Creating datasets with non-existing fileIDs should be impossible

[MalinAhlberg]

THIS BUG HAS BEEN FIXED
Just verify that it works properly and if it is just close the issue

Describe the bug
It is possible to create a dataset with incorrect file IDs:

sda-admin dataset create --dataset-id dataset1 non-existing-id1 bad-id2

This could happen if the user for example mistypes an ID. sda-admin reports no error but the services starts to malfunction, and reports this on the following requests:

Error: failed to [...], reason: failed to send request, reason: Post "http://localhost:8090/file/accession": EOF

Steps to reproduce
sda-admin dataset create --dataset-id dataset1 non-existing-id1 bad-id2

A/C

• Validate the files before mapping (if the file id exist) in sda-api

Expected behavior

Additional context
From #1026.

Estimation of size: medium

Estimation of priority: medium