Rules accumulator transform Icon Rules accumulator

Description

The Rules Accumulator collects incoming rows and executes them against a rule set. This may be useful to determine the answer to a question or otherwise analyze a dataset.

Drools is the present rule engine implementation and its rule language can be referenced for use by this transform. Drools documentation

Supported Engines

Hop Engine

Supported

Spark

Maybe Supported

Flink

Maybe Supported

Dataflow

Maybe Supported

Details

Once all incoming rows have been collected by the Rules Accumulator transform (e.g. - the previous transform shuts down) the rows are transformed into Rules.Row objects and passed into the rules engine to be executed against the given rule set.

Rules.Row is defined as a key / value Map of fields where key is the name of the field and value is the value of the field; as well as the externalSource boolean property to indicate whether the Rules.Row object was created in the rule set or injected from an external source.

Fields of a row are accessed as "Row (column["<fieldname>"])".

Rules Tab

The Rules tab is where you enter your rule definition or a reference to the rule file..

Option Description

Rules filename

Rules script to execute

Rules result Tab

Rules result tab defines the layout of the Rules output fields.

Option Description

Result column name

Result column type

Example

All Rule Definitions should contain "import org.apache.hop.pipeline.transforms.drools.Rules.Row;" to give access to the Rules.Row class.

For the input with a row meta: name (String), position (Integer), color (String); a Rules.Row object will be created for each row with a Map containing those fields.

Rules.Row→row (Map) Name Position Color

Fred

1

Blue

Fred

2

Red

Bob

1

Blue

Bob

1

Red

Rules can be defined and applied:

rule "Golfers problem"
    dialect "mvel"
    when

    # Define Fred
    $fred : Row ( externalSource == true,
        column["name"] == "Fred"
    )

    # Define Bob
    $bob : Row ( externalSource == true,
        column["name"] == "Bob",
        column["position"] != $fred.column["position"],
        column["color"] == "blue",
        column["color"] != $fred.column["color"],
    )

    then

      Row fredRow = new Row();
      Row bobRow = new Row();

      fredRow.addColumn("name", "Fred");
      fredRow.addColumn("position", $fred.column["position"]);
      fredRow.addColumn("color", $fred.column["color"]);

      bobRow.addColumn("name", "Bob");
      bobRow.addColumn("position", $bob.column["position"]);
      bobRow.addColumn("color", $bob.column["color"]);


      insert(fredRow);
      insert(bobRow);

end

Rules.Row objects can be checked as in the above’s left hand side.

Rules.Row objects can be created for the data stream as in the above’s right hand side.

The transform can be told what fields to pickup from the generated row objects by defining the field Map’s name in the transform’s Results tab. Type conversions can be applied by setting the Result column type as well.