#!/usr/bin/env python3
# -----------------------------------------------------------------------------
# House-Ka — fusion des candidats RealtyPress reste-du-Canada (2026-08-27)
# scripts/merge_canada_candidates.py : lit /tmp/rp_canada_candidates.json
#   (sortie du scout), filtre (volume, Québec/Ontario), et ajoute les nouveaux
#   sites à data/canada_agencies.json + data/sources.json.
# Usage : .venv/bin/python scripts/merge_canada_candidates.py [--min-cards 10]
# -----------------------------------------------------------------------------
import json
import re
import sys
from pathlib import Path

ROOT = Path(__file__).resolve().parent.parent
CANDIDATES = Path("/tmp/rp_canada_candidates.json")
MIN_CARDS = int(sys.argv[sys.argv.index("--min-cards") + 1]) if "--min-cards" in sys.argv else 10

cands = json.loads(CANDIDATES.read_text())
reg_p = ROOT / "data" / "canada_agencies.json"
src_p = ROOT / "data" / "sources.json"
registry = json.loads(reg_p.read_text())
sources = json.loads(src_p.read_text())
known_domains = {re.sub(r"^https?://(www\.)?", "", e["site"]).rstrip("/")
                 for e in registry}
known_ids = {s["id"] for s in sources["sources"]}

added = []
for c in sorted(cands, key=lambda x: -x.get("approx_volume_gte", 0)):
    dom = c["domain"]
    if dom in known_domains:
        continue
    provs = c.get("provinces", {})
    total = sum(provs.values()) or 1
    # province dominante observée sur la page 1
    top = max(provs, key=provs.get) if provs else c.get("province_hint", "")
    # hors périmètre : sites majoritairement québécois ou ontariens (l'Ontario
    # est déjà couvert par les 15 sources existantes — n'ajouter que si gros)
    if re.search(r"qu[ée]bec", top, re.I):
        continue
    if top == "Ontario" and c.get("approx_volume_gte", 0) < 5000:
        continue
    if c.get("cards_page1", 0) < MIN_CARDS:
        continue
    slug = re.sub(r"[^a-z0-9]+", "", dom.split(".")[0])[:24]
    sid = f"rp_ag_{slug}"
    if sid in known_ids:
        continue
    pages = max(10, min(1500, c.get("approx_pages_gte", 1) * 2))
    entry = {
        "id": sid,
        "name": f"{dom} ({top or 'Canada'})",
        "site": f"https://{dom}",
        "archive": c.get("archive", "listing"),
        "max_pages": pages,
        "province": top or "Ontario",
        "note": (f"Recensement CANADA 2026-08-27 : ~≥{c.get('approx_volume_gte', '?')} fiches, "
                 f"provinces page 1 : {provs}. RealtyPress/DDF."),
    }
    registry.append(entry)
    sources["sources"].append({
        "id": sid,
        "name": entry["name"],
        "url": entry["site"],
        "listing_url": f"{entry['site']}/{entry['archive']}/",
        "coverage": f"{top or 'Canada'} — ~≥{c.get('approx_volume_gte', '?')} fiches DDF",
        "connector": "realtypress",
        "status": "actif",
        "type": "agence",
        "note": "House-Ka — generic RealtyPress connector (registry data/canada_agencies.json). WordPress RealtyPress plugin on the CREA DDF feed. external_id ddf<id>: cross-site dedup by MIN(uid).",
    })
    known_ids.add(sid)
    known_domains.add(dom)
    added.append((sid, top, c.get("approx_volume_gte")))

reg_p.write_text(json.dumps(registry, ensure_ascii=False, indent=1))
src_p.write_text(json.dumps(sources, ensure_ascii=False, indent=2))
print(f"{len(added)} nouvelles sources :")
for sid, top, vol in added:
    print(f"  {sid:34s} {top:24s} ~≥{vol}")