HYC-1934 - Send download events to database (#1108)

* add hycdownloadstat model and migration

* update schema after running HycDownloadStat migration

* Update DownloadAnalyticsBehavior to push stats to local db

* logs, factorybot and logs in create_download_stat

* changed phrasing, add new test

* test for download tracking within database, rubocop

* test to verify download count is incremented for existing table entries

* removing repeated mocking in download controller tests and adding a function to return fileset ids

* removing some duplicate code, tests for new helper function

* specifying uid to avoid conflicts

* helper function tests, removing tests with duplicate coverage

* pr changes

app/controllers/concerns/hyc/download_analytics_behavior.rb

@@ -12,7 +12,7 @@ def track_download
           return
         end
         if Hyrax::Analytics.config.auth_token.present? && !request.url.match('thumbnail')
-          Rails.logger.debug("Recording download event for #{params[:id]}")
+          Rails.logger.debug("Recording download event for #{fileset_id}")
           medium = request.referrer.present? ? 'referral' : 'direct'
 
           client_ip = request.remote_ip
@@ -34,7 +34,7 @@ def track_download
             apiv: '1',
             e_a: 'DownloadIR',
             e_c: @admin_set_name,
-            e_n: params[:id] || 'Unknown',
+            e_n: fileset_id,
             e_v: medium,
             _id: client_id,
             cip: client_ip,
@@ -50,18 +50,53 @@ def track_download
           if response.code >= 300
             Rails.logger.error("DownloadAnalyticsBehavior received an error response #{response.code} for matomo query: #{uri}")
           end
+          # Send download events to db
+          create_download_stat
           Rails.logger.debug("DownloadAnalyticsBehavior request completed #{response.code}")
           response.code
         end
       end
 
+      def create_download_stat
+        record_id_value = record_id
+        admin_set_id_value = admin_set_id
+        work_type = fetch_record.dig(0, 'has_model_ssim', 0)
+        date = Date.today
+
+        Rails.logger.debug('Creating or updating hyc-download-stat database entry with the following attributes:')
+        Rails.logger.debug("fileset_id: #{fileset_id}, work_id: #{record_id_value}, admin_set_id: #{admin_set_id_value}, work_type: #{work_type}, date: #{date.beginning_of_month}")
+
+        stat = HycDownloadStat.find_or_initialize_by(
+          fileset_id: fileset_id,
+          work_id: record_id_value,
+          admin_set_id: admin_set_id_value,
+          work_type: work_type,
+          date: date.beginning_of_month
+        )
+        stat.download_count += 1
+        if stat.save
+          Rails.logger.debug("Database entry for fileset_id: #{fileset_id} successfully saved with download count: #{stat.download_count}.")
+        else
+          Rails.logger.error("Failed to update database entry for fileset_id: #{fileset_id}." \
+                             "Errors: #{stat.errors.full_messages}")
+        end
+      end
+
       def bot_request?(user_agent)
         browser = Browser.new(user_agent)
         browser.bot?
       end
 
       def fetch_record
-        @record ||= ActiveFedora::SolrService.get("file_set_ids_ssim:#{params[:id]}", rows: 1)['response']['docs']
+        @record ||= ActiveFedora::SolrService.get("file_set_ids_ssim:#{fileset_id}", rows: 1)['response']['docs']
+      end
+
+      def fetch_admin_set
+        @admin_set ||= ActiveFedora::SolrService.get("title_tesim:#{@admin_set_name}", rows: 1)['response']['docs']
+      end
+
+      def admin_set_id
+        @admin_set_id ||= fetch_admin_set.dig(0, 'id')
       end
 
       def record_id
@@ -72,6 +107,10 @@ def record_id
                        end
       end
 
+      def fileset_id
+        @fileset_id ||= params[:id] || 'Unknown'
+      end
+
       def record_title
         @record_title ||= if !fetch_record.blank? && fetch_record[0]['title_tesim']
                             fetch_record[0]['title_tesim'].first

app/models/hyc_download_stat.rb

@@ -0,0 +1,12 @@
+# frozen_string_literal: true
+class HycDownloadStat < ApplicationRecord
+  scope :with_fileset_id_and_date, ->(fileset_id, start_date, end_date) { where(fileset_id: fileset_id, date: start_date..end_date) }
+  scope :with_work_id_and_date, ->(work_id, start_date, end_date) { where(work_id: work_id, date: start_date..end_date) }
+  scope :with_admin_set_id, ->(admin_set_id) { where(admin_set_id: admin_set_id) }
+  scope :with_work_type, ->(work_type) { where(work_type: work_type) }
+
+    # Additional scopes for flexibility
+  scope :with_fileset_id, ->(fileset_id) { where(fileset_id: fileset_id) }
+  scope :with_work_id, ->(work_id) { where(work_id: work_id) }
+  scope :within_date_range, ->(start_date, end_date) { where(date: start_date..end_date) }
+  end

db/migrate/20240714234200_create_hyc_download_stats.rb

@@ -0,0 +1,20 @@
+# frozen_string_literal: true
+class CreateHycDownloadStats < ActiveRecord::Migration[6.1]
+  def change
+    create_table :hyc_download_stats do |t|
+      t.string :fileset_id, null: false
+      t.string :work_id, null: false
+      t.string :admin_set_id, null: false
+      t.string :work_type, null: false
+      t.date :date, null: false
+      t.integer :download_count, default: 0, null: false
+
+      t.timestamps
+    end
+
+    add_index :hyc_download_stats, [:fileset_id, :date]
+    add_index :hyc_download_stats, [:work_id, :date]
+    add_index :hyc_download_stats, :admin_set_id
+    add_index :hyc_download_stats, :work_type
+  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.define(version: 2023_06_08_153601) do
+ActiveRecord::Schema.define(version: 2024_07_14_234200) do
 
   # These are extensions that must be enabled in order to support this database
   enable_extension "plpgsql"
@@ -249,6 +249,21 @@
     t.index ["user_id"], name: "index_file_view_stats_on_user_id"
   end
 
+  create_table "hyc_download_stats", force: :cascade do |t|
+    t.string "fileset_id", null: false
+    t.string "work_id", null: false
+    t.string "admin_set_id", null: false
+    t.string "work_type", null: false
+    t.date "date", null: false
+    t.integer "download_count", default: 0, null: false
+    t.datetime "created_at", precision: 6, null: false
+    t.datetime "updated_at", precision: 6, null: false
+    t.index ["admin_set_id"], name: "index_hyc_download_stats_on_admin_set_id"
+    t.index ["fileset_id", "date"], name: "index_hyc_download_stats_on_fileset_id_and_date"
+    t.index ["work_id", "date"], name: "index_hyc_download_stats_on_work_id_and_date"
+    t.index ["work_type"], name: "index_hyc_download_stats_on_work_type"
+  end
+
   create_table "hyrax_collection_types", id: :serial, force: :cascade do |t|
     t.string "title"
     t.text "description"