Make utils.commons.parse_coordinates() work with astropy 5.0

[eerovaher]

The exception handling in astropy.coordinates changed in version 5.0. This pull request ensures astroquery.utils.commons.parse_coordinates() works with astropy 5.0 and also with the older versions.

Fixes #2194.

[codecov]

Codecov Report

Merging #2196 (34a72df) into main (75cfd69) will increase coverage by 0.01%.
The diff coverage is 94.44%.

@@            Coverage Diff             @@
##             main    #2196      +/-   ##
==========================================
+ Coverage   66.10%   66.11%   +0.01%     
==========================================
  Files         418      418              
  Lines       28167    28180      +13     
==========================================
+ Hits        18619    18631      +12     
- Misses       9548     9549       +1     

Impacted Files	Coverage Δ
astroquery/query.py	58.87% <80.00%> (+0.46%)	⬆️
astroquery/simbad/core.py	89.75% <100.00%> (+0.09%)	⬆️
astroquery/simbad/tests/test_simbad.py	98.35% <100.00%> (+0.02%)	⬆️
astroquery/utils/commons.py	78.78% <100.00%> (+0.10%)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update f18f02b...34a72df. Read the comment docs.

[keflavich]

Looks like this solves #2194?

[keflavich]

@bsipocz I'm pretty sure this does solve the broken CI issues.

@eerovaher thank you!

[bsipocz]

Thank you @eerovaher!

[astrofrog]

@bsipocz @keflavich - just to check, are you planning a release of astroquery soon? If possible I would like to hold back from releasing astropy 5.0 until there is acompatible astroquery version so that users have a pathway to upgrade.

[astrofrog]

Note that this still fails with 5.0rc2, I think because minversion doesn't include release candidates. So I think we are ok with 5.0rc2, but I wonder if it would be more robust to check the length of args instead of relying on the astropy version.

[bsipocz]

Yes, hopefully today, if not, then by the end of the week for sure.

[bsipocz]

If you or someone else could open a PR that doesn't check the version, please do so. I have a lot of smallish PRs that needs to be done for the release, so this is not at the top of my personal list.

[eerovaher]

Note that this still fails with 5.0rc2, I think because minversion doesn't include release candidates. So I think we are ok with 5.0rc2, but I wonder if it would be more robust to check the length of args instead of relying on the astropy version.

An alternative would be to replace

astroquery/astroquery/utils/commons.py

Line 60 in 34a72df

ASTROPY_LT_5_0 = not minversion('astropy', '5.0')

with

 ASTROPY_LT_5_0 = not minversion('astropy', '5.0rc1')