diff options
Diffstat (limited to 'Lib/test/test_time.py')
-rw-r--r-- | Lib/test/test_time.py | 598 |
1 files changed, 281 insertions, 317 deletions
diff --git a/Lib/test/test_time.py b/Lib/test/test_time.py index 76b894e..f224212 100644 --- a/Lib/test/test_time.py +++ b/Lib/test/test_time.py @@ -1,6 +1,8 @@ from test import support +import decimal import enum import locale +import math import platform import sys import sysconfig @@ -21,17 +23,27 @@ SIZEOF_INT = sysconfig.get_config_var('SIZEOF_INT') or 4 TIME_MAXYEAR = (1 << 8 * SIZEOF_INT - 1) - 1 TIME_MINYEAR = -TIME_MAXYEAR - 1 +SEC_TO_US = 10 ** 6 US_TO_NS = 10 ** 3 MS_TO_NS = 10 ** 6 SEC_TO_NS = 10 ** 9 +NS_TO_SEC = 10 ** 9 class _PyTime(enum.IntEnum): # Round towards minus infinity (-inf) ROUND_FLOOR = 0 # Round towards infinity (+inf) ROUND_CEILING = 1 + # Round to nearest with ties going to nearest even integer + ROUND_HALF_EVEN = 2 -ALL_ROUNDING_METHODS = (_PyTime.ROUND_FLOOR, _PyTime.ROUND_CEILING) +# Rounding modes supported by PyTime +ROUNDING_MODES = ( + # (PyTime rounding method, decimal rounding method) + (_PyTime.ROUND_FLOOR, decimal.ROUND_FLOOR), + (_PyTime.ROUND_CEILING, decimal.ROUND_CEILING), + (_PyTime.ROUND_HALF_EVEN, decimal.ROUND_HALF_EVEN), +) class TimeTestCase(unittest.TestCase): @@ -607,79 +619,6 @@ class TestStrftime4dyear(_TestStrftimeYear, _Test4dYear, unittest.TestCase): class TestPytime(unittest.TestCase): - def setUp(self): - self.invalid_values = ( - -(2 ** 100), 2 ** 100, - -(2.0 ** 100.0), 2.0 ** 100.0, - ) - - @support.cpython_only - def test_time_t(self): - from _testcapi import pytime_object_to_time_t - for obj, time_t, rnd in ( - # Round towards minus infinity (-inf) - (0, 0, _PyTime.ROUND_FLOOR), - (-1, -1, _PyTime.ROUND_FLOOR), - (-1.0, -1, _PyTime.ROUND_FLOOR), - (-1.9, -2, _PyTime.ROUND_FLOOR), - (1.0, 1, _PyTime.ROUND_FLOOR), - (1.9, 1, _PyTime.ROUND_FLOOR), - # Round towards infinity (+inf) - (0, 0, _PyTime.ROUND_CEILING), - (-1, -1, _PyTime.ROUND_CEILING), - (-1.0, -1, _PyTime.ROUND_CEILING), - (-1.9, -1, _PyTime.ROUND_CEILING), - (1.0, 1, _PyTime.ROUND_CEILING), - (1.9, 2, _PyTime.ROUND_CEILING), - ): - self.assertEqual(pytime_object_to_time_t(obj, rnd), time_t) - - rnd = _PyTime.ROUND_FLOOR - for invalid in self.invalid_values: - self.assertRaises(OverflowError, - pytime_object_to_time_t, invalid, rnd) - - @support.cpython_only - def test_timespec(self): - from _testcapi import pytime_object_to_timespec - for obj, timespec, rnd in ( - # Round towards minus infinity (-inf) - (0, (0, 0), _PyTime.ROUND_FLOOR), - (-1, (-1, 0), _PyTime.ROUND_FLOOR), - (-1.0, (-1, 0), _PyTime.ROUND_FLOOR), - (1e-9, (0, 1), _PyTime.ROUND_FLOOR), - (1e-10, (0, 0), _PyTime.ROUND_FLOOR), - (-1e-9, (-1, 999999999), _PyTime.ROUND_FLOOR), - (-1e-10, (-1, 999999999), _PyTime.ROUND_FLOOR), - (-1.2, (-2, 800000000), _PyTime.ROUND_FLOOR), - (0.9999999999, (0, 999999999), _PyTime.ROUND_FLOOR), - (1.1234567890, (1, 123456789), _PyTime.ROUND_FLOOR), - (1.1234567899, (1, 123456789), _PyTime.ROUND_FLOOR), - (-1.1234567890, (-2, 876543211), _PyTime.ROUND_FLOOR), - (-1.1234567891, (-2, 876543210), _PyTime.ROUND_FLOOR), - # Round towards infinity (+inf) - (0, (0, 0), _PyTime.ROUND_CEILING), - (-1, (-1, 0), _PyTime.ROUND_CEILING), - (-1.0, (-1, 0), _PyTime.ROUND_CEILING), - (1e-9, (0, 1), _PyTime.ROUND_CEILING), - (1e-10, (0, 1), _PyTime.ROUND_CEILING), - (-1e-9, (-1, 999999999), _PyTime.ROUND_CEILING), - (-1e-10, (0, 0), _PyTime.ROUND_CEILING), - (-1.2, (-2, 800000000), _PyTime.ROUND_CEILING), - (0.9999999999, (1, 0), _PyTime.ROUND_CEILING), - (1.1234567890, (1, 123456790), _PyTime.ROUND_CEILING), - (1.1234567899, (1, 123456790), _PyTime.ROUND_CEILING), - (-1.1234567890, (-2, 876543211), _PyTime.ROUND_CEILING), - (-1.1234567891, (-2, 876543211), _PyTime.ROUND_CEILING), - ): - with self.subTest(obj=obj, round=rnd, timespec=timespec): - self.assertEqual(pytime_object_to_timespec(obj, rnd), timespec) - - rnd = _PyTime.ROUND_FLOOR - for invalid in self.invalid_values: - self.assertRaises(OverflowError, - pytime_object_to_timespec, invalid, rnd) - @unittest.skipUnless(time._STRUCT_TM_ITEMS == 11, "needs tm_zone support") def test_localtime_timezone(self): @@ -734,266 +673,291 @@ class TestPytime(unittest.TestCase): self.assertIs(lt.tm_zone, None) -@unittest.skipUnless(_testcapi is not None, - 'need the _testcapi module') -class TestPyTime_t(unittest.TestCase): +@unittest.skipIf(_testcapi is None, 'need the _testcapi module') +class CPyTimeTestCase: + """ + Base class to test the C _PyTime_t API. + """ + OVERFLOW_SECONDS = None + + def setUp(self): + from _testcapi import SIZEOF_TIME_T + bits = SIZEOF_TIME_T * 8 - 1 + self.time_t_min = -2 ** bits + self.time_t_max = 2 ** bits - 1 + + def time_t_filter(self, seconds): + return (self.time_t_min <= seconds <= self.time_t_max) + + def _rounding_values(self, use_float): + "Build timestamps used to test rounding." + + units = [1, US_TO_NS, MS_TO_NS, SEC_TO_NS] + if use_float: + # picoseconds are only tested to pytime_converter accepting floats + units.append(1e-3) + + values = ( + # small values + 1, 2, 5, 7, 123, 456, 1234, + # 10^k - 1 + 9, + 99, + 999, + 9999, + 99999, + 999999, + # test half even rounding near 0.5, 1.5, 2.5, 3.5, 4.5 + 499, 500, 501, + 1499, 1500, 1501, + 2500, + 3500, + 4500, + ) + + ns_timestamps = [0] + for unit in units: + for value in values: + ns = value * unit + ns_timestamps.extend((-ns, ns)) + for pow2 in (0, 5, 10, 15, 22, 23, 24, 30, 33): + ns = (2 ** pow2) * SEC_TO_NS + ns_timestamps.extend(( + -ns-1, -ns, -ns+1, + ns-1, ns, ns+1 + )) + for seconds in (_testcapi.INT_MIN, _testcapi.INT_MAX): + ns_timestamps.append(seconds * SEC_TO_NS) + if use_float: + # numbers with an exact representation in IEEE 754 (base 2) + for pow2 in (3, 7, 10, 15): + ns = 2.0 ** (-pow2) + ns_timestamps.extend((-ns, ns)) + + # seconds close to _PyTime_t type limit + ns = (2 ** 63 // SEC_TO_NS) * SEC_TO_NS + ns_timestamps.extend((-ns, ns)) + + return ns_timestamps + + def _check_rounding(self, pytime_converter, expected_func, + use_float, unit_to_sec, value_filter=None): + + def convert_values(ns_timestamps): + if use_float: + unit_to_ns = SEC_TO_NS / float(unit_to_sec) + values = [ns / unit_to_ns for ns in ns_timestamps] + else: + unit_to_ns = SEC_TO_NS // unit_to_sec + values = [ns // unit_to_ns for ns in ns_timestamps] + + if value_filter: + values = filter(value_filter, values) + + # remove duplicates and sort + return sorted(set(values)) + + # test rounding + ns_timestamps = self._rounding_values(use_float) + valid_values = convert_values(ns_timestamps) + for time_rnd, decimal_rnd in ROUNDING_MODES : + context = decimal.getcontext() + context.rounding = decimal_rnd + + for value in valid_values: + debug_info = {'value': value, 'rounding': decimal_rnd} + try: + result = pytime_converter(value, time_rnd) + expected = expected_func(value) + except Exception as exc: + self.fail("Error on timestamp conversion: %s" % debug_info) + self.assertEqual(result, + expected, + debug_info) + + # test overflow + ns = self.OVERFLOW_SECONDS * SEC_TO_NS + ns_timestamps = (-ns, ns) + overflow_values = convert_values(ns_timestamps) + for time_rnd, _ in ROUNDING_MODES : + for value in overflow_values: + debug_info = {'value': value, 'rounding': time_rnd} + with self.assertRaises(OverflowError, msg=debug_info): + pytime_converter(value, time_rnd) + + def check_int_rounding(self, pytime_converter, expected_func, + unit_to_sec=1, value_filter=None): + self._check_rounding(pytime_converter, expected_func, + False, unit_to_sec, value_filter) + + def check_float_rounding(self, pytime_converter, expected_func, + unit_to_sec=1, value_filter=None): + self._check_rounding(pytime_converter, expected_func, + True, unit_to_sec, value_filter) + + def decimal_round(self, x): + d = decimal.Decimal(x) + d = d.quantize(1) + return int(d) + + +class TestCPyTime(CPyTimeTestCase, unittest.TestCase): + """ + Test the C _PyTime_t API. + """ + # _PyTime_t is a 64-bit signed integer + OVERFLOW_SECONDS = math.ceil((2**63 + 1) / SEC_TO_NS) + def test_FromSeconds(self): from _testcapi import PyTime_FromSeconds - for seconds in (0, 3, -456, _testcapi.INT_MAX, _testcapi.INT_MIN): - with self.subTest(seconds=seconds): - self.assertEqual(PyTime_FromSeconds(seconds), - seconds * SEC_TO_NS) + + # PyTime_FromSeconds() expects a C int, reject values out of range + def c_int_filter(secs): + return (_testcapi.INT_MIN <= secs <= _testcapi.INT_MAX) + + self.check_int_rounding(lambda secs, rnd: PyTime_FromSeconds(secs), + lambda secs: secs * SEC_TO_NS, + value_filter=c_int_filter) def test_FromSecondsObject(self): from _testcapi import PyTime_FromSecondsObject - # Conversion giving the same result for all rounding methods - for rnd in ALL_ROUNDING_METHODS: - for obj, ts in ( - # integers - (0, 0), - (1, SEC_TO_NS), - (-3, -3 * SEC_TO_NS), - - # float: subseconds - (0.0, 0), - (1e-9, 1), - (1e-6, 10 ** 3), - (1e-3, 10 ** 6), - - # float: seconds - (2.0, 2 * SEC_TO_NS), - (123.0, 123 * SEC_TO_NS), - (-7.0, -7 * SEC_TO_NS), - - # nanosecond are kept for value <= 2^23 seconds - (2**22 - 1e-9, 4194303999999999), - (2**22, 4194304000000000), - (2**22 + 1e-9, 4194304000000001), - (2**23 - 1e-9, 8388607999999999), - (2**23, 8388608000000000), - - # start losing precision for value > 2^23 seconds - (2**23 + 1e-9, 8388608000000002), - - # nanoseconds are lost for value > 2^23 seconds - (2**24 - 1e-9, 16777215999999998), - (2**24, 16777216000000000), - (2**24 + 1e-9, 16777216000000000), - (2**25 - 1e-9, 33554432000000000), - (2**25 , 33554432000000000), - (2**25 + 1e-9, 33554432000000000), - - # close to 2^63 nanoseconds (_PyTime_t limit) - (9223372036, 9223372036 * SEC_TO_NS), - (9223372036.0, 9223372036 * SEC_TO_NS), - (-9223372036, -9223372036 * SEC_TO_NS), - (-9223372036.0, -9223372036 * SEC_TO_NS), - ): - with self.subTest(obj=obj, round=rnd, timestamp=ts): - self.assertEqual(PyTime_FromSecondsObject(obj, rnd), ts) - - with self.subTest(round=rnd): - with self.assertRaises(OverflowError): - PyTime_FromSecondsObject(9223372037, rnd) - PyTime_FromSecondsObject(9223372037.0, rnd) - PyTime_FromSecondsObject(-9223372037, rnd) - PyTime_FromSecondsObject(-9223372037.0, rnd) - - # Conversion giving different results depending on the rounding method - FLOOR = _PyTime.ROUND_FLOOR - CEILING = _PyTime.ROUND_CEILING - for obj, ts, rnd in ( - # close to zero - ( 1e-10, 0, FLOOR), - ( 1e-10, 1, CEILING), - (-1e-10, -1, FLOOR), - (-1e-10, 0, CEILING), - - # test rounding of the last nanosecond - ( 1.1234567899, 1123456789, FLOOR), - ( 1.1234567899, 1123456790, CEILING), - (-1.1234567899, -1123456790, FLOOR), - (-1.1234567899, -1123456789, CEILING), - - # close to 1 second - ( 0.9999999999, 999999999, FLOOR), - ( 0.9999999999, 1000000000, CEILING), - (-0.9999999999, -1000000000, FLOOR), - (-0.9999999999, -999999999, CEILING), - ): - with self.subTest(obj=obj, round=rnd, timestamp=ts): - self.assertEqual(PyTime_FromSecondsObject(obj, rnd), ts) + self.check_int_rounding( + PyTime_FromSecondsObject, + lambda secs: secs * SEC_TO_NS) + + self.check_float_rounding( + PyTime_FromSecondsObject, + lambda ns: self.decimal_round(ns * SEC_TO_NS)) def test_AsSecondsDouble(self): from _testcapi import PyTime_AsSecondsDouble - for nanoseconds, seconds in ( - # near 1 nanosecond - ( 0, 0.0), - ( 1, 1e-9), - (-1, -1e-9), - - # near 1 second - (SEC_TO_NS + 1, 1.0 + 1e-9), - (SEC_TO_NS, 1.0), - (SEC_TO_NS - 1, 1.0 - 1e-9), - - # a few seconds - (123 * SEC_TO_NS, 123.0), - (-567 * SEC_TO_NS, -567.0), - - # nanosecond are kept for value <= 2^23 seconds - (4194303999999999, 2**22 - 1e-9), - (4194304000000000, 2**22), - (4194304000000001, 2**22 + 1e-9), - - # start losing precision for value > 2^23 seconds - (8388608000000002, 2**23 + 1e-9), - - # nanoseconds are lost for value > 2^23 seconds - (16777215999999998, 2**24 - 1e-9), - (16777215999999999, 2**24 - 1e-9), - (16777216000000000, 2**24 ), - (16777216000000001, 2**24 ), - (16777216000000002, 2**24 + 2e-9), - - (33554432000000000, 2**25 ), - (33554432000000002, 2**25 ), - (33554432000000004, 2**25 + 4e-9), - - # close to 2^63 nanoseconds (_PyTime_t limit) - (9223372036 * SEC_TO_NS, 9223372036.0), - (-9223372036 * SEC_TO_NS, -9223372036.0), - ): - with self.subTest(nanoseconds=nanoseconds, seconds=seconds): - self.assertEqual(PyTime_AsSecondsDouble(nanoseconds), - seconds) - - def test_timeval(self): + def float_converter(ns): + if abs(ns) % SEC_TO_NS == 0: + return float(ns // SEC_TO_NS) + else: + return float(ns) / SEC_TO_NS + + self.check_int_rounding(lambda ns, rnd: PyTime_AsSecondsDouble(ns), + float_converter, + NS_TO_SEC) + + def create_decimal_converter(self, denominator): + denom = decimal.Decimal(denominator) + + def converter(value): + d = decimal.Decimal(value) / denom + return self.decimal_round(d) + + return converter + + def test_AsTimeval(self): from _testcapi import PyTime_AsTimeval - for rnd in ALL_ROUNDING_METHODS: - for ns, tv in ( - # microseconds - (0, (0, 0)), - (1000, (0, 1)), - (-1000, (-1, 999999)), - - # seconds - (2 * SEC_TO_NS, (2, 0)), - (-3 * SEC_TO_NS, (-3, 0)), - ): - with self.subTest(nanoseconds=ns, timeval=tv, round=rnd): - self.assertEqual(PyTime_AsTimeval(ns, rnd), tv) - - FLOOR = _PyTime.ROUND_FLOOR - CEILING = _PyTime.ROUND_CEILING - for ns, tv, rnd in ( - # nanoseconds - (1, (0, 0), FLOOR), - (1, (0, 1), CEILING), - (-1, (-1, 999999), FLOOR), - (-1, (0, 0), CEILING), - - # seconds + nanoseconds - (1234567001, (1, 234567), FLOOR), - (1234567001, (1, 234568), CEILING), - (-1234567001, (-2, 765432), FLOOR), - (-1234567001, (-2, 765433), CEILING), - ): - with self.subTest(nanoseconds=ns, timeval=tv, round=rnd): - self.assertEqual(PyTime_AsTimeval(ns, rnd), tv) + + us_converter = self.create_decimal_converter(US_TO_NS) + + def timeval_converter(ns): + us = us_converter(ns) + return divmod(us, SEC_TO_US) + + if sys.platform == 'win32': + from _testcapi import LONG_MIN, LONG_MAX + + # On Windows, timeval.tv_sec type is a C long + def seconds_filter(secs): + return LONG_MIN <= secs <= LONG_MAX + else: + seconds_filter = self.time_t_filter + + self.check_int_rounding(PyTime_AsTimeval, + timeval_converter, + NS_TO_SEC, + value_filter=seconds_filter) @unittest.skipUnless(hasattr(_testcapi, 'PyTime_AsTimespec'), 'need _testcapi.PyTime_AsTimespec') - def test_timespec(self): + def test_AsTimespec(self): from _testcapi import PyTime_AsTimespec - for ns, ts in ( - # nanoseconds - (0, (0, 0)), - (1, (0, 1)), - (-1, (-1, 999999999)), - - # seconds - (2 * SEC_TO_NS, (2, 0)), - (-3 * SEC_TO_NS, (-3, 0)), - - # seconds + nanoseconds - (1234567890, (1, 234567890)), - (-1234567890, (-2, 765432110)), - ): - with self.subTest(nanoseconds=ns, timespec=ts): - self.assertEqual(PyTime_AsTimespec(ns), ts) - - def test_milliseconds(self): + + def timespec_converter(ns): + return divmod(ns, SEC_TO_NS) + + self.check_int_rounding(lambda ns, rnd: PyTime_AsTimespec(ns), + timespec_converter, + NS_TO_SEC, + value_filter=self.time_t_filter) + + def test_AsMilliseconds(self): from _testcapi import PyTime_AsMilliseconds - for rnd in ALL_ROUNDING_METHODS: - for ns, tv in ( - # milliseconds - (1 * MS_TO_NS, 1), - (-2 * MS_TO_NS, -2), - - # seconds - (2 * SEC_TO_NS, 2000), - (-3 * SEC_TO_NS, -3000), - ): - with self.subTest(nanoseconds=ns, timeval=tv, round=rnd): - self.assertEqual(PyTime_AsMilliseconds(ns, rnd), tv) - - FLOOR = _PyTime.ROUND_FLOOR - CEILING = _PyTime.ROUND_CEILING - for ns, ms, rnd in ( - # nanoseconds - (1, 0, FLOOR), - (1, 1, CEILING), - (-1, -1, FLOOR), - (-1, 0, CEILING), - - # seconds + nanoseconds - (1234 * MS_TO_NS + 1, 1234, FLOOR), - (1234 * MS_TO_NS + 1, 1235, CEILING), - (-1234 * MS_TO_NS - 1, -1235, FLOOR), - (-1234 * MS_TO_NS - 1, -1234, CEILING), - ): - with self.subTest(nanoseconds=ns, milliseconds=ms, round=rnd): - self.assertEqual(PyTime_AsMilliseconds(ns, rnd), ms) - - def test_microseconds(self): + + self.check_int_rounding(PyTime_AsMilliseconds, + self.create_decimal_converter(MS_TO_NS), + NS_TO_SEC) + + def test_AsMicroseconds(self): from _testcapi import PyTime_AsMicroseconds - for rnd in ALL_ROUNDING_METHODS: - for ns, tv in ( - # microseconds - (1 * US_TO_NS, 1), - (-2 * US_TO_NS, -2), - - # milliseconds - (1 * MS_TO_NS, 1000), - (-2 * MS_TO_NS, -2000), - - # seconds - (2 * SEC_TO_NS, 2000000), - (-3 * SEC_TO_NS, -3000000), - ): - with self.subTest(nanoseconds=ns, timeval=tv, round=rnd): - self.assertEqual(PyTime_AsMicroseconds(ns, rnd), tv) - - FLOOR = _PyTime.ROUND_FLOOR - CEILING = _PyTime.ROUND_CEILING - for ns, ms, rnd in ( - # nanoseconds - (1, 0, FLOOR), - (1, 1, CEILING), - (-1, -1, FLOOR), - (-1, 0, CEILING), - - # seconds + nanoseconds - (1234 * US_TO_NS + 1, 1234, FLOOR), - (1234 * US_TO_NS + 1, 1235, CEILING), - (-1234 * US_TO_NS - 1, -1235, FLOOR), - (-1234 * US_TO_NS - 1, -1234, CEILING), - ): - with self.subTest(nanoseconds=ns, milliseconds=ms, round=rnd): - self.assertEqual(PyTime_AsMicroseconds(ns, rnd), ms) + + self.check_int_rounding(PyTime_AsMicroseconds, + self.create_decimal_converter(US_TO_NS), + NS_TO_SEC) + + +class TestOldPyTime(CPyTimeTestCase, unittest.TestCase): + """ + Test the old C _PyTime_t API: _PyTime_ObjectToXXX() functions. + """ + + # time_t is a 32-bit or 64-bit signed integer + OVERFLOW_SECONDS = 2 ** 64 + + def test_object_to_time_t(self): + from _testcapi import pytime_object_to_time_t + + self.check_int_rounding(pytime_object_to_time_t, + lambda secs: secs, + value_filter=self.time_t_filter) + + self.check_float_rounding(pytime_object_to_time_t, + self.decimal_round, + value_filter=self.time_t_filter) + + def create_converter(self, sec_to_unit): + def converter(secs): + floatpart, intpart = math.modf(secs) + intpart = int(intpart) + floatpart *= sec_to_unit + floatpart = self.decimal_round(floatpart) + if floatpart < 0: + floatpart += sec_to_unit + intpart -= 1 + elif floatpart >= sec_to_unit: + floatpart -= sec_to_unit + intpart += 1 + return (intpart, floatpart) + return converter + + def test_object_to_timeval(self): + from _testcapi import pytime_object_to_timeval + + self.check_int_rounding(pytime_object_to_timeval, + lambda secs: (secs, 0), + value_filter=self.time_t_filter) + + self.check_float_rounding(pytime_object_to_timeval, + self.create_converter(SEC_TO_US), + value_filter=self.time_t_filter) + + def test_object_to_timespec(self): + from _testcapi import pytime_object_to_timespec + + self.check_int_rounding(pytime_object_to_timespec, + lambda secs: (secs, 0), + value_filter=self.time_t_filter) + + self.check_float_rounding(pytime_object_to_timespec, + self.create_converter(SEC_TO_NS), + value_filter=self.time_t_filter) if __name__ == "__main__": |