add %f format to datetime - issue 1158

Doc/library/datetime.rst

@@ -1489,9 +1489,31 @@ For :class:`time` objects, the format codes for year, month, and day should not
 be used, as time objects have no such values.  If they're used anyway, ``1900``
 is substituted for the year, and ``0`` for the month and day.
 
-For :class:`date` objects, the format codes for hours, minutes, and seconds
-should not be used, as :class:`date` objects have no such values.  If they're
-used anyway, ``0`` is substituted for them.
+For :class:`date` objects, the format codes for hours, minutes, seconds, and
+microseconds should not be used, as :class:`date` objects have no such
+values.  If they're used anyway, ``0`` is substituted for them.
+
+:class:`time` and :class:`datetime` objects support a ``%f`` format code
+which expands to the number of microseconds in the object, zero-padded on
+the left to six places.
+
+.. versionadded:: 2.6
+
+For a naive object, the ``%z`` and ``%Z`` format codes are replaced by empty
+strings.
+
+For an aware object:
+
+``%z``
+   :meth:`utcoffset` is transformed into a 5-character string of the form +HHMM or
+   -HHMM, where HH is a 2-digit string giving the number of UTC offset hours, and
+   MM is a 2-digit string giving the number of UTC offset minutes.  For example, if
+   :meth:`utcoffset` returns ``timedelta(hours=-3, minutes=-30)``, ``%z`` is
+   replaced with the string ``'-0330'``.
+
+``%Z``
+   If :meth:`tzname` returns ``None``, ``%Z`` is replaced by an empty string.
+   Otherwise ``%Z`` is replaced by the returned value, which must be a string.
 
 The full set of format codes supported varies across platforms, because Python
 calls the platform C library's :func:`strftime` function, and platform
@@ -1524,6 +1546,10 @@ platforms.  Regardless of platform, years before 1900 cannot be used.
 | ``%d``    | Day of the month as a decimal  |       |
 |           | number [01,31].                |       |
 +-----------+--------------------------------+-------+
+| ``%f``    | Microsecond as a decimal       | \(1)  |
+|           | number [0,999999], zero-padded |       |
+|           | on the left                    |       |
++-----------+--------------------------------+-------+
 | ``%H``    | Hour (24-hour clock) as a      |       |
 |           | decimal number [00,23].        |       |
 +-----------+--------------------------------+-------+
@@ -1539,13 +1565,13 @@ platforms.  Regardless of platform, years before 1900 cannot be used.
 | ``%M``    | Minute as a decimal number     |       |
 |           | [00,59].                       |       |
 +-----------+--------------------------------+-------+
-| ``%p``    | Locale's equivalent of either  | \(1)  |
+| ``%p``    | Locale's equivalent of either  | \(2)  |
 |           | AM or PM.                      |       |
 +-----------+--------------------------------+-------+
-| ``%S``    | Second as a decimal number     | \(2)  |
+| ``%S``    | Second as a decimal number     | \(3)  |
 |           | [00,61].                       |       |
 +-----------+--------------------------------+-------+
-| ``%U``    | Week number of the year        | \(3)  |
+| ``%U``    | Week number of the year        | \(4)  |
 |           | (Sunday as the first day of    |       |
 |           | the week) as a decimal number  |       |
 |           | [00,53].  All days in a new    |       |
@@ -1556,7 +1582,7 @@ platforms.  Regardless of platform, years before 1900 cannot be used.
 | ``%w``    | Weekday as a decimal number    |       |
 |           | [0(Sunday),6].                 |       |
 +-----------+--------------------------------+-------+
-| ``%W``    | Week number of the year        | \(3)  |
+| ``%W``    | Week number of the year        | \(4)  |
 |           | (Monday as the first day of    |       |
 |           | the week) as a decimal number  |       |
 |           | [00,53].  All days in a new    |       |
@@ -1576,7 +1602,7 @@ platforms.  Regardless of platform, years before 1900 cannot be used.
 | ``%Y``    | Year with century as a decimal |       |
 |           | number.                        |       |
 +-----------+--------------------------------+-------+
-| ``%z``    | UTC offset in the form +HHMM   | \(4)  |
+| ``%z``    | UTC offset in the form +HHMM   | \(5)  |
 |           | or -HHMM (empty string if the  |       |
 |           | the object is naive).          |       |
 +-----------+--------------------------------+-------+
@@ -1589,17 +1615,22 @@ platforms.  Regardless of platform, years before 1900 cannot be used.
 Notes:
 
 (1)
+   When used with the :func:`strptime` function, the ``%f`` directive
+   accepts from one to six digits and zero pads on the right.  ``%f`` is
+   an extension to the set of format characters in the C standard.
+
+(2)
    When used with the :func:`strptime` function, the ``%p`` directive only affects
    the output hour field if the ``%I`` directive is used to parse the hour.
 
-(2)
+(3)
    The range really is ``0`` to ``61``; this accounts for leap seconds and the
    (very rare) double leap seconds.
 
-(3)
+(4)
    When used with the :func:`strptime` function, ``%U`` and ``%W`` are only used in
    calculations when the day of the week and the year are specified.
 
-(4)
+(5)
    For example, if :meth:`utcoffset` returns ``timedelta(hours=-3, minutes=-30)``,
    ``%z`` is replaced with the string ``'-0330'``.

Lib/_strptime.py

@@ -22,7 +22,7 @@ try:
 except:
     from dummy_thread import allocate_lock as _thread_allocate_lock
 
-__all__ = ['strptime']
+__all__ = []
 
 def _getlang():
     # Figure out what the current language is set to.
@@ -190,6 +190,7 @@ class TimeRE(dict):
         base.__init__({
             # The " \d" part of the regex is to make %c from ANSI C work
             'd': r"(?P<d>3[0-1]|[1-2]\d|0[1-9]|[1-9]| [1-9])",
+            'f': r"(?P<f>[0-9]{1,6})",
             'H': r"(?P<H>2[0-3]|[0-1]\d|\d)",
             'I': r"(?P<I>1[0-2]|0[1-9]|[1-9])",
             'j': r"(?P<j>36[0-6]|3[0-5]\d|[1-2]\d\d|0[1-9]\d|00[1-9]|[1-9]\d|0[1-9]|[1-9])",
@@ -291,7 +292,7 @@ def _calc_julian_from_U_or_W(year, week_of_year, day_of_week, week_starts_Mon):
         return 1 + days_to_week + day_of_week
 
 
-def strptime(data_string, format="%a %b %d %H:%M:%S %Y"):
+def _strptime(data_string, format="%a %b %d %H:%M:%S %Y"):
     """Return a time struct based on the input string and the format string."""
     global _TimeRE_cache, _regex_cache
     with _cache_lock:
@@ -327,7 +328,7 @@ def strptime(data_string, format="%a %b %d %H:%M:%S %Y"):
                           data_string[found.end():])
     year = 1900
     month = day = 1
-    hour = minute = second = 0
+    hour = minute = second = fraction = 0
     tz = -1
     # Default to -1 to signify that values not known; not critical to have,
     # though
@@ -384,6 +385,11 @@ def strptime(data_string, format="%a %b %d %H:%M:%S %Y"):
             minute = int(found_dict['M'])
         elif group_key == 'S':
             second = int(found_dict['S'])
+        elif group_key == 'f':
+            s = found_dict['f']
+            # Pad to always return microseconds.
+            s += "0" * (6 - len(s))
+            fraction = int(s)
         elif group_key == 'A':
             weekday = locale_time.f_weekday.index(found_dict['A'].lower())
         elif group_key == 'a':
@@ -440,6 +446,9 @@ def strptime(data_string, format="%a %b %d %H:%M:%S %Y"):
         day = datetime_result.day
     if weekday == -1:
         weekday = datetime_date(year, month, day).weekday()
-    return time.struct_time((year, month, day,
-                             hour, minute, second,
-                             weekday, julian, tz))
+    return (time.struct_time((year, month, day,
+                              hour, minute, second,
+                              weekday, julian, tz)), fraction)
+
+def _strptime_time(data_string, format="%a %b %d %H:%M:%S %Y"):
+    return _strptime(data_string, format)[0]

Lib/test/test_datetime.py

@@ -1507,11 +1507,12 @@ class TestDateTime(TestDate):
         self.failUnless(abs(from_timestamp - from_now) <= tolerance)
 
     def test_strptime(self):
-        import time
+        import _strptime
 
-        string = '2004-12-01 13:02:47'
-        format = '%Y-%m-%d %H:%M:%S'
-        expected = self.theclass(*(time.strptime(string, format)[0:6]))
+        string = '2004-12-01 13:02:47.197'
+        format = '%Y-%m-%d %H:%M:%S.%f'
+        result, frac = _strptime._strptime(string, format)
+        expected = self.theclass(*(result[0:6]+(frac,)))
         got = self.theclass.strptime(string, format)
         self.assertEqual(expected, got)
 
@@ -1539,9 +1540,9 @@ class TestDateTime(TestDate):
 
     def test_more_strftime(self):
         # This tests fields beyond those tested by the TestDate.test_strftime.
-        t = self.theclass(2004, 12, 31, 6, 22, 33)
-        self.assertEqual(t.strftime("%m %d %y %S %M %H %j"),
-                                    "12 31 04 33 22 06 366")
+        t = self.theclass(2004, 12, 31, 6, 22, 33, 47)
+        self.assertEqual(t.strftime("%m %d %y %f %S %M %H %j"),
+                                    "12 31 04 000047 33 22 06 366")
 
     def test_extract(self):
         dt = self.theclass(2002, 3, 4, 18, 45, 3, 1234)
@@ -1814,7 +1815,7 @@ class TestTime(HarmlessMixedComparison, unittest.TestCase):
 
     def test_strftime(self):
         t = self.theclass(1, 2, 3, 4)
-        self.assertEqual(t.strftime('%H %M %S'), "01 02 03")
+        self.assertEqual(t.strftime('%H %M %S %f'), "01 02 03 000004")
         # A naive object replaces %z and %Z with empty strings.
         self.assertEqual(t.strftime("'%z' '%Z'"), "'' ''")
 

Modules/datetimemodule.c

@@ -1130,10 +1130,24 @@ format_utcoffset(char *buf, size_t buflen, const char *sep,
 	return 0;
 }
 
+static PyObject *
+make_freplacement(PyObject *object)
+{
+	char freplacement[7];
+	if (PyTime_Check(object))
+	    sprintf(freplacement, "%06d", TIME_GET_MICROSECOND(object));
+	else if (PyDateTime_Check(object))
+	    sprintf(freplacement, "%06d", DATE_GET_MICROSECOND(object));
+	else
+	    sprintf(freplacement, "%06d", 0);
+
+	return PyString_FromStringAndSize(freplacement, strlen(freplacement));
+}
+
 /* I sure don't want to reproduce the strftime code from the time module,
  * so this imports the module and calls it.  All the hair is due to
- * giving special meanings to the %z and %Z format codes via a preprocessing
- * step on the format string.
+ * giving special meanings to the %z, %Z and %f format codes via a
+ * preprocessing step on the format string.
  * tzinfoarg is the argument to pass to the object's tzinfo method, if
  * needed.
  */
@@ -1145,6 +1159,7 @@ wrap_strftime(PyObject *object, PyObject *format, PyObject *timetuple,
 
 	PyObject *zreplacement = NULL;	/* py string, replacement for %z */
 	PyObject *Zreplacement = NULL;	/* py string, replacement for %Z */
+	PyObject *freplacement = NULL;	/* py string, replacement for %f */
 
 	char *pin;	/* pointer to next char in input format */
 	char ch;	/* next char in input format */
@@ -1186,11 +1201,11 @@ wrap_strftime(PyObject *object, PyObject *format, PyObject *timetuple,
 		}
 	}
 
-	/* Scan the input format, looking for %z and %Z escapes, building
+	/* Scan the input format, looking for %z/%Z/%f escapes, building
 	 * a new format.  Since computing the replacements for those codes
 	 * is expensive, don't unless they're actually used.
 	 */
-	totalnew = PyString_Size(format) + 1;	/* realistic if no %z/%Z */
+	totalnew = PyString_Size(format) + 1;	/* realistic if no %z/%Z/%f */
 	newfmt = PyString_FromStringAndSize(NULL, totalnew);
 	if (newfmt == NULL) goto Done;
 	pnew = PyString_AsString(newfmt);
@@ -1272,6 +1287,18 @@ wrap_strftime(PyObject *object, PyObject *format, PyObject *timetuple,
 			ptoappend = PyString_AS_STRING(Zreplacement);
 			ntoappend = PyString_GET_SIZE(Zreplacement);
 		}
+		else if (ch == 'f') {
+			/* format microseconds */
+			if (freplacement == NULL) {
+				freplacement = make_freplacement(object);
+				if (freplacement == NULL)
+					goto Done;
+			}
+			assert(freplacement != NULL);
+			assert(PyString_Check(freplacement));
+			ptoappend = PyString_AS_STRING(freplacement);
+			ntoappend = PyString_GET_SIZE(freplacement);
+		}
 		else {
 			/* percent followed by neither z nor Z */
 			ptoappend = pin - 2;
@@ -1313,6 +1340,7 @@ wrap_strftime(PyObject *object, PyObject *format, PyObject *timetuple,
 		Py_DECREF(time);
     	}
  Done:
+	Py_XDECREF(freplacement);
 	Py_XDECREF(zreplacement);
 	Py_XDECREF(Zreplacement);
 	Py_XDECREF(newfmt);
@@ -3853,43 +3881,69 @@ datetime_utcfromtimestamp(PyObject *cls, PyObject *args)
 static PyObject *
 datetime_strptime(PyObject *cls, PyObject *args)
 {
-	PyObject *result = NULL, *obj, *module;
+	static PyObject *module = NULL;
+	PyObject *result = NULL, *obj, *st = NULL, *frac = NULL;
 	const char *string, *format;
 
 	if (!PyArg_ParseTuple(args, "ss:strptime", &string, &format))
 		return NULL;
 
-	if ((module = PyImport_ImportModuleNoBlock("time")) == NULL)
+	if (module == NULL &&
+	    (module = PyImport_ImportModuleNoBlock("_strptime")) == NULL)
 		return NULL;
-	obj = PyObject_CallMethod(module, "strptime", "ss", string, format);
-	Py_DECREF(module);
 
+	/* _strptime._strptime returns a two-element tuple.  The first
+	   element is a time.struct_time object.  The second is the
+	   microseconds (which are not defined for time.struct_time). */
+	obj = PyObject_CallMethod(module, "_strptime", "ss", string, format);
 	if (obj != NULL) {
 		int i, good_timetuple = 1;
-		long int ia[6];
-		if (PySequence_Check(obj) && PySequence_Size(obj) >= 6)
-			for (i=0; i < 6; i++) {
-				PyObject *p = PySequence_GetItem(obj, i);
-				if (p == NULL) {
-					Py_DECREF(obj);
-					return NULL;
+		long int ia[7];
+		if (PySequence_Check(obj) && PySequence_Size(obj) == 2) {
+			st = PySequence_GetItem(obj, 0);
+			frac = PySequence_GetItem(obj, 1);
+			if (st == NULL || frac == NULL)
+				good_timetuple = 0;
+			/* copy y/m/d/h/m/s values out of the
+			   time.struct_time */
+			if (good_timetuple &&
+			    PySequence_Check(st) &&
+			    PySequence_Size(st) >= 6) {
+				for (i=0; i < 6; i++) {
+					PyObject *p = PySequence_GetItem(st, i);
+					if (p == NULL) {
+						good_timetuple = 0;
+						break;
+					}
+					if (PyInt_Check(p))
+						ia[i] = PyInt_AsLong(p);
+					else
+						good_timetuple = 0;
+					Py_DECREF(p);
 				}
-				if (PyInt_Check(p))
-					ia[i] = PyInt_AsLong(p);
-				else
-					good_timetuple = 0;
-				Py_DECREF(p);
 			}
+			else
+				good_timetuple = 0;
+			/* follow that up with a little dose of microseconds */
+			if (PyInt_Check(frac))
+				ia[6] = PyInt_AsLong(frac);
+			else
+				good_timetuple = 0;
+		}
 		else
 			good_timetuple = 0;
 		if (good_timetuple)
-			result = PyObject_CallFunction(cls, "iiiiii",
-				ia[0], ia[1], ia[2], ia[3], ia[4], ia[5]);
+			result = PyObject_CallFunction(cls, "iiiiiii",
+						       ia[0], ia[1], ia[2],
+						       ia[3], ia[4], ia[5],
+						       ia[6]);
 		else
 			PyErr_SetString(PyExc_ValueError,
-				"unexpected value from time.strptime");
-		Py_DECREF(obj);
+				"unexpected value from _strptime._strptime");
 	}
+	Py_XDECREF(obj);
+	Py_XDECREF(st);
+	Py_XDECREF(frac);
 	return result;
 }
 

Modules/timemodule.c

@@ -520,7 +520,7 @@ time_strptime(PyObject *self, PyObject *args)
 
     if (!strptime_module)
         return NULL;
-    strptime_result = PyObject_CallMethod(strptime_module, "strptime", "O", args);
+    strptime_result = PyObject_CallMethod(strptime_module, "_strptime_time", "O", args);
     Py_DECREF(strptime_module);
     return strptime_result;
 }