cpython/Lib/test/test_unicode.py

1524 lines
66 KiB
Python

""" Test script for the Unicode implementation.
Written by Marc-Andre Lemburg (mal@lemburg.com).
(c) Copyright CNRI, All Rights Reserved. NO WARRANTY.
"""#"
import codecs
import struct
import sys
import unittest
import warnings
from test import support, string_tests
# Error handling (bad decoder return)
def search_function(encoding):
def decode1(input, errors="strict"):
return 42 # not a tuple
def encode1(input, errors="strict"):
return 42 # not a tuple
def encode2(input, errors="strict"):
return (42, 42) # no unicode
def decode2(input, errors="strict"):
return (42, 42) # no unicode
if encoding=="test.unicode1":
return (encode1, decode1, None, None)
elif encoding=="test.unicode2":
return (encode2, decode2, None, None)
else:
return None
codecs.register(search_function)
class UnicodeTest(string_tests.CommonTest,
string_tests.MixinStrUnicodeUserStringTest,
string_tests.MixinStrUnicodeTest):
type2test = str
def checkequalnofix(self, result, object, methodname, *args):
method = getattr(object, methodname)
realresult = method(*args)
self.assertEqual(realresult, result)
self.assertTrue(type(realresult) is type(result))
# if the original is returned make sure that
# this doesn't happen with subclasses
if realresult is object:
class usub(str):
def __repr__(self):
return 'usub(%r)' % str.__repr__(self)
object = usub(object)
method = getattr(object, methodname)
realresult = method(*args)
self.assertEqual(realresult, result)
self.assertTrue(object is not realresult)
def test_literals(self):
self.assertEqual('\xff', '\u00ff')
self.assertEqual('\uffff', '\U0000ffff')
self.assertRaises(SyntaxError, eval, '\'\\Ufffffffe\'')
self.assertRaises(SyntaxError, eval, '\'\\Uffffffff\'')
self.assertRaises(SyntaxError, eval, '\'\\U%08x\'' % 0x110000)
# raw strings should not have unicode escapes
self.assertNotEqual(r"\u0020", " ")
def test_ascii(self):
if not sys.platform.startswith('java'):
# Test basic sanity of repr()
self.assertEqual(ascii('abc'), "'abc'")
self.assertEqual(ascii('ab\\c'), "'ab\\\\c'")
self.assertEqual(ascii('ab\\'), "'ab\\\\'")
self.assertEqual(ascii('\\c'), "'\\\\c'")
self.assertEqual(ascii('\\'), "'\\\\'")
self.assertEqual(ascii('\n'), "'\\n'")
self.assertEqual(ascii('\r'), "'\\r'")
self.assertEqual(ascii('\t'), "'\\t'")
self.assertEqual(ascii('\b'), "'\\x08'")
self.assertEqual(ascii("'\""), """'\\'"'""")
self.assertEqual(ascii("'\""), """'\\'"'""")
self.assertEqual(ascii("'"), '''"'"''')
self.assertEqual(ascii('"'), """'"'""")
latin1repr = (
"'\\x00\\x01\\x02\\x03\\x04\\x05\\x06\\x07\\x08\\t\\n\\x0b\\x0c\\r"
"\\x0e\\x0f\\x10\\x11\\x12\\x13\\x14\\x15\\x16\\x17\\x18\\x19\\x1a"
"\\x1b\\x1c\\x1d\\x1e\\x1f !\"#$%&\\'()*+,-./0123456789:;<=>?@ABCDEFGHI"
"JKLMNOPQRSTUVWXYZ[\\\\]^_`abcdefghijklmnopqrstuvwxyz{|}~\\x7f"
"\\x80\\x81\\x82\\x83\\x84\\x85\\x86\\x87\\x88\\x89\\x8a\\x8b\\x8c\\x8d"
"\\x8e\\x8f\\x90\\x91\\x92\\x93\\x94\\x95\\x96\\x97\\x98\\x99\\x9a\\x9b"
"\\x9c\\x9d\\x9e\\x9f\\xa0\\xa1\\xa2\\xa3\\xa4\\xa5\\xa6\\xa7\\xa8\\xa9"
"\\xaa\\xab\\xac\\xad\\xae\\xaf\\xb0\\xb1\\xb2\\xb3\\xb4\\xb5\\xb6\\xb7"
"\\xb8\\xb9\\xba\\xbb\\xbc\\xbd\\xbe\\xbf\\xc0\\xc1\\xc2\\xc3\\xc4\\xc5"
"\\xc6\\xc7\\xc8\\xc9\\xca\\xcb\\xcc\\xcd\\xce\\xcf\\xd0\\xd1\\xd2\\xd3"
"\\xd4\\xd5\\xd6\\xd7\\xd8\\xd9\\xda\\xdb\\xdc\\xdd\\xde\\xdf\\xe0\\xe1"
"\\xe2\\xe3\\xe4\\xe5\\xe6\\xe7\\xe8\\xe9\\xea\\xeb\\xec\\xed\\xee\\xef"
"\\xf0\\xf1\\xf2\\xf3\\xf4\\xf5\\xf6\\xf7\\xf8\\xf9\\xfa\\xfb\\xfc\\xfd"
"\\xfe\\xff'")
testrepr = ascii(''.join(map(chr, range(256))))
self.assertEqual(testrepr, latin1repr)
# Test ascii works on wide unicode escapes without overflow.
self.assertEqual(ascii("\U00010000" * 39 + "\uffff" * 4096),
ascii("\U00010000" * 39 + "\uffff" * 4096))
class WrongRepr:
def __repr__(self):
return b'byte-repr'
self.assertRaises(TypeError, ascii, WrongRepr())
def test_repr(self):
if not sys.platform.startswith('java'):
# Test basic sanity of repr()
self.assertEqual(repr('abc'), "'abc'")
self.assertEqual(repr('ab\\c'), "'ab\\\\c'")
self.assertEqual(repr('ab\\'), "'ab\\\\'")
self.assertEqual(repr('\\c'), "'\\\\c'")
self.assertEqual(repr('\\'), "'\\\\'")
self.assertEqual(repr('\n'), "'\\n'")
self.assertEqual(repr('\r'), "'\\r'")
self.assertEqual(repr('\t'), "'\\t'")
self.assertEqual(repr('\b'), "'\\x08'")
self.assertEqual(repr("'\""), """'\\'"'""")
self.assertEqual(repr("'\""), """'\\'"'""")
self.assertEqual(repr("'"), '''"'"''')
self.assertEqual(repr('"'), """'"'""")
latin1repr = (
"'\\x00\\x01\\x02\\x03\\x04\\x05\\x06\\x07\\x08\\t\\n\\x0b\\x0c\\r"
"\\x0e\\x0f\\x10\\x11\\x12\\x13\\x14\\x15\\x16\\x17\\x18\\x19\\x1a"
"\\x1b\\x1c\\x1d\\x1e\\x1f !\"#$%&\\'()*+,-./0123456789:;<=>?@ABCDEFGHI"
"JKLMNOPQRSTUVWXYZ[\\\\]^_`abcdefghijklmnopqrstuvwxyz{|}~\\x7f"
"\\x80\\x81\\x82\\x83\\x84\\x85\\x86\\x87\\x88\\x89\\x8a\\x8b\\x8c\\x8d"
"\\x8e\\x8f\\x90\\x91\\x92\\x93\\x94\\x95\\x96\\x97\\x98\\x99\\x9a\\x9b"
"\\x9c\\x9d\\x9e\\x9f\\xa0\xa1\xa2\xa3\xa4\xa5\xa6\xa7\xa8\xa9"
"\xaa\xab\xac\\xad\xae\xaf\xb0\xb1\xb2\xb3\xb4\xb5\xb6\xb7"
"\xb8\xb9\xba\xbb\xbc\xbd\xbe\xbf\xc0\xc1\xc2\xc3\xc4\xc5"
"\xc6\xc7\xc8\xc9\xca\xcb\xcc\xcd\xce\xcf\xd0\xd1\xd2\xd3"
"\xd4\xd5\xd6\xd7\xd8\xd9\xda\xdb\xdc\xdd\xde\xdf\xe0\xe1"
"\xe2\xe3\xe4\xe5\xe6\xe7\xe8\xe9\xea\xeb\xec\xed\xee\xef"
"\xf0\xf1\xf2\xf3\xf4\xf5\xf6\xf7\xf8\xf9\xfa\xfb\xfc\xfd"
"\xfe\xff'")
testrepr = repr(''.join(map(chr, range(256))))
self.assertEqual(testrepr, latin1repr)
# Test repr works on wide unicode escapes without overflow.
self.assertEqual(repr("\U00010000" * 39 + "\uffff" * 4096),
repr("\U00010000" * 39 + "\uffff" * 4096))
class WrongRepr:
def __repr__(self):
return b'byte-repr'
self.assertRaises(TypeError, repr, WrongRepr())
def test_iterators(self):
# Make sure unicode objects have an __iter__ method
it = "\u1111\u2222\u3333".__iter__()
self.assertEqual(next(it), "\u1111")
self.assertEqual(next(it), "\u2222")
self.assertEqual(next(it), "\u3333")
self.assertRaises(StopIteration, next, it)
def test_count(self):
string_tests.CommonTest.test_count(self)
# check mixed argument types
self.checkequalnofix(3, 'aaa', 'count', 'a')
self.checkequalnofix(0, 'aaa', 'count', 'b')
self.checkequalnofix(3, 'aaa', 'count', 'a')
self.checkequalnofix(0, 'aaa', 'count', 'b')
self.checkequalnofix(0, 'aaa', 'count', 'b')
self.checkequalnofix(1, 'aaa', 'count', 'a', -1)
self.checkequalnofix(3, 'aaa', 'count', 'a', -10)
self.checkequalnofix(2, 'aaa', 'count', 'a', 0, -1)
self.checkequalnofix(0, 'aaa', 'count', 'a', 0, -10)
def test_find(self):
self.checkequalnofix(0, 'abcdefghiabc', 'find', 'abc')
self.checkequalnofix(9, 'abcdefghiabc', 'find', 'abc', 1)
self.checkequalnofix(-1, 'abcdefghiabc', 'find', 'def', 4)
self.assertRaises(TypeError, 'hello'.find)
self.assertRaises(TypeError, 'hello'.find, 42)
def test_rfind(self):
string_tests.CommonTest.test_rfind(self)
# check mixed argument types
self.checkequalnofix(9, 'abcdefghiabc', 'rfind', 'abc')
self.checkequalnofix(12, 'abcdefghiabc', 'rfind', '')
self.checkequalnofix(12, 'abcdefghiabc', 'rfind', '')
def test_index(self):
string_tests.CommonTest.test_index(self)
self.checkequalnofix(0, 'abcdefghiabc', 'index', '')
self.checkequalnofix(3, 'abcdefghiabc', 'index', 'def')
self.checkequalnofix(0, 'abcdefghiabc', 'index', 'abc')
self.checkequalnofix(9, 'abcdefghiabc', 'index', 'abc', 1)
self.assertRaises(ValueError, 'abcdefghiabc'.index, 'hib')
self.assertRaises(ValueError, 'abcdefghiab'.index, 'abc', 1)
self.assertRaises(ValueError, 'abcdefghi'.index, 'ghi', 8)
self.assertRaises(ValueError, 'abcdefghi'.index, 'ghi', -1)
def test_rindex(self):
string_tests.CommonTest.test_rindex(self)
self.checkequalnofix(12, 'abcdefghiabc', 'rindex', '')
self.checkequalnofix(3, 'abcdefghiabc', 'rindex', 'def')
self.checkequalnofix(9, 'abcdefghiabc', 'rindex', 'abc')
self.checkequalnofix(0, 'abcdefghiabc', 'rindex', 'abc', 0, -1)
self.assertRaises(ValueError, 'abcdefghiabc'.rindex, 'hib')
self.assertRaises(ValueError, 'defghiabc'.rindex, 'def', 1)
self.assertRaises(ValueError, 'defghiabc'.rindex, 'abc', 0, -1)
self.assertRaises(ValueError, 'abcdefghi'.rindex, 'ghi', 0, 8)
self.assertRaises(ValueError, 'abcdefghi'.rindex, 'ghi', 0, -1)
def test_maketrans_translate(self):
# these work with plain translate()
self.checkequalnofix('bbbc', 'abababc', 'translate',
{ord('a'): None})
self.checkequalnofix('iiic', 'abababc', 'translate',
{ord('a'): None, ord('b'): ord('i')})
self.checkequalnofix('iiix', 'abababc', 'translate',
{ord('a'): None, ord('b'): ord('i'), ord('c'): 'x'})
self.checkequalnofix('c', 'abababc', 'translate',
{ord('a'): None, ord('b'): ''})
self.checkequalnofix('xyyx', 'xzx', 'translate',
{ord('z'): 'yy'})
# this needs maketrans()
self.checkequalnofix('abababc', 'abababc', 'translate',
{'b': '<i>'})
tbl = self.type2test.maketrans({'a': None, 'b': '<i>'})
self.checkequalnofix('<i><i><i>c', 'abababc', 'translate', tbl)
# test alternative way of calling maketrans()
tbl = self.type2test.maketrans('abc', 'xyz', 'd')
self.checkequalnofix('xyzzy', 'abdcdcbdddd', 'translate', tbl)
self.assertRaises(TypeError, self.type2test.maketrans)
self.assertRaises(ValueError, self.type2test.maketrans, 'abc', 'defg')
self.assertRaises(TypeError, self.type2test.maketrans, 2, 'def')
self.assertRaises(TypeError, self.type2test.maketrans, 'abc', 2)
self.assertRaises(TypeError, self.type2test.maketrans, 'abc', 'def', 2)
self.assertRaises(ValueError, self.type2test.maketrans, {'xy': 2})
self.assertRaises(TypeError, self.type2test.maketrans, {(1,): 2})
self.assertRaises(TypeError, 'hello'.translate)
self.assertRaises(TypeError, 'abababc'.translate, 'abc', 'xyz')
def test_split(self):
string_tests.CommonTest.test_split(self)
# Mixed arguments
self.checkequalnofix(['a', 'b', 'c', 'd'], 'a//b//c//d', 'split', '//')
self.checkequalnofix(['a', 'b', 'c', 'd'], 'a//b//c//d', 'split', '//')
self.checkequalnofix(['endcase ', ''], 'endcase test', 'split', 'test')
def test_join(self):
string_tests.MixinStrUnicodeUserStringTest.test_join(self)
class MyWrapper:
def __init__(self, sval): self.sval = sval
def __str__(self): return self.sval
# mixed arguments
self.checkequalnofix('a b c d', ' ', 'join', ['a', 'b', 'c', 'd'])
self.checkequalnofix('abcd', '', 'join', ('a', 'b', 'c', 'd'))
self.checkequalnofix('w x y z', ' ', 'join', string_tests.Sequence('wxyz'))
self.checkequalnofix('a b c d', ' ', 'join', ['a', 'b', 'c', 'd'])
self.checkequalnofix('a b c d', ' ', 'join', ['a', 'b', 'c', 'd'])
self.checkequalnofix('abcd', '', 'join', ('a', 'b', 'c', 'd'))
self.checkequalnofix('w x y z', ' ', 'join', string_tests.Sequence('wxyz'))
self.checkraises(TypeError, ' ', 'join', ['1', '2', MyWrapper('foo')])
self.checkraises(TypeError, ' ', 'join', ['1', '2', '3', bytes()])
self.checkraises(TypeError, ' ', 'join', [1, 2, 3])
self.checkraises(TypeError, ' ', 'join', ['1', '2', 3])
def test_replace(self):
string_tests.CommonTest.test_replace(self)
# method call forwarded from str implementation because of unicode argument
self.checkequalnofix('one@two!three!', 'one!two!three!', 'replace', '!', '@', 1)
self.assertRaises(TypeError, 'replace'.replace, "r", 42)
def test_bytes_comparison(self):
with support.check_warnings():
warnings.simplefilter('ignore', BytesWarning)
self.assertEqual('abc' == b'abc', False)
self.assertEqual('abc' != b'abc', True)
self.assertEqual('abc' == bytearray(b'abc'), False)
self.assertEqual('abc' != bytearray(b'abc'), True)
def test_comparison(self):
# Comparisons:
self.assertEqual('abc', 'abc')
self.assertTrue('abcd' > 'abc')
self.assertTrue('abc' < 'abcd')
if 0:
# Move these tests to a Unicode collation module test...
# Testing UTF-16 code point order comparisons...
# No surrogates, no fixup required.
self.assertTrue('\u0061' < '\u20ac')
# Non surrogate below surrogate value, no fixup required
self.assertTrue('\u0061' < '\ud800\udc02')
# Non surrogate above surrogate value, fixup required
def test_lecmp(s, s2):
self.assertTrue(s < s2)
def test_fixup(s):
s2 = '\ud800\udc01'
test_lecmp(s, s2)
s2 = '\ud900\udc01'
test_lecmp(s, s2)
s2 = '\uda00\udc01'
test_lecmp(s, s2)
s2 = '\udb00\udc01'
test_lecmp(s, s2)
s2 = '\ud800\udd01'
test_lecmp(s, s2)
s2 = '\ud900\udd01'
test_lecmp(s, s2)
s2 = '\uda00\udd01'
test_lecmp(s, s2)
s2 = '\udb00\udd01'
test_lecmp(s, s2)
s2 = '\ud800\ude01'
test_lecmp(s, s2)
s2 = '\ud900\ude01'
test_lecmp(s, s2)
s2 = '\uda00\ude01'
test_lecmp(s, s2)
s2 = '\udb00\ude01'
test_lecmp(s, s2)
s2 = '\ud800\udfff'
test_lecmp(s, s2)
s2 = '\ud900\udfff'
test_lecmp(s, s2)
s2 = '\uda00\udfff'
test_lecmp(s, s2)
s2 = '\udb00\udfff'
test_lecmp(s, s2)
test_fixup('\ue000')
test_fixup('\uff61')
# Surrogates on both sides, no fixup required
self.assertTrue('\ud800\udc02' < '\ud84d\udc56')
def test_islower(self):
string_tests.MixinStrUnicodeUserStringTest.test_islower(self)
self.checkequalnofix(False, '\u1FFc', 'islower')
def test_isupper(self):
string_tests.MixinStrUnicodeUserStringTest.test_isupper(self)
if not sys.platform.startswith('java'):
self.checkequalnofix(False, '\u1FFc', 'isupper')
def test_istitle(self):
string_tests.MixinStrUnicodeUserStringTest.test_title(self)
self.checkequalnofix(True, '\u1FFc', 'istitle')
self.checkequalnofix(True, 'Greek \u1FFcitlecases ...', 'istitle')
def test_isspace(self):
string_tests.MixinStrUnicodeUserStringTest.test_isspace(self)
self.checkequalnofix(True, '\u2000', 'isspace')
self.checkequalnofix(True, '\u200a', 'isspace')
self.checkequalnofix(False, '\u2014', 'isspace')
def test_isalpha(self):
string_tests.MixinStrUnicodeUserStringTest.test_isalpha(self)
self.checkequalnofix(True, '\u1FFc', 'isalpha')
def test_isdecimal(self):
self.checkequalnofix(False, '', 'isdecimal')
self.checkequalnofix(False, 'a', 'isdecimal')
self.checkequalnofix(True, '0', 'isdecimal')
self.checkequalnofix(False, '\u2460', 'isdecimal') # CIRCLED DIGIT ONE
self.checkequalnofix(False, '\xbc', 'isdecimal') # VULGAR FRACTION ONE QUARTER
self.checkequalnofix(True, '\u0660', 'isdecimal') # ARABIC-INDIC DIGIT ZERO
self.checkequalnofix(True, '0123456789', 'isdecimal')
self.checkequalnofix(False, '0123456789a', 'isdecimal')
self.checkraises(TypeError, 'abc', 'isdecimal', 42)
def test_isdigit(self):
string_tests.MixinStrUnicodeUserStringTest.test_isdigit(self)
self.checkequalnofix(True, '\u2460', 'isdigit')
self.checkequalnofix(False, '\xbc', 'isdigit')
self.checkequalnofix(True, '\u0660', 'isdigit')
def test_isnumeric(self):
self.checkequalnofix(False, '', 'isnumeric')
self.checkequalnofix(False, 'a', 'isnumeric')
self.checkequalnofix(True, '0', 'isnumeric')
self.checkequalnofix(True, '\u2460', 'isnumeric')
self.checkequalnofix(True, '\xbc', 'isnumeric')
self.checkequalnofix(True, '\u0660', 'isnumeric')
self.checkequalnofix(True, '0123456789', 'isnumeric')
self.checkequalnofix(False, '0123456789a', 'isnumeric')
self.assertRaises(TypeError, "abc".isnumeric, 42)
def test_isidentifier(self):
self.assertTrue("a".isidentifier())
self.assertTrue("Z".isidentifier())
self.assertTrue("_".isidentifier())
self.assertTrue("b0".isidentifier())
self.assertTrue("bc".isidentifier())
self.assertTrue("b_".isidentifier())
self.assertTrue("µ".isidentifier())
self.assertFalse(" ".isidentifier())
self.assertFalse("[".isidentifier())
self.assertFalse("©".isidentifier())
self.assertFalse("0".isidentifier())
def test_isprintable(self):
self.assertTrue("".isprintable())
self.assertTrue(" ".isprintable())
self.assertTrue("abcdefg".isprintable())
self.assertFalse("abcdefg\n".isprintable())
# some defined Unicode character
self.assertTrue("\u0374".isprintable())
# undefined character
self.assertFalse("\u0378".isprintable())
# single surrogate character
self.assertFalse("\ud800".isprintable())
def test_contains(self):
# Testing Unicode contains method
self.assertIn('a', 'abdb')
self.assertIn('a', 'bdab')
self.assertIn('a', 'bdaba')
self.assertIn('a', 'bdba')
self.assertNotIn('a', 'bdb')
self.assertIn('a', 'bdba')
self.assertIn('a', ('a',1,None))
self.assertIn('a', (1,None,'a'))
self.assertIn('a', ('a',1,None))
self.assertIn('a', (1,None,'a'))
self.assertNotIn('a', ('x',1,'y'))
self.assertNotIn('a', ('x',1,None))
self.assertNotIn('abcd', 'abcxxxx')
self.assertIn('ab', 'abcd')
self.assertIn('ab', 'abc')
self.assertIn('ab', (1,None,'ab'))
self.assertIn('', 'abc')
self.assertIn('', '')
self.assertIn('', 'abc')
self.assertNotIn('\0', 'abc')
self.assertIn('\0', '\0abc')
self.assertIn('\0', 'abc\0')
self.assertIn('a', '\0abc')
self.assertIn('asdf', 'asdf')
self.assertNotIn('asdf', 'asd')
self.assertNotIn('asdf', '')
self.assertRaises(TypeError, "abc".__contains__)
def test_format(self):
self.assertEqual(''.format(), '')
self.assertEqual('a'.format(), 'a')
self.assertEqual('ab'.format(), 'ab')
self.assertEqual('a{{'.format(), 'a{')
self.assertEqual('a}}'.format(), 'a}')
self.assertEqual('{{b'.format(), '{b')
self.assertEqual('}}b'.format(), '}b')
self.assertEqual('a{{b'.format(), 'a{b')
# examples from the PEP:
import datetime
self.assertEqual("My name is {0}".format('Fred'), "My name is Fred")
self.assertEqual("My name is {0[name]}".format(dict(name='Fred')),
"My name is Fred")
self.assertEqual("My name is {0} :-{{}}".format('Fred'),
"My name is Fred :-{}")
d = datetime.date(2007, 8, 18)
self.assertEqual("The year is {0.year}".format(d),
"The year is 2007")
# classes we'll use for testing
class C:
def __init__(self, x=100):
self._x = x
def __format__(self, spec):
return spec
class D:
def __init__(self, x):
self.x = x
def __format__(self, spec):
return str(self.x)
# class with __str__, but no __format__
class E:
def __init__(self, x):
self.x = x
def __str__(self):
return 'E(' + self.x + ')'
# class with __repr__, but no __format__ or __str__
class F:
def __init__(self, x):
self.x = x
def __repr__(self):
return 'F(' + self.x + ')'
# class with __format__ that forwards to string, for some format_spec's
class G:
def __init__(self, x):
self.x = x
def __str__(self):
return "string is " + self.x
def __format__(self, format_spec):
if format_spec == 'd':
return 'G(' + self.x + ')'
return object.__format__(self, format_spec)
class I(datetime.date):
def __format__(self, format_spec):
return self.strftime(format_spec)
class J(int):
def __format__(self, format_spec):
return int.__format__(self * 2, format_spec)
self.assertEqual(''.format(), '')
self.assertEqual('abc'.format(), 'abc')
self.assertEqual('{0}'.format('abc'), 'abc')
self.assertEqual('{0:}'.format('abc'), 'abc')
# self.assertEqual('{ 0 }'.format('abc'), 'abc')
self.assertEqual('X{0}'.format('abc'), 'Xabc')
self.assertEqual('{0}X'.format('abc'), 'abcX')
self.assertEqual('X{0}Y'.format('abc'), 'XabcY')
self.assertEqual('{1}'.format(1, 'abc'), 'abc')
self.assertEqual('X{1}'.format(1, 'abc'), 'Xabc')
self.assertEqual('{1}X'.format(1, 'abc'), 'abcX')
self.assertEqual('X{1}Y'.format(1, 'abc'), 'XabcY')
self.assertEqual('{0}'.format(-15), '-15')
self.assertEqual('{0}{1}'.format(-15, 'abc'), '-15abc')
self.assertEqual('{0}X{1}'.format(-15, 'abc'), '-15Xabc')
self.assertEqual('{{'.format(), '{')
self.assertEqual('}}'.format(), '}')
self.assertEqual('{{}}'.format(), '{}')
self.assertEqual('{{x}}'.format(), '{x}')
self.assertEqual('{{{0}}}'.format(123), '{123}')
self.assertEqual('{{{{0}}}}'.format(), '{{0}}')
self.assertEqual('}}{{'.format(), '}{')
self.assertEqual('}}x{{'.format(), '}x{')
# weird field names
self.assertEqual("{0[foo-bar]}".format({'foo-bar':'baz'}), 'baz')
self.assertEqual("{0[foo bar]}".format({'foo bar':'baz'}), 'baz')
self.assertEqual("{0[ ]}".format({' ':3}), '3')
self.assertEqual('{foo._x}'.format(foo=C(20)), '20')
self.assertEqual('{1}{0}'.format(D(10), D(20)), '2010')
self.assertEqual('{0._x.x}'.format(C(D('abc'))), 'abc')
self.assertEqual('{0[0]}'.format(['abc', 'def']), 'abc')
self.assertEqual('{0[1]}'.format(['abc', 'def']), 'def')
self.assertEqual('{0[1][0]}'.format(['abc', ['def']]), 'def')
self.assertEqual('{0[1][0].x}'.format(['abc', [D('def')]]), 'def')
# strings
self.assertEqual('{0:.3s}'.format('abc'), 'abc')
self.assertEqual('{0:.3s}'.format('ab'), 'ab')
self.assertEqual('{0:.3s}'.format('abcdef'), 'abc')
self.assertEqual('{0:.0s}'.format('abcdef'), '')
self.assertEqual('{0:3.3s}'.format('abc'), 'abc')
self.assertEqual('{0:2.3s}'.format('abc'), 'abc')
self.assertEqual('{0:2.2s}'.format('abc'), 'ab')
self.assertEqual('{0:3.2s}'.format('abc'), 'ab ')
self.assertEqual('{0:x<0s}'.format('result'), 'result')
self.assertEqual('{0:x<5s}'.format('result'), 'result')
self.assertEqual('{0:x<6s}'.format('result'), 'result')
self.assertEqual('{0:x<7s}'.format('result'), 'resultx')
self.assertEqual('{0:x<8s}'.format('result'), 'resultxx')
self.assertEqual('{0: <7s}'.format('result'), 'result ')
self.assertEqual('{0:<7s}'.format('result'), 'result ')
self.assertEqual('{0:>7s}'.format('result'), ' result')
self.assertEqual('{0:>8s}'.format('result'), ' result')
self.assertEqual('{0:^8s}'.format('result'), ' result ')
self.assertEqual('{0:^9s}'.format('result'), ' result ')
self.assertEqual('{0:^10s}'.format('result'), ' result ')
self.assertEqual('{0:10000}'.format('a'), 'a' + ' ' * 9999)
self.assertEqual('{0:10000}'.format(''), ' ' * 10000)
self.assertEqual('{0:10000000}'.format(''), ' ' * 10000000)
# format specifiers for user defined type
self.assertEqual('{0:abc}'.format(C()), 'abc')
# !r, !s and !a coercions
self.assertEqual('{0!s}'.format('Hello'), 'Hello')
self.assertEqual('{0!s:}'.format('Hello'), 'Hello')
self.assertEqual('{0!s:15}'.format('Hello'), 'Hello ')
self.assertEqual('{0!s:15s}'.format('Hello'), 'Hello ')
self.assertEqual('{0!r}'.format('Hello'), "'Hello'")
self.assertEqual('{0!r:}'.format('Hello'), "'Hello'")
self.assertEqual('{0!r}'.format(F('Hello')), 'F(Hello)')
self.assertEqual('{0!r}'.format('\u0378'), "'\\u0378'") # nonprintable
self.assertEqual('{0!r}'.format('\u0374'), "'\u0374'") # printable
self.assertEqual('{0!r}'.format(F('\u0374')), 'F(\u0374)')
self.assertEqual('{0!a}'.format('Hello'), "'Hello'")
self.assertEqual('{0!a}'.format('\u0378'), "'\\u0378'") # nonprintable
self.assertEqual('{0!a}'.format('\u0374'), "'\\u0374'") # printable
self.assertEqual('{0!a:}'.format('Hello'), "'Hello'")
self.assertEqual('{0!a}'.format(F('Hello')), 'F(Hello)')
self.assertEqual('{0!a}'.format(F('\u0374')), 'F(\\u0374)')
# test fallback to object.__format__
self.assertEqual('{0}'.format({}), '{}')
self.assertEqual('{0}'.format([]), '[]')
self.assertEqual('{0}'.format([1]), '[1]')
self.assertEqual('{0:d}'.format(G('data')), 'G(data)')
self.assertEqual('{0!s}'.format(G('data')), 'string is data')
msg = 'object.__format__ with a non-empty format string is deprecated'
with support.check_warnings((msg, PendingDeprecationWarning)):
self.assertEqual('{0:^10}'.format(E('data')), ' E(data) ')
self.assertEqual('{0:^10s}'.format(E('data')), ' E(data) ')
self.assertEqual('{0:>15s}'.format(G('data')), ' string is data')
self.assertEqual("{0:date: %Y-%m-%d}".format(I(year=2007,
month=8,
day=27)),
"date: 2007-08-27")
# test deriving from a builtin type and overriding __format__
self.assertEqual("{0}".format(J(10)), "20")
# string format specifiers
self.assertEqual('{0:}'.format('a'), 'a')
# computed format specifiers
self.assertEqual("{0:.{1}}".format('hello world', 5), 'hello')
self.assertEqual("{0:.{1}s}".format('hello world', 5), 'hello')
self.assertEqual("{0:.{precision}s}".format('hello world', precision=5), 'hello')
self.assertEqual("{0:{width}.{precision}s}".format('hello world', width=10, precision=5), 'hello ')
self.assertEqual("{0:{width}.{precision}s}".format('hello world', width='10', precision='5'), 'hello ')
# test various errors
self.assertRaises(ValueError, '{'.format)
self.assertRaises(ValueError, '}'.format)
self.assertRaises(ValueError, 'a{'.format)
self.assertRaises(ValueError, 'a}'.format)
self.assertRaises(ValueError, '{a'.format)
self.assertRaises(ValueError, '}a'.format)
self.assertRaises(IndexError, '{0}'.format)
self.assertRaises(IndexError, '{1}'.format, 'abc')
self.assertRaises(KeyError, '{x}'.format)
self.assertRaises(ValueError, "}{".format)
self.assertRaises(ValueError, "abc{0:{}".format)
self.assertRaises(ValueError, "{0".format)
self.assertRaises(IndexError, "{0.}".format)
self.assertRaises(ValueError, "{0.}".format, 0)
self.assertRaises(IndexError, "{0[}".format)
self.assertRaises(ValueError, "{0[}".format, [])
self.assertRaises(KeyError, "{0]}".format)
self.assertRaises(ValueError, "{0.[]}".format, 0)
self.assertRaises(ValueError, "{0..foo}".format, 0)
self.assertRaises(ValueError, "{0[0}".format, 0)
self.assertRaises(ValueError, "{0[0:foo}".format, 0)
self.assertRaises(KeyError, "{c]}".format)
self.assertRaises(ValueError, "{{ {{{0}}".format, 0)
self.assertRaises(ValueError, "{0}}".format, 0)
self.assertRaises(KeyError, "{foo}".format, bar=3)
self.assertRaises(ValueError, "{0!x}".format, 3)
self.assertRaises(ValueError, "{0!}".format, 0)
self.assertRaises(ValueError, "{0!rs}".format, 0)
self.assertRaises(ValueError, "{!}".format)
self.assertRaises(IndexError, "{:}".format)
self.assertRaises(IndexError, "{:s}".format)
self.assertRaises(IndexError, "{}".format)
big = "23098475029384702983476098230754973209482573"
self.assertRaises(ValueError, ("{" + big + "}").format)
self.assertRaises(ValueError, ("{[" + big + "]}").format, [0])
# issue 6089
self.assertRaises(ValueError, "{0[0]x}".format, [None])
self.assertRaises(ValueError, "{0[0](10)}".format, [None])
# can't have a replacement on the field name portion
self.assertRaises(TypeError, '{0[{1}]}'.format, 'abcdefg', 4)
# exceed maximum recursion depth
self.assertRaises(ValueError, "{0:{1:{2}}}".format, 'abc', 's', '')
self.assertRaises(ValueError, "{0:{1:{2:{3:{4:{5:{6}}}}}}}".format,
0, 1, 2, 3, 4, 5, 6, 7)
# string format spec errors
self.assertRaises(ValueError, "{0:-s}".format, '')
self.assertRaises(ValueError, format, "", "-")
self.assertRaises(ValueError, "{0:=s}".format, '')
# Alternate formatting is not supported
self.assertRaises(ValueError, format, '', '#')
self.assertRaises(ValueError, format, '', '#20')
def test_format_map(self):
self.assertEqual(''.format_map({}), '')
self.assertEqual('a'.format_map({}), 'a')
self.assertEqual('ab'.format_map({}), 'ab')
self.assertEqual('a{{'.format_map({}), 'a{')
self.assertEqual('a}}'.format_map({}), 'a}')
self.assertEqual('{{b'.format_map({}), '{b')
self.assertEqual('}}b'.format_map({}), '}b')
self.assertEqual('a{{b'.format_map({}), 'a{b')
# using mappings
class Mapping(dict):
def __missing__(self, key):
return key
self.assertEqual('{hello}'.format_map(Mapping()), 'hello')
self.assertEqual('{a} {world}'.format_map(Mapping(a='hello')), 'hello world')
class InternalMapping:
def __init__(self):
self.mapping = {'a': 'hello'}
def __getitem__(self, key):
return self.mapping[key]
self.assertEqual('{a}'.format_map(InternalMapping()), 'hello')
class C:
def __init__(self, x=100):
self._x = x
def __format__(self, spec):
return spec
self.assertEqual('{foo._x}'.format_map({'foo': C(20)}), '20')
# test various errors
self.assertRaises(TypeError, '{'.format_map)
self.assertRaises(TypeError, '}'.format_map)
self.assertRaises(TypeError, 'a{'.format_map)
self.assertRaises(TypeError, 'a}'.format_map)
self.assertRaises(TypeError, '{a'.format_map)
self.assertRaises(TypeError, '}a'.format_map)
def test_format_auto_numbering(self):
class C:
def __init__(self, x=100):
self._x = x
def __format__(self, spec):
return spec
self.assertEqual('{}'.format(10), '10')
self.assertEqual('{:5}'.format('s'), 's ')
self.assertEqual('{!r}'.format('s'), "'s'")
self.assertEqual('{._x}'.format(C(10)), '10')
self.assertEqual('{[1]}'.format([1, 2]), '2')
self.assertEqual('{[a]}'.format({'a':4, 'b':2}), '4')
self.assertEqual('a{}b{}c'.format(0, 1), 'a0b1c')
self.assertEqual('a{:{}}b'.format('x', '^10'), 'a x b')
self.assertEqual('a{:{}x}b'.format(20, '#'), 'a0x14b')
# can't mix and match numbering and auto-numbering
self.assertRaises(ValueError, '{}{1}'.format, 1, 2)
self.assertRaises(ValueError, '{1}{}'.format, 1, 2)
self.assertRaises(ValueError, '{:{1}}'.format, 1, 2)
self.assertRaises(ValueError, '{0:{}}'.format, 1, 2)
# can mix and match auto-numbering and named
self.assertEqual('{f}{}'.format(4, f='test'), 'test4')
self.assertEqual('{}{f}'.format(4, f='test'), '4test')
self.assertEqual('{:{f}}{g}{}'.format(1, 3, g='g', f=2), ' 1g3')
self.assertEqual('{f:{}}{}{g}'.format(2, 4, f=1, g='g'), ' 14g')
def test_formatting(self):
string_tests.MixinStrUnicodeUserStringTest.test_formatting(self)
# Testing Unicode formatting strings...
self.assertEqual("%s, %s" % ("abc", "abc"), 'abc, abc')
self.assertEqual("%s, %s, %i, %f, %5.2f" % ("abc", "abc", 1, 2, 3), 'abc, abc, 1, 2.000000, 3.00')
self.assertEqual("%s, %s, %i, %f, %5.2f" % ("abc", "abc", 1, -2, 3), 'abc, abc, 1, -2.000000, 3.00')
self.assertEqual("%s, %s, %i, %f, %5.2f" % ("abc", "abc", -1, -2, 3.5), 'abc, abc, -1, -2.000000, 3.50')
self.assertEqual("%s, %s, %i, %f, %5.2f" % ("abc", "abc", -1, -2, 3.57), 'abc, abc, -1, -2.000000, 3.57')
self.assertEqual("%s, %s, %i, %f, %5.2f" % ("abc", "abc", -1, -2, 1003.57), 'abc, abc, -1, -2.000000, 1003.57')
if not sys.platform.startswith('java'):
self.assertEqual("%r, %r" % (b"abc", "abc"), "b'abc', 'abc'")
self.assertEqual("%r" % ("\u1234",), "'\u1234'")
self.assertEqual("%a" % ("\u1234",), "'\\u1234'")
self.assertEqual("%(x)s, %(y)s" % {'x':"abc", 'y':"def"}, 'abc, def')
self.assertEqual("%(x)s, %(\xfc)s" % {'x':"abc", '\xfc':"def"}, 'abc, def')
self.assertEqual('%c' % 0x1234, '\u1234')
self.assertEqual('%c' % 0x21483, '\U00021483')
self.assertRaises(OverflowError, "%c".__mod__, (0x110000,))
self.assertEqual('%c' % '\U00021483', '\U00021483')
self.assertRaises(TypeError, "%c".__mod__, "aa")
self.assertRaises(ValueError, "%.1\u1032f".__mod__, (1.0/3))
# formatting jobs delegated from the string implementation:
self.assertEqual('...%(foo)s...' % {'foo':"abc"}, '...abc...')
self.assertEqual('...%(foo)s...' % {'foo':"abc"}, '...abc...')
self.assertEqual('...%(foo)s...' % {'foo':"abc"}, '...abc...')
self.assertEqual('...%(foo)s...' % {'foo':"abc"}, '...abc...')
self.assertEqual('...%(foo)s...' % {'foo':"abc",'def':123}, '...abc...')
self.assertEqual('...%(foo)s...' % {'foo':"abc",'def':123}, '...abc...')
self.assertEqual('...%s...%s...%s...%s...' % (1,2,3,"abc"), '...1...2...3...abc...')
self.assertEqual('...%%...%%s...%s...%s...%s...%s...' % (1,2,3,"abc"), '...%...%s...1...2...3...abc...')
self.assertEqual('...%s...' % "abc", '...abc...')
self.assertEqual('%*s' % (5,'abc',), ' abc')
self.assertEqual('%*s' % (-5,'abc',), 'abc ')
self.assertEqual('%*.*s' % (5,2,'abc',), ' ab')
self.assertEqual('%*.*s' % (5,3,'abc',), ' abc')
self.assertEqual('%i %*.*s' % (10, 5,3,'abc',), '10 abc')
self.assertEqual('%i%s %*.*s' % (10, 3, 5, 3, 'abc',), '103 abc')
self.assertEqual('%c' % 'a', 'a')
class Wrapper:
def __str__(self):
return '\u1234'
self.assertEqual('%s' % Wrapper(), '\u1234')
# issue 3382
NAN = float('nan')
INF = float('inf')
self.assertEqual('%f' % NAN, 'nan')
self.assertEqual('%F' % NAN, 'NAN')
self.assertEqual('%f' % INF, 'inf')
self.assertEqual('%F' % INF, 'INF')
@support.run_with_locale('LC_ALL', 'de_DE', 'fr_FR')
def test_format_float(self):
# should not format with a comma, but always with C locale
self.assertEqual('1.0', '%.1f' % 1.0)
def test_constructor(self):
# unicode(obj) tests (this maps to PyObject_Unicode() at C level)
self.assertEqual(
str('unicode remains unicode'),
'unicode remains unicode'
)
class UnicodeSubclass(str):
pass
self.assertEqual(
str(UnicodeSubclass('unicode subclass becomes unicode')),
'unicode subclass becomes unicode'
)
self.assertEqual(
str('strings are converted to unicode'),
'strings are converted to unicode'
)
class StringCompat:
def __init__(self, x):
self.x = x
def __str__(self):
return self.x
self.assertEqual(
str(StringCompat('__str__ compatible objects are recognized')),
'__str__ compatible objects are recognized'
)
# unicode(obj) is compatible to str():
o = StringCompat('unicode(obj) is compatible to str()')
self.assertEqual(str(o), 'unicode(obj) is compatible to str()')
self.assertEqual(str(o), 'unicode(obj) is compatible to str()')
for obj in (123, 123.45, 123):
self.assertEqual(str(obj), str(str(obj)))
# unicode(obj, encoding, error) tests (this maps to
# PyUnicode_FromEncodedObject() at C level)
if not sys.platform.startswith('java'):
self.assertRaises(
TypeError,
str,
'decoding unicode is not supported',
'utf-8',
'strict'
)
self.assertEqual(
str(b'strings are decoded to unicode', 'utf-8', 'strict'),
'strings are decoded to unicode'
)
if not sys.platform.startswith('java'):
self.assertEqual(
str(
memoryview(b'character buffers are decoded to unicode'),
'utf-8',
'strict'
),
'character buffers are decoded to unicode'
)
self.assertRaises(TypeError, str, 42, 42, 42)
def test_codecs_utf7(self):
utfTests = [
('A\u2262\u0391.', b'A+ImIDkQ.'), # RFC2152 example
('Hi Mom -\u263a-!', b'Hi Mom -+Jjo--!'), # RFC2152 example
('\u65E5\u672C\u8A9E', b'+ZeVnLIqe-'), # RFC2152 example
('Item 3 is \u00a31.', b'Item 3 is +AKM-1.'), # RFC2152 example
('+', b'+-'),
('+-', b'+--'),
('+?', b'+-?'),
('\?', b'+AFw?'),
('+?', b'+-?'),
(r'\\?', b'+AFwAXA?'),
(r'\\\?', b'+AFwAXABc?'),
(r'++--', b'+-+---'),
('\U000abcde', b'+2m/c3g-'), # surrogate pairs
('/', b'/'),
]
for (x, y) in utfTests:
self.assertEqual(x.encode('utf-7'), y)
# Unpaired surrogates not supported
self.assertRaises(UnicodeError, str, b'+3ADYAA-', 'utf-7')
self.assertEqual(str(b'+3ADYAA-', 'utf-7', 'replace'), '\ufffd\ufffd')
# Issue #2242: crash on some Windows/MSVC versions
self.assertEqual(b'+\xc1'.decode('utf-7'), '\xc1')
# Direct encoded characters
set_d = "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789'(),-./:?"
# Optional direct characters
set_o = '!"#$%&*;<=>@[]^_`{|}'
for c in set_d:
self.assertEqual(c.encode('utf7'), c.encode('ascii'))
self.assertEqual(c.encode('ascii').decode('utf7'), c)
for c in set_o:
self.assertEqual(c.encode('ascii').decode('utf7'), c)
def test_codecs_utf8(self):
self.assertEqual(''.encode('utf-8'), b'')
self.assertEqual('\u20ac'.encode('utf-8'), b'\xe2\x82\xac')
if sys.maxunicode == 65535:
self.assertEqual('\ud800\udc02'.encode('utf-8'), b'\xf0\x90\x80\x82')
self.assertEqual('\ud84d\udc56'.encode('utf-8'), b'\xf0\xa3\x91\x96')
self.assertEqual('\ud800'.encode('utf-8', 'surrogatepass'), b'\xed\xa0\x80')
self.assertEqual('\udc00'.encode('utf-8', 'surrogatepass'), b'\xed\xb0\x80')
if sys.maxunicode == 65535:
self.assertEqual(
('\ud800\udc02'*1000).encode('utf-8'),
b'\xf0\x90\x80\x82'*1000)
self.assertEqual(
'\u6b63\u78ba\u306b\u8a00\u3046\u3068\u7ffb\u8a33\u306f'
'\u3055\u308c\u3066\u3044\u307e\u305b\u3093\u3002\u4e00'
'\u90e8\u306f\u30c9\u30a4\u30c4\u8a9e\u3067\u3059\u304c'
'\u3001\u3042\u3068\u306f\u3067\u305f\u3089\u3081\u3067'
'\u3059\u3002\u5b9f\u969b\u306b\u306f\u300cWenn ist das'
' Nunstuck git und'.encode('utf-8'),
b'\xe6\xad\xa3\xe7\xa2\xba\xe3\x81\xab\xe8\xa8\x80\xe3\x81'
b'\x86\xe3\x81\xa8\xe7\xbf\xbb\xe8\xa8\xb3\xe3\x81\xaf\xe3'
b'\x81\x95\xe3\x82\x8c\xe3\x81\xa6\xe3\x81\x84\xe3\x81\xbe'
b'\xe3\x81\x9b\xe3\x82\x93\xe3\x80\x82\xe4\xb8\x80\xe9\x83'
b'\xa8\xe3\x81\xaf\xe3\x83\x89\xe3\x82\xa4\xe3\x83\x84\xe8'
b'\xaa\x9e\xe3\x81\xa7\xe3\x81\x99\xe3\x81\x8c\xe3\x80\x81'
b'\xe3\x81\x82\xe3\x81\xa8\xe3\x81\xaf\xe3\x81\xa7\xe3\x81'
b'\x9f\xe3\x82\x89\xe3\x82\x81\xe3\x81\xa7\xe3\x81\x99\xe3'
b'\x80\x82\xe5\xae\x9f\xe9\x9a\x9b\xe3\x81\xab\xe3\x81\xaf'
b'\xe3\x80\x8cWenn ist das Nunstuck git und'
)
# UTF-8 specific decoding tests
self.assertEqual(str(b'\xf0\xa3\x91\x96', 'utf-8'), '\U00023456' )
self.assertEqual(str(b'\xf0\x90\x80\x82', 'utf-8'), '\U00010002' )
self.assertEqual(str(b'\xe2\x82\xac', 'utf-8'), '\u20ac' )
# Other possible utf-8 test cases:
# * strict decoding testing for all of the
# UTF8_ERROR cases in PyUnicode_DecodeUTF8
def test_utf8_decode_valid_sequences(self):
sequences = [
# single byte
(b'\x00', '\x00'), (b'a', 'a'), (b'\x7f', '\x7f'),
# 2 bytes
(b'\xc2\x80', '\x80'), (b'\xdf\xbf', '\u07ff'),
# 3 bytes
(b'\xe0\xa0\x80', '\u0800'), (b'\xed\x9f\xbf', '\ud7ff'),
(b'\xee\x80\x80', '\uE000'), (b'\xef\xbf\xbf', '\uffff'),
# 4 bytes
(b'\xF0\x90\x80\x80', '\U00010000'),
(b'\xf4\x8f\xbf\xbf', '\U0010FFFF')
]
for seq, res in sequences:
self.assertEqual(seq.decode('utf-8'), res)
def test_utf8_decode_invalid_sequences(self):
# continuation bytes in a sequence of 2, 3, or 4 bytes
continuation_bytes = [bytes([x]) for x in range(0x80, 0xC0)]
# start bytes of a 2-byte sequence equivalent to codepoints < 0x7F
invalid_2B_seq_start_bytes = [bytes([x]) for x in range(0xC0, 0xC2)]
# start bytes of a 4-byte sequence equivalent to codepoints > 0x10FFFF
invalid_4B_seq_start_bytes = [bytes([x]) for x in range(0xF5, 0xF8)]
invalid_start_bytes = (
continuation_bytes + invalid_2B_seq_start_bytes +
invalid_4B_seq_start_bytes + [bytes([x]) for x in range(0xF7, 0x100)]
)
for byte in invalid_start_bytes:
self.assertRaises(UnicodeDecodeError, byte.decode, 'utf-8')
for sb in invalid_2B_seq_start_bytes:
for cb in continuation_bytes:
self.assertRaises(UnicodeDecodeError, (sb+cb).decode, 'utf-8')
for sb in invalid_4B_seq_start_bytes:
for cb1 in continuation_bytes[:3]:
for cb3 in continuation_bytes[:3]:
self.assertRaises(UnicodeDecodeError,
(sb+cb1+b'\x80'+cb3).decode, 'utf-8')
for cb in [bytes([x]) for x in range(0x80, 0xA0)]:
self.assertRaises(UnicodeDecodeError,
(b'\xE0'+cb+b'\x80').decode, 'utf-8')
self.assertRaises(UnicodeDecodeError,
(b'\xE0'+cb+b'\xBF').decode, 'utf-8')
# surrogates
for cb in [bytes([x]) for x in range(0xA0, 0xC0)]:
self.assertRaises(UnicodeDecodeError,
(b'\xED'+cb+b'\x80').decode, 'utf-8')
self.assertRaises(UnicodeDecodeError,
(b'\xED'+cb+b'\xBF').decode, 'utf-8')
for cb in [bytes([x]) for x in range(0x80, 0x90)]:
self.assertRaises(UnicodeDecodeError,
(b'\xF0'+cb+b'\x80\x80').decode, 'utf-8')
self.assertRaises(UnicodeDecodeError,
(b'\xF0'+cb+b'\xBF\xBF').decode, 'utf-8')
for cb in [bytes([x]) for x in range(0x90, 0xC0)]:
self.assertRaises(UnicodeDecodeError,
(b'\xF4'+cb+b'\x80\x80').decode, 'utf-8')
self.assertRaises(UnicodeDecodeError,
(b'\xF4'+cb+b'\xBF\xBF').decode, 'utf-8')
def test_issue8271(self):
# Issue #8271: during the decoding of an invalid UTF-8 byte sequence,
# only the start byte and the continuation byte(s) are now considered
# invalid, instead of the number of bytes specified by the start byte.
# See http://www.unicode.org/versions/Unicode5.2.0/ch03.pdf (page 95,
# table 3-8, Row 2) for more information about the algorithm used.
FFFD = '\ufffd'
sequences = [
# invalid start bytes
(b'\x80', FFFD), # continuation byte
(b'\x80\x80', FFFD*2), # 2 continuation bytes
(b'\xc0', FFFD),
(b'\xc0\xc0', FFFD*2),
(b'\xc1', FFFD),
(b'\xc1\xc0', FFFD*2),
(b'\xc0\xc1', FFFD*2),
# with start byte of a 2-byte sequence
(b'\xc2', FFFD), # only the start byte
(b'\xc2\xc2', FFFD*2), # 2 start bytes
(b'\xc2\xc2\xc2', FFFD*3), # 2 start bytes
(b'\xc2\x41', FFFD+'A'), # invalid continuation byte
# with start byte of a 3-byte sequence
(b'\xe1', FFFD), # only the start byte
(b'\xe1\xe1', FFFD*2), # 2 start bytes
(b'\xe1\xe1\xe1', FFFD*3), # 3 start bytes
(b'\xe1\xe1\xe1\xe1', FFFD*4), # 4 start bytes
(b'\xe1\x80', FFFD), # only 1 continuation byte
(b'\xe1\x41', FFFD+'A'), # invalid continuation byte
(b'\xe1\x41\x80', FFFD+'A'+FFFD), # invalid cb followed by valid cb
(b'\xe1\x41\x41', FFFD+'AA'), # 2 invalid continuation bytes
(b'\xe1\x80\x41', FFFD+'A'), # only 1 valid continuation byte
(b'\xe1\x80\xe1\x41', FFFD*2+'A'), # 1 valid and the other invalid
(b'\xe1\x41\xe1\x80', FFFD+'A'+FFFD), # 1 invalid and the other valid
# with start byte of a 4-byte sequence
(b'\xf1', FFFD), # only the start byte
(b'\xf1\xf1', FFFD*2), # 2 start bytes
(b'\xf1\xf1\xf1', FFFD*3), # 3 start bytes
(b'\xf1\xf1\xf1\xf1', FFFD*4), # 4 start bytes
(b'\xf1\xf1\xf1\xf1\xf1', FFFD*5), # 5 start bytes
(b'\xf1\x80', FFFD), # only 1 continuation bytes
(b'\xf1\x80\x80', FFFD), # only 2 continuation bytes
(b'\xf1\x80\x41', FFFD+'A'), # 1 valid cb and 1 invalid
(b'\xf1\x80\x41\x41', FFFD+'AA'), # 1 valid cb and 1 invalid
(b'\xf1\x80\x80\x41', FFFD+'A'), # 2 valid cb and 1 invalid
(b'\xf1\x41\x80', FFFD+'A'+FFFD), # 1 invalid cv and 1 valid
(b'\xf1\x41\x80\x80', FFFD+'A'+FFFD*2), # 1 invalid cb and 2 invalid
(b'\xf1\x41\x80\x41', FFFD+'A'+FFFD+'A'), # 2 invalid cb and 1 invalid
(b'\xf1\x41\x41\x80', FFFD+'AA'+FFFD), # 1 valid cb and 1 invalid
(b'\xf1\x41\xf1\x80', FFFD+'A'+FFFD),
(b'\xf1\x41\x80\xf1', FFFD+'A'+FFFD*2),
(b'\xf1\xf1\x80\x41', FFFD*2+'A'),
(b'\xf1\x41\xf1\xf1', FFFD+'A'+FFFD*2),
# with invalid start byte of a 4-byte sequence (rfc2279)
(b'\xf5', FFFD), # only the start byte
(b'\xf5\xf5', FFFD*2), # 2 start bytes
(b'\xf5\x80', FFFD*2), # only 1 continuation byte
(b'\xf5\x80\x80', FFFD*3), # only 2 continuation byte
(b'\xf5\x80\x80\x80', FFFD*4), # 3 continuation bytes
(b'\xf5\x80\x41', FFFD*2+'A'), # 1 valid cb and 1 invalid
(b'\xf5\x80\x41\xf5', FFFD*2+'A'+FFFD),
(b'\xf5\x41\x80\x80\x41', FFFD+'A'+FFFD*2+'A'),
# with invalid start byte of a 5-byte sequence (rfc2279)
(b'\xf8', FFFD), # only the start byte
(b'\xf8\xf8', FFFD*2), # 2 start bytes
(b'\xf8\x80', FFFD*2), # only one continuation byte
(b'\xf8\x80\x41', FFFD*2 + 'A'), # 1 valid cb and 1 invalid
(b'\xf8\x80\x80\x80\x80', FFFD*5), # invalid 5 bytes seq with 5 bytes
# with invalid start byte of a 6-byte sequence (rfc2279)
(b'\xfc', FFFD), # only the start byte
(b'\xfc\xfc', FFFD*2), # 2 start bytes
(b'\xfc\x80\x80', FFFD*3), # only 2 continuation bytes
(b'\xfc\x80\x80\x80\x80\x80', FFFD*6), # 6 continuation bytes
# invalid start byte
(b'\xfe', FFFD),
(b'\xfe\x80\x80', FFFD*3),
# other sequences
(b'\xf1\x80\x41\x42\x43', '\ufffd\x41\x42\x43'),
(b'\xf1\x80\xff\x42\x43', '\ufffd\ufffd\x42\x43'),
(b'\xf1\x80\xc2\x81\x43', '\ufffd\x81\x43'),
(b'\x61\xF1\x80\x80\xE1\x80\xC2\x62\x80\x63\x80\xBF\x64',
'\x61\uFFFD\uFFFD\uFFFD\x62\uFFFD\x63\uFFFD\uFFFD\x64'),
]
for n, (seq, res) in enumerate(sequences):
self.assertRaises(UnicodeDecodeError, seq.decode, 'utf-8', 'strict')
self.assertEqual(seq.decode('utf-8', 'replace'), res)
self.assertEqual((seq+b'b').decode('utf-8', 'replace'), res+'b')
self.assertEqual(seq.decode('utf-8', 'ignore'),
res.replace('\uFFFD', ''))
def test_codecs_idna(self):
# Test whether trailing dot is preserved
self.assertEqual("www.python.org.".encode("idna"), b"www.python.org.")
def test_codecs_errors(self):
# Error handling (encoding)
self.assertRaises(UnicodeError, 'Andr\202 x'.encode, 'ascii')
self.assertRaises(UnicodeError, 'Andr\202 x'.encode, 'ascii','strict')
self.assertEqual('Andr\202 x'.encode('ascii','ignore'), b"Andr x")
self.assertEqual('Andr\202 x'.encode('ascii','replace'), b"Andr? x")
self.assertEqual('Andr\202 x'.encode('ascii', 'replace'),
'Andr\202 x'.encode('ascii', errors='replace'))
self.assertEqual('Andr\202 x'.encode('ascii', 'ignore'),
'Andr\202 x'.encode(encoding='ascii', errors='ignore'))
# Error handling (decoding)
self.assertRaises(UnicodeError, str, b'Andr\202 x', 'ascii')
self.assertRaises(UnicodeError, str, b'Andr\202 x', 'ascii', 'strict')
self.assertEqual(str(b'Andr\202 x', 'ascii', 'ignore'), "Andr x")
self.assertEqual(str(b'Andr\202 x', 'ascii', 'replace'), 'Andr\uFFFD x')
# Error handling (unknown character names)
self.assertEqual(b"\\N{foo}xx".decode("unicode-escape", "ignore"), "xx")
# Error handling (truncated escape sequence)
self.assertRaises(UnicodeError, b"\\".decode, "unicode-escape")
self.assertRaises(TypeError, b"hello".decode, "test.unicode1")
self.assertRaises(TypeError, str, b"hello", "test.unicode2")
self.assertRaises(TypeError, "hello".encode, "test.unicode1")
self.assertRaises(TypeError, "hello".encode, "test.unicode2")
# executes PyUnicode_Encode()
import imp
self.assertRaises(
ImportError,
imp.find_module,
"non-existing module",
["non-existing dir"]
)
# Error handling (wrong arguments)
self.assertRaises(TypeError, "hello".encode, 42, 42, 42)
# Error handling (lone surrogate in PyUnicode_TransformDecimalToASCII())
self.assertRaises(UnicodeError, int, "\ud800")
self.assertRaises(UnicodeError, int, "\udf00")
self.assertRaises(UnicodeError, float, "\ud800")
self.assertRaises(UnicodeError, float, "\udf00")
self.assertRaises(UnicodeError, complex, "\ud800")
self.assertRaises(UnicodeError, complex, "\udf00")
def test_codecs(self):
# Encoding
self.assertEqual('hello'.encode('ascii'), b'hello')
self.assertEqual('hello'.encode('utf-7'), b'hello')
self.assertEqual('hello'.encode('utf-8'), b'hello')
self.assertEqual('hello'.encode('utf8'), b'hello')
self.assertEqual('hello'.encode('utf-16-le'), b'h\000e\000l\000l\000o\000')
self.assertEqual('hello'.encode('utf-16-be'), b'\000h\000e\000l\000l\000o')
self.assertEqual('hello'.encode('latin-1'), b'hello')
# Roundtrip safety for BMP (just the first 1024 chars)
for c in range(1024):
u = chr(c)
for encoding in ('utf-7', 'utf-8', 'utf-16', 'utf-16-le',
'utf-16-be', 'raw_unicode_escape',
'unicode_escape', 'unicode_internal'):
self.assertEqual(str(u.encode(encoding),encoding), u)
# Roundtrip safety for BMP (just the first 256 chars)
for c in range(256):
u = chr(c)
for encoding in ('latin-1',):
self.assertEqual(str(u.encode(encoding),encoding), u)
# Roundtrip safety for BMP (just the first 128 chars)
for c in range(128):
u = chr(c)
for encoding in ('ascii',):
self.assertEqual(str(u.encode(encoding),encoding), u)
# Roundtrip safety for non-BMP (just a few chars)
u = '\U00010001\U00020002\U00030003\U00040004\U00050005'
for encoding in ('utf-8', 'utf-16', 'utf-16-le', 'utf-16-be',
#'raw_unicode_escape',
'unicode_escape', 'unicode_internal'):
self.assertEqual(str(u.encode(encoding),encoding), u)
# UTF-8 must be roundtrip safe for all UCS-2 code points
# This excludes surrogates: in the full range, there would be
# a surrogate pair (\udbff\udc00), which gets converted back
# to a non-BMP character (\U0010fc00)
u = ''.join(map(chr, list(range(0,0xd800)) +
list(range(0xe000,0x10000))))
for encoding in ('utf-8',):
self.assertEqual(str(u.encode(encoding),encoding), u)
def test_codecs_charmap(self):
# 0-127
s = bytes(range(128))
for encoding in (
'cp037', 'cp1026',
'cp437', 'cp500', 'cp720', 'cp737', 'cp775', 'cp850',
'cp852', 'cp855', 'cp858', 'cp860', 'cp861', 'cp862',
'cp863', 'cp865', 'cp866',
'iso8859_10', 'iso8859_13', 'iso8859_14', 'iso8859_15',
'iso8859_2', 'iso8859_3', 'iso8859_4', 'iso8859_5', 'iso8859_6',
'iso8859_7', 'iso8859_9', 'koi8_r', 'latin_1',
'mac_cyrillic', 'mac_latin2',
'cp1250', 'cp1251', 'cp1252', 'cp1253', 'cp1254', 'cp1255',
'cp1256', 'cp1257', 'cp1258',
'cp856', 'cp857', 'cp864', 'cp869', 'cp874',
'mac_greek', 'mac_iceland','mac_roman', 'mac_turkish',
'cp1006', 'iso8859_8',
### These have undefined mappings:
#'cp424',
### These fail the round-trip:
#'cp875'
):
self.assertEqual(str(s, encoding).encode(encoding), s)
# 128-255
s = bytes(range(128, 256))
for encoding in (
'cp037', 'cp1026',
'cp437', 'cp500', 'cp720', 'cp737', 'cp775', 'cp850',
'cp852', 'cp855', 'cp858', 'cp860', 'cp861', 'cp862',
'cp863', 'cp865', 'cp866',
'iso8859_10', 'iso8859_13', 'iso8859_14', 'iso8859_15',
'iso8859_2', 'iso8859_4', 'iso8859_5',
'iso8859_9', 'koi8_r', 'latin_1',
'mac_cyrillic', 'mac_latin2',
### These have undefined mappings:
#'cp1250', 'cp1251', 'cp1252', 'cp1253', 'cp1254', 'cp1255',
#'cp1256', 'cp1257', 'cp1258',
#'cp424', 'cp856', 'cp857', 'cp864', 'cp869', 'cp874',
#'iso8859_3', 'iso8859_6', 'iso8859_7',
#'mac_greek', 'mac_iceland','mac_roman', 'mac_turkish',
### These fail the round-trip:
#'cp1006', 'cp875', 'iso8859_8',
):
self.assertEqual(str(s, encoding).encode(encoding), s)
def test_concatenation(self):
self.assertEqual(("abc" "def"), "abcdef")
self.assertEqual(("abc" "def"), "abcdef")
self.assertEqual(("abc" "def"), "abcdef")
self.assertEqual(("abc" "def" "ghi"), "abcdefghi")
self.assertEqual(("abc" "def" "ghi"), "abcdefghi")
def test_printing(self):
class BitBucket:
def write(self, text):
pass
out = BitBucket()
print('abc', file=out)
print('abc', 'def', file=out)
print('abc', 'def', file=out)
print('abc', 'def', file=out)
print('abc\n', file=out)
print('abc\n', end=' ', file=out)
print('abc\n', end=' ', file=out)
print('def\n', file=out)
print('def\n', file=out)
def test_ucs4(self):
x = '\U00100000'
y = x.encode("raw-unicode-escape").decode("raw-unicode-escape")
self.assertEqual(x, y)
y = br'\U00100000'
x = y.decode("raw-unicode-escape").encode("raw-unicode-escape")
self.assertEqual(x, y)
y = br'\U00010000'
x = y.decode("raw-unicode-escape").encode("raw-unicode-escape")
self.assertEqual(x, y)
try:
br'\U11111111'.decode("raw-unicode-escape")
except UnicodeDecodeError as e:
self.assertEqual(e.start, 0)
self.assertEqual(e.end, 10)
else:
self.fail("Should have raised UnicodeDecodeError")
def test_conversion(self):
# Make sure __unicode__() works properly
class Foo0:
def __str__(self):
return "foo"
class Foo1:
def __str__(self):
return "foo"
class Foo2(object):
def __str__(self):
return "foo"
class Foo3(object):
def __str__(self):
return "foo"
class Foo4(str):
def __str__(self):
return "foo"
class Foo5(str):
def __str__(self):
return "foo"
class Foo6(str):
def __str__(self):
return "foos"
def __str__(self):
return "foou"
class Foo7(str):
def __str__(self):
return "foos"
def __str__(self):
return "foou"
class Foo8(str):
def __new__(cls, content=""):
return str.__new__(cls, 2*content)
def __str__(self):
return self
class Foo9(str):
def __str__(self):
return "not unicode"
self.assertEqual(str(Foo0()), "foo")
self.assertEqual(str(Foo1()), "foo")
self.assertEqual(str(Foo2()), "foo")
self.assertEqual(str(Foo3()), "foo")
self.assertEqual(str(Foo4("bar")), "foo")
self.assertEqual(str(Foo5("bar")), "foo")
self.assertEqual(str(Foo6("bar")), "foou")
self.assertEqual(str(Foo7("bar")), "foou")
self.assertEqual(str(Foo8("foo")), "foofoo")
self.assertEqual(str(Foo9("foo")), "not unicode")
def test_unicode_repr(self):
class s1:
def __repr__(self):
return '\\n'
class s2:
def __repr__(self):
return '\\n'
self.assertEqual(repr(s1()), '\\n')
self.assertEqual(repr(s2()), '\\n')
def test_printable_repr(self):
self.assertEqual(repr('\U00010000'), "'%c'" % (0x10000,)) # printable
self.assertEqual(repr('\U00014000'), "'\\U00014000'") # nonprintable
def test_expandtabs_overflows_gracefully(self):
# This test only affects 32-bit platforms because expandtabs can only take
# an int as the max value, not a 64-bit C long. If expandtabs is changed
# to take a 64-bit long, this test should apply to all platforms.
if sys.maxsize > (1 << 32) or struct.calcsize('P') != 4:
return
self.assertRaises(OverflowError, 't\tt\t'.expandtabs, sys.maxsize)
def test_raiseMemError(self):
# Ensure that the freelist contains a consistent object, even
# when a string allocation fails with a MemoryError.
# This used to crash the interpreter,
# or leak references when the number was smaller.
charwidth = 4 if sys.maxunicode >= 0x10000 else 2
# Note: sys.maxsize is half of the actual max allocation because of
# the signedness of Py_ssize_t.
alloc = lambda: "a" * (sys.maxsize // charwidth * 2)
self.assertRaises(MemoryError, alloc)
self.assertRaises(MemoryError, alloc)
def test_format_subclass(self):
class S(str):
def __str__(self):
return '__str__ overridden'
s = S('xxx')
self.assertEqual("%s" % s, '__str__ overridden')
self.assertEqual("{}".format(s), '__str__ overridden')
# Test PyUnicode_FromFormat()
def test_from_format(self):
support.import_module('ctypes')
from ctypes import pythonapi, py_object
if sys.maxunicode == 65535:
name = "PyUnicodeUCS2_FromFormat"
else:
name = "PyUnicodeUCS4_FromFormat"
_PyUnicode_FromFormat = getattr(pythonapi, name)
_PyUnicode_FromFormat.restype = py_object
def PyUnicode_FromFormat(format, *args):
cargs = tuple(
py_object(arg) if isinstance(arg, str) else arg
for arg in args)
return _PyUnicode_FromFormat(format, *cargs)
# ascii format, non-ascii argument
text = PyUnicode_FromFormat(b'ascii\x7f=%U', 'unicode\xe9')
self.assertEqual(text, 'ascii\x7f=unicode\xe9')
# non-ascii format, ascii argument: ensure that PyUnicode_FromFormatV()
# raises an error
self.assertRaisesRegex(ValueError,
'^PyUnicode_FromFormatV\(\) expects an ASCII-encoded format '
'string, got a non-ASCII byte: 0xe9$',
PyUnicode_FromFormat, b'unicode\xe9=%s', 'ascii')
# other tests
text = PyUnicode_FromFormat(b'%%A:%A', 'abc\xe9\uabcd\U0010ffff')
self.assertEqual(text, r"%A:'abc\xe9\uabcd\U0010ffff'")
# Test PyUnicode_AsWideChar()
def test_aswidechar(self):
from _testcapi import unicode_aswidechar
support.import_module('ctypes')
from ctypes import c_wchar, sizeof
wchar, size = unicode_aswidechar('abcdef', 2)
self.assertEqual(size, 2)
self.assertEqual(wchar, 'ab')
wchar, size = unicode_aswidechar('abc', 3)
self.assertEqual(size, 3)
self.assertEqual(wchar, 'abc')
wchar, size = unicode_aswidechar('abc', 4)
self.assertEqual(size, 3)
self.assertEqual(wchar, 'abc\0')
wchar, size = unicode_aswidechar('abc', 10)
self.assertEqual(size, 3)
self.assertEqual(wchar, 'abc\0')
wchar, size = unicode_aswidechar('abc\0def', 20)
self.assertEqual(size, 7)
self.assertEqual(wchar, 'abc\0def\0')
nonbmp = chr(0x10ffff)
if sizeof(c_wchar) == 2:
buflen = 3
nchar = 2
else: # sizeof(c_wchar) == 4
buflen = 2
nchar = 1
wchar, size = unicode_aswidechar(nonbmp, buflen)
self.assertEqual(size, nchar)
self.assertEqual(wchar, nonbmp + '\0')
# Test PyUnicode_AsWideCharString()
def test_aswidecharstring(self):
from _testcapi import unicode_aswidecharstring
support.import_module('ctypes')
from ctypes import c_wchar, sizeof
wchar, size = unicode_aswidecharstring('abc')
self.assertEqual(size, 3)
self.assertEqual(wchar, 'abc\0')
wchar, size = unicode_aswidecharstring('abc\0def')
self.assertEqual(size, 7)
self.assertEqual(wchar, 'abc\0def\0')
nonbmp = chr(0x10ffff)
if sizeof(c_wchar) == 2:
nchar = 2
else: # sizeof(c_wchar) == 4
nchar = 1
wchar, size = unicode_aswidecharstring(nonbmp)
self.assertEqual(size, nchar)
self.assertEqual(wchar, nonbmp + '\0')
def test_main():
support.run_unittest(__name__)
if __name__ == "__main__":
test_main()