OpenSearch/docs/reference/ingest/processors/csv.asciidoc

34 lines
1.5 KiB
Plaintext

[[csv-processor]]
=== CSV Processor
Extracts fields from CSV line out of a single text field within a document. Any empty field in CSV will be skipped.
[[csv-options]]
.CSV Options
[options="header"]
|======
| Name | Required | Default | Description
| `field` | yes | - | The field to extract data from
| `target_fields` | yes | - | The array of fields to assign extracted values to
| `separator` | no | , | Separator used in CSV, has to be single character string
| `quote` | no | " | Quote used in CSV, has to be single character string
| `ignore_missing` | no | `true` | If `true` and `field` does not exist, the processor quietly exits without modifying the document
| `trim` | no | `false` | Trim whitespaces in unquoted fields
include::common-options.asciidoc[]
|======
[source,js]
--------------------------------------------------
{
"csv": {
"field": "my_field",
"target_fields": ["field1, field2"],
}
}
--------------------------------------------------
// NOTCONSOLE
If the `trim` option is enabled then any whitespace in the beginning and in the end of each unquoted field will be trimmed.
For example with configuration above, a value of `A, B` will result in field `field2`
having value `{nbsp}B` (with space at the beginning). If `trim` is enabled `A, B` will result in field `field2`
having value `B` (no whitespace). Quoted fields will be left untouched.