<a id="flatten"></a>

# Kafka Connect Flatten SMT Usage Reference for Confluent Platform

The `Flatten` SMT (`org.apache.kafka.connect.transforms.Flatten`)
flattens a nested Apache Kafka® record key or value into a flat map by
concatenating field names with a configurable delimiter.

## Description

Flatten a nested data structure, generating names for each field by
concatenating the field names at each level with a configurable delimiter
character. Applies to a Struct when a schema is present, or a Map for
schemaless data.

The default delimiter is the period (`.`) character.

Use the specific transformation type designed for the record key
(`org.apache.kafka.connect.transforms.Flatten$Key`) or value
(`org.apache.kafka.connect.transforms.Flatten$Value`).

#### NOTE
- The `Flatten` SMT does not flatten arrays. Arrays are preserved as-is.
- When using an Avro schema, set `delimiter` to `_` because Avro
  field names allow only alphanumeric characters and underscore. For
  the full configuration, see [Flatten with the underscore delimiter (for Avro schemas)](#flatten-avro-example).

<a id="flatten-properties"></a>

## Properties

| Name        | Description                                                                                                      | Type   | Default   | Valid values                     | Importance   |
|-------------|------------------------------------------------------------------------------------------------------------------|--------|-----------|----------------------------------|--------------|
| `delimiter` | Delimiter to insert between field names from the input record when generating field names for the output record. | string | `.`       | Any string (commonly `.` or `_`) | medium       |

## Flatten with the period delimiter (for JSON schemas)

The following configuration snippet shows how to use `Flatten` to
concatenate field names with the period `.` delimiter character.

```json
"transforms": "flatten",
"transforms.flatten.type": "org.apache.kafka.connect.transforms.Flatten$Value",
"transforms.flatten.delimiter": "."
```

Before:

```none
{
  "content": {
    "id": 42,
    "name": {
      "first": "David",
      "middle": null,
      "last": "Wong"
    }
  }
}
```

After:

```none
{
  "content.id": 42,
  "content.name.first": "David",
  "content.name.middle": null,
  "content.name.last": "Wong"
}
```

<a id="flatten-avro-example"></a>

## Flatten with the underscore delimiter (for Avro schemas)

Set `delimiter` to `_` when using Avro schemas. The Avro schema
specification only allows alphanumeric characters and underscore in field
names. The following configuration snippet shows how to use `Flatten` to
concatenate field names with the underscore `_` delimiter.

```json
"transforms": "flatten",
"transforms.flatten.type": "org.apache.kafka.connect.transforms.Flatten$Value",
"transforms.flatten.delimiter": "_"
```

Before:

```none
{
  "content": {
    "id": 42,
    "name": {
      "first": "David",
      "middle": null,
      "last": "Wong"
    }
  }
}
```

After:

```none
{
  "content_id": 42,
  "content_name_first": "David",
  "content_name_middle": null,
  "content_name_last": "Wong"
 }
```

## Predicates

Configure transformations with predicates to ensure they process only records
that satisfy a particular condition. You can also use predicates in a
transformation chain with the [Kafka Connect Filter (Kafka) SMT Usage Reference for Confluent Platform](filter-ak.md#ak-filter) to conditionally filter specific
records. For more information, see [Predicates](filter-ak.md#predicates).