nx_func_extract.nx

nx_extract.nx -- function-body extraction + normalization + hashing. Substrate-native primitive for semantic-equivalence verification. Foundation for the dedup-then-consolidate workflow: extract a named function from a file, normalize away comments and whitespace, compute a deterministic hash, compare across files. Capabilities provided: - nx_extract_func_body : locate `func NAME(...)` and return its body byte range [body_start, body_end) - nx_extract_normalize : copy a buffer slice with line comments removed and runs of whitespace collapsed to single spaces; strings preserved verbatim - nx_extract_hash : FNV-1a 64-bit hash of a byte range - nx_extract_func_hash : convenience -- extract + normalize + hash All pure-NishiLang. No reliance on external grep/awk/sed. This is the substrate's answer to "diff two functions for equivalence" without leaving NishiLang. genealogy_id: fowler_noll_vo_1991_fnv1a + baker_1995_clone_detection lineage_id: syntax_directed_function_extraction + content_hash axioms: NX_AX_REL_REFLEXIVITY (hash(a) = hash(a)) + NX_AX_REL_SYMMETRY (hash(a)=hash(b) <-> hash(b)=hash(a)) + NX_AX_REL_TRANSITIVITY (a~b, b~c -> a~c) together: hash-equality is an equivalence relation; collisions are theoretically possible (Pigeonhole) but FNV-1a 64-bit collision probability is negligible at our function-population scale (thousands, not billions).