REVISITED - df.cdfepoch.encode_tt2000 and cdf.cdfepoch.breakdown_tt2000 - unexpected return values during or just after leap second

[EsaRiihonen]

I closed earlier issue as it seemed that the recent version corrected the problem – alas not. The output changed but the results are still erroneous.

cdflib-1.3.3
Python-3.9.7

Below are the results from a test around the latest leap second.

cdfepoch.parse seems to work correctly as the resulting tt2000 epoch values are as expected. However cdfepoch.encode has a problem during the leap second and cdfepoch.breakdown in addition also during the first second after the leap second.

TEST

import cdflib as cdf

tt2000_check = cdf.cdfepoch.parse(['2016-12-31T23:59:59.000000000',
                                   '2016-12-31T23:59:60.000000000',
                                   '2016-12-31T23:59:60.500000000',
                                   '2017-01-01T00:00:00.000000000',
                                   '2017-01-01T00:00:01.000000000',])

for tt2000 in tt2000_check:
    print(tt2000)

> 536500867184000000
> 536500868184000000
> 536500868684000000
> 536500869184000000
> 536500870184000000

cdf.cdfepoch.encode_tt2000(tt2000_check)

> ['2016-12-31T23:59:59.000000000',    <--- OK
>  '2016-12-31T23:60:00.000000000',    <--- ??? 
>  '2016-12-31T23:60:00.500000000',    <--- ???
>  '2017-01-01T00:00:00.000000000',    <--- OK
>  '2017-01-01T00:00:01.000000000']    <--- OK

cdf.cdfepoch.breakdown_tt2000(tt2000_check)

> array([[2016,   12,   31,   23,   59,   59,    0,    0,    0],    <--- OK
>        [2016,   12,   31,   23,   60,    0,    0,    0,    0],    <--- ???
>        [2016,   12,   31,   23,   60,    0,  500,    0,    0],    <--- ???
>        [2016,   12,   31,   23,   60,    0,    0,    0,    0],    <--- ???
>        [2017,    1,    1,    0,    0,    1,    0,    0,    0]])   <--- OK