Adding HAPCut functionality to astroquery.mast.Cutouts tool

[jaymedina]

This PR will expand astroquery.mast's cutout functionality to support making Hubble Advanced Products (HAP) cutouts via MAST's HAPCut API.

! Timeline: ASAB would like to see this merged before the end of Q1. Ideally before Dec 21st if possible.

Changes include:

• New HapcutClass in cutouts.py
• Unit testing for HapcutClass
• Updated documentation for HapcutClass functionality
• Updated release notes

Example download_cutouts call with sample cutout shown:

Example get_cutouts call with sample cutout shown

(default cutout size is 5 x 5 pixels):

[codecov]

Codecov Report

Merging #2613 (10f335e) into main (5057cbc) will decrease coverage by 4.64%.
The diff coverage is 14.54%.

❗ Current head 10f335e differs from pull request most recent head e773121. Consider uploading reports for the commit e773121 to get more accurate results

@@            Coverage Diff             @@
##             main    #2613      +/-   ##
==========================================
- Coverage   68.82%   64.17%   -4.65%     
==========================================
  Files         294      130     -164     
  Lines       22215    16891    -5324     
==========================================
- Hits        15289    10840    -4449     
+ Misses       6926     6051     -875     

Impacted Files	Coverage Δ
astroquery/mast/cutouts.py	65.38% <14.54%> (-13.64%)	⬇️
astroquery/heasarc/core.py	73.79% <0.00%> (-1.70%)	⬇️
astroquery/mast/__init__.py
astroquery/ibe/__init__.py
astroquery/astrometry_net/__init__.py
astroquery/dace/__init__.py
astroquery/vo_conesearch/__init__.py
astroquery/esa/iso/tests/test_iso.py
astroquery/utils/tap/conn/tests/test_conn.py
astroquery/esa/iso/tests/setup_package.py
... and 157 more

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

[jaymedina]

This question came up during one of our stand-ups: When will the next release of astroquery be?

[bsipocz]

The next tagged release is not planned before we finish most of the API cleanups we planned to do, so maybe Q1 (calendar year) next year. In the meantime I can push dev releases to pypi without any problems.

[bsipocz]

I have some big picture comments, which certainly points beyond this current PR.:

• I wonder at what point will it make sense to make a generic mast.Cutout class, that takes the coordinate and returns all available cutouts, or maybe also takes a missing name? Even if it's not exposed to the public API, I think we reached a point where having mast cutout base class would be beneficial as some of the methods share significant amount of code (e.g. enhancements and bugfixes, or any maintenance will be easier with a baseclass).

Otherwise I have minor comments, so would only like to see a few small things to be fixed.

[bsipocz]

side note: the doctest are failing in an earlier example, so these has not been run during my testing

[jaymedina]

I wonder at what point will it make sense to make a generic mast.Cutout class, that takes the coordinate and returns all available cutouts, or maybe also takes a missing name? Even if it's not exposed to the public API, I think we reached a point where having mast cutout base class would be beneficial as some of the methods share significant amount of code (e.g. enhancements and bugfixes, or any maintenance will be easier with a baseclass).

This has actually been a topic of discussion recently, and we do have plans to consolidate all the cutouts classes into a single mission-agnostic cutout tool. I'm not sure what the timeline will be on this, though. We have a few more things we want to update astroquery.mast.Cutouts with, including updating TesscutClass with support for a high level science product called TICA, as well as pointing Cutouts to the cloud, so consolidating all the classes into one will likely happen after that.

[bsipocz]

This has actually been a topic of discussion recently ...

Sounds all great. It certainly points beyond this current PR, and any timeline will work that works for you.
Also re: cloud - as part of fornax (so mast is in the know), we're factoring out the cloud download logic from the mast module and upstream it into pyvo for a more generic audience, so it would be useful to communicate any plans for enhancements if it affects the bottom infrastructure.

[jaymedina]

Good to know, I'll relay this to the team

[bsipocz]

Could you squash out some of the later debugging/incremental commits? I plan to come back and go through it again in the evening, but don't think there is any more things to address for merge.

[jaymedina]

Yeah, I'm waiting for a science review on this from one group, once I hear back from them I'll squash this down so it's ready for merge.

[jaymedina]

Squashed! @bsipocz

[bsipocz]

Thanks! No need to squash to one, it's really more about keeping the incremental fixes at the later parts into more logical chunks.
Either way, this is now good to go and I'll do it later today, but first will reuse a PR for an experiment for the gh command line client.

[bsipocz]

Thank you @jaymedina!