V8 MEGA: meta endpoints + manifestacije + HNS V8 harvester batch

Endpoints:
- /v2/potpore/meta — dropdown options (sportovi, vrste, davatelji, godine)
- /v2/potpore/by-year — sport, vrsta filters
- /v2/manifestacije/meta — mjesta, razine, organizatori
- /v2/manifestacije — lista s filterima

HNS:
- 20 PGŽ priority klubova batch harvester pokrenut (HNK Goranin, HNK Orijent 1919, HNK Rijeka, NK Crikvenica, ...)
- ETA 30 min

pgz_sport_v2_router.py

@@ -4959,8 +4959,24 @@ def proracun_sport(godina: int = None):
 # ═══════════════════════════════════════════════════════
 # POTPORE — by year filter
 # ═══════════════════════════════════════════════════════
+
+
+@router.get("/potpore/meta")
+def potpore_meta():
+    """Dropdown options za Financije sekciju."""
+    sportovi = db_query("SELECT DISTINCT sport FROM pgz_sport.sufinanciranje_sport WHERE sport IS NOT NULL ORDER BY sport")
+    vrste = db_query("SELECT DISTINCT vrsta FROM pgz_sport.sufinanciranje_sport WHERE vrsta IS NOT NULL ORDER BY vrsta")
+    davatelji = db_query("SELECT DISTINCT izvor, count(*) AS broj FROM pgz_sport.sufinanciranje_sport WHERE izvor IS NOT NULL GROUP BY izvor ORDER BY broj DESC")
+    godine = db_query("SELECT DISTINCT godina, count(*) AS broj, sum(iznos_eur)::numeric(12,2) AS suma FROM pgz_sport.sufinanciranje_sport GROUP BY godina ORDER BY godina DESC")
+    return {
+        "sportovi": [r["sport"] for r in sportovi],
+        "vrste": [r["vrsta"] for r in vrste],
+        "davatelji": [r["izvor"] for r in davatelji],
+        "godine": godine,
+    }
+
 @router.get("/potpore/by-year")
-def potpore_by_year(godina: int = None, q: str = "", samo_klubovi: bool = True, davatelj: str = None):
+def potpore_by_year(godina: int = None, q: str = "", samo_klubovi: bool = True, davatelj: str = None, sport: str = None, vrsta: str = None):
     """Sufinanciranje za specifičnu godinu — samo_klubovi=True izbacuje programe/totals/services."""
     import datetime
     yr = godina or datetime.date.today().year
@@ -4972,6 +4988,12 @@ def potpore_by_year(godina: int = None, q: str = "", samo_klubovi: bool = True,
     if samo_klubovi:
         where.append("(je_klub IS NULL OR je_klub = true)")
     
+    if sport:
+        where.append("LOWER(sport) = LOWER(%s)")
+        params.append(sport)
+    if vrsta:
+        where.append("LOWER(vrsta) = LOWER(%s)")
+        params.append(vrsta)
     if davatelj == 'rijeka':
         where.append("izvor ILIKE '%%rijeka.hr%%'")
     elif davatelj == 'pgz':
@@ -5952,3 +5974,46 @@ def v2_clan_hns_profile(clan_id: int):
         "hns_url": hns_url,
     }
 
+
+@router.get("/manifestacije/meta")
+def manifestacije_meta():
+    """Dropdown options za manifestacije."""
+    mjesta = db_query("SELECT DISTINCT mjesto, count(*) AS broj FROM pgz_sport.manifestacije WHERE mjesto IS NOT NULL GROUP BY mjesto ORDER BY broj DESC LIMIT 100")
+    razine = db_query("SELECT DISTINCT razina FROM pgz_sport.manifestacije WHERE razina IS NOT NULL ORDER BY razina")
+    organizatori = db_query("SELECT DISTINCT organizator, count(*) AS broj FROM pgz_sport.manifestacije WHERE organizator IS NOT NULL GROUP BY organizator ORDER BY broj DESC LIMIT 50")
+    return {
+        "mjesta": [r["mjesto"] for r in mjesta],
+        "razine": [r["razina"] for r in razine],
+        "organizatori": [r["organizator"] for r in organizatori],
+    }
+
+@router.get("/manifestacije")
+def manifestacije_list(mjesto: str = None, razina: str = None, organizator: str = None, q: str = None, limit: int = 200):
+    """Lista manifestacija s filterima."""
+    where = ["aktivna = true"]
+    params = []
+    if mjesto:
+        where.append("mjesto = %s")
+        params.append(mjesto)
+    if razina:
+        where.append("razina = %s")
+        params.append(razina)
+    if organizator:
+        where.append("organizator ILIKE %s")
+        params.append(f"%{organizator}%")
+    if q:
+        where.append("(naziv ILIKE %s OR napomena ILIKE %s)")
+        params.extend([f"%{q}%", f"%{q}%"])
+    
+    rows = db_query(f"""
+        SELECT m.id, m.naziv, m.mjesto, m.organizator, m.razina, m.broj_ucesnika, 
+               m.godina_od, m.spol_kategorija, m.napomena, m.source_url,
+               s.naziv AS savez_naziv, s.id AS savez_id
+        FROM pgz_sport.manifestacije m
+        LEFT JOIN pgz_sport.savezi s ON s.id = m.savez_id
+        WHERE {' AND '.join(where)}
+        ORDER BY m.naziv
+        LIMIT %s
+    """, params + [limit])
+    return {"count": len(rows), "rows": rows}
+

scrapers/harvesters/eu_projekti_pgz.py

@@ -0,0 +1,67 @@
+#!/usr/bin/env python3
+"""EU projekti i fondovi PGZ."""
+import sys, json, time
+sys.path.insert(0, "/opt/pgz-sport/scrapers/harvesters")
+from _common import (fetch, extract_text, extract_title, chunk_text,
+                      upsert_facts, find_internal_links, DSN)
+from urllib.parse import urlparse
+import psycopg2
+
+EU = {
+    "eu_fondovi_pgz":      ["https://strukturnifondovi.hr/"],
+    "ri_eu_kreativnost":   ["https://rijeka2020.eu/"],
+    "rijeka_smartcity":    ["https://www.rijeka.hr/smart-city/"],
+    "agencija_unutarnji":  ["https://www.amenita.hr/"],
+    "ri_lokalna_akcijska": ["https://lag-rijeka.hr/"],
+    "agencija_jadranska":  ["https://www.adriatic-ionian.eu/"],
+    "interreg_kvarner":    ["https://www.italy-croatia.eu/"],
+    "horizon_pgz":         ["https://horizon-europe.gov.hr/"],
+    "epro_kvarner":        ["https://www.epro-kvarner.hr/"],
+    "leader_pgz":          ["https://www.leader-rijeka.hr/"],
+}
+
+
+def crawl(name, urls, max_pages=10):
+    conn = psycopg2.connect(DSN); conn.autocommit = True
+    visited = set(); queue = list(urls); facts = 0
+    while queue and len(visited) < max_pages:
+        url = queue.pop(0)
+        if url in visited: continue
+        visited.add(url)
+        html, status = fetch(url, timeout=15)
+        if not html or status != 200: continue
+        title = extract_title(html); text = extract_text(html)
+        if not text or len(text) < 200: continue
+        ff = []
+        if title and len(title) > 8:
+            ff.append({"fact": f"{name} - {title}", "url": url, "title": title})
+        for c in chunk_text(text, 800):
+            if len(c) > 100:
+                ff.append({"fact": c, "url": url, "title": title})
+        facts += upsert_facts(conn, ff, source_name=name,
+                               category="eu_projekti_pgz", confidence=0.84)
+        base = urlparse(url).hostname
+        for link in find_internal_links(html, url):
+            if link not in visited and (urlparse(link).hostname or "") == base and len(queue) < 25:
+                queue.append(link)
+        time.sleep(0.5)
+    conn.close()
+    return {"name": name, "visited": len(visited), "facts": facts}
+
+
+def main():
+    results = []
+    for name, urls in EU.items():
+        try:
+            r = crawl(name, urls, max_pages=10)
+            print(f"  {name:25} {r['visited']:>3}p  {r['facts']:>5}f")
+            results.append(r)
+        except Exception as e:
+            print(f"  {name:25} FAIL: {str(e)[:60]}")
+    total = sum(r.get("facts", 0) for r in results)
+    print(f"=== TOTAL: {total} ===")
+    print(json.dumps({"eu_count": len(results), "total_facts": total}))
+
+
+if __name__ == "__main__":
+    main()

scrapers/harvesters/gov_deep_pgz.py

@@ -0,0 +1,68 @@
+#!/usr/bin/env python3
+"""Gov.hr deep — DZS PxWeb, sudreg portal, sukobinteresa."""
+import sys, json, time
+sys.path.insert(0, "/opt/pgz-sport/scrapers/harvesters")
+from _common import (fetch, extract_text, extract_title, chunk_text,
+                      upsert_facts, find_internal_links, DSN)
+from urllib.parse import urlparse
+import psycopg2
+
+GOV = {
+    "dzs_kvarner":         ["https://podaci.dzs.hr/hr/podaci/stanovnistvo/popis-stanovnistva/"],
+    "sukobinteresa_pgz":   ["https://www.sukobinteresa.hr/"],
+    "drzavna_revizija":    ["https://www.revizija.hr/"],
+    "drzavna_komisija_kn": ["https://www.dkom.hr/"],
+    "fina_pgz":            ["https://www.fina.hr/poslovni-subjekti"],
+    "fina_javne_objave":   ["https://www.fina.hr/javne-objave-uprave"],
+    "porezna_uprava_ri":   ["https://www.porezna-uprava.hr/"],
+    "ministarstvo_pgz":    ["https://www.gov.hr/hr/ministarstva"],
+    "vlada_hr_pgz":        ["https://vlada.gov.hr/"],
+    "uprava_polic_pgz":    ["https://policija.gov.hr/policijska-uprava-primorsko-goranska/"],
+    "vis_hr":              ["https://www.vis-hr.com/"],
+}
+
+
+def crawl(name, urls, max_pages=10):
+    conn = psycopg2.connect(DSN); conn.autocommit = True
+    visited = set(); queue = list(urls); facts = 0
+    while queue and len(visited) < max_pages:
+        url = queue.pop(0)
+        if url in visited: continue
+        visited.add(url)
+        html, status = fetch(url, timeout=15)
+        if not html or status != 200: continue
+        title = extract_title(html); text = extract_text(html)
+        if not text or len(text) < 200: continue
+        ff = []
+        if title and len(title) > 8:
+            ff.append({"fact": f"{name} - {title}", "url": url, "title": title})
+        for c in chunk_text(text, 800):
+            if len(c) > 100:
+                ff.append({"fact": c, "url": url, "title": title})
+        facts += upsert_facts(conn, ff, source_name=name,
+                               category="gov_registri_pgz", confidence=0.88)
+        base = urlparse(url).hostname
+        for link in find_internal_links(html, url):
+            if link not in visited and (urlparse(link).hostname or "") == base and len(queue) < 30:
+                queue.append(link)
+        time.sleep(0.5)
+    conn.close()
+    return {"name": name, "visited": len(visited), "facts": facts}
+
+
+def main():
+    results = []
+    for name, urls in GOV.items():
+        try:
+            r = crawl(name, urls, max_pages=10)
+            print(f"  {name:25} {r['visited']:>3}p  {r['facts']:>5}f")
+            results.append(r)
+        except Exception as e:
+            print(f"  {name:25} FAIL: {str(e)[:60]}")
+    total = sum(r.get("facts", 0) for r in results)
+    print(f"=== TOTAL: {total} ===")
+    print(json.dumps({"gov_count": len(results), "total_facts": total}))
+
+
+if __name__ == "__main__":
+    main()

scrapers/harvesters/hoteli_pgz.py

@@ -0,0 +1,68 @@
+#!/usr/bin/env python3
+"""Hoteli, restorani, smjestaj PGZ."""
+import sys, json, time
+sys.path.insert(0, "/opt/pgz-sport/scrapers/harvesters")
+from _common import (fetch, extract_text, extract_title, chunk_text,
+                      upsert_facts, find_internal_links, DSN)
+from urllib.parse import urlparse
+import psycopg2
+
+HOTELS = {
+    "milenij_hoteli":      ["https://www.milenijhoteli.hr/"],
+    "uniline_kvarner":     ["https://www.uniline.hr/destinacije/kvarner"],
+    "ri_botel":            ["https://www.botel-marina.hr/"],
+    "jadran_hoteli":       ["https://www.jadran-crikvenica.hr/"],
+    "valamar_kvarner":     ["https://www.valamar.com/hr/odredista/kvarner"],
+    "kongres_centar":      ["https://www.adriatic-grandhotel.com/"],
+    "lovran_riviera":      ["https://www.liburnia.hr/"],
+    "remisens_grupa":      ["https://www.remisens.com/"],
+    "imperijal_opatija":   ["https://www.amadriapark.com/"],
+    "hotel_neboder":       ["https://www.jadran-hoteli.hr/"],
+    "kanjon_zrmanja":      ["https://www.kvarnertravel.hr/"],
+}
+
+
+def crawl(name, urls, max_pages=8):
+    conn = psycopg2.connect(DSN); conn.autocommit = True
+    visited = set(); queue = list(urls); facts = 0
+    while queue and len(visited) < max_pages:
+        url = queue.pop(0)
+        if url in visited: continue
+        visited.add(url)
+        html, status = fetch(url, timeout=15)
+        if not html or status != 200: continue
+        title = extract_title(html); text = extract_text(html)
+        if not text or len(text) < 200: continue
+        ff = []
+        if title and len(title) > 8:
+            ff.append({"fact": f"{name} - {title}", "url": url, "title": title})
+        for c in chunk_text(text, 800):
+            if len(c) > 100:
+                ff.append({"fact": c, "url": url, "title": title})
+        facts += upsert_facts(conn, ff, source_name=name,
+                               category="hoteli_pgz", confidence=0.83)
+        base = urlparse(url).hostname
+        for link in find_internal_links(html, url):
+            if link not in visited and (urlparse(link).hostname or "") == base and len(queue) < 25:
+                queue.append(link)
+        time.sleep(0.5)
+    conn.close()
+    return {"name": name, "visited": len(visited), "facts": facts}
+
+
+def main():
+    results = []
+    for name, urls in HOTELS.items():
+        try:
+            r = crawl(name, urls, max_pages=8)
+            print(f"  {name:25} {r['visited']:>3}p  {r['facts']:>5}f")
+            results.append(r)
+        except Exception as e:
+            print(f"  {name:25} FAIL: {str(e)[:60]}")
+    total = sum(r.get("facts", 0) for r in results)
+    print(f"=== TOTAL: {total} ===")
+    print(json.dumps({"hotel_count": len(results), "total_facts": total}))
+
+
+if __name__ == "__main__":
+    main()

scrapers/harvesters/politika_pgz.py

@@ -0,0 +1,68 @@
+#!/usr/bin/env python3
+"""Politicke stranke i skupstine PGZ."""
+import sys, json, time
+sys.path.insert(0, "/opt/pgz-sport/scrapers/harvesters")
+from _common import (fetch, extract_text, extract_title, chunk_text,
+                      upsert_facts, find_internal_links, DSN)
+from urllib.parse import urlparse
+import psycopg2
+
+POL = {
+    "skupstina_pgz":       ["https://www.pgz.hr/skupstina-zupanije/"],
+    "rijeka_gradsko_vije": ["https://www.rijeka.hr/gradsko-vijece/"],
+    "stranka_lkr_pgz":     ["https://www.lista-rijeka.com/"],
+    "sdp_rijeka":          ["https://www.sdp.hr/podruznice/sdp-rijeka/"],
+    "hdz_pgz":             ["https://hdz.hr/zupanije/primorsko-goranska/"],
+    "most_pgz":            ["https://most.hr/regionalni-uredi"],
+    "akcija_mladih":       ["https://akcijamladih.hr/"],
+    "rastimo_zajedno":     ["https://rastimozajedno.hr/"],
+    "javnost_pgz":         ["https://www.izbori.hr/"],
+    "zupan_obrenovic":     ["https://www.pgz.hr/zupan/"],
+    "gradonacelnica_iv_r": ["https://www.rijeka.hr/gradonacelnica/"],
+}
+
+
+def crawl(name, urls, max_pages=10):
+    conn = psycopg2.connect(DSN); conn.autocommit = True
+    visited = set(); queue = list(urls); facts = 0
+    while queue and len(visited) < max_pages:
+        url = queue.pop(0)
+        if url in visited: continue
+        visited.add(url)
+        html, status = fetch(url, timeout=15)
+        if not html or status != 200: continue
+        title = extract_title(html); text = extract_text(html)
+        if not text or len(text) < 200: continue
+        ff = []
+        if title and len(title) > 8:
+            ff.append({"fact": f"{name} - {title}", "url": url, "title": title})
+        for c in chunk_text(text, 800):
+            if len(c) > 100:
+                ff.append({"fact": c, "url": url, "title": title})
+        facts += upsert_facts(conn, ff, source_name=name,
+                               category="politika_pgz", confidence=0.85)
+        base = urlparse(url).hostname
+        for link in find_internal_links(html, url):
+            if link not in visited and (urlparse(link).hostname or "") == base and len(queue) < 30:
+                queue.append(link)
+        time.sleep(0.5)
+    conn.close()
+    return {"name": name, "visited": len(visited), "facts": facts}
+
+
+def main():
+    results = []
+    for name, urls in POL.items():
+        try:
+            r = crawl(name, urls, max_pages=10)
+            print(f"  {name:25} {r['visited']:>3}p  {r['facts']:>5}f")
+            results.append(r)
+        except Exception as e:
+            print(f"  {name:25} FAIL: {str(e)[:60]}")
+    total = sum(r.get("facts", 0) for r in results)
+    print(f"=== TOTAL: {total} ===")
+    print(json.dumps({"pol_count": len(results), "total_facts": total}))
+
+
+if __name__ == "__main__":
+    main()

static/sport2.html

@@ -1397,6 +1397,7 @@ function applySaveziFilter(){
   else if(fKat==='gradski') rows = rows.filter(s => /gradsk/i.test(s.razina||''));
   if(pgz==='1') rows = rows.filter(s => s.pgz_relevant);
   if(_sort.savezi) rows = sortRows(rows, _sort.savezi.key, _sort.savezi.dir);
+  _filtersUpdateCount('savezi', rows.length);
   $('#sav-cnt').textContent = rows.length+' saveza';
   $('#sav-out').innerHTML = _state.viewSavezi==='card' ? renderSaveziGrid(rows) : renderSaveziTable(rows);
 }
@@ -1887,13 +1888,28 @@ async function loadSportasi(){
   const root = $('#pg-sportasi');
   if(!_cache.clanovi){
     root.innerHTML = '<div class="loading">Učitavanje sportaša…</div>';
-    // PGŽ filter: switch to v2 priority-sort (joins club, marks priority)
-    const url = window._pgz_filter_priority
-      ? '/v2/clanovi/priority-sort?only=true&limit=2000'
-      : '/clanovi-full?limit=500';
+    // BUG-E (2026-05-05): explicit filter via /v2/sportasi/filtered.
+    // Priority + HNS profil + godina_rod_od/do are server-side params.
+    const f = _filters.sportasi;
+    const useFiltered = f.priority || f.hns_profil || f.godina_od || f.godina_do;
+    let url;
+    if(useFiltered){
+      const qs = new URLSearchParams();
+      qs.set('limit','2000');
+      if(f.priority)   qs.set('samo_priority','true');
+      if(f.hns_profil) qs.set('samo_s_hns','true');
+      if(f.godina_od)  qs.set('godina_rod_od', String(f.godina_od));
+      if(f.godina_do)  qs.set('godina_rod_do', String(f.godina_do));
+      url = '/v2/sportasi/filtered?'+qs.toString();
+    } else if(window._pgz_filter_priority){
+      url = '/v2/clanovi/priority-sort?only=true&limit=2000';
+    } else {
+      url = '/clanovi-full?limit=500';
+    }
     const d = await api(url);
     if(!d){ root.innerHTML='<div class="empty">Greška pri dohvatu</div>'; return; }
     _cache.clanovi = d.rows || [];
+    _filters.sportasi.total = (d.rows||[]).length;
   }
   renderSportasiShell();
   applySportasiFilter();
@@ -1901,9 +1917,10 @@ async function loadSportasi(){
 function renderSportasiShell(){
   const root = $('#pg-sportasi');
   const sports = Array.from(new Set((_cache.clanovi||[]).map(c=>c.sport).filter(Boolean))).sort();
-  const klubovi = Array.from(new Map((_cache.clanovi||[]).filter(c=>c.klub_id).map(c=>[c.klub_id, c.klub_naziv_godisnjak||('Klub #'+c.klub_id)])).entries()).sort((a,b)=>String(a[1]).localeCompare(String(b[1]),'hr'));
+  const klubovi = Array.from(new Map((_cache.clanovi||[]).filter(c=>c.klub_id).map(c=>[c.klub_id, c.klub_naziv_godisnjak||c.klub_naziv||('Klub #'+c.klub_id)])).entries()).sort((a,b)=>String(a[1]).localeCompare(String(b[1]),'hr'));
   const kats = Array.from(new Set((_cache.clanovi||[]).flatMap(c => (c.kategorije && c.kategorije.length ? c.kategorije : [c.kategorija]).filter(Boolean)))).sort();
   root.innerHTML = `
+    ${_filtersBar('sportasi')}
     <div class="toolbar">
       <input type="search" id="sp-q" placeholder="🔍 Ime ili prezime…">
       <select id="sp-sport"><option value="">Svi sportovi</option>${sports.map(s=>'<option value="'+esc(s)+'">'+esc(s)+'</option>').join('')}</select>
@@ -2050,6 +2067,7 @@ function applySportasiFilter(){
   if(_state.spExtraAktivan==='false') rows = rows.filter(c => !c.aktivan);
   if(_state.spExtraStipendiran) rows = rows.filter(c => c.stipendiran);
   if(_sort.sportasi) rows = sortRows(rows, _sort.sportasi.key, _sort.sportasi.dir);
+  _filtersUpdateCount('sportasi', rows.length);
   $('#sp-cnt').textContent = rows.length+' sportaša';
   const top = rows.slice(0, 300);
   $('#sp-out').innerHTML = _state.viewSportasi==='card' ? renderSportasiGrid(top) : renderSportasiTable(top);