Automatic removal of stale Sphinx documents during Realtime Indexing

sphinx_indexing.rb

# Improve indexing to remove records whose sphinx document id is stale due to changes to the number/order of indices
module SphinxIndexing
  # Delete existing record before inserting new ones. This avoids the need to rebuild the index after adding or removing
  # an index. Without this, the old indexed record would not be deleted since Sphinx's calculation of that record's id
  # would not match now that the number/order of indices has changed.
  module Transcriber
    def copy(*instances)
      return unless instances.present?

      ids = instances.map(&:id).compact
      sphinxql = <<~SQL
        DELETE FROM #{@index.name} WHERE sphinx_internal_id IN (#{ids.join(', ')})
      SQL

      ThinkingSphinx::Logger.log :query, sphinxql do
        ThinkingSphinx::Connection.take do |connection|
          connection.execute sphinxql
        end
      end

      super
    end
  end

  # Delete record using their ActiveRecord id instead of their sphinx document id. This avoids the need to rebuild the
  # index after adding or removing an index. Without this, the old indexed record would not be deleted since Sphinx's
  # calculation of that record's id would not match now that the number/order of indices has changed.
  module RealtimeDeletion
    def perform
      return unless callbacks_enabled?

      sphinxql = <<~SQL
        DELETE FROM #{@index.name} WHERE sphinx_internal_id IN (#{ids.join(', ')})
      SQL

      execute sphinxql
    end
  end
end

ThinkingSphinx::RealTime::Transcriber.prepend(SphinxIndexing::Transcriber)
ThinkingSphinx::Deletion::RealtimeDeletion.prepend(SphinxIndexing::RealtimeDeletion)