Issue #18818: The "encodingname" part of PYTHONIOENCODING is now optional.
diff --git a/Doc/using/cmdline.rst b/Doc/using/cmdline.rst
index 908a17c..e8a329e 100644
--- a/Doc/using/cmdline.rst
+++ b/Doc/using/cmdline.rst
@@ -538,13 +538,16 @@
.. envvar:: PYTHONIOENCODING
If this is set before running the interpreter, it overrides the encoding used
- for stdin/stdout/stderr, in the syntax ``encodingname:errorhandler``. The
- ``:errorhandler`` part is optional and has the same meaning as in
- :func:`str.encode`.
+ for stdin/stdout/stderr, in the syntax ``encodingname:errorhandler``. Both
+ the ``encodingname`` and the ``:errorhandler`` parts are optional and have
+ the same meaning as in :func:`str.encode`.
For stderr, the ``:errorhandler`` part is ignored; the handler will always be
``'backslashreplace'``.
+ .. versionchanged:: 3.4
+ The ``encodingname`` part is now optional.
+
.. envvar:: PYTHONNOUSERSITE
diff --git a/Lib/test/test_sys.py b/Lib/test/test_sys.py
index 4044aa3..30eca4a 100644
--- a/Lib/test/test_sys.py
+++ b/Lib/test/test_sys.py
@@ -544,6 +544,42 @@
out = p.communicate()[0].strip()
self.assertEqual(out, b'?')
+ env["PYTHONIOENCODING"] = "ascii"
+ p = subprocess.Popen([sys.executable, "-c", 'print(chr(0xa2))'],
+ stdout=subprocess.PIPE, stderr=subprocess.PIPE,
+ env=env)
+ out, err = p.communicate()
+ self.assertEqual(out, b'')
+ self.assertIn(b'UnicodeEncodeError:', err)
+ self.assertIn(rb"'\xa2'", err)
+
+ env["PYTHONIOENCODING"] = "ascii:"
+ p = subprocess.Popen([sys.executable, "-c", 'print(chr(0xa2))'],
+ stdout=subprocess.PIPE, stderr=subprocess.PIPE,
+ env=env)
+ out, err = p.communicate()
+ self.assertEqual(out, b'')
+ self.assertIn(b'UnicodeEncodeError:', err)
+ self.assertIn(rb"'\xa2'", err)
+
+ env["PYTHONIOENCODING"] = ":surrogateescape"
+ p = subprocess.Popen([sys.executable, "-c", 'print(chr(0xdcbd))'],
+ stdout=subprocess.PIPE, env=env)
+ out = p.communicate()[0].strip()
+ self.assertEqual(out, b'\xbd')
+
+ @unittest.skipUnless(test.support.FS_NONASCII,
+ 'requires OS support of non-ASCII encodings')
+ def test_ioencoding_nonascii(self):
+ env = dict(os.environ)
+
+ env["PYTHONIOENCODING"] = ""
+ p = subprocess.Popen([sys.executable, "-c",
+ 'print(%a)' % test.support.FS_NONASCII],
+ stdout=subprocess.PIPE, env=env)
+ out = p.communicate()[0].strip()
+ self.assertEqual(out, os.fsencode(test.support.FS_NONASCII))
+
@unittest.skipIf(sys.base_prefix != sys.prefix,
'Test is not venv-compatible')
def test_executable(self):
diff --git a/Misc/NEWS b/Misc/NEWS
index 9f6871a..33658a7 100644
--- a/Misc/NEWS
+++ b/Misc/NEWS
@@ -7,6 +7,8 @@
Core and Builtins
-----------------
+- Issue #18818: The "encodingname" part of PYTHONIOENCODING is now optional.
+
Library
-------
diff --git a/Python/pythonrun.c b/Python/pythonrun.c
index cbd62aa..c2ca563 100644
--- a/Python/pythonrun.c
+++ b/Python/pythonrun.c
@@ -1056,7 +1056,7 @@
PyObject *std = NULL;
int status = 0, fd;
PyObject * encoding_attr;
- char *encoding = NULL, *errors;
+ char *pythonioencoding = NULL, *encoding, *errors;
/* Hack to avoid a nasty recursion issue when Python is invoked
in verbose mode: pre-import the Latin-1 and UTF-8 codecs */
@@ -1088,19 +1088,23 @@
}
Py_DECREF(wrapper);
- encoding = Py_GETENV("PYTHONIOENCODING");
- errors = NULL;
- if (encoding) {
- encoding = _PyMem_Strdup(encoding);
- if (encoding == NULL) {
+ pythonioencoding = Py_GETENV("PYTHONIOENCODING");
+ encoding = errors = NULL;
+ if (pythonioencoding) {
+ pythonioencoding = _PyMem_Strdup(pythonioencoding);
+ if (pythonioencoding == NULL) {
PyErr_NoMemory();
goto error;
}
- errors = strchr(encoding, ':');
+ errors = strchr(pythonioencoding, ':');
if (errors) {
*errors = '\0';
errors++;
+ if (!*errors)
+ errors = NULL;
}
+ if (*pythonioencoding)
+ encoding = pythonioencoding;
}
/* Set sys.stdin */
@@ -1180,7 +1184,7 @@
status = -1;
}
- PyMem_Free(encoding);
+ PyMem_Free(pythonioencoding);
Py_XDECREF(bimod);
Py_XDECREF(iomod);
return status;