I need to append a string in already existing field in a document without using script.
My document has a field "Field1" which currently has value1. I want to update the document by appending value2 to the "Field1", such that the field's value is updated to "value1value2".
This can be easily done by using script. But is there any other option by which this can be achieved. Also I want to use only update/update_by_query statement.
As @joe suggested you need to do it at ingest time, however, you can create a pipeline as a suggestion. I did a local test using the append and join processors. First I appended field1's value and field2's value, then use the join processor on the array created with name append_fields.
I ingested the following values in the index "append_index",
"field1" : "value1", "field2" : "value2"
"field1": "value10", "field2": "value11"
I suggest you create the following pipeline(append_values) and then run it using the _update_by_query:
PUT _ingest/pipeline/append_values
{
"processors": [
{
"append": {
"field": "append_fields",
"value": [
"{{field1}}",
"{{field2}}"
]
}
},
{
"join": {
"field": "append_fields",
"separator": ""
}
}
]
}
To use a pipeline in the _update__by_query, use the following:
POST append_index/_update_by_query?pipeline=append_values
The response would be:
"hits" : [
{
"_index" : "append_index",
"_type" : "_doc",
"_id" : "_9ovaHQBsTCl1BZv7qhZ",
"_score" : 1.0,
"_source" : {
"field1" : "value1",
"append_fields" : "value1value2",
"field2" : "value2"
}
},
{
"_index" : "append_index",
"_type" : "_doc",
"_id" : "AdovaHQBsTCl1BZv9ak1",
"_score" : 1.0,
"_source" : {
"field1" : "value10",
"append_fields" : "value10value11",
"field2" : "value11"
}
}
]
Links: https://www.elastic.co/guide/en/elasticsearch/reference/7.9/append-processor.html https://www.elastic.co/guide/en/elasticsearch/reference/7.9/join-processor.html https://www.elastic.co/guide/en/elasticsearch/reference/7.9/pipeline.html https://www.elastic.co/guide/en/elasticsearch/reference/current/docs-update-by-query.html
Hope my suggestion helps :)
If you need help to understand what the append processor do, here is the output of executing the append processor without using the join processor:
"hits" : [
{
"_index" : "append_index",
"_type" : "_doc",
"_id" : "G9o9aHQBsTCl1BZvMcv6",
"_score" : 1.0,
"_source" : {
"field1" : "value1",
"append_fields" : [
"value1",
"value2"
],
"field2" : "value2"
}
},
{
"_index" : "append_index",
"_type" : "_doc",
"_id" : "edo9aHQBsTCl1BZvP8vT",
"_score" : 1.0,
"_source" : {
"field1" : "value10",
"append_fields" : [
"value10",
"value11"
],
"field2" : "value11"
}
}
]
Not that I know of.
The only (indirect) alternative is appending those values at ingest time. Though that approach also utilizes a script...
What's the reasoning behind not wanting to script?
