* "Exact" path match.
                        * "Prefix" path match with largest number of characters.
                        * Method match.
                        * Largest number of header matches.
                        * Largest number of query param matches.

def TF_TopKV2Op : TF_Op<"TopKV2", [Pure]> {
  let summary = [{
Finds values and indices of the `k` largest elements for the last dimension.
  }];

  let description = [{
If the input is a vector (rank-1), finds the `k` largest entries in the vector
and outputs their values and indices as vectors.  Thus `values[j]` is the
`j`-th largest entry in `input`, and its index is `indices[j]`.

                  group:
                    description: group is the group of the target resource.
                    maxLength: 253
                    pattern: ^$|^[a-z0-9]([-a-z0-9]*[a-z0-9])?(\.[a-z0-9]([-a-z0-9]*[a-z0-9])?)*$
                    type: string
                  kind:
                    description: kind is kind of the target resource.
                    maxLength: 63
                    minLength: 1

                  group:
                    description: group is the group of the target resource.
                    maxLength: 253
                    pattern: ^$|^[a-z0-9]([-a-z0-9]*[a-z0-9])?(\.[a-z0-9]([-a-z0-9]*[a-z0-9])?)*$
                    type: string
                  kind:
                    description: kind is kind of the target resource.
                    maxLength: 63
                    minLength: 1

*   Add `is_cpu_target_available`, which indicates whether or not TensorFlow was
    built with support for a given CPU target. This can be useful for skipping
    target-specific tests if a target is not supported.

*   `tf.data`
    * Support `data.experimental.distribued_save`. `distribued_save` uses
      tf.data service

                  group:
                    description: group is the group of the target resource.
                    type: string
                  kind:
                    description: kind is kind of the target resource.
                    type: string
                  name:
                    description: name is the name of the target resource.
                    type: string
                  namespace:

            "type": "string"
          },
          "iqn": {
            "default": "",
            "description": "iqn is the target iSCSI Qualified Name.",
            "type": "string"
          },
          "iscsiInterface": {

    specifically bugging me now because they make the action graph harder to
    analyze programmatically. By using aliases here, we would need to handle
    potentially aliased go_binary targets and dereference to the effective
    target.
  
    The comment references an issue with `pure = select(...)` which appears
    to be resolved considering this now builds.
  

		{"Name.Space", Field, 0},
		{"NewDecoder", Func, 0},
		{"NewEncoder", Func, 0},
		{"NewTokenDecoder", Func, 10},
		{"ProcInst", Type, 0},
		{"ProcInst.Inst", Field, 0},
		{"ProcInst.Target", Field, 0},
		{"StartElement", Type, 0},
		{"StartElement.Attr", Field, 0},
		{"StartElement.Name", Field, 0},
		{"SyntaxError", Type, 0},
		{"SyntaxError.Line", Field, 0},
		{"SyntaxError.Msg", Field, 0},

        "properties": {
          "resourceVersion": {
            "description": "Specifies the target ResourceVersion",
            "type": "string"
          },
          "uid": {
            "description": "Specifies the target UID.",
            "type": "string"
          }
        },
        "type": "object"
      },