There can possibly be different functions, useful for some specific
operations. It is possible to keep them all in the individual files, but
it is not very comfortable to use (one have to type something like
`from pyDKB.common.my_function_name import my_function_name`, or in
`__init__` file have a list of imports like this:
`from my_function_name import my_function_name`).

So I think it will be fine to put them all to a common file ('utils' or
'misc' or whatever) and use like `from pyDKB.common.utils import
my_function_name`). Looks a little better (for me, at least).

Function like this seem to be needed in multiple places, so why not to
have it in the common library.

If `--es-config FILE` parameter is specified, use DKB ES storage
as a backup metadata source in case that in primary source (AMI)
information was removed.

I am not sure under what curcumstances information can be removed from
AMI, so for now we just check if there are empty/missed fields in the
data taken from AMI and then, if such firlds found, check ES for
(possibly) already known values.

The problem is that there are quite a lot of "missed" values, so almost
for every record we have to check both AMI and ES. Maybe there is more
delicate trigger, like "No data at all", or "if dataset property
'deleted' is set to True", or...

Or maybe AMI just doesn't remove data at all?..

In theory it should make the request slightly faster.