cpython/Lib/test/test_format.py
Mark Dickinson 895c9c1d43
bpo-40780: Fix failure of _Py_dg_dtoa to remove trailing zeros (GH-20435)
* Fix failure of _Py_dg_dtoa to remove trailing zeros

* Add regression test and news entry

* Add explanation about why it's safe to strip trailing zeros

* Make code safer, clean up comments, add change note at top of file

* Nitpick: avoid implicit int-to-float conversion in tests
2020-05-29 14:23:57 +01:00

501 lines
23 KiB
Python

from test.support import verbose, TestFailed
import locale
import sys
import test.support as support
import unittest
maxsize = support.MAX_Py_ssize_t
# test string formatting operator (I am not sure if this is being tested
# elsewhere but, surely, some of the given cases are *not* tested because
# they crash python)
# test on bytes object as well
def testformat(formatstr, args, output=None, limit=None, overflowok=False):
if verbose:
if output:
print("{!a} % {!a} =? {!a} ...".format(formatstr, args, output),
end=' ')
else:
print("{!a} % {!a} works? ...".format(formatstr, args), end=' ')
try:
result = formatstr % args
except OverflowError:
if not overflowok:
raise
if verbose:
print('overflow (this is fine)')
else:
if output and limit is None and result != output:
if verbose:
print('no')
raise AssertionError("%r %% %r == %r != %r" %
(formatstr, args, result, output))
# when 'limit' is specified, it determines how many characters
# must match exactly; lengths must always match.
# ex: limit=5, '12345678' matches '12345___'
# (mainly for floating point format tests for which an exact match
# can't be guaranteed due to rounding and representation errors)
elif output and limit is not None and (
len(result)!=len(output) or result[:limit]!=output[:limit]):
if verbose:
print('no')
print("%s %% %s == %s != %s" % \
(repr(formatstr), repr(args), repr(result), repr(output)))
else:
if verbose:
print('yes')
def testcommon(formatstr, args, output=None, limit=None, overflowok=False):
# if formatstr is a str, test str, bytes, and bytearray;
# otherwise, test bytes and bytearray
if isinstance(formatstr, str):
testformat(formatstr, args, output, limit, overflowok)
b_format = formatstr.encode('ascii')
else:
b_format = formatstr
ba_format = bytearray(b_format)
b_args = []
if not isinstance(args, tuple):
args = (args, )
b_args = tuple(args)
if output is None:
b_output = ba_output = None
else:
if isinstance(output, str):
b_output = output.encode('ascii')
else:
b_output = output
ba_output = bytearray(b_output)
testformat(b_format, b_args, b_output, limit, overflowok)
testformat(ba_format, b_args, ba_output, limit, overflowok)
def test_exc(formatstr, args, exception, excmsg):
try:
testformat(formatstr, args)
except exception as exc:
if str(exc) == excmsg:
if verbose:
print("yes")
else:
if verbose: print('no')
print('Unexpected ', exception, ':', repr(str(exc)))
except:
if verbose: print('no')
print('Unexpected exception')
raise
else:
raise TestFailed('did not get expected exception: %s' % excmsg)
def test_exc_common(formatstr, args, exception, excmsg):
# test str and bytes
test_exc(formatstr, args, exception, excmsg)
test_exc(formatstr.encode('ascii'), args, exception, excmsg)
class FormatTest(unittest.TestCase):
def test_common_format(self):
# test the format identifiers that work the same across
# str, bytes, and bytearrays (integer, float, oct, hex)
testcommon("%%", (), "%")
testcommon("%.1d", (1,), "1")
testcommon("%.*d", (sys.maxsize,1), overflowok=True) # expect overflow
testcommon("%.100d", (1,), '00000000000000000000000000000000000000'
'000000000000000000000000000000000000000000000000000000'
'00000001', overflowok=True)
testcommon("%#.117x", (1,), '0x00000000000000000000000000000000000'
'000000000000000000000000000000000000000000000000000000'
'0000000000000000000000000001',
overflowok=True)
testcommon("%#.118x", (1,), '0x00000000000000000000000000000000000'
'000000000000000000000000000000000000000000000000000000'
'00000000000000000000000000001',
overflowok=True)
testcommon("%f", (1.0,), "1.000000")
# these are trying to test the limits of the internal magic-number-length
# formatting buffer, if that number changes then these tests are less
# effective
testcommon("%#.*g", (109, -1.e+49/3.))
testcommon("%#.*g", (110, -1.e+49/3.))
testcommon("%#.*g", (110, -1.e+100/3.))
# test some ridiculously large precision, expect overflow
testcommon('%12.*f', (123456, 1.0))
# check for internal overflow validation on length of precision
# these tests should no longer cause overflow in Python
# 2.7/3.1 and later.
testcommon("%#.*g", (110, -1.e+100/3.))
testcommon("%#.*G", (110, -1.e+100/3.))
testcommon("%#.*f", (110, -1.e+100/3.))
testcommon("%#.*F", (110, -1.e+100/3.))
# Formatting of integers. Overflow is not ok
testcommon("%x", 10, "a")
testcommon("%x", 100000000000, "174876e800")
testcommon("%o", 10, "12")
testcommon("%o", 100000000000, "1351035564000")
testcommon("%d", 10, "10")
testcommon("%d", 100000000000, "100000000000")
big = 123456789012345678901234567890
testcommon("%d", big, "123456789012345678901234567890")
testcommon("%d", -big, "-123456789012345678901234567890")
testcommon("%5d", -big, "-123456789012345678901234567890")
testcommon("%31d", -big, "-123456789012345678901234567890")
testcommon("%32d", -big, " -123456789012345678901234567890")
testcommon("%-32d", -big, "-123456789012345678901234567890 ")
testcommon("%032d", -big, "-0123456789012345678901234567890")
testcommon("%-032d", -big, "-123456789012345678901234567890 ")
testcommon("%034d", -big, "-000123456789012345678901234567890")
testcommon("%034d", big, "0000123456789012345678901234567890")
testcommon("%0+34d", big, "+000123456789012345678901234567890")
testcommon("%+34d", big, " +123456789012345678901234567890")
testcommon("%34d", big, " 123456789012345678901234567890")
testcommon("%.2d", big, "123456789012345678901234567890")
testcommon("%.30d", big, "123456789012345678901234567890")
testcommon("%.31d", big, "0123456789012345678901234567890")
testcommon("%32.31d", big, " 0123456789012345678901234567890")
testcommon("%d", float(big), "123456________________________", 6)
big = 0x1234567890abcdef12345 # 21 hex digits
testcommon("%x", big, "1234567890abcdef12345")
testcommon("%x", -big, "-1234567890abcdef12345")
testcommon("%5x", -big, "-1234567890abcdef12345")
testcommon("%22x", -big, "-1234567890abcdef12345")
testcommon("%23x", -big, " -1234567890abcdef12345")
testcommon("%-23x", -big, "-1234567890abcdef12345 ")
testcommon("%023x", -big, "-01234567890abcdef12345")
testcommon("%-023x", -big, "-1234567890abcdef12345 ")
testcommon("%025x", -big, "-0001234567890abcdef12345")
testcommon("%025x", big, "00001234567890abcdef12345")
testcommon("%0+25x", big, "+0001234567890abcdef12345")
testcommon("%+25x", big, " +1234567890abcdef12345")
testcommon("%25x", big, " 1234567890abcdef12345")
testcommon("%.2x", big, "1234567890abcdef12345")
testcommon("%.21x", big, "1234567890abcdef12345")
testcommon("%.22x", big, "01234567890abcdef12345")
testcommon("%23.22x", big, " 01234567890abcdef12345")
testcommon("%-23.22x", big, "01234567890abcdef12345 ")
testcommon("%X", big, "1234567890ABCDEF12345")
testcommon("%#X", big, "0X1234567890ABCDEF12345")
testcommon("%#x", big, "0x1234567890abcdef12345")
testcommon("%#x", -big, "-0x1234567890abcdef12345")
testcommon("%#27x", big, " 0x1234567890abcdef12345")
testcommon("%#-27x", big, "0x1234567890abcdef12345 ")
testcommon("%#027x", big, "0x00001234567890abcdef12345")
testcommon("%#.23x", big, "0x001234567890abcdef12345")
testcommon("%#.23x", -big, "-0x001234567890abcdef12345")
testcommon("%#27.23x", big, " 0x001234567890abcdef12345")
testcommon("%#-27.23x", big, "0x001234567890abcdef12345 ")
testcommon("%#027.23x", big, "0x00001234567890abcdef12345")
testcommon("%#+.23x", big, "+0x001234567890abcdef12345")
testcommon("%# .23x", big, " 0x001234567890abcdef12345")
testcommon("%#+.23X", big, "+0X001234567890ABCDEF12345")
# next one gets two leading zeroes from precision, and another from the
# 0 flag and the width
testcommon("%#+027.23X", big, "+0X0001234567890ABCDEF12345")
testcommon("%# 027.23X", big, " 0X0001234567890ABCDEF12345")
# same, except no 0 flag
testcommon("%#+27.23X", big, " +0X001234567890ABCDEF12345")
testcommon("%#-+27.23x", big, "+0x001234567890abcdef12345 ")
testcommon("%#- 27.23x", big, " 0x001234567890abcdef12345 ")
big = 0o12345670123456701234567012345670 # 32 octal digits
testcommon("%o", big, "12345670123456701234567012345670")
testcommon("%o", -big, "-12345670123456701234567012345670")
testcommon("%5o", -big, "-12345670123456701234567012345670")
testcommon("%33o", -big, "-12345670123456701234567012345670")
testcommon("%34o", -big, " -12345670123456701234567012345670")
testcommon("%-34o", -big, "-12345670123456701234567012345670 ")
testcommon("%034o", -big, "-012345670123456701234567012345670")
testcommon("%-034o", -big, "-12345670123456701234567012345670 ")
testcommon("%036o", -big, "-00012345670123456701234567012345670")
testcommon("%036o", big, "000012345670123456701234567012345670")
testcommon("%0+36o", big, "+00012345670123456701234567012345670")
testcommon("%+36o", big, " +12345670123456701234567012345670")
testcommon("%36o", big, " 12345670123456701234567012345670")
testcommon("%.2o", big, "12345670123456701234567012345670")
testcommon("%.32o", big, "12345670123456701234567012345670")
testcommon("%.33o", big, "012345670123456701234567012345670")
testcommon("%34.33o", big, " 012345670123456701234567012345670")
testcommon("%-34.33o", big, "012345670123456701234567012345670 ")
testcommon("%o", big, "12345670123456701234567012345670")
testcommon("%#o", big, "0o12345670123456701234567012345670")
testcommon("%#o", -big, "-0o12345670123456701234567012345670")
testcommon("%#38o", big, " 0o12345670123456701234567012345670")
testcommon("%#-38o", big, "0o12345670123456701234567012345670 ")
testcommon("%#038o", big, "0o000012345670123456701234567012345670")
testcommon("%#.34o", big, "0o0012345670123456701234567012345670")
testcommon("%#.34o", -big, "-0o0012345670123456701234567012345670")
testcommon("%#38.34o", big, " 0o0012345670123456701234567012345670")
testcommon("%#-38.34o", big, "0o0012345670123456701234567012345670 ")
testcommon("%#038.34o", big, "0o000012345670123456701234567012345670")
testcommon("%#+.34o", big, "+0o0012345670123456701234567012345670")
testcommon("%# .34o", big, " 0o0012345670123456701234567012345670")
testcommon("%#+38.34o", big, " +0o0012345670123456701234567012345670")
testcommon("%#-+38.34o", big, "+0o0012345670123456701234567012345670 ")
testcommon("%#- 38.34o", big, " 0o0012345670123456701234567012345670 ")
testcommon("%#+038.34o", big, "+0o00012345670123456701234567012345670")
testcommon("%# 038.34o", big, " 0o00012345670123456701234567012345670")
# next one gets one leading zero from precision
testcommon("%.33o", big, "012345670123456701234567012345670")
# base marker added in spite of leading zero (different to Python 2)
testcommon("%#.33o", big, "0o012345670123456701234567012345670")
# reduce precision, and base marker is always added
testcommon("%#.32o", big, "0o12345670123456701234567012345670")
# one leading zero from precision, plus two from "0" flag & width
testcommon("%035.33o", big, "00012345670123456701234567012345670")
# base marker shouldn't change the size
testcommon("%0#35.33o", big, "0o012345670123456701234567012345670")
# Some small ints, in both Python int and flavors).
testcommon("%d", 42, "42")
testcommon("%d", -42, "-42")
testcommon("%d", 42.0, "42")
testcommon("%#x", 1, "0x1")
testcommon("%#X", 1, "0X1")
testcommon("%#o", 1, "0o1")
testcommon("%#o", 0, "0o0")
testcommon("%o", 0, "0")
testcommon("%d", 0, "0")
testcommon("%#x", 0, "0x0")
testcommon("%#X", 0, "0X0")
testcommon("%x", 0x42, "42")
testcommon("%x", -0x42, "-42")
testcommon("%o", 0o42, "42")
testcommon("%o", -0o42, "-42")
# alternate float formatting
testcommon('%g', 1.1, '1.1')
testcommon('%#g', 1.1, '1.10000')
if verbose:
print('Testing exceptions')
test_exc_common('%', (), ValueError, "incomplete format")
test_exc_common('% %s', 1, ValueError,
"unsupported format character '%' (0x25) at index 2")
test_exc_common('%d', '1', TypeError,
"%d format: a number is required, not str")
test_exc_common('%d', b'1', TypeError,
"%d format: a number is required, not bytes")
test_exc_common('%x', '1', TypeError,
"%x format: an integer is required, not str")
test_exc_common('%x', 3.14, TypeError,
"%x format: an integer is required, not float")
def test_str_format(self):
testformat("%r", "\u0378", "'\\u0378'") # non printable
testformat("%a", "\u0378", "'\\u0378'") # non printable
testformat("%r", "\u0374", "'\u0374'") # printable
testformat("%a", "\u0374", "'\\u0374'") # printable
# Test exception for unknown format characters, etc.
if verbose:
print('Testing exceptions')
test_exc('abc %b', 1, ValueError,
"unsupported format character 'b' (0x62) at index 5")
#test_exc(unicode('abc %\u3000','raw-unicode-escape'), 1, ValueError,
# "unsupported format character '?' (0x3000) at index 5")
test_exc('%g', '1', TypeError, "must be real number, not str")
test_exc('no format', '1', TypeError,
"not all arguments converted during string formatting")
test_exc('%c', -1, OverflowError, "%c arg not in range(0x110000)")
test_exc('%c', sys.maxunicode+1, OverflowError,
"%c arg not in range(0x110000)")
#test_exc('%c', 2**128, OverflowError, "%c arg not in range(0x110000)")
test_exc('%c', 3.14, TypeError, "%c requires int or char")
test_exc('%c', 'ab', TypeError, "%c requires int or char")
test_exc('%c', b'x', TypeError, "%c requires int or char")
if maxsize == 2**31-1:
# crashes 2.2.1 and earlier:
try:
"%*d"%(maxsize, -127)
except MemoryError:
pass
else:
raise TestFailed('"%*d"%(maxsize, -127) should fail')
def test_bytes_and_bytearray_format(self):
# %c will insert a single byte, either from an int in range(256), or
# from a bytes argument of length 1, not from a str.
testcommon(b"%c", 7, b"\x07")
testcommon(b"%c", b"Z", b"Z")
testcommon(b"%c", bytearray(b"Z"), b"Z")
testcommon(b"%5c", 65, b" A")
testcommon(b"%-5c", 65, b"A ")
# %b will insert a series of bytes, either from a type that supports
# the Py_buffer protocol, or something that has a __bytes__ method
class FakeBytes(object):
def __bytes__(self):
return b'123'
fb = FakeBytes()
testcommon(b"%b", b"abc", b"abc")
testcommon(b"%b", bytearray(b"def"), b"def")
testcommon(b"%b", fb, b"123")
testcommon(b"%b", memoryview(b"abc"), b"abc")
# # %s is an alias for %b -- should only be used for Py2/3 code
testcommon(b"%s", b"abc", b"abc")
testcommon(b"%s", bytearray(b"def"), b"def")
testcommon(b"%s", fb, b"123")
testcommon(b"%s", memoryview(b"abc"), b"abc")
# %a will give the equivalent of
# repr(some_obj).encode('ascii', 'backslashreplace')
testcommon(b"%a", 3.14, b"3.14")
testcommon(b"%a", b"ghi", b"b'ghi'")
testcommon(b"%a", "jkl", b"'jkl'")
testcommon(b"%a", "\u0544", b"'\\u0544'")
# %r is an alias for %a
testcommon(b"%r", 3.14, b"3.14")
testcommon(b"%r", b"ghi", b"b'ghi'")
testcommon(b"%r", "jkl", b"'jkl'")
testcommon(b"%r", "\u0544", b"'\\u0544'")
# Test exception for unknown format characters, etc.
if verbose:
print('Testing exceptions')
test_exc(b'%g', '1', TypeError, "float argument required, not str")
test_exc(b'%g', b'1', TypeError, "float argument required, not bytes")
test_exc(b'no format', 7, TypeError,
"not all arguments converted during bytes formatting")
test_exc(b'no format', b'1', TypeError,
"not all arguments converted during bytes formatting")
test_exc(b'no format', bytearray(b'1'), TypeError,
"not all arguments converted during bytes formatting")
test_exc(b"%c", -1, OverflowError,
"%c arg not in range(256)")
test_exc(b"%c", 256, OverflowError,
"%c arg not in range(256)")
test_exc(b"%c", 2**128, OverflowError,
"%c arg not in range(256)")
test_exc(b"%c", b"Za", TypeError,
"%c requires an integer in range(256) or a single byte")
test_exc(b"%c", "Y", TypeError,
"%c requires an integer in range(256) or a single byte")
test_exc(b"%c", 3.14, TypeError,
"%c requires an integer in range(256) or a single byte")
test_exc(b"%b", "Xc", TypeError,
"%b requires a bytes-like object, "
"or an object that implements __bytes__, not 'str'")
test_exc(b"%s", "Wd", TypeError,
"%b requires a bytes-like object, "
"or an object that implements __bytes__, not 'str'")
if maxsize == 2**31-1:
# crashes 2.2.1 and earlier:
try:
"%*d"%(maxsize, -127)
except MemoryError:
pass
else:
raise TestFailed('"%*d"%(maxsize, -127) should fail')
def test_nul(self):
# test the null character
testcommon("a\0b", (), 'a\0b')
testcommon("a%cb", (0,), 'a\0b')
testformat("a%sb", ('c\0d',), 'ac\0db')
testcommon(b"a%sb", (b'c\0d',), b'ac\0db')
def test_non_ascii(self):
testformat("\u20ac=%f", (1.0,), "\u20ac=1.000000")
self.assertEqual(format("abc", "\u2007<5"), "abc\u2007\u2007")
self.assertEqual(format(123, "\u2007<5"), "123\u2007\u2007")
self.assertEqual(format(12.3, "\u2007<6"), "12.3\u2007\u2007")
self.assertEqual(format(0j, "\u2007<4"), "0j\u2007\u2007")
self.assertEqual(format(1+2j, "\u2007<8"), "(1+2j)\u2007\u2007")
self.assertEqual(format("abc", "\u2007>5"), "\u2007\u2007abc")
self.assertEqual(format(123, "\u2007>5"), "\u2007\u2007123")
self.assertEqual(format(12.3, "\u2007>6"), "\u2007\u200712.3")
self.assertEqual(format(1+2j, "\u2007>8"), "\u2007\u2007(1+2j)")
self.assertEqual(format(0j, "\u2007>4"), "\u2007\u20070j")
self.assertEqual(format("abc", "\u2007^5"), "\u2007abc\u2007")
self.assertEqual(format(123, "\u2007^5"), "\u2007123\u2007")
self.assertEqual(format(12.3, "\u2007^6"), "\u200712.3\u2007")
self.assertEqual(format(1+2j, "\u2007^8"), "\u2007(1+2j)\u2007")
self.assertEqual(format(0j, "\u2007^4"), "\u20070j\u2007")
def test_locale(self):
try:
oldloc = locale.setlocale(locale.LC_ALL)
locale.setlocale(locale.LC_ALL, '')
except locale.Error as err:
self.skipTest("Cannot set locale: {}".format(err))
try:
localeconv = locale.localeconv()
sep = localeconv['thousands_sep']
point = localeconv['decimal_point']
text = format(123456789, "n")
self.assertIn(sep, text)
self.assertEqual(text.replace(sep, ''), '123456789')
text = format(1234.5, "n")
self.assertIn(sep, text)
self.assertIn(point, text)
self.assertEqual(text.replace(sep, ''), '1234' + point + '5')
finally:
locale.setlocale(locale.LC_ALL, oldloc)
@support.cpython_only
def test_optimisations(self):
text = "abcde" # 5 characters
self.assertIs("%s" % text, text)
self.assertIs("%.5s" % text, text)
self.assertIs("%.10s" % text, text)
self.assertIs("%1s" % text, text)
self.assertIs("%5s" % text, text)
self.assertIs("{0}".format(text), text)
self.assertIs("{0:s}".format(text), text)
self.assertIs("{0:.5s}".format(text), text)
self.assertIs("{0:.10s}".format(text), text)
self.assertIs("{0:1s}".format(text), text)
self.assertIs("{0:5s}".format(text), text)
self.assertIs(text % (), text)
self.assertIs(text.format(), text)
def test_precision(self):
f = 1.2
self.assertEqual(format(f, ".0f"), "1")
self.assertEqual(format(f, ".3f"), "1.200")
with self.assertRaises(ValueError) as cm:
format(f, ".%sf" % (sys.maxsize + 1))
c = complex(f)
self.assertEqual(format(c, ".0f"), "1+0j")
self.assertEqual(format(c, ".3f"), "1.200+0.000j")
with self.assertRaises(ValueError) as cm:
format(c, ".%sf" % (sys.maxsize + 1))
@support.cpython_only
def test_precision_c_limits(self):
from _testcapi import INT_MAX
f = 1.2
with self.assertRaises(ValueError) as cm:
format(f, ".%sf" % (INT_MAX + 1))
c = complex(f)
with self.assertRaises(ValueError) as cm:
format(c, ".%sf" % (INT_MAX + 1))
def test_g_format_has_no_trailing_zeros(self):
# regression test for bugs.python.org/issue40780
self.assertEqual("%.3g" % 1505.0, "1.5e+03")
self.assertEqual("%#.3g" % 1505.0, "1.50e+03")
self.assertEqual(format(1505.0, ".3g"), "1.5e+03")
self.assertEqual(format(1505.0, "#.3g"), "1.50e+03")
self.assertEqual(format(12300050.0, ".6g"), "1.23e+07")
self.assertEqual(format(12300050.0, "#.6g"), "1.23000e+07")
if __name__ == "__main__":
unittest.main()