For matter of terminology only, the Pipeline is processing "Pages". A Page is an abstraction (object) representing one crawled URI, the retrieved response, related metadata, and facts collected via processing (for instance, extracted text).
The Pipeline consumes Crawler output (usually huge WARC (Web ARChive) files), and merges the contained data into these abstract Pages. In case of CommonCrawl data, it takes the related request, response and metadata WARC-records from one WARC archive, additional metadata from a "WAT" WARC, and extracted text from a "WET" WARC: five parts from three files into one abstract object presented to the Task.
When you are creating a Task, you do not need to know anything about the interface which the Page object offers: you will submit Filter rules, Extraction wishes, and Packaging instructions. This is an iterative cooperation with the developers and maintainers of the Pipeline.
- Pipeline Tasks;
- Pipeline Pages;
- Data License