Multiline codec and JDBC

Hello everybody,
I'm trying to use the multiline codec with the ids of rows from a table (Database).
i've an input plugin with JDBC which give me some rows with same id (because of the SQL query).
I would like to process these rows by block which have the same id because i want to create a document in ElasticSearch not for each row but for each id.
Thanks in advance for yours answers.

Maybe if you can provide an example of your config and data we can help.

Hello warkolm,
Yes I can give more details about my problem. So, here is an exemple of my .conf file with Logstash :

input {
	jdbc {
            jdbc_connection_string => "jdbc:oracle:thin:@..."
            jdbc_user => "..."
    	    jdbc_password => "..."
            jdbc_driver_library => "ojdbc7.jar"
            jdbc_driver_class => "Java::oracle.jdbc.driver.OracleDriver"
            statement_filepath => "query.sql"

filter {

output {
    elasticsearch {
        index => "..."
        hosts => "..."
        action => "..."
        document_type => "..."
        document_id => "%{id}"

Because of relations in my database, the query give me some rows with the same identifiant (it's an exemple):
id1, data1, data2
id1, data1', data2'
id1, data1'', data2''
id2, data1, data2
id2, data1', data2'

Each row is an event but I would like process the rows with the same identifiant together (by block). In my mind, I can't use a pattern because I look for an exact value.