Issue #17106: Fix a segmentation fault in io.TextIOWrapper when an underlying
stream or a decoder produces data of an unexpected type (i.e. when
io.TextIOWrapper initialized with text stream or use bytes-to-bytes codec).
diff --git a/Lib/test/test_io.py b/Lib/test/test_io.py
index 249dc30..a7e596a 100644
--- a/Lib/test/test_io.py
+++ b/Lib/test/test_io.py
@@ -2481,6 +2481,30 @@
txt.write('5')
self.assertEqual(b''.join(raw._write_stack), b'123\n45')
+ def test_read_nonbytes(self):
+ # Issue #17106
+ # Crash when underlying read() returns non-bytes
+ t = self.TextIOWrapper(self.StringIO('a'))
+ self.assertRaises(TypeError, t.read, 1)
+ t = self.TextIOWrapper(self.StringIO('a'))
+ self.assertRaises(TypeError, t.readline)
+ t = self.TextIOWrapper(self.StringIO('a'))
+ self.assertRaises(TypeError, t.read)
+
+ def test_illegal_decoder(self):
+ # Issue #17106
+ # Crash when decoder returns non-string
+ t = self.TextIOWrapper(self.BytesIO(b'aaaaaa'), newline='\n',
+ encoding='quopri_codec')
+ self.assertRaises(TypeError, t.read, 1)
+ t = self.TextIOWrapper(self.BytesIO(b'aaaaaa'), newline='\n',
+ encoding='quopri_codec')
+ self.assertRaises(TypeError, t.readline)
+ t = self.TextIOWrapper(self.BytesIO(b'aaaaaa'), newline='\n',
+ encoding='quopri_codec')
+ self.assertRaises(TypeError, t.read)
+
+
class CTextIOWrapperTest(TextIOWrapperTest):
def test_initialization(self):
diff --git a/Misc/NEWS b/Misc/NEWS
index 382cd7f..53b88ac 100644
--- a/Misc/NEWS
+++ b/Misc/NEWS
@@ -212,6 +212,10 @@
Library
-------
+- Issue #17106: Fix a segmentation fault in io.TextIOWrapper when an underlying
+ stream or a decoder produces data of an unexpected type (i.e. when
+ io.TextIOWrapper initialized with text stream or use bytes-to-bytes codec).
+
- Issue #15633: httplib.HTTPResponse is now mark closed when the server
sends less than the advertised Content-Length.
diff --git a/Modules/_io/textio.c b/Modules/_io/textio.c
index 7d48388..a8231be 100644
--- a/Modules/_io/textio.c
+++ b/Modules/_io/textio.c
@@ -236,6 +236,21 @@
Py_TYPE(self)->tp_free((PyObject *)self);
}
+static int
+check_decoded(PyObject *decoded)
+{
+ if (decoded == NULL)
+ return -1;
+ if (!PyUnicode_Check(decoded)) {
+ PyErr_Format(PyExc_TypeError,
+ "decoder should return a string result, not '%.200s'",
+ Py_TYPE(decoded)->tp_name);
+ Py_DECREF(decoded);
+ return -1;
+ }
+ return 0;
+}
+
#define SEEN_CR 1
#define SEEN_LF 2
#define SEEN_CRLF 4
@@ -265,15 +280,9 @@
Py_INCREF(output);
}
- if (output == NULL)
+ if (check_decoded(output) < 0)
return NULL;
- if (!PyUnicode_Check(output)) {
- PyErr_SetString(PyExc_TypeError,
- "decoder should return a string result");
- goto error;
- }
-
output_len = PyUnicode_GET_SIZE(output);
if (self->pendingcr && (final || output_len > 0)) {
Py_UNICODE *out;
@@ -1454,7 +1463,13 @@
Py_DECREF(chunk_size);
if (input_chunk == NULL)
goto fail;
- assert(PyBytes_Check(input_chunk));
+ if (!PyBytes_Check(input_chunk)) {
+ PyErr_Format(PyExc_TypeError,
+ "underlying %s() should have returned a bytes object, "
+ "not '%.200s'", (self->has_read1 ? "read1": "read"),
+ Py_TYPE(input_chunk)->tp_name);
+ goto fail;
+ }
eof = (PyBytes_Size(input_chunk) == 0);
@@ -1467,8 +1482,7 @@
_PyIO_str_decode, input_chunk, eof ? Py_True : Py_False, NULL);
}
- /* TODO sanity check: isinstance(decoded_chars, unicode) */
- if (decoded_chars == NULL)
+ if (check_decoded(decoded_chars) < 0)
goto fail;
textiowrapper_set_decoded_chars(self, decoded_chars);
if (PyUnicode_GET_SIZE(decoded_chars) > 0)
@@ -1481,7 +1495,14 @@
PyObject *next_input = PyNumber_Add(dec_buffer, input_chunk);
if (next_input == NULL)
goto fail;
- assert (PyBytes_Check(next_input));
+ if (!PyBytes_Check(next_input)) {
+ PyErr_Format(PyExc_TypeError,
+ "decoder getstate() should have returned a bytes "
+ "object, not '%.200s'",
+ Py_TYPE(next_input)->tp_name);
+ Py_DECREF(next_input);
+ goto fail;
+ }
Py_DECREF(dec_buffer);
Py_CLEAR(self->snapshot);
self->snapshot = Py_BuildValue("NN", dec_flags, next_input);
@@ -1525,7 +1546,7 @@
decoded = PyObject_CallMethodObjArgs(self->decoder, _PyIO_str_decode,
bytes, Py_True, NULL);
Py_DECREF(bytes);
- if (decoded == NULL)
+ if (check_decoded(decoded) < 0)
goto fail;
result = textiowrapper_get_decoded_chars(self, -1);
@@ -2123,7 +2144,14 @@
if (input_chunk == NULL)
goto fail;
- assert (PyBytes_Check(input_chunk));
+ if (!PyBytes_Check(input_chunk)) {
+ PyErr_Format(PyExc_TypeError,
+ "underlying read() should have returned a bytes "
+ "object, not '%.200s'",
+ Py_TYPE(input_chunk)->tp_name);
+ Py_DECREF(input_chunk);
+ goto fail;
+ }
self->snapshot = Py_BuildValue("iN", cookie.dec_flags, input_chunk);
if (self->snapshot == NULL) {
@@ -2134,7 +2162,7 @@
decoded = PyObject_CallMethod(self->decoder, "decode",
"Oi", input_chunk, (int)cookie.need_eof);
- if (decoded == NULL)
+ if (check_decoded(decoded) < 0)
goto fail;
textiowrapper_set_decoded_chars(self, decoded);
@@ -2257,9 +2285,8 @@
PyObject *decoded = PyObject_CallMethod(
self->decoder, "decode", "y#", input, 1);
- if (decoded == NULL)
+ if (check_decoded(decoded) < 0)
goto fail;
- assert (PyUnicode_Check(decoded));
chars_decoded += PyUnicode_GET_SIZE(decoded);
Py_DECREF(decoded);
@@ -2291,9 +2318,8 @@
/* We didn't get enough decoded data; signal EOF to get more. */
PyObject *decoded = PyObject_CallMethod(
self->decoder, "decode", "yi", "", /* final = */ 1);
- if (decoded == NULL)
+ if (check_decoded(decoded) < 0)
goto fail;
- assert (PyUnicode_Check(decoded));
chars_decoded += PyUnicode_GET_SIZE(decoded);
Py_DECREF(decoded);
cookie.need_eof = 1;