Improvements to catalogue record fetching

[watucker]

We might need a better way to handle not being able to fetch a catalogue record, and probably shouldn't need to try more than one API endpoint (the MOLES catalogue) to get a cachable result. See:

archive_browser/browser/views.py

Line 72 in aaeaaad

def moles_record(path):

Unsure what the best way to handle a failure is, but at least we should remove "catalogue.ceda.ac.uk" from the list of CAT_URLS.

[watucker]

It's been noted that a relatively new addition to the "moles_record" function above may have caused some unexpected problems. Specifically, the code within this if block may need better error handling:

archive_browser/browser/views.py

Line 77 in 35052b1

if data.get("record_type") == 'Dataset':