Parse MPC Epoch & Tp values as TT instead of UTC

[dschlekat]

Fixes #109 .

This PR updates MPC asteroid/comet orbit parsing so that epoch and Tp is interpreted as TT, matching the MPC orbit-element specifications.

Previously these fields were parsed as UTC, which shifted epoch and peri_time later by TT−UTC before conversion to kete's internal JD time representation. For current MPCORB rows this is about 69.184 seconds.

Changes:

• Parse MPC asteroid Epoch and Tp as TT, as documented in the MPCORB manual.
• Apply the same correction to comet orbit parsing, which required adding scaling to the Time.from_ymd method, to ensure backwards compatibility for previous UTC-only functionality.
• Add a regression test confirming that MPC JD-TT epochs are preserved.
• Updated the leap_second.dat file (only changes are the number/date of the IERS Bulletin and the file expiration date).

All tests pass with the command

pytest --cov-report term-missing --cov=kete

[dahlend]

Unless you have a strong argument for adding this, this a major footgun for users.

[dschlekat]

I don't, it was to maintain the Time.from_ymd() method used in the mpc.py file. I agree with your comment about it being a major footgun.

Upon reflection it would be easier (and far less invasive) to parse the YMD data from the MPC comet file to JD, then create a Time object from that with "TT" scaling.

[dschlekat]

Pushed a new version of the fix, addressing your concerns to do with changes to Time.from_ymd(). Adds _ymd_to_jd()helper to mpc.py. Normally I'd look to place this in a utils.py file, but as one didn't exist I just left it in the normal file. This helper is then called within fetch_known_comet_orbit_data.py to create Time objects with TT scaling.

Also, I want to make it clear it is totally fine with me if you choose to go a completely different route for the fix or not use my changes! Just wanted to try and help if I could.

[dahlend]

I hate to do this, but do this instead (and please remove the helper function that was added).

        obj = dict(
            desig=name,
            group_name=f"Comet {comet['Orbit_type']}",
            peri_dist=comet["Perihelion_dist"],
            ecc=comet["e"],
            incl=comet["i"],
            lon_node=comet["Node"],
            peri_arg=comet["Peri"],
            # MPC comet files store YMD as TT time.
            # Support for this will not be added into kete, so here is a workaround.
            peri_time=Time.from_jd(Time.from_ymd(*peri_time).utc_jd, scaling="tt").jd,
            epoch=Time.from_jd(Time.from_ymd(*epoch_time).utc_jd, scaling="tt").jd,
        )

[dschlekat]

Still significantly cleaner than the helper function.
I pushed, only change was I used Time() not Time.from_jd() (default Time new function takes in jd).

[dahlend]

Thanks!