Fix download error due to no dataURL in MAST result

[orionlee]

Close #1377 , #1381

---

Changelog:

Fixed SearchResult.download() / download_all()'s misuse of dataURL. [#1380]

[orionlee]

A small, unrelated commit, 7693e28, was added to update the test codes. It handles an unrelated changes in MAST search result for Kepler products, to handle the additional KBONUS-BKG product, which seems to be related to https://archive.stsci.edu/hlsp/kbonus-apexba

I don't think anything needs to be changed on lightkurve side. But I am not familiar with Kepler products, so I call it out here just in case.

import lightkurve as lk
# insert code here ...

sr = lk.search_lightcurve("Kepler-10", mission="Kepler", cadence="long")
print(sr)

SearchResult containing 16 data products.

 #       mission      year   author   exptime         target_name          distance proposal_id
                                         s                                  arcsec             
--- ----------------- ---- ---------- ------- ---------------------------- -------- -----------
  0   Kepler Quarter  2009 KBONUS-BKG    1765 Gaia DR3 2132155017099178624      0.0         N/A
  1 Kepler Quarter 00 2009     Kepler    1800                kplr011904151      0.0          EX
  2 Kepler Quarter 01 2009     Kepler    1800                kplr011904151      0.0          EX
...
15 Kepler Quarter 17 2013     Kepler    1800                kplr011904151      0.0          EX

[scfleming]

Just a quick note, that is a new HLSP update we just announced today (shortly, I'm working on the MAST announcement as we speak.) MAST homepage: https://archive.stsci.edu/hlsp/kbonus-bkg

[Nschanche]

We could sort the search results by author to return 'mission products' at the top and HLSPs below if that would be preferable?

[christinahedges]

@orionlee thanks for opening this PR, we're all just coming up against this new change now. I've updated two things on your PR, I've added the URL for this HLSP which @scfleming provided, and I've changed the ordering to make sure that this HLSP is not the first product that comes up (the mission product should be first in the list.) (seeing now @Nschanche has made the same comment above!)

@jorgemarpa and I will open a new PR to make sure there's a specific reader for this new HLSP.

@orionlee I'm going to release a new version ASAP to make sure people can pip upgrade to get around this issue. Let's prepare for lots of new issues to open about this...!

[orionlee]

@christinahedges

1. The CHANGES.rst in main is out of sync with the v2.4.x version, essentially part of the issue branch v2.4.x and main is of of sync that could cause merge headache #1378 [*]. It probably needs to be addressed before release.

2. For KBONUS-BKG support, I did a quick look yesterday, and suspect the following might need to be considered:

• each lightcurve file contains data stitched across multiple quarters. It could be surprising to users when they search for data in specific quarter, the KBONUS-BKG lightcurve will have results across multiple quarters.

---

[*] The current main CHANGES.rst is out of sync with the released 2.4.1 . It should be something like:

2.4.2dev (unreleased)
=====================

- Updated plot style to default to higher resolution [#1311]
- Fixed download issue due to MAST API change [#1380]

2.4.1 (2023-09-06)
==================

- Updated interact features to work with JupyterHub (e.g. TiKE) [#1349]
- Exposed niters parameter in the PCA function of design matrix
- Fixed the aperture parsing functions inside TPFs to be compliant with `numpy` v1.25 [#1360]

[christinahedges]

@orionlee great catch, I lost track of this with all the MAST changes, I'll remedy it.

For the new HLSP, @jorgemarpa and I are the authors of the release so I'm working on putting in a fix now.