def test_ctrl_z(self):
with open('CONIN$', 'rb', buffering=0) as stdin:
source = '\xC4\x1A\r\n'.encode('utf-16-le')
expected = '\xC4'.encode('utf-8')
write_input(stdin, source)
a, b = stdin.read(1), stdin.readall()
self.assertEqual(expected[0:1], a)
self.assertEqual(expected[1:], b)
def test_ctrl_z(self):
with open('CONIN$', 'rb', buffering=0) as stdin:
source = '\xC4\x1A\r\n'.encode('utf-16-le')
expected = '\xC4'.encode('utf-8')
write_input(stdin, source)
a, b = stdin.read(1), stdin.readall()
self.assertEqual(expected[0:1], a)
self.assertEqual(expected[1:], b)
def assertStdinRoundTrip(self, text):
stdin = open('CONIN$', 'r')
old_stdin = sys.stdin
try:
sys.stdin = stdin
write_input(stdin.buffer.raw,
(text + '\r\n').encode('utf-16-le', 'surrogatepass'))
actual = input()
finally:
sys.stdin = old_stdin
self.assertEqual(actual, text)
def test_partial_surrogate_reads(self):
source = '\U00101fff\U00101001\r\n'.encode('utf-16-le')
expected = '\U00101fff\U00101001\r\n'.encode('utf-8')
for read_count in range(1, 16):
with open('CONIN$', 'rb', buffering=0) as stdin:
write_input(stdin, source)
actual = b''
while not actual.endswith(b'\n'):
b = stdin.read(read_count)
actual += b
self.assertEqual(actual, expected, 'stdin.read({})'.format(
read_count))
def assertStdinRoundTrip(self, text):
stdin = open('CONIN$', 'r')
old_stdin = sys.stdin
try:
sys.stdin = stdin
write_input(
stdin.buffer.raw,
(text + '\r\n').encode('utf-16-le', 'surrogatepass')
)
actual = input()
finally:
sys.stdin = old_stdin
self.assertEqual(actual, text)
def test_partial_reads(self):
# Test that reading less than 1 full character works when stdin
# contains multibyte UTF-8 sequences
source = 'ϼўТλФЙ\r\n'.encode('utf-16-le')
expected = 'ϼўТλФЙ\r\n'.encode('utf-8')
for read_count in range(1, 16):
with open('CONIN$', 'rb', buffering=0) as stdin:
write_input(stdin, source)
actual = b''
while not actual.endswith(b'\n'):
b = stdin.read(read_count)
actual += b
self.assertEqual(actual, expected, 'stdin.read({})'.format(read_count))
def test_partial_reads(self):
# Test that reading less than 1 full character works when stdin
# contains multibyte UTF-8 sequences
source = 'ϼўТλФЙ\r\n'.encode('utf-16-le')
expected = 'ϼўТλФЙ\r\n'.encode('utf-8')
for read_count in range(1, 16):
with open('CONIN$', 'rb', buffering=0) as stdin:
write_input(stdin, source)
actual = b''
while not actual.endswith(b'\n'):
b = stdin.read(read_count)
actual += b
self.assertEqual(actual, expected, 'stdin.read({})'.format(read_count))
def test_partial_surrogate_reads(self):
# Test that reading less than 1 full character works when stdin
# contains surrogate pairs that cannot be decoded to UTF-8 without
# reading an extra character.
source = '\U00101FFF\U00101001\r\n'.encode('utf-16-le')
expected = '\U00101FFF\U00101001\r\n'.encode('utf-8')
for read_count in range(1, 16):
with open('CONIN$', 'rb', buffering=0) as stdin:
write_input(stdin, source)
actual = b''
while not actual.endswith(b'\n'):
b = stdin.read(read_count)
actual += b
self.assertEqual(actual, expected, 'stdin.read({})'.format(read_count))
def test_partial_surrogate_reads(self):
# Test that reading less than 1 full character works when stdin
# contains surrogate pairs that cannot be decoded to UTF-8 without
# reading an extra character.
source = '\U00101FFF\U00101001\r\n'.encode('utf-16-le')
expected = '\U00101FFF\U00101001\r\n'.encode('utf-8')
for read_count in range(1, 16):
with open('CONIN$', 'rb', buffering=0) as stdin:
write_input(stdin, source)
actual = b''
while not actual.endswith(b'\n'):
b = stdin.read(read_count)
actual += b
self.assertEqual(actual, expected, 'stdin.read({})'.format(read_count))
