By default the str()
conversion of a timedelta
will always include the days
portion. Internally, the value is always normalised as a number of days, seconds and microseconds, there is no point in trying to 'convert' days to hours because no separate hour component is tracked.
If you want to format a timedelta()
object differently, you can easily do so manually:
def format_timedelta(td):
minutes, seconds = divmod(td.seconds + td.days * 86400, 60)
hours, minutes = divmod(minutes, 60)
return '{:d}:{:02d}:{:02d}'.format(hours, minutes, seconds)
This ignores any microseconds portion, but that is trivially added:
return '{:d}:{:02d}:{:02d}.{:06d}'.format(hours, minutes, seconds, td.microseconds)
Demo:
>>> format_timedelta(timedelta(days=2, hours=10, minutes=20, seconds=3))
'58:20:03'
>>> format_timedelta(timedelta(hours=10, minutes=20, seconds=3))
'10:20:03'
与恶龙缠斗过久,自身亦成为恶龙;凝视深渊过久,深渊将回以凝视…