damir
f07fdad919
DB: - pgz_sport.sufinanciranje_sport.je_klub flag (RSS programi/totals false) - pgz_sport.sufinanciranje_sport.klub_id matched Endpoints: - /v2/potpore/by-year: samo_klubovi=True default + davatelj filter Frontend: - sport2.html PANEL FORCE HIDE CSS (right:-100vw default) - crm_v2.html: redirect to /login only on actual 401, not on page load
66 lines
2.7 KiB
Python
66 lines
2.7 KiB
Python
#!/usr/bin/env python3
|
|
"""Media deep crawl — full pages of local portals."""
|
|
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
|
|
|
|
MEDIA = {
|
|
"novilist_rijeka": ["https://www.novilist.hr/rijeka/", "https://www.novilist.hr/regija/"],
|
|
"rijekadanas_full": ["https://rijekadanas.com/category/rijeka/", "https://rijekadanas.com/category/pgz/"],
|
|
"rijekain_full": ["https://rijekain.hr/category/rijeka/"],
|
|
"primorske_full": ["https://primorskenovice.hr/"],
|
|
"rkc_blog": ["https://www.rkcrijeka.hr/blog/"],
|
|
"rijeka2020_arhiva": ["https://rijeka2020.eu/category/news/"],
|
|
"kulturpunkt_ri": ["https://www.kulturpunkt.hr/tag/rijeka"],
|
|
"5portala_hr_pgz": ["https://www.5portala.hr/regije/primorsko-goranska/"],
|
|
}
|
|
|
|
|
|
def crawl(name, urls, max_pages=20):
|
|
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) < 300: continue
|
|
ff = []
|
|
if title and len(title) > 12:
|
|
ff.append({"fact": f"[{name}] {title}", "url": url, "title": title})
|
|
for c in chunk_text(text, 800):
|
|
if len(c) > 120:
|
|
ff.append({"fact": c, "url": url, "title": title})
|
|
facts += upsert_facts(conn, ff, source_name=f"media_{name}",
|
|
category="media_pgz_deep", confidence=0.82)
|
|
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) < 60:
|
|
queue.append(link)
|
|
time.sleep(0.6)
|
|
conn.close()
|
|
return {"name": name, "visited": len(visited), "facts": facts}
|
|
|
|
|
|
def main():
|
|
results = []
|
|
for name, urls in MEDIA.items():
|
|
try:
|
|
r = crawl(name, urls, max_pages=18)
|
|
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({"media_count": len(results), "total_facts": total}))
|
|
|
|
|
|
if __name__ == "__main__":
|
|
main()
|