Add audit model

app/controllers/sites_controller.rb

@@ -4,7 +4,7 @@ class SitesController < ApplicationController
 
   # GET /sites
   def index
-    @sites = Site.all
+    @pagy, @sites = pagy Site.includes(:audit)
   end
 
   # GET /sites/1
@@ -18,7 +18,7 @@ def edit; end
 
   # POST /sites
   def create
-    @site = Site.new(site_params)
+    @site = Site.find_or_create_by_url(site_params)
     if @site.save
       redirect_to @site, notice: t(".notice")
     else
@@ -44,7 +44,7 @@ def destroy
   private
 
   def set_site
-    @site = params[:id].present? ? Site.friendly.find(params.expect(:id)) : Site.new
+    @site = params[:id].present? ? Site.friendly.find(params.expect(:id)) : Site.new_with_audit
   end
 
   def redirect_old_slugs

app/models/audit.rb

@@ -0,0 +1,32 @@
+class Audit < ApplicationRecord
+  MAX_ATTEMPTS = 3
+  MAX_RUNTIME = 1.hour.freeze
+
+  belongs_to :site, touch: true
+
+  validates :url, presence: true, url: true
+  normalizes :url, with: ->(url) { URI.parse(url).normalize.to_s }
+
+  enum :status, ["pending", "running", "passed", "retryable", "failed"].index_by(&:itself), validate: true, default: :pending
+
+  scope :sort_by_newest, -> { order(created_at: :desc) }
+  scope :sort_by_url, -> { order(Arel.sql("REGEXP_REPLACE(url, '^https?://(www\.)?', '') ASC")) }
+  scope :due, -> { pending.where("run_at <= now()") }
+  scope :past, -> { where(status: [:passed, :failed]) }
+  scope :scheduled, -> { where("run_at > now()") }
+  scope :retryable, -> { failed.where(attempts: ...MAX_ATTEMPTS) }
+  scope :to_run, -> { due.or(retryable) }
+  scope :clean, -> { passed.where(attempts: 0) }
+  scope :late, -> { pending.where("run_at <= ?", 1.hour.ago) }
+  scope :retried, -> { passed.where(attempts: 1..) }
+  scope :stalled, -> { running.where("run_at < ?", MAX_RUNTIME.ago) }
+  scope :crashed, -> { failed.where(attempts: MAX_ATTEMPTS) }
+
+  delegate :hostname, :path, to: :parsed_url
+
+  def run_at = super || Time.zone.now
+  def due? = pending? && run_at <= Time.zone.now
+  def runnable? = due? || retryable?
+  def parsed_url = @parsed_url ||= URI.parse(url).normalize
+  def url_without_scheme = @url_without_scheme ||= [hostname, path == "/" ? nil : path].compact.join(nil)
+end

app/models/site.rb

@@ -1,7 +1,49 @@
 class Site < ApplicationRecord
   extend FriendlyId
 
-  friendly_id :url, use: [:slugged, :history]
+  has_many :audits, dependent: :destroy
+  has_one_of_many :audit, -> { order("audits.created_at DESC") }, class_name: "Audit"
 
-  attribute :url, :string
+  friendly_id :url_without_scheme, use: [:slugged, :history]
+
+  delegate :url, :url_without_scheme, to: :audit
+
+  class << self
+    def find_or_create_by_url(attributes)
+      url = attributes.to_h.delete(:url)
+      attributes.to_h.delete(:name) if attributes[:name].blank?
+      # Ignore http/https duplicates when searching
+      normalized_url = [url, url.sub(/^https?/, url.start_with?("https") ? "http" : "https")]
+      joins(:audits).find_by(audits: { url: normalized_url })&.tap { it.update(attributes) } \
+        || create_with_audit(url:, **attributes)
+    end
+
+    def create_with_audit(url: nil, **attributes)
+      new_with_audit(url:, **attributes).tap(&:save)
+    end
+
+    def new_with_audit(url: nil, **attributes)
+      new(attributes).tap { |site| site.audits.build(url:) }
+    end
+  end
+
+  def new(attributes = nil)
+    return super if attributes.nil?
+
+    attributes = attributes.to_h.symbolize_keys
+
+    if url = attributes.delete(:url)
+      self.class.find_or_create_by(url:, **attributes)
+    else
+      super
+    end
+  end
+
+  def url=(new_url)
+    audit = audits.build(url: new_url)
+  end
+
+  def to_title = url_without_scheme
+  def audit = super || audits.last || audits.build
+  def should_generate_new_friendly_id? = new_record? || (audit && slug != url_without_scheme.parameterize)
 end

app/views/sites/index.html.erb

@@ -9,7 +9,7 @@
   <% t.with_body do %>
     <% @sites.each do |site| %>
       <tr>
-        <td><%= link_to site.name, site %></td>
+        <td><%= link_to site.url_without_scheme, site %></td>
         <td>
           <%= l site.updated_at, format: :compact %>
           <br>

db/migrate/20250213152147_create_audits.rb

@@ -0,0 +1,18 @@
+class CreateAudits < ActiveRecord::Migration[8.0]
+  def change
+    create_table :audits do |t|
+      t.belongs_to :site, null: false, foreign_key: true
+      t.string :url, null: false
+      t.string :status, null: false
+      t.integer :attempts, null: false, default: 0
+      t.datetime :run_at, null: false, default: -> { "CURRENT_TIMESTAMP" }
+
+      t.timestamps
+
+      t.index :url
+      t.index [:status, :run_at], name: "index_audits_on_status_and_run_at"
+      t.index "REGEXP_REPLACE(url, '^https?://(www\.)?', '')", name: "index_audits_on_normalized_url"
+      t.index :attempts, where: "status = 'failed' AND attempts > 0", name: "index_audits_on_retryable"
+    end
+  end
+end

spec/factories/audits.rb

@@ -0,0 +1,22 @@
+FactoryBot.define do
+  factory :audit do
+    site { association :site, url:, strategy: :build }
+    url { "https://example.com" }
+
+    trait :passed do
+      status { "passed" }
+    end
+
+    trait :failed do
+      status { "failed" }
+    end
+
+    trait :pending do
+      status { "pending" }
+    end
+
+    trait :running do
+      status { "running" }
+    end
+  end
+end

spec/factories/sites.rb

@@ -1,5 +1,5 @@
 FactoryBot.define do
   factory :site do
-    sequence(:slug) { |n| "www.example-#{n}.com" }
+    sequence(:url) { |n| "https://www.example-#{n}.com/" }
   end
 end