A plugin input uses data from a plugin as input for a data set. Plugin input is created when the factory is run. For more information about factory plugins, see Factory Plugin Configuration File Overview.
Plugin input can use an existing data set. This allows you to use data from an existing data set, and incorporate it into a new data set as the source, or setup, data. This transforms the data within a factory build, while using data already processed by another data set.
NOTE: This data set cannot be generated before the factory process because the data is created during the factory process.
Here are some examples of when to use Plugin input:
-
You want one data set to be a summarized version of another. For example, you have Sales data in a cBase with multiple records for each Order Date, and you want rules that only apply at the Order Date level. This means you need a data set with one record per Order Date. You can use a pre-factory ETL to squash the data and feed the resulting cBase into the factory, or you can use a Plugin input to generate the smaller data set and base it on the Sales data already defined in another data set in the factory.
-
You have a flag rule that marks certain Products as "critical" in a factory data set and you want to extract the data for all "critical" Products from the external Parts database. Using a Plugin input, you can use the Products data set as input to an Integrator script in the Plugin. After filtering on the "critical" flag rule, the Integrator script could generate "where_clauses" for the SQL query and then pull the data from the Parts database and return it to the factory using the Plugin as the input for a new data set.
The Plugin input requires something from the factory itself, which means it cannot be calculated beforehand.
For information about using a Plugin file as input for a data set, see Using an Existing Data Set as Data Input.
For information about the difference between plugin and plugin-input, see Plugin Versus Plugin-Input.