data-dwh-dbt-project/models/intermediate/core/schema.yaml

version: 2

models:
  - name: int_core__duplicate_bookings
    description: |
      A list of bookings which are considered duplicates of other bookings.

      We currently consider two bookings to be duplicate if they have the same:
        - Guest user id
        - Accomodation id
        - Check-in date

      Bear in mind these bookings do have different booking ids.

      Out of a duplicated tuple of 2 or more bookings:
        - Our logic will consider the oldest one to be the "original", not duplicate one.
        - This table will contain only the duplicates, and not the original.
    columns:
      - name: id_booking
        data_type: bigint
        description: The unique, Superhog generated id for this booking.

      - name: is_duplicate_booking
        data_type: boolean
        description: |
          True if the booking is duplicate.

          If you are thinking that this is redundant, you are right. All 
          records in this table will be true. But we keep this field to 
          make your life easier when joining with other tables.

      - name: is_duplicating_booking_with_id
        data_type: bigint
        description: |

          Indicates what's the original booking being duplicated.

          If there is a tuple of duplicate bookings {A, B, C}, where A is the 
          original and the others are the duplicates:
            - B and C will appear in this table, A will not.
            - The value of this field for both B and C will be A's id.
  
  - name: int_core__booking_charge_events
    description: |
      
      Booking charge events is a fancy word for saying: a booking happened, 
      the related host had a booking fee set up at the right time, hence we 
      need to charge him.

      The table contains one record per booking and shows the associated
      booking fee, as well as the point in time in which the charge event was
      considered.

      Be wary of the booking fees: they don't have an associated currency. 
      Crazy, I know, but we currently don't store that information in the 
      backend.

      As for the charge dates: the exact point in time at which we consider 
      that we  should be charging a fee depends on billing details of the host
      customer. For some bookings, this will be the check-in. For others, its
      when the guest begins the verification process.

      Not all bookings appear here since we don't charge a fee for all 
      bookings.

    columns:
      - name: id_booking
        data_type: bigint
        description: The unique, Superhog generated id for this booking.

      - name: id_price_plan
        data_type: bigint
        description: The id of the price plan that relates to this booking.

      - name: booking_fee_local
        data_type: numeric
        description: The fee to apply to the booking, in host currency.

      - name: booking_fee_charge_at_utc
        data_type: timestamp without time zone
        description: |
          The point in time in which the booking should be invoiced.

          This could be the check-in date of the booking or the date in which the guest verification
          started, depending on the billing settings of the host.

      - name: booking_fee_charge_date_utc
        data_type: date
        description: |
          The date in which the booking should be invoiced.

          This could be the check-in date of the booking or the date in which the guest verification
          started, depending on the billing settings of the host.
  
  - name: int_core__check_in_cover_prices
    description: |
      
      This table shows the active price and cover for the Check-In Hero 
      product.

      The prices are obtained through a gross `GROUP BY` thrown at the payment
      validation sets table. It works this way because the price settings of 
      this product were done with a terrible backend data model design.

      How could the prices be changed remains a mystery, and the current design
      does not support any kind of history tracking. When the time comes to 
      adjust prices, we will have a lot of careful work to do to make sure that
      we keep history and that no downstream dependencies of this model blow 
      up.

    columns:
      - name: local_currency_iso_4217
        data_type: character varying
        description: A currency code.

      - name: checkin_cover_guest_fee_local_curr
        data_type: numeric
        description: | 
          The fee that the guest user must pay if he wants to purchase the 
          cover.

      - name: checkin_cover_cover_amount_local_curr
        data_type: numeric
        description: |
          The amount for which the guest user is covered if he faces problems
          during check-in.