Logstash config showing how to create a document identifier built from MD5 hash prefixed by hex formatted epoch date

epoch_prefixed_md5_identifier.conf

input {
  generator {
    lines => ['2011-04-19T03:44:01.103Z testlog1',
              '2011-04-19T03:44:02.035Z testlog2',
              '2011-04-19T03:44:03.654Z testlog3',
              '2011-04-19T03:44:03.654Z testlog3']
    count => 1
  }
}

filter {
  dissect {
    mapping => {
      "message" => "%{ts} %{msg}"
    }
  }

  date {
    match => [ "ts", "ISO8601" ]
  }

  fingerprint {
    source => "message"
    target => "[@metadata][fingerprint]"
    method => "MD5"
    key => "test"
  }

  ruby {
    code => "event.set('@metadata[prefix]', event.get('@timestamp').to_i.to_s(16))"
  }
}

output {
  elasticsearch {
    document_id => "%{[@metadata][prefix]}%{[@metadata][fingerprint]}"
  }
  stdout { codec => rubydebug { metadata => true }}
}

I get the following Error when attempting to use the code
[2019-05-22T13:56:55,964][ERROR][logstash.filters.ruby ] Ruby exception occurred: Invalid FieldReference: '@metadata[prefix]'

adding brackets to the set metadata variable seems to resolve the issue:
ruby { code => 'event.set("[@metadata][prefix]", event.get("@timestamp").to_i.to_s(16))' }