class Listing < ApplicationRecord
  # We used to use both "classified listing" and "listing" throughout the app.
  # We standardized on the latter, but keeping the table name was easier.
  self.table_name = "classified_listings"

  include PgSearch::Model

  attr_accessor :action

  # NOTE: categories were hardcoded at first and the model was only added later.
  # The foreign_key and inverse_of options are used because of legacy table names.
  belongs_to :listing_category, inverse_of: :listings, foreign_key: :classified_listing_category_id
  belongs_to :user
  belongs_to :organization, optional: true
  before_validation :modify_inputs
  before_save :evaluate_markdown
  before_create :create_slug
  acts_as_taggable_on :tags
  has_many :credits, as: :purchase, inverse_of: :purchase, dependent: :nullify

  validates :user_id, presence: true
  validates :organization_id, presence: true, unless: :user_id?

  validates :title, presence: true, length: { maximum: 128 }
  validates :body_markdown, presence: true, length: { maximum: 400 }
  validates :location, length: { maximum: 32 }
  validate :restrict_markdown_input
  validate :validate_tags

  pg_search_scope :search_listings,
                  against: %i[body_markdown cached_tag_list location slug title],
                  using: { tsearch: { prefix: true } }

  scope :published, -> { where(published: true) }

  # NOTE: we still need to use the old column name for the join query
  scope :in_category, lambda { |slug|
    joins(:listing_category).where("classified_listing_categories.slug" => slug)
  }

  delegate :cost, to: :listing_category

  # Wrapping the column accessor names for consistency. Aliasing did not work.
  def listing_category_id
    classified_listing_category_id
  end

  def listing_category_id=(id)
    self.classified_listing_category_id = id
  end

  def category
    listing_category&.slug
  end

  def author
    organization || user
  end

  def path
    "/listings/#{category}/#{slug}"
  end

  def natural_expiration_date
    (bumped_at || created_at) + 30.days
  end

  private

  def evaluate_markdown
    self.processed_html = MarkdownProcessor::Parser.new(body_markdown).evaluate_listings_markdown
  end

  def modify_inputs
    temp_tags = tag_list
    self.tag_list = [] # overwrite any existing tag with those from the front matter
    tag_list.add(temp_tags, parser: ActsAsTaggableOn::TagParser)
    self.body_markdown = body_markdown.to_s.gsub(/\r\n/, "\n")
  end

  def restrict_markdown_input
    markdown_string = body_markdown.to_s
    if markdown_string.scan(/(?=\n)/).count > 12
      errors.add(:body_markdown,
                 "has too many linebreaks. No more than 12 allowed.")
    end
    errors.add(:body_markdown, "is not allowed to include images.") if markdown_string.include?("![")
    errors.add(:body_markdown, "is not allowed to include liquid tags.") if markdown_string.include?("{% ")
  end

  def validate_tags
    errors.add(:tag_list, "exceed the maximum of 8 tags") if tag_list.length > 8
  end

  def create_slug
    self.slug = "#{title.downcase.parameterize.delete('_')}-#{rand(100_000).to_s(26)}"
  end
end