使用Ruby on Rails导入CSV

Question

I am want to Import an Outlook CSV file and check for duplicates based on Email (child) ideally before creating or saving any of the models (either the Contact - parent or Email- Child)

Steps (how it currently works but is a flawed solution )

1. Import File - I save the file
2. Parse each Row to fields
3. Perform Checks for uniqueness of email address (which I am saving email record to do this before saving parent record- Contact).

*ISSUES:
4. SO at the point I save Email address record, it is missing the Contact Id/ owner- potentially creating Orphan situation if the process fails before Contact is created) as it is saving the child(Email) before the Contact is saved (and I don't think or want to be doing this)

5. However, if I save the Contact just based on name (eg David Smith), I may have:

   • and check based on name - there are scenarios where I know 2 people with same name (eg David Smith and will then be appending 2 different people together)
   • If I save all Contacts (and then check Email uniqueness), I will have created a lot of extra Contacts.

6. As it currently works, the check for duplicate Email is on my entire database because I don't have the contact_id (to associated with the user_id aka owner_id)

7. I tried saving the Contact first but then realized this is causing me to have a lot of extra records (very messy).

Here is my code to initial process the row

---

  def process_row(smart_row)
    new_contact, existing_records = smart_row.to_contact

    self.contact = ContactMergingService.new(csv_file.user, new_contact, existing_records).perform
    log_processed_contacts new_contact
    init_contact_info self.contact
    self.contact.required_salutations_to_set = true # will be used for envelope/letter saluation
    if contact.first_name || contact.last_name || contact.email_addresses.first || contact.phone_numbers.first
      self.contact.save!
      csv_file.increment!(:total_imported_records)
    end
  end

---

This is the first method called above (to save Email before saving Contact)

---

   def to_contact
      existing_emails = existing_phone_numbers = nil
      contact = Contact.new.tap do |contact|
        initiate_instance(contact, CONTACT_MAPPING)
        address = initiate_instance(Address.new, ADDRESS_MAPPING)
        contact.addresses << address if address
        email_addresses, existing_emails = initialize_emails(EMAIL_ADDRESS_FIELDS)
        contact.email_addresses << email_addresses
        phone_numbers, existing_phone_numbers = initialize_phone_numbers(PHONE_TYPE_MAPPINGS)
        contact.phone_numbers << phone_numbers
        contact
      end
      existing_records = []
      existing_records << existing_emails
      existing_records << existing_phone_numbers
      existing_records.flatten!
      existing_records.compact!
      [contact,  existing_records]
    end

Here is my code when I save the Email Address (after this I then am saving the Contact)

---

def initialize_emails email_fields
  email_addresses = []
  email_fields.each do |field|
    value = evaluate_value field
    if value.present?
      new_email = EmailAddress.find_or_create_by(email: value, primary: (primary_email_field?(field)))
      if new_email.save
        email_addresses << new_email
      end
    end
  end
  existing_emails = email_addresses.select{ |email_address| email_address.owner_id.present?}
  [email_addresses, existing_emails]
end

---

I have 3 models:

User (has many)
  has_many :contacts
  has_many :email_campaigns  has_many :email_messages

Contacts:  First Name and Last Name
  belongs_to :user, counter_cache: true
  has_many :addresses, as: :owner, dependent: :destroy
  has_many :phone_numbers, as: :owner, dependent: :destroy
  has_many :email_addresses, as: :owner, dependent: :destroy
  accepts_nested_attributes_for :email_addresses, allow_destroy: true
``
Email:  email_address - polymorphic
  belongs_to :owner, polymorphic: true, touch: true

So my questions is:

• Do i need to save the records (either contact or email) to be able to do a check for duplicates?
• Is there a way I can process the CSV file in such a way that I can check for duplicates based on email_address before creating either the Contact record or the Email_address record? I want to check for duplicates against my existing database and the other records in the file based on contact.first_name, contact.last_name, email.address

Any thoughts? many thanks. Annie

Answer 1

Do i need to save the records (either contact or email) to be able to do a check for duplicates?

No, just store them in a variable

Is there a way I can process the CSV file in such a way that I can check for duplicates based on email_address before creating either the Contact record or the Email_address record? I want to check for duplicates against my existing database and the other records in the file based on contact.first_name, contact.last_name, email.address

find_or_create_by is meant to do specifically this. From the docs:

Finds the first record with the given attributes, or creates a record with the attributes if one is not found

That is, you can pass an email address and the method will either find a record with that value or create a new one.