pgz-sport/scrapers/wiki_pgz_sport.py

#!/usr/bin/env python3
# ═══════════════════════════════════════════════════════════════════
# Fajl: wiki_pgz_sport.py | v1.0.0 | 05.05.2026
# Lokacija: /opt/pgz-sport/scrapers/wiki_pgz_sport.py
# Svrha: Wikipedia HR/EN scrape — PGŽ sport klubovi + sportaši
#   - Iterate kroz sve known PGŽ klubove
#   - Wiki API → page extract
#   - Plus historical match results od Wikipedia season tables
# ═══════════════════════════════════════════════════════════════════
"""Wikipedia PGŽ sport corpus."""
import os, sys, re, time, hashlib, logging, json
import urllib.request, urllib.parse
import psycopg2
from psycopg2.extras import execute_batch

logging.basicConfig(level=logging.INFO, format="%(asctime)s [wiki_sport] %(message)s")
log = logging.getLogger("wiki_sport")

DSN = "host=10.10.0.2 port=6432 dbname=rinet_v3 user=rinet password=R1net2026!SecureDB#v7"
UA = "Ri.NET Civic Bot 1.0 (contact: dradulic@outlook.com)"
API_HR = "https://hr.wikipedia.org/w/api.php"
API_EN = "https://en.wikipedia.org/w/api.php"


def wiki_extract(api, title, sentences=None):
    """Get plain text extract for a Wikipedia page."""
    params = {
        "action": "query", "prop": "extracts", "explaintext": "1",
        "redirects": "1", "format": "json", "titles": title,
    }
    if sentences:
        params["exsentences"] = str(sentences)
    
    url = api + "?" + urllib.parse.urlencode(params)
    req = urllib.request.Request(url, headers={"User-Agent": UA})
    try:
        with urllib.request.urlopen(req, timeout=15) as r:
            d = json.loads(r.read())
        pages = d.get("query", {}).get("pages", {})
        for pid, p in pages.items():
            if pid == "-1": return None  # not found
            return p.get("extract", "")
    except Exception as e:
        log.warning(f"wiki err {title}: {e}")
        return None


def wiki_search(api, query, limit=5):
    """Search Wikipedia for related pages."""
    params = {"action": "query", "list": "search", "srsearch": query,
              "format": "json", "srlimit": str(limit)}
    url = api + "?" + urllib.parse.urlencode(params)
    req = urllib.request.Request(url, headers={"User-Agent": UA})
    try:
        with urllib.request.urlopen(req, timeout=10) as r:
            d = json.loads(r.read())
        return [p["title"] for p in d.get("query", {}).get("search", [])]
    except Exception:
        return []


def get_pgz_clubs(conn):
    """Fetch active PGŽ clubs from DB."""
    cur = conn.cursor()
    cur.execute("""
        SELECT naziv, COALESCE(skraceni_naziv, '') 
        FROM pgz_sport.klubovi 
        WHERE COALESCE(aktivan, true) = true
        ORDER BY naziv
    """)
    out = [(r[0], r[1]) for r in cur.fetchall()]
    cur.close()
    return out


def chunk(text, max_len=700):
    if len(text) <= max_len: return [text] if text else []
    out = []; start = 0
    while start < len(text):
        end = min(start + max_len, len(text))
        if end < len(text):
            for sep in [". ", "! ", "? ", "\n"]:
                p = text.rfind(sep, start, end)
                if p > start + max_len // 2:
                    end = p + len(sep); break
        out.append(text[start:end].strip())
        start = end
    return [c for c in out if len(c) > 80]


def upsert(conn, facts):
    if not facts: return 0
    cur = conn.cursor()
    rows = []
    for f in facts:
        h = hashlib.md5(f["fact"].encode()).hexdigest()
        rows.append((f["fact"], "wikipedia_pgz_sport", "pgz_sport_wiki",
                     f.get("confidence", 0.84), h,
                     json.dumps({"page": f.get("page", ""), "lang": f.get("lang", "hr")})))
    sql = """INSERT INTO dabi.knowledge (fact, source, category, confidence, data_hash, source_refs)
             VALUES (%s, %s, %s, %s, %s, %s::jsonb) ON CONFLICT (data_hash) DO NOTHING"""
    try:
        execute_batch(cur, sql, rows, page_size=50)
        n = cur.rowcount; cur.close()
        return n
    except Exception as e:
        log.error(f"upsert: {e}")
        return 0


def main():
    conn = psycopg2.connect(DSN); conn.autocommit = True
    clubs = get_pgz_clubs(conn)
    log.info(f"PGŽ active clubs: {len(clubs)}")
    
    total_facts = 0
    found_pages = 0
    
    for naziv, kraci in clubs[:200]:  # limit first run
        # Try direct page first
        text = wiki_extract(API_HR, naziv)
        if not text:
            # Try search
            candidates = wiki_search(API_HR, naziv, limit=3)
            for c in candidates:
                if any(t.lower() in c.lower() for t in [naziv.split()[-1], "Rijeka", "Opatija", "Krk"]):
                    text = wiki_extract(API_HR, c)
                    if text:
                        break
        
        if text and len(text) > 200:
            found_pages += 1
            facts = [{"fact": c, "page": naziv, "lang": "hr", "confidence": 0.85}
                     for c in chunk(text, 700)]
            total_facts += upsert(conn, facts)
        
        time.sleep(0.5)
        if found_pages % 20 == 0 and found_pages > 0:
            log.info(f"  progress: pages {found_pages}, facts {total_facts}")
    
    log.info(f"=== DONE: pages={found_pages} facts={total_facts} ===")
    print(json.dumps({"pages": found_pages, "facts": total_facts}))
    conn.close()


if __name__ == "__main__":
    main()