code wiki / crawl

topic: crawl

19 modules sharing the crawl name family (derived from the tree's prefix discipline).

The 'crawl' topic family in the Nishi sovereign ecosystem enables systematic and secure web traversal, indexing, and processing. The nx_crawl_bfs module drives autonomous link-following using breadth-first search, while nx_crawl_burst manages distributed crawling with budget control and shard integration. nx_crawl_https ensures robust, validated HTTPS navigation, forming the backbone of reliable data ingestion. Together, these modules support a cohesive pipeline from frontier extraction to document processing, ensuring comprehensive and trustworthy web coverage.

auto-narrated by the local model from this topic's module headers; links verified against the wiki index.

narrated overview -- maintained by the narration lane, module links verified against this wiki.

moduledescriptionlinesfuncs
nx_crawl_bfs.nxautonomous link-following crawler (frontier BFS).1637
nx_crawl_burst.nxCoordinates distributed web crawling with governed budgets, lease atomicity, and shard merging into a unified index.964
nx_crawl_callee_probe.nxPER-CALLEE VmSize PROBE for the crawler's per-page pipeline (debt 1787082132).2036
nx_crawl_coverage_census.nxHONEST access-coverage census for the sovereign crawler (operator 2026-07-04:975
nx_crawl_doc.nxper-page processing unit for the sovereign crawler.422
nx_crawl_doc_test.nxKAT for the crawl back-half on real CONTENT.761
nx_crawl_frontier.nxoutbound-link (frontier) extraction for the crawler.1155
nx_crawl_frontier_test.nxKAT for outbound-link extraction.431
nx_crawl_https.nxthe hardened crawler on the REAL VALIDATED HTTPS web.1386
nx_crawl_main.nxthe deployable Nishi crawl+serve daemon.2568
nx_crawl_pace.nxSOTA polite crawler pacing (operator 2026-07-05: "pacing is a great callout on all the44024
nx_crawl_pace_gate.nxdeterministic gate for the pacing POLICY (no network): honor-Retry-After,3164
nx_crawl_polite.nxhardened, polite crawler (don't get blacklisted).1627
nx_crawl_preserve.nxthe CRAWL DRIVER: sweep a list of URLs, render-fetch each as a browser (browser headers724
nx_crawl_run.nxend-to-end crawl: fetch (real HTTP) -> extract -> dedup1096
nx_crawl_sufficiency.nxADAPTIVE STOP: "do we know enough to stop fetching?" -- the /compare/webscraping39115
nx_crawl_sufficiency_cli.nxCLI over nx_crawl_sufficiency (the /compare/webscraping R5 lib): judge a set of745
nx_crawl_sufficiency_gate.nxGATE for nx_crawl_sufficiency (the /compare/webscraping R5 contract1624
nx_crawl_web.nxREAL-WEB BFS crawler: "crawl whatever is out there", sovereign + polite. Combines what the19813