When flag `USE_DEDUPLICATION_MODEL` is turn on, use this logic.

Validate request

1. 1. Validate mandatory query params
      
      1. first_name
      2. last_name
      3. birth_date
   2. Validate optional query params
      1. second_name
      2. tax_id
      3. birth_certificate
      4. phone_number (phone_nu,ber or auth_number)

Search existing MPI entity

Search only active persons - MPI.persons.is_active=true.

We are looking for such parameters:

1. tax_id and phone_number
2. tax_id and last_name( by MetaPhoneRu)
3. birth_certificate and phone_number
4. birth_certificate and last_name( by MetaPhoneRu)
5. phone_number and last_name( by MetaPhoneRu)
6. birth_date and last_name( by MetaPhoneRu)  and first_name( by MetaPhoneRu)

After the search we get the clusters (maximum 6) and display them one by one - persons from the first cluster, then from the second, etc.

And within each cluster we sort:

1. last_name_distance == 0
2. last_name_distance + first_name_distance + tax_id_distance (Ascending)

Return empty array if no data found

Return only requested params (if equal tax_id, birth_certificate, birth_date), last_name(if match by metaphone ) id, first_name, second_name, birth_place and merged_ids if data found.

Tests - https://e-health-ua.atlassian.net/wiki/x/MgHfIg