Question on ingest pipeline split operation

[pk.241011]

For a given string {a}{b}{c} I want output as {a} {b} {c} as seperate outputs.
But what I get is {a, b and }c.

The pipeline

PUT _ingest/pipeline/my_pipeline
{
  "processors": [
	{
	  "split": {
		"field": "raw_string",
		"separator": "\\}\\{"
	  }
	}
  ]
}

The simuate part:

POST _ingest/pipeline/my_pipeline/_simulate
{
  "docs":[
	{
	  "_source":{
		"raw_string" : "{s}{a}{d}"
	  }
	}
	]
}

The sad out put I get:

{
  "docs" : [
	{
	  "doc" : {
		"_index" : "_index",
		"_type" : "_type",
		"_id" : "_id",
		"_source" : {
		  "raw_string" : [
			"{s",
			"a",
			"d}"
		  ]
		},
		"_ingest" : {
		  "timestamp" : "2019-03-04T06:00:28.739Z"
		}
	  }
	}
  ]
}

Any ideas?

[glenacota (Guido Lena Cota)]

Hi,
if you don't want to create a custom script for this, here is a very simple idea:

PUT _ingest/pipeline/my_pipeline
{
  "processors": [
    {
      "gsub": {
        "field": "raw_string",
        "pattern": "\\{",
        "replacement": ""
      }
    },
    {
      "split": {
        "field": "raw_string",
        "separator": "}"
      }
    }
  ]
}

[dadoonet (David Pilato)]

May be this?

POST _ingest/pipeline/_simulate
{
  "pipeline": {
    "description": "test",
    "processors": [
      {
        "split": {
          "field": "raw_string",
          "separator": "\\}\\{|\\{|\\}"
        }
      }
    ]
  },
  "docs": [
    {
      "_index": "index",
      "_type": "_doc",
      "_id": "id",
      "_source": {
        "raw_string": "{s}{a}{d}"
      }
    }
  ]
}

It gives:

{
  "docs" : [
    {
      "doc" : {
        "_index" : "index",
        "_type" : "_doc",
        "_id" : "id",
        "_source" : {
          "raw_string" : [
            "",
            "s",
            "a",
            "d"
          ]
        },
        "_ingest" : {
          "timestamp" : "2019-03-04T18:07:40.378708Z"
        }
      }
    }
  ]
}

[pk.241011]

Thanks for the reply. Got me started off.

[pk.241011]

Thanks for reply. It is working now.