Fix astroquery.utils.system_tools.in_ipynb()

[eerovaher]

The purpose of astroquery.utils.system_tools.in_ipynb() is to recognize if it is being called from a .ipynb notebook or not, but the current function seems to always return False. The function in this pull request returns False if it is called from ipython, but True if called from jupyter notebook or jupyter lab, at least for the versions I'm using.

ipython --version
7.19.0
jupyter-notebook --version
6.1.5
jupyter-lab --version
3.0.16

[keflavich]

I don't know if it's safe to rely on an environmental variable for this sort of check... but clearly the "robust-looking" approach we took wasn't. I'm OK with trying this.

[codecov]

Codecov Report

Merging #2289 (8cec4e5) into main (08e57d7) will increase coverage by 0.02%.
The diff coverage is 0.00%.

@@            Coverage Diff             @@
##             main    #2289      +/-   ##
==========================================
+ Coverage   62.75%   62.77%   +0.02%     
==========================================
  Files         130      130              
  Lines       16835    16828       -7     
==========================================
  Hits        10564    10564              
+ Misses       6271     6264       -7     

Impacted Files	Coverage Δ
astroquery/utils/system_tools.py	75.00% <0.00%> (+22.82%)	⬆️

---

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 08e57d7...8cec4e5. Read the comment docs.

[bsipocz]

I wonder if we need this as a separate function or could just use the suggested workaround directly in query.py?

[eerovaher]

I don't know if it's safe to rely on an environmental variable for this sort of check...

An alternative might be

try:
    return 'connection_file' in get_ipython().config['IPKernelApp']
except NameError:
    return False

But I don't know if there's actually any difference in practice.

[bsipocz]

I still feel this may be now an overshoot for a separate function, but in either case the fix is working and an improvement itself.

[bsipocz]

Thanks @eerovaher!