timespan parsing is broken in dataframe_from_result_table when days and seconds co-exists

[Intaik]

the following lambda fails to parse timespan correctly on when days and seconds co-exists.
ex> Kusto timespan: 7.04:44:01.5115511 (7 days, 4 hrs, 44 min, 1.5115511 seconds) will results in "7 days 04:44:01 days 5115511" string passed to pd.to_timedelta, which will produce "67 days 09:42:31" incorrectly. only the first dot should be replaced with "days"

if col_type.lower() == "timespan":

        frame[col_name] = pandas.to_timedelta(

            frame[col_name].apply(lambda t: t.replace(".", " days ") if t and "." in t.split(":")[0] else t)

        )

[toshetah]

Thanks Intaik for reporting this. If you encounter this issue again please reopen it.