Skip to main content

Concepts

This section explains how Norconex Crawler works conceptually. Understanding these ideas will help you configure and extend the crawler effectively.

Topics in this section

TopicWhat it covers
Crawl PipelineHow documents move from source to destination
Crawler FlowThe same journey stage by stage, in interactive diagrams
Sessions and RunsResuming, deduplication, recrawl policy, and scheduling
Configuration SemanticsDefaults, null/empty behavior, variables, and fragments
Document ProcessingThe Import module: parsing, enrichment, metadata
Extending the CrawlerCustom components, SPI, event listeners

The big picture

Raw content enters from the left (amber), moves through the crawler's own stages (blue), and lands structured in your destinations (green).

Every document — whether a web page or a file — passes through the same three-stage pipeline. The crawl type (Web vs. File System) only affects the Crawl stage. Everything downstream is identical.

Configuration model

All configuration lives in a single file (XML, YAML, or JSON).

File System Fetcher Start References

For filesystem crawling, the most common first issue is choosing the right start reference scheme for the fetcher you want to use.

Use FS Fetchers Quickstart for a compact table covering all supported filesystem fetchers, including start reference examples and the first configuration step for each.