ENH: Add AltAzTarget and support time-dependent targets

[michaelbaisch]

• Add AltAzTarget for targets defined in horizontal coordinates at a fixed EarthLocation, evaluated to time-dependent ICRS SkyCoord when needed; support vacuum (pressure=0) and apparent (pressure>0) interpretations via stored atmospheric parameters.
• Export AltAzTarget and make get_skycoord(time-aware): accept times, evaluate time-dependent targets via get_skycoord(times), and broadcast coords.
• Pass time into get_skycoord from Observer._preprocess_inputs to enable time-dependent target evaluation during constraint calculations.
• Update Constraint.call to delegate gridding and time-dependent evaluation to Observer._preprocess_inputs.
• In scheduling add target_type/target_info columns in Schedule.to_table.
• Add tests.

#566 was updated to be compatible with this PR.

[codecov]

Codecov Report

❌ Patch coverage is 83.00654% with 26 lines in your changes missing coverage. Please review.
✅ Project coverage is 74.60%. Comparing base (a4d89a0) to head (1a1eadb).

Files with missing lines	Patch %	Lines
astroplan/scheduling.py	60.00%	16 Missing ⚠️
astroplan/target.py	91.00%	9 Missing ⚠️
astroplan/observer.py	90.90%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #634      +/-   ##
==========================================
+ Coverage   72.90%   74.60%   +1.69%     
==========================================
  Files          14       14              
  Lines        1849     1953     +104     
==========================================
+ Hits         1348     1457     +109     
+ Misses        501      496       -5     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[m4cd4r4]

Solid addition to astroplan. The AltAzTarget class fills a real gap for users working with local pointing directions rather than celestial coordinates. Test coverage is thorough (208 lines).

A few things I noticed:

Breaking change in Schedule.to_table():
The ra and dec columns are replaced with target type and target info. Any downstream code that accesses table['ra'] or table['dec'] will break silently. Would it be possible to keep the old columns (populated where applicable) alongside the new ones? Or add a deprecation warning for one release cycle?

Time-dependent detection heuristic:
The _is_time_dependent check (has get_skycoord callable and no coord attribute) is duck-typing based. If someone later adds a get_skycoord method to FixedTarget, this logic breaks silently. A dedicated is_time_dependent property on the Target base class would be more explicit and safer against future changes.

Atmospheric parameter defaults:
The docstring says pressure defaults to 0 hPa (vacuum), but the code passes pressure=None to AltAz. Depending on the astropy version, AltAz(pressure=None) could use a different default. Setting pressure = 0 * u.hPa explicitly in __init__ would guarantee vacuum behavior regardless of astropy version.

Minor:

• The except Exception blocks in _format_target_info are broad. Narrowing to AttributeError or ConvertError would avoid masking real bugs.
• The error message when times=None for a time-dependent target could mention that the caller needs to provide times (currently just raises a bare ValueError).

The refactoring of _preprocess_inputs to handle time-dependent gridding is clean, and the score_key rounding in the scheduler is a good catch. Nice work.