sul-dlss/exhibits

# frozen_string_literal: true

require 'digest'
require 'faraday'
require_relative 'extraction'

module Traject
  module Macros
    # IIIF canvas extraction
    module Canvas
      def extract_parent_manifest_iiif_id
        lambda do |_record, accumulator, context|
          return if context.output_hash['related_document_id_ssim'].blank?

          accumulator << format(Settings.purl.iiif_manifest_url,
                                druid: context.output_hash['related_document_id_ssim'].first.to_s)
        end
      end

      def extract_canvas_id
        lambda do |record, accumulator, _context|
          accumulator << "canvas-#{Digest::MD5.hexdigest(record['@id'].to_s)}"
        end
      end

      def extract_canvas_iiif_id
        lambda do |record, accumulator, _context|
          accumulator << record['@id'].to_s
        end
      end

      ##
      # Note: This method assumes an "enhanced" canvas with additional properties
      # added beyond the IIIF Canvas model.
      def extract_canvas_label
        lambda do |record, accumulator, _context|
          labels = [record['label'].to_s, record['manifest_label'].to_s].reject(&:empty?)
          accumulator << labels.join(': ')
        end
      end

      ##
      # Note: This method assumes an "enhanced" canvas with additional properties
      # added beyond the IIIF Canvas model.
      def extract_canvas_label_sort
        lambda do |record, accumulator, _context|
          labels = [record['manifest_label'].to_s, record['label'].to_s].reject(&:empty?)
          accumulator << labels.join(': ')
        end
      end

      def extract_canvas_parent_manuscript_number
        lambda do |record, accumulator, _context|
          return if record['parent_manuscript_number'].blank?

          accumulator.push(*record['parent_manuscript_number'].map(&:to_s))
        end
      end

      ##
      # Note: This method assumes an "enhanced" canvas with additional properties
      # added beyond the IIIF Canvas model.
      def extract_canvas_range_labels
        lambda do |record, accumulator, _context|
          return if record['range_labels'].blank?

          accumulator.push(*record['range_labels'].map(&:to_s))
        end
      end

      def extract_canvas_annotation_list_urls
        lambda do |record, accumulator, _context|
          return if record['otherContent'].blank?

          record['otherContent'].each do |link|
            next unless link['@type'] == 'sc:AnnotationList'

            accumulator << link['@id'].to_s
          end
        end
      end

      def extract_canvas_annotations
        lambda do |record, accumulator, _context|
          return if record['otherContent'].blank?

          record['otherContent'].each do |link|
            next unless link['@type'] == 'sc:AnnotationList'

            extract_annotations_from_list(accumulator, link['@id'].to_s)
          end
        end
      end

      # Druids are kept as part of the canvas-id
      def extract_canvas_related_document_ids
        lambda do |record, accumulator, _context|
          match = record['@id'][Exhibits::Application.config.druid_regex]
          return if match.blank?

          accumulator << match
        end
      end

      private

      def extract_annotations_from_list(accumulator, url)
        annotation_list = JSON.parse(Faraday.get(url).body)
        return unless annotation_list['@type'] == 'sc:AnnotationList' && annotation_list['resources']

        annotation_list['resources'].each do |resource|
          next unless resource['@type'] == 'oa:Annotation'
          next if resource['resource']['chars'].blank?

          accumulator << resource['resource']['chars'].to_s
        end
      end
    end
  end
end