The XMLRecordSetWriter Controller Service writes record objects to XML. The Controller Service must be configured with a schema that describes the structure of the record objects. Multiple records are wrapped by a root node. The name of the root node can be configured via property. If no root node is configured, the writer expects only one record for each FlowFile (that will not be wrapped). As Avro does not support defining attributes for records, this writer currently does not support writing XML attributes.

Example: Simple records 

                RecordSet (
                  Record (
                    Field "name1" = "value1",
                    Field "name2" = 42
                  ),
                  Record (
                    Field "name1" = "value2",
                    Field "name2" = 84
                  )
                )

This record can be described by the following schema: 

                {
                  "name": "test",
                  "namespace": "nifi",
                  "type": "record",
                  "fields": [
                    { "name": "name1", "type": "string" },
                    { "name": "name2", "type": "int" }
                  ]
                }

Assuming that "root_name" has been configured as the name for the root node and "record_name" has been configured as the name for the record nodes, the writer will write the following XML: 

                <root_name>
                  <record_name>
                    <name1>value1</name1>
                    <name2>42</name2>
                  </record_name>
                  <record_name>
                    <name1>value2</name1>
                    <name2>84</name2>
                  </record_name>
                </root_name>

The writer furthermore can be configured how to treat null or missing values in records: 

                RecordSet (
                  Record (
                    Field "name1" = "value1",
                    Field "name2" = null
                  ),
                  Record (
                    Field "name1" = "value2",
                  )
                )

If the writer is configured always to suppress missing or null values, only the field of name "name1" will appear in the XML. If the writer ist configured only to suppress missing values, the field of name "name2" will appear in the XML as a node without content for the first record. If the writer is configured never to suppress anything, the field of name "name2" will appear in the XML as a node without content for both records.

Example: Arrays

The writer furthermore can be configured how to write arrays: 

                RecordSet (
                  Record (
                    Field "name1" = "value1",
                    Field "array_field" = [ 1, 2, 3 ]
                  )
                )

This record can be described by the following schema: 

                {
                  "name": "test",
                  "namespace": "nifi",
                  "type": "record",
                  "fields": [
                    { "name": "array_field", "type":
                      { "type": "array", "items": int }
                    },
                    { "name": "name1", "type": "string" }
                  ]
                }

If the writer is configured not to wrap arrays, it will transform the record to the following XML: 

                <root_name>
                  <record_name>
                    <name1>value1</name1>
                    <array_field>1</array_field>
                    <array_field>2</array_field>
                    <array_field>3</array_field>
                  </record_name>
                </root_name>

If the writer is configured to wrap arrays using the field name as wrapper and "elem" as tag name for element nodes, it will transform the record to the following XML: 

                <root_name>
                  <record_name>
                    <name1>value1</field2>
                    <array_field>
                      <elem>1</elem>
                      <elem>2</elem>
                      <elem>3</elem>
                    </array_field>
                  </record_name>
                </root_name>

If the writer is configured to wrap arrays using "wrap" as wrapper and the field name as tag name for element nodes, it will transform the record to the following XML: 

                <root_name>
                  <record_name>
                    <name1>value1</field2>
                    <wrap>
                      <array_field>1</array_field>
                      <array_field>2</array_field>
                      <array_field>3</array_field>
                    </wrap>
                  </record_name>
                </root_name>