Conversion type: Connected, active
Filter conversions allow you to filter rows in a map. Pass all the rows from the source transformation through the filter transformation, and then enter the filter criteria for the transformation. The ports in all filter conversions are input/output ports, and only qualifying rows can be converted through a filter.
In some cases, the data needs to be filtered based on one or more criteria before it is written to the target. For example, if you have a human resources goal that includes information about your current employees, you may want to filter out part-time employees and hourly workers.
The following mapping passes the row through a filter transformation from the Human resources table that contains employee data. Filters only allow employees who pay for $30,000 or higher to pass.
With SALARY > 30000 filters, only data rows where employees pay more than $30,000 can reach the target.
As an active conversion, the filter transformation can change the number of rows passed through it. The filter condition returns TRUE or FALSE for each line that passes the transformation, depending on whether the row meets the specified criteria. Only rows that return TRUE can pass this conversion. The discarded rows do not appear in the session log or in the Reject file.
For session performance optimization, you should make the filter transform as close to the source in the map as possible. Instead of passing the rows you are about to discard to the map, you filter out unwanted data from the source-to-destination data stream later.
You cannot connect multiple converted ports to a filter transform. The input port of the filter must come from a single conversion. The filter conversion does not allow setting the output default value.
Overview of the Informatica Common components filter