code wiki / (root) / nx_research_extract.nx

nx_research_extract.nx

buildroot/runtime/nx_research_extract.nx

2648 B59 linesdepth 2pulls 2 transitivereach 114 importersview sourcekind librarytopic research

docs dependencies structs consts functions

about

nx_research_extract.nx -- the Researcher's EXTRACT stage: parse FACTS from fetched text by PATTERN (Claude was wrong to mark this an LLM-gap). Pulling "53" out of "53 percent of users abandon" is substring + integer scan -- pure mechanizable string work the team owns. Only DEEP semantic synthesis needs the LLM; fact/number/citation extraction does not. Pairs with nx_browse_text (the team's sovereign fetch) so the Researcher does fetch+extract end to end. license_tier: ORIGINAL.

dependencies 1 imports · 45 importers

diagram shows first 10 each side; +0 more imports, +35 more importers in the complete lists below.

imports: nx_syscalls.nx

imported by: nx_bm25.nxnx_bm25_naive.nxnx_bm25_test.nxnx_book_search.nxnx_book_search_lib.nxnx_capability_census.nxnx_cms_census.nxnx_competitive_census.nxnx_corroborate.nxnx_corroborate_test.nxnx_deck_gate.nxnx_engineer_wire_sclass_gate.nxnx_engineer_wire_test.nxnx_galx_census.nxnx_ge_synth.nxnx_grep.nxnx_grep_hb.nxnx_html_extract_data.nxnx_html_extract_data_test.nxnx_inbook_search.nxnx_kernel_adoption.nxnx_kernel_census.nxnx_library_search.nxnx_param_census.nxnx_ppmi_svd.nxnx_ppmi_svd_test.nxnx_referee_test.nxnx_rel_measure.nxnx_research_fetch_test.nxnx_research_landscape.nxnx_research_landscape_test.nxnx_research_live_landscape.nxnx_research_relation.nxnx_research_relation_clean.nxnx_research_relation_clean_test.nxnx_research_relation_test.nxnx_research_relations.nxnx_research_relations_test.nxnx_research_resilience_test.nxnx_research_sovereign_lens.nxnx_research_sovereign_lens_test.nxnx_semantic_retrieval.nxnx_semantic_retrieval_test.nxnx_sourced_census_test.nxnx_voxel_census.nx

structs

none

consts

none

functions

9	func re_strlen(s: *u8) -> i64 { var n: i64 = 0; while s[n] != (0 as u8) { n = n + 1 } return n } called by 19: main main main main score_point hed_field+13
12	func re_find(text: u8, n: i64, needle: u8) -> i64 called by 12: score_point grep_line_matches grep_line_matches hed_field hed_meta ib_search+6 calls 1: re_strlen
26	func re_int_from(text: *u8, n: i64, pos: i64) -> i64 called by 1: re_stat_after
40	func re_stat_after(text: u8, n: i64, keyword: u8) -> i64 called by 3: co_value main facts_ok calls 3: re_find re_int_from re_strlen
47	func re_has(text: u8, n: i64, term: u8) -> i64 { if re_find(text, n, term) >= 0 { return 1 } return 0 } called by 27: bm_df bm_df main cc_comp_attest main cc_one+21 calls 1: re_find
50	func re_count(text: u8, n: i64, term: u8) -> i64 called by 13: bm_score bm_naive_tfidf bm_score bm_naive_tfidf bmf_df bmf_score+7 calls 1: re_strlen