ENH: Expand CDMS testing

[keflavich]

Following #3094, it is clear that there is a range of untested parameters in the CDMS module that require tests.

TODO includes:

• Regression test for 3094 (check that 100501 can be parsed)
• Systematic check that all molecules can be queried

[keflavich]

The systematic search will begin with something like:

from astroquery.linelists.cdms import CDMS
from astropy import units as u

sptbl = CDMS.get_species_table()

for row in sptbl:
    result = CDMS.query_lines(1*u.GHz, 1*u.THz, molecule=str(row['tag']), parse_name_locally=False)
    print(tag, row['molecule'], len(result))

but either CDMS is down right now or they're blocking the airport IP.

[kmaitreys]

Yeah, I tried something like this but some of my requests got timed out as well. Although I still managed to fetch transitions for most of molecules in the database. While doing that, I encountered one more bug

ValueError: invalid literal for int() with base 10: '1 1 2' for molecule 028528 

Here 028528 is another molecule which has some formatting bug. I will add a fix for this as well soon on my #3094 PR.

EDIT: I am not sure that only these two TAG values were problematic though, so a more exhaustive check will be nice.

[bsipocz]

Yeap, adding just those one or two to #3094 would be enough, and this issue can be addressed in a follow-up PR that does a the exhaustive checks.