Source names query breaks when not all sources are found

[ddobie]

Kindly pointed out by @Rachel-Berdugo.

I suspect we just need to drop sources earlier, but will look into this post-workshop.

[ddobie]

This issue has two causes.

1. We don't handle the simbad query properly and just assume that each requested name will be successfully resolved, and only resolved into one source. Anything other than a one-to-one match results in things breaking.
2. Rachel's specific problem was being caused by one of her objects having the name "AT". Querying simbad for this name (along with at least one other real name) results in it returning every object with an "AT" prefix, of which there is ~3000.

I think the solution is to just drop all sources that don't have a TYPED_ID.

[ddobie]

As an aside, there's a lot of claptrap that we have in query.py from before we were able to extract the originally queried name. Most of that can be removed now that TYPED_ID is returned.

[ddobie]

Also weirdly enough, this bug only occurs when you query "AT" along with at least one other name. Querying "AT" by itself results in the expected behaviour (a single source with no match).

[ddobie]

The same behaviour occurs for "GRB", but not for "SN" or "M". Querying "NGC" results in a timeout.