Support for ML Inference API

[italodirenzo93]

Is your feature request related to a problem? Please describe.
I'm attempting to implement an ingest pipeline for my index that generated dense_vectors using the inference pipeline processor. This is currently not possible as both the Inference API and the Inference processor are not available in the Elasticsearch Terraform provider.

Describe the resource you would like to have implemented.
Broad support for the Inference API but for my use case specifically, the following endpoints as resources:

• Create an inference endpoint
• Update an inference endpoint

As well as the following as data objects:

• Inference pipeline processor

Pseudo-code example

resource "elasticstack_elasticsearch_inference_endpoint" "my_endpoint" {
  service = "openai"
  service_settings = {
    api_key = "..."
    model_id = "text-embedding-3-small"
  }
}

data "elasticstack_elasticsearch_ingest_processor_inference" "my_proc" {
  model_id = "my_endpoint"
  input_output = {
    input_field = "foo"
    output_field = "bar"
  }
}

resource "elasticstack_elasticsearch_ingest_pipeline" "my_pipeline" {
  name = "..."
  processors= [data.elasticstack_elasticsearch_ingest_processor_append.my_proc.json]
}

resource "elasticstack_elasticsearch_index" "my_index" {
  name = "..."
  default_pipeline = elasticstack_elasticsearch_ingest_pipeline.my_pipeline.name
}