removing duplicates from guesty and removing tests for edeposit_users

models/intermediate/edeposit/int_edeposit__guesty_verifications.sql

@@ -4,14 +4,23 @@
 with
     int_edeposit__verifications as (
         select * from {{ ref("int_edeposit__verifications") }}
+    ),
+    -- CTE to rank verifications by updated_at_utc per id_booking
+    ranked_verifications as (
+        select
+            v.*,
+            row_number() over (
+                partition by v.id_booking order by v.updated_at_utc desc
+            ) as rn
+        from int_edeposit__verifications v
+        where v.version = 'V1' and v.id_booking is not null
     )
 select
     v.id_verification,
     v.id_booking,
     v.verification_status,
     v.is_cancelled,
-    -- when number_nights = 0 (booking's checkin and checkout are on the same day)
-    -- it's charged for just 1 night
+    -- Charge for 1 night if number_nights = 0
     case
         when v.number_nights = 0 and v.verification_status = '{{ ok_status }}'
         then {{ cost_per_night }}
@@ -21,5 +30,7 @@ select
     end as ok_status_fee_in_gbp,
     v.created_date_utc,
     v.checkout_date_utc
-from int_edeposit__verifications v
-where v.version = 'V1' and v.id_booking is not null
+from ranked_verifications v
+where
+    -- Select only the most recent verification for each id_booking
+    v.rn = 1

models/intermediate/edeposit/schema.yml

@@ -319,10 +319,11 @@ models:
       - name: id_booking
         data_type: text
         description: "unique Superhog generated id for a booking.
-          note that this could be duplicated and both will be charged,
-          it's up to the user to not generate or cancel duplicate verifications"
+          note that there might be duplicate bookings on the original data
+          but we remove them keeping only the verification with the most recent update."
         tests:
           - not_null
+          - unique
 
       - name: verification_status
         data_type: text