"""Zitierfähigkeits-Join (Wake-up #2): logischer norm_id-Join auf legal_basis.

KB-v2-Konvention (Board 2026-07-01, Compliance/KB-v2): `EU-<ACT>-Anhang<ROM>` (Annex-Ebene, grob) ·
`EU-<ACT>-Art<N>` (Artikel, in KB-v2 noch zu verifizieren) · Kapitel = TBD-Konvention.
Namensvariante: `EU-MaschVO-*` (NICHT MaschinenVO). Kein char-Span nötig — logischer Join auf norm_id.
Fügt `norm_ids` (Liste) je legal_basis + `norm_id_status` hinzu; setzt obligation.citation_status
auf `norm_id_linked` (annex-grob). KEINE neue Klasse (Attribut). Freeze-safe.
"""
from __future__ import annotations

import glob
import json
import re

ACT = {"CRA": "CRA", "MaschVO": "MaschVO", "MaschinenVO": "MaschVO"}
FILES = sorted(glob.glob("obligations/cra*.json"))


def derive(source: str, anchor: str) -> tuple[list[str], str]:
    act = ACT.get(source, source)
    ids: list[str] = []
    for rom in re.findall(r"An(?:hang|nex)\s+([IVX]+)", anchor, re.I):
        ids.append(f"EU-{act}-Anhang{rom.upper()}")
    articles = re.findall(r"\bArt(?:icle|\.)?\s*(\d+)", anchor)
    chapters = re.findall(r"Kapitel\s+([IVX/]+)", anchor, re.I)
    verify: list[str] = []
    for n in articles:
        verify.append(f"EU-{act}-Art{n}")
    for grp in chapters:
        for rom in grp.split("/"):
            rom = rom.strip()
            if rom:
                verify.append(f"EU-{act}-Kapitel{rom.upper()}")
    # dedup, Annexe zuerst (confirmed), dann verify
    seen: set[str] = set()
    ordered = [x for x in ids + verify if not (x in seen or seen.add(x))]
    status = "annex_confirmed" if ids else ("verify_pending" if verify else "unparsed")
    return ordered, status


def main() -> None:
    total_lb = linked = unparsed = 0
    obl_linked = 0
    for f in FILES:
        d = json.load(open(f, encoding="utf-8"))
        d.setdefault("norm_id_contract", {
            "convention": "EU-<ACT>-Anhang<ROM> (Annex-Ebene) / EU-<ACT>-Art<N> (verify) — KB-v2 bp_compliance_kb_2026_1_build",
            "act_naming": "EU-MaschVO-* (NICHT MaschinenVO)",
            "granularity": "annex-grob — 'Annex I Part II (1)' -> EU-CRA-AnhangI; Part/Punkt = KB-Enhancement TBD",
            "article_status": "EU-<ACT>-Art<N> in KB-v2 noch zu verifizieren; Annex-IDs confirmed",
            "source": "Board Compliance/KB-v2 2026-07-01",
        })
        for o in d.get("obligations", []):
            got = False
            for b in o.get("legal_basis", []):
                total_lb += 1
                nids, st = derive(b.get("source", ""), b.get("anchor", ""))
                b["norm_ids"] = nids
                b["norm_id_status"] = st
                if nids:
                    linked += 1
                    got = True
                if st == "unparsed":
                    unparsed += 1
                    print(f"  UNPARSED: {b.get('source')} \"{b.get('anchor')}\"")
            if got:
                o["citation_status"] = "norm_id_linked"
                obl_linked += 1
        json.dump(d, open(f, "w", encoding="utf-8"), ensure_ascii=False, indent=1)
    print(f"legal_basis gesamt {total_lb} | mit norm_ids {linked} | unparsed {unparsed}")
    print(f"Obligations citation_status -> norm_id_linked: {obl_linked}")


if __name__ == "__main__":
    main()