Ingest: Align syntax to use _ingest._value across ingest processors

[fkelbert]

Different ingest processors require different syntax to access the value of _ingest._value ("_ingest._value" vs "{{_ingest._value}}").

Instead, syntax across different processors should be the same for better user experience.

E.g., append vs uppercase:

POST /_ingest/pipeline/_simulate
{
  "pipeline": {
    "processors": [
      {
        "foreach": {
          "field": "recipients",
          "processor": {
            "append": {
              "field": "test",
              "value": [ "{{_ingest._value}}" ]
            }
          }
        }
      }
    ]
  },
  "docs": [
    {
      "_source": {
        "recipients": [
          "Bar@foo.com"
        ]
      }
    }
  ]
}

vs

POST /_ingest/pipeline/_simulate
{
  "pipeline": {
    "processors": [
      {
        "foreach": {
          "field": "recipients",
          "processor": {
            "uppercase": {
              "field": "_ingest._value"
            }
          }
        }
      }
    ]
  },
  "docs": [
    {
      "_source": {
        "recipients": [
          "Bar@foo.com"
        ]
      }
    }
  ]
}

[elasticmachine]

Pinging @elastic/es-core-features (:Core/Features/Ingest)

[danhermann]

@fkelbert, the uppercase processor takes only a field name and not literal values (it would make no sense to uppercase a literal value) and therefore does not require the template snippet syntax. The append processor, on the other hand, accepts a value rather than a field name. The template snippet syntax is available should you wish the append processor to use the value of a field. These are intentionally different behaviors (and their parameters are intentionally named field and value).

[fkelbert]

@danhermann Thanks for the clarification. The difference does make sense.

And yet, I am wondering if this can be simplified for users. At the very least, we should provide documentation explaining the difference and in which context which syntax needs to be used.