# -----------------------------------------------------------------------------
# Home-Ka — US real-estate aggregator (Groupe KA)
# Author: Simon-Pierre Boucher — contact@spboucher.ai
# brokerages.py : brokerage registry — seed loading + scoring refresh.
# The seed (data/brokerages_seed.json) is the initial base of 500+ US
# brokerages worth pursuing for direct feeds; the discovery pipeline
# (discovery.py) then inspects, classifies and scores them.
# -----------------------------------------------------------------------------
from __future__ import annotations

import json
import time
from pathlib import Path

from . import db
from .discovery import priority

SEED_PATH = Path(__file__).resolve().parent.parent / "data" / "brokerages_seed.json"


def load_seed(path: Path | str | None = None) -> dict:
    """Upsert the seed file into the brokerages table (idempotent, keyed on
    website). Never overwrites discovery results or partnership statuses."""
    p = Path(path) if path else SEED_PATH
    data = json.loads(p.read_text(encoding="utf-8"))
    rows = data["brokerages"] if isinstance(data, dict) else data
    con = db.connect()
    now = time.time()
    added = updated = 0
    for b in rows:
        website = (b.get("website") or "").strip().rstrip("/")
        if not website or not b.get("name"):
            continue
        states = b.get("states") or []
        prio = priority(b.get("estimated_listings"), b.get("estimated_agents"),
                        30, len(states))   # 30 = unknown feed prob. before inspection
        existing = con.execute("SELECT id FROM brokerages WHERE website=?",
                               (website,)).fetchone()
        if existing is None:
            con.execute(
                """INSERT INTO brokerages (name, website, states, cities,
                   estimated_agents, estimated_listings, mls_affiliations,
                   possible_feed_type, priority_score, partnership_status,
                   created, updated)
                   VALUES (?,?,?,?,?,?,?,?,?,'prospect',?,?)""",
                (b["name"], website,
                 json.dumps(states, ensure_ascii=False),
                 json.dumps(b.get("cities") or [], ensure_ascii=False),
                 b.get("estimated_agents"), b.get("estimated_listings"),
                 json.dumps(b.get("mls_affiliations") or [], ensure_ascii=False),
                 b.get("possible_feed_type") or "crawl", prio, now, now))
            added += 1
        else:
            con.execute(
                """UPDATE brokerages SET name=?,
                     states=CASE WHEN states='[]' THEN ? ELSE states END,
                     cities=CASE WHEN cities='[]' THEN ? ELSE cities END,
                     estimated_agents=COALESCE(estimated_agents, ?),
                     estimated_listings=COALESCE(estimated_listings, ?),
                     updated=? WHERE id=?""",
                (b["name"], json.dumps(states, ensure_ascii=False),
                 json.dumps(b.get("cities") or [], ensure_ascii=False),
                 b.get("estimated_agents"), b.get("estimated_listings"),
                 now, existing["id"]))
            updated += 1
    con.commit()
    n = con.execute("SELECT COUNT(*) c FROM brokerages").fetchone()["c"]
    con.close()
    return {"added": added, "updated": updated, "total": n}


def rescore(con=None) -> int:
    """Recompute priority scores from current inspection data."""
    own = con is None
    con = con or db.connect()
    n = 0
    for r in con.execute("SELECT id, estimated_listings, estimated_agents,"
                         " feed_probability_score, states FROM brokerages"):
        states = json.loads(r["states"] or "[]")
        prio = priority(r["estimated_listings"], r["estimated_agents"],
                        r["feed_probability_score"] or 30, len(states))
        con.execute("UPDATE brokerages SET priority_score=? WHERE id=?",
                    (prio, r["id"]))
        n += 1
    con.commit()
    if own:
        con.close()
    return n