Processors that operate on documents flowing through a pipeline.

Their input is a batch of pipeline messages where each message carries a serialized document. They are typically used in file-processing pipelines that run after a connector fetches a document.