7724600eb8
update pre-commit workflow versions
2023-10-03 09:06:48 -05:00
b5e7f3eff6
tune when mypy is run
2023-10-03 09:03:03 -05:00
f280d36d26
Add pre-commit
2023-10-03 08:57:23 -05:00
0f62a55154
test across oses
2023-10-03 08:56:31 -05:00
0c1bf1c6dd
ietf no longer serves leapseconddata, use alternate sources
2023-10-03 08:53:27 -05:00
053543c9c8
Merge pull request #8 from jepler/is-tai-improvement
...
Is tai improvement
2022-11-05 18:25:06 -05:00
168426a4c6
Make the 'is tai' check not depend on object identity
...
.. because the identity of the 'tai' object can be lost when
the multiprocessing module gets involved.
2022-11-05 18:22:28 -05:00
16198dc6b7
remove option that disappeared from pylint (thanks guys)
2022-11-05 17:34:58 -05:00
78cceb8302
Merge pull request #7 from jepler/use-dataclasses
...
Switch from NamedTuple to dataclass
2022-10-09 12:46:05 -05:00
c2b821a72d
Switch from NamedTuple to dataclass
2022-10-09 12:41:48 -05:00
98549c8943
Merge pull request #6 from jepler/doc-badge
...
show doc badge
2022-10-09 12:39:43 -05:00
190d66ba17
show doc badge
2022-10-09 12:34:40 -05:00
e19fe327e0
Merge pull request #5 from jepler/fold-ls
...
Set the 'fold' property of the leap second in tai_to_utc
2022-10-09 12:27:22 -05:00
def8d7dc41
modernize the pre-commit
2022-10-09 12:24:57 -05:00
059e715a71
must now ensure fold=True for leap seconds
2022-10-09 12:24:50 -05:00
8dc8bcb37e
Set the 'fold' property of the leap second in tai_to_utc
2022-10-09 12:24:49 -05:00
a83d6a4b86
Merge pull request #4 from jepler/py.types
...
Add PEP0561 py.typed, convert to package
2022-03-23 11:18:16 -05:00
15135d024b
Merge pull request #3 from jepler/issue2
...
Get leap seconds from network if local file doesn't exist
2022-03-23 11:16:37 -05:00
dd958418e4
Convert to package and add py.typed
2022-03-23 11:10:36 -05:00
de87dda19f
use future-annotations to simplify annotations (no need to stringify)
2022-03-23 10:58:27 -05:00
9835ebcd00
fix typing and coverage
2022-03-23 10:56:27 -05:00
15ea91682e
Skip a source if it can't be retrieved
...
This fixes #2 , by going on to a network source if the local file
isn't present, such as on CentOS7.
2022-03-23 10:50:39 -05:00
9f6820dabf
Allow the 'from_url' method to return None, go to next source
2022-03-23 10:50:04 -05:00
e6dc7f1cc2
This file is on debian linux, but not CentOS 7
2022-03-23 10:49:39 -05:00
66185789ff
doc fixes
2021-10-20 16:23:30 -05:00
47f044b3a4
Merge pull request #1 from jepler/rtd
...
Add sphinx / read-the-docs documentation
2021-10-20 15:07:59 -05:00
96bf05b0de
3.7ism
2021-10-20 15:00:26 -05:00
f157185120
fix workflow
2021-10-20 14:54:17 -05:00
4fb3e9ea7f
Add sphinx / rtd documentation
2021-10-20 14:53:02 -05:00
8007b02dcc
Add more TAI-wrangling
...
* leapseconddata.tai is a timezone object. BEWARE, because of reasons,
it compares equal to datetime.datetime.utc!
* tai_offset can take either kind of timestamp
* to_tai, tai_to_utc: convert to/from the TAI timescale.
* is_leap_second: return True if the given second is a leap second.
In the UTC timescale, the ":59" second returns True even though
"only" the second repetition of :59 is a leap second
2021-10-20 14:44:34 -05:00
2dddcdf282
Update README.md
2021-10-18 19:48:42 -05:00
507255e040
fix workflow
2021-10-18 19:31:46 -05:00
632139c299
Add unittest and improve coverage
2021-10-18 19:29:14 -05:00
8a25adb764
improvements
...
* Our own error types (squee, such cute little classes)
* check_validity of a timestamp without exceptions
* tai_offset always returns a timedelta now
* add from_standard_source, searches standard linux & freebsd locs, then
goes to the network for ietf
* reformat some test-run messages so they don't trigger my editor's
"error on file and line" regular expression matching
* lightly use logging
2021-10-18 14:15:57 -05:00
30d6bb889b
version numbers important
2021-10-18 10:45:38 -05:00
59da27383f
don't run coverage during release
2021-10-18 10:34:57 -05:00
9b2c42b376
package-ize
2021-10-18 10:22:33 -05:00
78d51b3d71
initial commit
2021-10-18 09:20:59 -05:00
