#!/usr/bin/env python3
"""data.gov.hr — Open Data PGZ."""
import sys, json, time
sys.path.insert(0, "/opt/pgz-sport/scrapers/harvesters")
from _common import upsert_facts, DSN, UA
import urllib.request
import psycopg2

API = "https://data.gov.hr/api/3/action"


def search(query, rows=50):
    url = f"{API}/package_search?q={urllib.parse.quote(query)}&rows={rows}"
    try:
        req = urllib.request.Request(url, headers={"User-Agent": UA})
        with urllib.request.urlopen(req, timeout=30) as r:
            d = json.loads(r.read())
        return d.get("result", {}).get("results", [])
    except Exception as e:
        print(f"search err: {e}")
        return []

import urllib.parse


def main():
    conn = psycopg2.connect(DSN); conn.autocommit = True
    
    queries = [
        "Primorsko-goranska", "Rijeka", "Opatija", "Crikvenica", "Krk",
        "Cres", "Lošinj", "Rab", "Delnice", "Bakar", "Kvarner",
    ]
    
    total_inserted = 0
    seen = set()
    
    for q in queries:
        results = search(q, rows=50)
        ff = []
        for pkg in results:
            pkg_id = pkg.get("id", "")
            if pkg_id in seen: continue
            seen.add(pkg_id)
            
            title = pkg.get("title", "")
            notes = pkg.get("notes", "")[:600]
            org = pkg.get("organization", {}).get("title", "")
            tags = ", ".join([t.get("name", "") for t in pkg.get("tags", [])])
            
            fact = f"[OpenData] {title} | Org: {org} | {notes} | Tags: {tags}"[:1200]
            if len(fact) > 50:
                ff.append({
                    "fact": fact,
                    "url": f"https://data.gov.hr/dataset/{pkg.get('name', '')}",
                    "title": title,
                })
        
        n = upsert_facts(conn, ff, source_name="data_gov_hr_pgz",
                          category="opendata_pgz", confidence=0.85)
        total_inserted += n
        print(f"  query='{q}' -> {len(results)} results, {n} new facts")
        time.sleep(1)
    
    conn.close()
    print(f"=== TOTAL: {total_inserted} ===")
    print(json.dumps({"queries": len(queries), "total_facts": total_inserted}))


if __name__ == "__main__":
    main()