Alexander Belopolsky <belopol...@users.sourceforge.net> added the comment:
On Sat, May 8, 2010 at 10:07 AM, Mark Dickinson <rep...@bugs.python.org> wrote:
..
> I definitely think this should go into 2.7 as well; Â total_seconds has only
> seen the light of day
> in the beta releases so far, so it's unlikely we'd break anyone's code with
> this change.
> Â (And IMO this is a bugfix, not a feature.)
Tracker web site seems to be down, so I am trying my luck attaching a
patch to an e-mail reply. It did work in the past.
The true division by 10**6 could probably be replaced with division by
1e6 in the tests and docs, but I am not sure if the two are always
equivalent.
----------
Added file: http://bugs.python.org/file17266/issue8644.diff
_______________________________________
Python tracker <rep...@bugs.python.org>
<http://bugs.python.org/issue8644>
_______________________________________
Index: Misc/NEWS
===================================================================
--- Misc/NEWS (revision 81011)
+++ Misc/NEWS (working copy)
@@ -70,6 +70,8 @@
inheritance with the underlying socket object.
The cheap inheritance has been deprecated.
+- Issue #8644: td.total_seconds() is now equivalent to td /
timedelta(seconds=1).
+
- Issue #4265: shutil.copyfile() was leaking file descriptors when disk fills.
Patch by Tres Seaver.
@@ -200,6 +202,8 @@
Extension Modules
-----------------
+- Issue #8644: Improved accuracy of timedelta.total_seconds().
+
- Use Clang 2.7's static analyzer to find places to clean up some code.
- Build the ossaudio extension on GNU/kFreeBSD.
Index: Doc/library/datetime.rst
===================================================================
--- Doc/library/datetime.rst (revision 81011)
+++ Doc/library/datetime.rst (working copy)
@@ -270,9 +270,13 @@
.. method:: timedelta.total_seconds()
- Return the total number of seconds contained in the duration. Equivalent to
- ``td.microseconds / 1000000 + td.seconds + td.days * 24 * 3600``.
+ Return the total number of seconds contained in the duration.
+ Equivalent to ``(td.microseconds + (td.seconds + td.days * 24 *
+ 3600) * 10**6) / 10**6`` computed with true division enabled.
+ Note that for very large time intervals (greater than 270 years on
+ most platforms) this method will lose microsecond accuracy.
+
.. versionadded:: 2.7
Index: Lib/test/test_datetime.py
===================================================================
--- Lib/test/test_datetime.py (revision 81011)
+++ Lib/test/test_datetime.py (working copy)
@@ -2,7 +2,7 @@
See http://www.zope.org/Members/fdrake/DateTimeWiki/TestCases
"""
-
+from __future__ import division
import os
import pickle
import cPickle
@@ -269,6 +269,13 @@
for total_seconds in [123456.789012, -123456.789012, 0.123456, 0, 1e6]:
td = timedelta(seconds=total_seconds)
self.assertEqual(td.total_seconds(), total_seconds)
+ # Issue8644: Test that td.total_seconds() has the same
+ # accuracy as td / timedelta(seconds=1).
+ for ms in [-1, -2, -123]:
+ td = timedelta(microseconds=ms)
+ self.assertEqual(td.total_seconds(),
+ ((24*3600*td.days + td.seconds)*10**6
+ + td.microseconds)/10**6)
def test_carries(self):
t1 = timedelta(days=100,
Index: Modules/datetimemodule.c
===================================================================
--- Modules/datetimemodule.c (revision 81011)
+++ Modules/datetimemodule.c (working copy)
@@ -2096,9 +2096,25 @@
static PyObject *
delta_total_seconds(PyObject *self)
{
- return PyFloat_FromDouble(GET_TD_MICROSECONDS(self) / 1000000.0 +
- GET_TD_SECONDS(self) +
- GET_TD_DAYS(self) * 24.0 * 3600.0);
+ PyObject *total_seconds;
+ PyObject *total_microseconds;
+ PyObject *one_million;
+
+ total_microseconds = delta_to_microseconds((PyDateTime_Delta *)self);
+ if (total_microseconds == NULL)
+ return NULL;
+
+ one_million = PyLong_FromLong(1000000L);
+ if (one_million == NULL) {
+ Py_DECREF(total_microseconds);
+ return NULL;
+ }
+
+ total_seconds = PyNumber_TrueDivide(total_microseconds, one_million);
+
+ Py_DECREF(total_microseconds);
+ Py_DECREF(one_million);
+ return total_seconds;
}
static PyObject *
_______________________________________________
Python-bugs-list mailing list
Unsubscribe:
http://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
