From b8b3e6afc0a48c3cbb7c36d2f73e332edcd6058c Mon Sep 17 00:00:00 2001 From: Serhiy Storchaka Date: Tue, 25 Jul 2023 14:15:14 +0300 Subject: [PATCH] [3.11] gh-99612: Fix PyUnicode_DecodeUTF8Stateful() for ASCII-only data (GH-99613) (GH-107224) Previously *consumed was not set in this case. (cherry picked from commit f08e52ccb027f6f703302b8c1a82db9fd3934270) --- Lib/test/test_capi/test_codecs.py | 54 +++++++++++++++++++ ...2-11-20-09-52-50.gh-issue-99612.eBHksg.rst | 2 + Modules/_testcapimodule.c | 37 ++++++++++++- Objects/unicodeobject.c | 3 ++ 4 files changed, 95 insertions(+), 1 deletion(-) create mode 100644 Lib/test/test_capi/test_codecs.py create mode 100644 Misc/NEWS.d/next/C API/2022-11-20-09-52-50.gh-issue-99612.eBHksg.rst diff --git a/Lib/test/test_capi/test_codecs.py b/Lib/test/test_capi/test_codecs.py new file mode 100644 index 00000000000..e46726192aa --- /dev/null +++ b/Lib/test/test_capi/test_codecs.py @@ -0,0 +1,54 @@ +import unittest +from test.support import import_helper + +_testcapi = import_helper.import_module('_testcapi') + + +class CAPITest(unittest.TestCase): + + def test_decodeutf8(self): + """Test PyUnicode_DecodeUTF8()""" + decodeutf8 = _testcapi.unicode_decodeutf8 + + for s in ['abc', '\xa1\xa2', '\u4f60\u597d', 'a\U0001f600']: + b = s.encode('utf-8') + self.assertEqual(decodeutf8(b), s) + self.assertEqual(decodeutf8(b, 'strict'), s) + + self.assertRaises(UnicodeDecodeError, decodeutf8, b'\x80') + self.assertRaises(UnicodeDecodeError, decodeutf8, b'\xc0') + self.assertRaises(UnicodeDecodeError, decodeutf8, b'\xff') + self.assertRaises(UnicodeDecodeError, decodeutf8, b'a\xf0\x9f') + self.assertEqual(decodeutf8(b'a\xf0\x9f', 'replace'), 'a\ufffd') + self.assertEqual(decodeutf8(b'a\xf0\x9fb', 'replace'), 'a\ufffdb') + + self.assertRaises(LookupError, decodeutf8, b'a\x80', 'foo') + # TODO: Test PyUnicode_DecodeUTF8() with NULL as data and + # negative size. + + def test_decodeutf8stateful(self): + """Test PyUnicode_DecodeUTF8Stateful()""" + decodeutf8stateful = _testcapi.unicode_decodeutf8stateful + + for s in ['abc', '\xa1\xa2', '\u4f60\u597d', 'a\U0001f600']: + b = s.encode('utf-8') + self.assertEqual(decodeutf8stateful(b), (s, len(b))) + self.assertEqual(decodeutf8stateful(b, 'strict'), (s, len(b))) + + self.assertRaises(UnicodeDecodeError, decodeutf8stateful, b'\x80') + self.assertRaises(UnicodeDecodeError, decodeutf8stateful, b'\xc0') + self.assertRaises(UnicodeDecodeError, decodeutf8stateful, b'\xff') + self.assertEqual(decodeutf8stateful(b'a\xf0\x9f'), ('a', 1)) + self.assertEqual(decodeutf8stateful(b'a\xf0\x9f', 'replace'), ('a', 1)) + self.assertRaises(UnicodeDecodeError, decodeutf8stateful, b'a\xf0\x9fb') + self.assertEqual(decodeutf8stateful(b'a\xf0\x9fb', 'replace'), ('a\ufffdb', 4)) + + self.assertRaises(LookupError, decodeutf8stateful, b'a\x80', 'foo') + # TODO: Test PyUnicode_DecodeUTF8Stateful() with NULL as data and + # negative size. + # TODO: Test PyUnicode_DecodeUTF8Stateful() with NULL as the address of + # "consumed". + + +if __name__ == "__main__": + unittest.main() diff --git a/Misc/NEWS.d/next/C API/2022-11-20-09-52-50.gh-issue-99612.eBHksg.rst b/Misc/NEWS.d/next/C API/2022-11-20-09-52-50.gh-issue-99612.eBHksg.rst new file mode 100644 index 00000000000..40e3c8db540 --- /dev/null +++ b/Misc/NEWS.d/next/C API/2022-11-20-09-52-50.gh-issue-99612.eBHksg.rst @@ -0,0 +1,2 @@ +Fix :c:func:`PyUnicode_DecodeUTF8Stateful` for ASCII-only data: +``*consumed`` was not set. diff --git a/Modules/_testcapimodule.c b/Modules/_testcapimodule.c index b40368eb4b3..b29a919a832 100644 --- a/Modules/_testcapimodule.c +++ b/Modules/_testcapimodule.c @@ -2307,6 +2307,40 @@ unicode_asutf8andsize(PyObject *self, PyObject *args) return Py_BuildValue("(Nn)", result, utf8_len); } +/* Test PyUnicode_DecodeUTF8() */ +static PyObject * +unicode_decodeutf8(PyObject *self, PyObject *args) +{ + const char *data; + Py_ssize_t size; + const char *errors = NULL; + + if (!PyArg_ParseTuple(args, "y#|z", &data, &size, &errors)) + return NULL; + + return PyUnicode_DecodeUTF8(data, size, errors); +} + +/* Test PyUnicode_DecodeUTF8Stateful() */ +static PyObject * +unicode_decodeutf8stateful(PyObject *self, PyObject *args) +{ + const char *data; + Py_ssize_t size; + const char *errors = NULL; + Py_ssize_t consumed = 123456789; + PyObject *result; + + if (!PyArg_ParseTuple(args, "y#|z", &data, &size, &errors)) + return NULL; + + result = PyUnicode_DecodeUTF8Stateful(data, size, errors, &consumed); + if (!result) { + return NULL; + } + return Py_BuildValue("(Nn)", result, consumed); +} + static PyObject * unicode_findchar(PyObject *self, PyObject *args) { @@ -6562,7 +6596,8 @@ static PyMethodDef TestMethods[] = { {"unicode_asucs4", unicode_asucs4, METH_VARARGS}, {"unicode_asutf8", unicode_asutf8, METH_VARARGS}, {"unicode_asutf8andsize", unicode_asutf8andsize, METH_VARARGS}, - {"unicode_findchar", unicode_findchar, METH_VARARGS}, + {"unicode_decodeutf8", unicode_decodeutf8, METH_VARARGS}, + {"unicode_decodeutf8stateful",unicode_decodeutf8stateful, METH_VARARGS}, {"unicode_findchar", unicode_findchar, METH_VARARGS}, {"unicode_copycharacters", unicode_copycharacters, METH_VARARGS}, #if USE_UNICODE_WCHAR_CACHE {"unicode_legacy_string", unicode_legacy_string, METH_VARARGS}, diff --git a/Objects/unicodeobject.c b/Objects/unicodeobject.c index 50deefe0adc..ffa8b982bdc 100644 --- a/Objects/unicodeobject.c +++ b/Objects/unicodeobject.c @@ -5120,6 +5120,9 @@ unicode_decode_utf8(const char *s, Py_ssize_t size, } s += ascii_decode(s, end, PyUnicode_1BYTE_DATA(u)); if (s == end) { + if (consumed) { + *consumed = size; + } return u; }