module TensorStream
  # Evaluator base module
  module Evaluator
    class OutputGroup
      include Enumerable

      attr_accessor :outputs, :data_types
      def initialize(outputs = [], data_types = [])
        @outputs = outputs
        @data_types = data_types
      end

      def each
        @outputs.map { |output| yield output }
      end
    end

    class UnsupportedOp < RuntimeError
      def initialize(tensor)
        @tensor = tensor
      end

      def message
        "unsupported op #{@tensor.operation}"
      end
    end

    ##
    # Evaluator base class
    #
    # Base class to be used by all tensor_stream evaluators, provides support functions
    class BaseEvaluator
      def initialize(session, _device, thread_pool: nil, log_intermediates: false)
        @session = session
        @log_intermediates = log_intermediates
        @thread_pool = thread_pool || Concurrent::ImmediateExecutor.new
        @context[:compute_history] = [] if log_intermediates
      end

      ##
      # Query all supported devices
      def self.query_supported_devices
        [Device.new("cpu", :cpu, self)]
      end

      ##
      # Select the best device available in the system for this evaluator
      def self.default_device
        Device.new("cpu", :cpu, self)
      end

      ##
      # Selects the best device with the specified query, query can
      # be evaluator specific
      def self.fetch_device(_query = [])
        Device.new("cpu", :cpu, self)
      end

      ##
      # Select device using uri
      def self.query_device(query)
        return default_device if query.nil? || query == :default

        all_devices = query_supported_devices
        substrs = query.split("/")
        substrs.each do |q|
          components = q.split(":")
          next if components.size.zero?

          if components[0] == "device" # use tensorflow convention
            device_type = components[1]
            select_index = components[2].to_i

            devices = all_devices.select { |d| d.type == device_type.downcase.to_sym }
            return nil if devices.empty?

            select_index = [devices.size - 1, select_index].min
            return devices[select_index]
          elsif %w[cpu gpu].include?(components[0])
            device_type = components[0].to_sym
            select_index = components[1].to_i

            devices = all_devices.select { |d| d.type == device_type.downcase.to_sym }
            return nil if devices.empty?

            select_index = [devices.size - 1, select_index].min
            return devices[select_index]
          elsif components[0] == "ts" # tensorstream specific
            evaluator_class = TensorStream::Evaluator.evaluators[components[1]][:class]
            return nil unless self == evaluator_class
            return evaluator_class.fetch_device(components[2..components.size]) if evaluator_class.respond_to?(:fetch_device)

            return nil
          end
        end
      end

      ##
      # registers an op for the current evaluator class
      def self.register_op(opcode, options = {}, &block)
        @ops ||= {}
        if opcode.is_a?(Array)
          opcode.each do |op|
            @ops[op.to_sym] = {options: options, block: block}
          end
        else
          @ops[opcode.to_sym] = {options: options, block: block}
        end
      end

      ##
      # gets all supported ops for this Evaluator class
      def self.ops
        @ops ||= {}
      end

      def invoke(tensor, execution_context)
        return eval_tensor(tensor, execution_context) unless tensor.is_a?(Operation)
        raise UnsupportedOp.new(tensor), "op #{tensor.operation} is not yet supported" unless self.class.ops.key?(tensor.operation.to_sym)

        op = self.class.ops[tensor.operation.to_sym]
        op_options = op[:options]

        resolved_inputs = tensor.inputs.map { |i|
          next if i.nil?
          next i if op_options[:noop]

          if i.is_a?(Array)
            next i.collect { |sub_item| sub_item.is_a?(Tensor) ? global_eval(tensor, sub_item, execution_context) : sub_item }
          end

          global_eval(tensor, i, execution_context, op_options)
        }

        start_time = if profile_enabled?
          time = Time.now
          time.to_i * (10**9) + time.nsec
        end

        instance_exec(execution_context, tensor, resolved_inputs, &op[:block]).tap do |result|
          if profile_enabled?
            time = Time.now
            end_time = time.to_i * (10**9) + time.nsec
            @context[:profile] ||= {step: 0, operations: {}}
            @context[:profile][:step] += 1
            @context[:profile][:operations][tensor.name] = {op: tensor.operation,
                                                             step: @context[:profile][:step],
                                                             eval_time: end_time - start_time,
                                                             shape: tensor.shape ? tensor.shape.shape : nil,
                                                             data_type: tensor.data_type,
                                                             tensor: tensor,}
          end
        end
      end

      protected

      def profile_enabled?
        @context[:_options][:profile_enabled]
      end

      ##
      # called when passing control to another evaluator
      def perform_transition(tensor, input, _next_evaluator, execution_context)
        cache_key = "#{tensor.graph.object_id}_#{input.name}:#{object_id}"
        return @context[:_cache][cache_key] if @context[:_cache].key?(cache_key)

        result = @session.delegate_to_evaluator(input, @context, execution_context)
        convert_from_buffer(input, result).tap do |buffer|
          @context[:_cache][cache_key] = buffer if input.is_const
        end
      end

      def global_eval(tensor, input, execution_context, op_options = {})
        return nil unless input
        return input unless input.is_a?(Tensor)

        # puts "global eval #{tensor.name}"
        @context[:_cache][:placement][input.name] = @session.assign_evaluator(input) if @context[:_cache][:placement][input.name].nil?
        if !on_same_device?(input) # tensor is on another device or evaluator
          # puts "transition #{object_id} -> #{@context[:_cache][:placement][input.name][1].object_id}"
          perform_transition(tensor, input, @context[:_cache][:placement][input.name][1], execution_context)
        else
          prepare_input(input, execution_context, op_options)
        end
      end

      def on_same_device?(tensor)
        object_id == @context[:_cache][:placement][tensor.name][1].object_id
      end

      def get_broadcast_gradient_args(input_a, input_b)
        return [[], []] if input_a == input_b

        input_a_args = []
        input_b_args = []

        input_a = Array.new(input_b.size) { |i| i < input_a.size ? input_a[i] : nil }.reverse if input_a.size < input_b.size
        input_b = Array.new(input_a.size) { |i| i < input_b.size ? input_b[i] : nil }.reverse if input_a.size > input_b.size

        input_a.reverse.zip(input_b.reverse).each_with_index do |item, index|
          a, b = item

          if a.nil? || b && (a < b)
            input_a_args << input_b.size - index - 1
          elsif b.nil? || a && (a > b)
            input_b_args << input_a.size - index - 1
          end
        end

        [input_a_args.reverse, input_b_args.reverse]
      end

      ##
      # converts from a ruby Buffer object to the evaluator's native buffer format
      def convert_from_buffer(_tensor, _result)
        raise "need implementation"
      end

      def prepare_input(_tensor, _context, _options = {})
        raise "need implementation"
      end
    end

    def self.evaluators
      @evaluators ||= {}
    end

    def self.register_evaluator(klass, name, index = 0)
      @evaluators ||= {}
      @evaluators[name] = {name: name, class: klass, index: index}
    end

    def self.default_evaluators
      evaluators.values.sort { |v| v[:index] }.reverse.map { |v| v[:class] }
    end
  end
end