneralFloatCases(unittest.TestCase): # extra long strings should not be a problem float(b'.' + b'1'*1000) float('.' + '1'*1000) + # Invalid unicode string + # See bpo-34087 + self.assertRaises(ValueError, float, '\u3053\u3093\u306b\u3061\u306f') def test_underscores(self): for lit in VALID_UNDERSCORE_LITERALS: diff --git a/Lib/test/test_long.py b/Lib/test/test_long.py index cc48259e35f..8472889d48b 100644 --- a/Lib/test/test_long.py +++ b/Lib/test/test_long.py @@ -373,6 +373,10 @@ class LongTest(unittest.TestCase): for base in invalid_bases: self.assertRaises(ValueError, int, '42', base) + # Invalid unicode string + # See bpo-34087 + self.assertRaises(ValueError, int, '\u3053\u3093\u306b\u3061\u306f') + def test_conversion(self): diff --git a/Misc/NEWS.d/next/Core and Builtins/2018-07-13-22-09-55.bpo-34087.I1Bxfc.rst b/Misc/NEWS.d/next/Core and Builtins/2018-07-13-22-09-55.bpo-34087.I1Bxfc.rst new file mode 100644 index 00000000000..5147395fa21 --- /dev/null +++ b/Misc/NEWS.d/next/Core and Builtins/2018-07-13-22-09-55.bpo-34087.I1Bxfc.rst @@ -0,0 +1 @@ +Fix buffer overflow while converting unicode to numeric values. diff --git a/Objects/unicodeobject.c b/Objects/unicodeobject.c index 80d1bba1e9b..2b06f15f6c6 100644 --- a/Objects/unicodeobject.c +++ b/Objects/unicodeobject.c @@ -9072,6 +9072,7 @@ _PyUnicode_TransformDecimalAndSpaceToASCII(PyObject *unicode) int decimal = Py_UNICODE_TODECIMAL(ch); if (decimal < 0) { out[i] = '?'; + out[i+1] = '\0'; _PyUnicode_LENGTH(result) = i + 1; break; } @@ -9079,6 +9080,7 @@ _PyUnicode_TransformDecimalAndSpaceToASCII(PyObject *unicode) } } + assert(_PyUnicode_CheckConsistency(result, 1)); return result; } diff --git a/Python/pystrtod.c b/Python/pystrtod.c index 3546d44c842..461e8dcb5e0 100644 --- a/Python/pystrtod.c +++ b/Python/pystrtod.c @@ -391,6 +391,8 @@ _Py_string_to_number_with_underscores( char *dup, *end; PyObject *result; + assert(s[orig_len] == '\0'); + if (strchr(s, '_') == NULL) { return innerfunc(s, orig_len, arg); }