for (Value input : island.getControlInputs()) {
      Operation* def = input.getDefiningOp();
      DCHECK_EQ(def->getParentOp(), graph);
      try_update_current_candidate(def);
    }

    // Check island data operands.
    island.walk([&](Operation* op) {
      for (Value input : op->getOperands()) {
        Operation* def = input.getDefiningOp();

[[sec:task_in_out]]
== Task inputs and outputs

Actionable tasks have inputs and outputs.
Inputs and outputs can be files, directories, or variables.

In actionable tasks:

- *Inputs* consist of a collection of files, folders, and/or configuration data. +

		if c.Namespace == "" {
			c.Namespace = "default"
		}
		configs[i] = c
	}
	configCache[input] = configs
	k8sConfigCache[input] = k8sTypes
	return configs, k8sTypes
}

func parseKubernetesTypes(inputs string) (string, int) {
	matches := 0
	sb := strings.Builder{}
	for _, text := range strings.Split(inputs, "\n---") {
		if strings.Contains(text, "kind: Secret") {
			sb.WriteString(text + "\n---\n")

        }
    })
    return properties
}


private
fun collectRegisteredInputsOf(task: Task): List<RegisteredProperty> {

    val properties = mutableListOf<RegisteredProperty>()

    (task.inputs as TaskInputsInternal).visitRegisteredProperties(object : PropertyVisitor {

        override fun visitInputFileProperty(
            propertyName: String,
            optional: Boolean,

		gp2xer1xer  = regInfo{inputs: []regMask{gp | sp | sb, gp | sp | sb, xer}, outputs: []regMask{gp, xer}, clobbers: xer}
		gp31        = regInfo{inputs: []regMask{gp | sp | sb, gp | sp | sb, gp | sp | sb}, outputs: []regMask{gp}}
		gp1cr       = regInfo{inputs: []regMask{gp | sp | sb}}
		gp2cr       = regInfo{inputs: []regMask{gp | sp | sb, gp | sp | sb}}
		crgp        = regInfo{inputs: nil, outputs: []regMask{gp}}

  // Whether to trigger quantization passses for post-training quantization.
  // If true, the model input doesn't require user specified input ranges.
  bool post_training_quantization = false;

  // Whether to allow dynamic range quantization. This is the easiest
  // quantization mode which doesn't require QAT or sample inputs.
  // This option only targets `DT_HALF` and `DT_QINT8` inference type.
  bool weight_quantization = false;

  """Constructs a feed_dict from input data.

  Note: This function should only be used in graph mode.

  This is a helper function that converts an 'input key -> input value' mapping
  to a feed dict. A feed dict is an 'input tensor name -> input value' mapping
  and can be directly passed to the `feed_dict` argument of `sess.run()`.

  Args:
    input_data: Input key -> input value mapping. The input keys should match

    return new FunctionForMapNoDefault<>(map);
  }

  /**
   * Returns a function which performs a map lookup with a default value. The function created by
   * this method returns {@code defaultValue} for all inputs that do not belong to the map's key
   * set. See also {@link #forMap(Map)}, which throws an exception in this case.
   *
   * <p><b>Java 8+ users:</b> you can just write the lambda expression {@code k ->

                      arg_types, arg_locs);

  IRMapping mapper;
  // stablehlo_func_op has 1 extra arg for platform index.
  for (auto [input, stablehlo_func_arg] : llvm::zip_equal(
           inputs, stablehlo_func_op.getArguments().take_back(inputs.size()))) {
    mapper.map(input, stablehlo_func_arg);
  }

  for (Operation* subgraph_op : llvm::reverse(reverse_subgraph)) {

    TF_EXCLUSIVE_LOCKS_REQUIRED(fn_body->mu) {
  input_tensors->reserve(ninputs);
  for (int i = 0; i < ninputs; ++i) {
    Node* node = inputs[i].oper ? &inputs[i].oper->node : nullptr;
    int idx = inputs[i].index;

    TF_RETURN_WITH_CONTEXT_IF_ERROR(
        fn_body->graph.IsValidOutputTensor(node, idx),
        "Encountered while processing input ", i, " into function '", fn_name,
        "'");