ruby code to direct all users to a particular location instead of multiple locations and delete duplicates - mysql

Question

I am having a User model and a Location Model. Each user belongs to a particular location in the Location model.
I am having duplicate locations in the Location table.
and User belongs to Location.
how can i remove duplicate rows in the location table and keep one row and make all users belong to that single row using ruby. Both the tables are connected through location_ID attribute.
I tried to do this through migration:
def dedupe(model, *key_attrs)
model.select(key_attrs).group(key_attrs).having('count(*) > 1').each { |duplicates|
dup_rows = model.where(duplicates.attributes.slice(key_attrs)).to_a
# the first one we want to keep right?
first_one = dup_rows.shift #stored the first one
dup_rows.each{ |double| double.destroy } # duplicates can now be destroyed
}
end
But there is foreign key constraint of User not letting the migration to run. How can I achieve this?
Current Models are :
User
user_id name location_id
1 tim 1
2 adam 2
3 Joy 3
Location
location_id name
1 NewYork
2 NewYork
3 NewYork
Expected Ouput:
User
user_id name location_id
1 tim 1
2 adam 1
3 Joy 1
Location
location_id name
1 NewYork

Answer 1

Kinda ugly, but you can use a subquery:
First, grab the first occurrence of all records which are duplicates;
original_duplicate_locations = Location.select("MIN(id) AS id, name, user_id").group(:name, :user_id).having("COUNT(id) > 1")
The extra duplicates are defined as locations having the same name, and user_id but not the same id:
duplicates_not_including_originals = Location.joins("JOIN (#{duplicates.to_sql}) dupes ON locations.name = dupes.name AND locations.user_id = dupes.user_id AND locations.id <> dupes.id")

Answer 2

Hey you can try this way:
1)First Update all entries with location first entry in location table using
User.joins(:location).update_all("location_id = select id from locations as l2 where l2.name = locations.name limit 1")
Note: you can also use order by id here if sub query not return first entry from table.
2)Destroy All entries from location table excluding first entry-
Before this make sure all your data get updated with first entry in location table properly means first id of repeated location is updated or not. because after deletion it is not possible to recover your data again. then just destroy all your repeated entries excluding first entry using
Location.where("id not in (?)", Location.select("min(id) as id").group("name").map(&:id)).destroy_all