Fix mistake converting string date during leap second to jd1, jd2

[taldcroft]

Description

The function convert_string_to_jd1_jd2 mistakenly used the TT scale for conversion. I do not understand where that came from, but UTC is the right answer because all the Chandra string-based date formats with fast converter support are UTC.

This only impacts conversions for times within a leap second. The last leap second was 2016-12-31, so this issue has almost no impact, but still good to fix.

Interface impacts

Testing

Unit tests

• Mac

(ska3) ➜  cxotime git:(fix-fast-convert-leap-sec) git rev-parse --short HEAD                   
c7d7cbe
(ska3) ➜  cxotime git:(fix-fast-convert-leap-sec) pytest
================================================== test session starts ===================================================
platform darwin -- Python 3.12.8, pytest-8.3.4, pluggy-1.5.0
rootdir: /Volumes/git
configfile: pytest.ini
plugins: anyio-4.7.0, timeout-2.3.1
collected 290 items                                                                                                      

cxotime/tests/test_cxotime.py .................................................................................... [ 28%]
.................................................................................................................. [ 68%]
.............................................................................                                      [ 94%]
cxotime/tests/test_cxotime_linspace.py ...............                                                             [100%]

================================================== 290 passed in 2.47s ===================================================

Independent check of unit tests by Jean

• Linux

ska3-jeanconn-fido> git rev-parse HEAD
c7d7cbe865ce9491d02c58550ebf0d1f5d40348a
ska3-jeanconn-fido> pytest
============================= test session starts ==============================
platform linux -- Python 3.12.8, pytest-8.3.4, pluggy-1.5.0
rootdir: /proj/sot/ska/jeanproj/git
configfile: pytest.ini
plugins: anyio-4.7.0, timeout-2.3.1
collected 290 items                                                                                                       

cxotime/tests/test_cxotime.py .......................................... [ 14%]
........................................................................ [ 39%]
........................................................................ [ 64%]
........................................................................ [ 88%]
.................                                                                                                   [ 94%]
cxotime/tests/test_cxotime_linspace.py ...............                                                              [100%]

=================================================== 290 passed in 5.58

Functional tests

New unit test.

[jeanconn]

By "because all the Chandra date formats are UTC" do you mean everything except Chandra seconds or do you not consider that a date format or?

[taldcroft]

By "because all the Chandra date formats are UTC" do you mean everything except Chandra seconds or do you not consider that a date format or?

I updated the description "because all the Chandra string-based date formats with fast converter support are UTC."

[jeanconn]

Looks good to me. I remain not an expert on leap second handling in erfa, but this fix also makes sense in that it repairs the symmetry with convert_string_to_jd2_jd2 (which explicitly uses dtf2d with UTC scale). I also explicitly confirmed that the new test fails in master as expected.