ATLAS-5399 Add Apache Nutch metadata model and CrawlDb/HostDB import bridge

[lewismc]

What changes were proposed in this pull request?

Catalog Apache Nutch crawls in Atlas without creating one entity per URL (ATLAS-5399).

Adds a Nutch metadata model (addons/models/7000-Nutch/) and a CrawlDb/HostDB import bridge (addons/nutch-bridge). The bridge creates crawl, seedlist, segment, domain, and host entities, plus per-crawl host metrics on nutch_crawl_hosts. Hosts are preferred from Nutch HostDB when {crawl}/hostdb/current exists; otherwise they are rolled up from CrawlDb. There is no nutch_url type.

Indexing lineage is not emitted here. The model includes nutch_index_process; the Nutch Atlas IndexWriter creates that process and a generic index DataSet (NUTCH-3210). The two sides share atlas.cluster.name and crawlId.

Import CLI:

hook-bin/import-nutch.sh -c -d [-H ] [-s ] [-g ]

The distro nutch-hook package ships the bridge and its runtime dependencies so import does not need Nutch on the classpath. Docs: docs/src/documents/Hook/HookNutch.md.

Out of scope: URL/page entities, Nutch REST/JobManager types (NUTCH-3165), indexer-kafka / ATLAS_HOOK, inject/fetch/parse process types.

How was this patch tested?

• mvn -pl addons/nutch-bridge test (CrawlDb MapFile partitions, HostDB SequenceFile, qualified names, host rollup, model JSON)
• Live dev-support/atlas-docker (postgres) with seed https://nutch.apache.org/: import created crawl, seedlist, segment, apache.org, nutch.apache.org, and crawl-host metrics; Nutch IndexingJob then created the index DataSet and nutch_index_process
• Lineage in the Atlas UI: crawl + segment to nutch_index_process to index

See attached screenshot of the lineage tab for the demo indexing process.