Add duplicate claims check

We're soon going to want to generate a report for the ops team of the
duplicate claims. This commit introduces a new claim verifier that
records the duplicate claims, if any.
Once existing duplicates on production have been back filled we can
replace other uses of `MatchingAttributeFinder#matching_claims` with
`claim.duplicates` and also join claims to their duplicates when
generating the duplicate claims report.

app/models/automated_checks/claim_verifiers/duplicate_claims_check.rb

@@ -0,0 +1,37 @@
+module AutomatedChecks
+  module ClaimVerifiers
+    class DuplicateClaimsCheck
+      def initialize(claim:)
+        @claim = claim
+      end
+
+      def perform
+        matching_attribute_finder.matching_claims.each do |existing_claim|
+          if existing_claim.created_at < claim.created_at
+            original_claim = existing_claim
+            duplicate_claim = claim
+          else
+            original_claim = claim
+            duplicate_claim = existing_claim
+          end
+
+          unless Claims::ClaimDuplicate.exists?(original_claim: original_claim, duplicate_claim: duplicate_claim)
+            Claims::ClaimDuplicate.create!(
+              original_claim: original_claim,
+              duplicate_claim: duplicate_claim,
+              matching_attributes: matching_attribute_finder.matching_attributes(existing_claim)
+            )
+          end
+        end
+      end
+
+      private
+
+      attr_reader :claim
+
+      def matching_attribute_finder
+        @matching_attribute_finder ||= Claim::MatchingAttributeFinder.new(claim)
+      end
+    end
+  end
+end

app/models/claim.rb

@@ -142,6 +142,26 @@ class Claim < ApplicationRecord
     inverse_of: :assigned_claims,
     optional: true
 
+  has_many :claim_duplicates_as_original_claim,
+    class_name: "Claims::ClaimDuplicate",
+    foreign_key: :original_claim_id,
+    dependent: :destroy
+
+  has_many :claim_duplicates_as_duplicate_claim,
+    class_name: "Claims::ClaimDuplicate",
+    foreign_key: :duplicate_claim_id,
+    dependent: :destroy
+
+  has_many :duplicates,
+    through: :claim_duplicates_as_original_claim,
+    source: :duplicate_claim,
+    class_name: "Claim"
+
+  has_many :originals,
+    through: :claim_duplicates_as_duplicate_claim,
+    source: :original_claim,
+    class_name: "Claim"
+
   enum :payroll_gender, {
     dont_know: 0,
     female: 1,

app/models/claims.rb

@@ -0,0 +1,5 @@
+module Claims
+  def self.table_name_prefix
+    "claims_"
+  end
+end

app/models/claims/claim_duplicate.rb

@@ -0,0 +1,28 @@
+module Claims
+  class ClaimDuplicate < ApplicationRecord
+    belongs_to :original_claim, class_name: "Claim"
+    belongs_to :duplicate_claim, class_name: "Claim"
+
+    validates :duplicate_claim, uniqueness: {
+      scope: :original_claim,
+      message: "has already been registered as a duplicate"
+    }
+    validate :claims_are_not_the_same, if: -> { original_claim && duplicate_claim }
+    validate :original_claim_is_older, if: -> { original_claim && duplicate_claim }
+    validates :matching_attributes, presence: true
+
+    private
+
+    def claims_are_not_the_same
+      return unless original_claim == duplicate_claim
+
+      errors.add(:duplicate_claim, "can't be the same as the original claim")
+    end
+
+    def original_claim_is_older
+      return unless original_claim.created_at > duplicate_claim.created_at
+
+      errors.add(:original_claim, "must be older than the duplicate claim")
+    end
+  end
+end

app/models/policies/early_career_payments.rb

@@ -20,7 +20,8 @@ module EarlyCareerPayments
       AutomatedChecks::ClaimVerifiers::CensusSubjectsTaught,
       AutomatedChecks::ClaimVerifiers::Employment,
       AutomatedChecks::ClaimVerifiers::StudentLoanPlan,
-      AutomatedChecks::ClaimVerifiers::FraudRisk
+      AutomatedChecks::ClaimVerifiers::FraudRisk,
+      AutomatedChecks::ClaimVerifiers::DuplicateClaimsCheck
     ].freeze
 
     POLICY_START_YEAR = AcademicYear.new(2021).freeze

app/models/policies/early_years_payments.rb

@@ -18,7 +18,8 @@ module EarlyYearsPayments
 
     VERIFIERS = [
       AutomatedChecks::ClaimVerifiers::StudentLoanPlan,
-      AutomatedChecks::ClaimVerifiers::EarlyYearsPayments::Identity
+      AutomatedChecks::ClaimVerifiers::EarlyYearsPayments::Identity,
+      AutomatedChecks::ClaimVerifiers::DuplicateClaimsCheck
     ]
 
     # Attributes to delete from claims submitted before the current academic

app/models/policies/further_education_payments.rb

@@ -22,7 +22,8 @@ module FurtherEducationPayments
       AutomatedChecks::ClaimVerifiers::ProviderVerification,
       AutomatedChecks::ClaimVerifiers::Employment,
       AutomatedChecks::ClaimVerifiers::StudentLoanPlan,
-      AutomatedChecks::ClaimVerifiers::FraudRisk
+      AutomatedChecks::ClaimVerifiers::FraudRisk,
+      AutomatedChecks::ClaimVerifiers::DuplicateClaimsCheck
     ]
 
     # Options shown to admins when rejecting a claim

app/models/policies/international_relocation_payments.rb

@@ -4,7 +4,8 @@ module InternationalRelocationPayments
     extend self
 
     VERIFIERS = [
-      AutomatedChecks::ClaimVerifiers::FraudRisk
+      AutomatedChecks::ClaimVerifiers::FraudRisk,
+      AutomatedChecks::ClaimVerifiers::DuplicateClaimsCheck
     ].freeze
 
     ELIGIBILITY_MATCHING_ATTRIBUTES = [["passport_number"]].freeze

app/models/policies/levelling_up_premium_payments.rb

@@ -10,7 +10,8 @@ module LevellingUpPremiumPayments
       AutomatedChecks::ClaimVerifiers::CensusSubjectsTaught,
       AutomatedChecks::ClaimVerifiers::Employment,
       AutomatedChecks::ClaimVerifiers::StudentLoanPlan,
-      AutomatedChecks::ClaimVerifiers::FraudRisk
+      AutomatedChecks::ClaimVerifiers::FraudRisk,
+      AutomatedChecks::ClaimVerifiers::DuplicateClaimsCheck
     ].freeze
 
     # Used in

app/models/policies/student_loans.rb

@@ -19,7 +19,8 @@ module StudentLoans
       AutomatedChecks::ClaimVerifiers::CensusSubjectsTaught,
       AutomatedChecks::ClaimVerifiers::Employment,
       AutomatedChecks::ClaimVerifiers::StudentLoanAmount,
-      AutomatedChecks::ClaimVerifiers::FraudRisk
+      AutomatedChecks::ClaimVerifiers::FraudRisk,
+      AutomatedChecks::ClaimVerifiers::DuplicateClaimsCheck
     ].freeze
 
     POLICY_START_YEAR = AcademicYear.new(2013).freeze

db/migrate/20241219115223_create_claims_claim_duplicates.rb

@@ -0,0 +1,33 @@
+class CreateClaimsClaimDuplicates < ActiveRecord::Migration[8.0]
+  def change
+    create_table :claims_claim_duplicates, id: :uuid do |t|
+      t.belongs_to(
+        :original_claim,
+        null: false,
+        foreign_key: {
+          to_table: :claims
+        },
+        type: :uuid
+      )
+
+      t.belongs_to(
+        :duplicate_claim,
+        null: false,
+        foreign_key: {
+          to_table: :claims
+        },
+        type: :uuid
+      )
+
+      t.jsonb :matching_attributes, default: []
+
+      t.timestamps
+    end
+
+    add_index(
+      :claims_claim_duplicates,
+      [:original_claim_id, :duplicate_claim_id],
+      unique: true
+    )
+  end
+end

db/schema.rb

@@ -10,7 +10,7 @@
 #
 # It's strongly recommended that you check this file into your version control system.
 
-ActiveRecord::Schema[8.0].define(version: 2024_11_26_105650) do
+ActiveRecord::Schema[8.0].define(version: 2024_12_19_115223) do
   # These are extensions that must be enabled in order to support this database
   enable_extension "citext"
   enable_extension "pg_catalog.plpgsql"
@@ -123,6 +123,17 @@
     t.index ["submitted_at"], name: "index_claims_on_submitted_at"
   end
 
+  create_table "claims_claim_duplicates", id: :uuid, default: -> { "gen_random_uuid()" }, force: :cascade do |t|
+    t.uuid "original_claim_id", null: false
+    t.uuid "duplicate_claim_id", null: false
+    t.jsonb "matching_attributes", default: []
+    t.datetime "created_at", null: false
+    t.datetime "updated_at", null: false
+    t.index ["duplicate_claim_id"], name: "index_claims_claim_duplicates_on_duplicate_claim_id"
+    t.index ["original_claim_id", "duplicate_claim_id"], name: "idx_on_original_claim_id_duplicate_claim_id_5ce2c01567", unique: true
+    t.index ["original_claim_id"], name: "index_claims_claim_duplicates_on_original_claim_id"
+  end
+
   create_table "decisions", id: :uuid, default: -> { "gen_random_uuid()" }, force: :cascade do |t|
     t.integer "result"
     t.uuid "claim_id"
@@ -585,6 +596,8 @@
   add_foreign_key "claim_payments", "claims"
   add_foreign_key "claim_payments", "payments"
   add_foreign_key "claims", "journeys_sessions"
+  add_foreign_key "claims_claim_duplicates", "claims", column: "duplicate_claim_id"
+  add_foreign_key "claims_claim_duplicates", "claims", column: "original_claim_id"
   add_foreign_key "decisions", "dfe_sign_in_users", column: "created_by_id"
   add_foreign_key "early_career_payments_eligibilities", "schools", column: "current_school_id"
   add_foreign_key "eligible_ey_providers", "local_authorities"

spec/models/automated_checks/claim_verifiers/duplicate_claims_check_spec.rb

@@ -0,0 +1,82 @@
+require "rails_helper"
+
+RSpec.describe AutomatedChecks::ClaimVerifiers::DuplicateClaimsCheck do
+  describe "#perform" do
+    context "when the claim has duplicates" do
+      it "marks the claim as having duplicates" do
+        existing_claim = create(
+          :claim,
+          :submitted,
+          email_address: "[email protected]",
+          first_name: "one"
+        )
+
+        new_claim = create(
+          :claim,
+          :submitted,
+          email_address: "[email protected]",
+          first_name: "two"
+        )
+
+        described_class.new(claim: new_claim).perform
+
+        expect(existing_claim.reload.duplicates).to include(new_claim)
+        expect(new_claim.reload.originals).to include(existing_claim)
+
+        claim_duplicate = new_claim.claim_duplicates_as_duplicate_claim.last
+
+        expect(claim_duplicate.original_claim).to eq(existing_claim)
+        expect(claim_duplicate.duplicate_claim).to eq(new_claim)
+        expect(claim_duplicate.matching_attributes).to eq(["email_address"])
+      end
+
+      it "is idempotent" do
+        existing_claim = create(
+          :claim,
+          :submitted,
+          email_address: "[email protected]",
+          first_name: "one"
+        )
+
+        new_claim = create(
+          :claim,
+          :submitted,
+          email_address: "[email protected]",
+          first_name: "two"
+        )
+
+        Claims::ClaimDuplicate.create!(
+          original_claim: existing_claim,
+          duplicate_claim: new_claim,
+          matching_attributes: ["email_address"]
+        )
+
+        expect { described_class.new(claim: new_claim).perform }.not_to(
+          change(existing_claim.duplicates, :count)
+        )
+      end
+    end
+
+    context "when the claim has no duplicates" do
+      it "does not mark the claim as having duplicates" do
+        existing_claim = create(
+          :claim,
+          :submitted,
+          email_address: "[email protected]",
+          first_name: "one"
+        )
+
+        new_claim = create(
+          :claim,
+          :submitted,
+          email_address: "[email protected]",
+          first_name: "two"
+        )
+
+        described_class.new(claim: new_claim).perform
+
+        expect(existing_claim.reload.duplicates).not_to include(new_claim)
+      end
+    end
+  end
+end