Spaces:
Sleeping
Sleeping
Create services/google_news.py
Browse files- services/google_news.py +106 -0
services/google_news.py
ADDED
|
@@ -0,0 +1,106 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
"""
|
| 2 |
+
services/google_news.py
|
| 3 |
+
Ambil berita Google News berdasarkan keyword menggunakan SerpApi.
|
| 4 |
+
|
| 5 |
+
Data yang diambil:
|
| 6 |
+
- title β judul berita
|
| 7 |
+
- snippet β ringkasan/deskripsi berita
|
| 8 |
+
- source β nama media
|
| 9 |
+
- date β tanggal terbit
|
| 10 |
+
"""
|
| 11 |
+
|
| 12 |
+
import os
|
| 13 |
+
|
| 14 |
+
# ββ CONFIG ββ
|
| 15 |
+
SERPAPI_KEY = os.getenv("SERPAPI_KEY") # set di HF Spaces Secrets
|
| 16 |
+
MAX_RESULTS = 30 # jumlah berita per query
|
| 17 |
+
|
| 18 |
+
|
| 19 |
+
# ββββββββββββββββββββββββββββββββββββββββββββ
|
| 20 |
+
# INTERNAL HELPERS
|
| 21 |
+
# ββββββββββββββββββββββββββββββββββββββββββββ
|
| 22 |
+
|
| 23 |
+
def _extract_texts(news_results: list) -> list[str]:
|
| 24 |
+
"""
|
| 25 |
+
Ekstrak teks dari hasil Google News SerpApi.
|
| 26 |
+
|
| 27 |
+
Struktur item:
|
| 28 |
+
item["title"] β judul berita
|
| 29 |
+
item["snippet"] β ringkasan berita
|
| 30 |
+
item["source"]["name"] β nama media (opsional)
|
| 31 |
+
item["stories"] β sub-artikel terkait (opsional)
|
| 32 |
+
.title / .snippet
|
| 33 |
+
"""
|
| 34 |
+
texts = []
|
| 35 |
+
|
| 36 |
+
for item in news_results:
|
| 37 |
+
# Judul
|
| 38 |
+
title = (item.get("title") or "").strip()
|
| 39 |
+
if title and len(title) > 5:
|
| 40 |
+
texts.append(title)
|
| 41 |
+
|
| 42 |
+
# Snippet / ringkasan
|
| 43 |
+
snippet = (item.get("snippet") or item.get("description") or "").strip()
|
| 44 |
+
if snippet and len(snippet) > 10:
|
| 45 |
+
texts.append(snippet)
|
| 46 |
+
|
| 47 |
+
# Sub-stories (Google News kadang mengelompokkan berita terkait)
|
| 48 |
+
for story in item.get("stories", []):
|
| 49 |
+
if isinstance(story, dict):
|
| 50 |
+
st = (story.get("title") or "").strip()
|
| 51 |
+
ss = (story.get("snippet") or "").strip()
|
| 52 |
+
if st and len(st) > 5:
|
| 53 |
+
texts.append(st)
|
| 54 |
+
if ss and len(ss) > 10:
|
| 55 |
+
texts.append(ss)
|
| 56 |
+
|
| 57 |
+
return texts
|
| 58 |
+
|
| 59 |
+
|
| 60 |
+
# ββββββββββββββββββββββββββββββββββββββββββββ
|
| 61 |
+
# PUBLIC INTERFACE
|
| 62 |
+
# ββββββββββββββββββββββββββββββββββββββββββββ
|
| 63 |
+
|
| 64 |
+
def get_google_news(keyword: str) -> list[str]:
|
| 65 |
+
"""
|
| 66 |
+
Scrape Google News berdasarkan keyword via SerpApi.
|
| 67 |
+
Return list of string (judul + snippet berita).
|
| 68 |
+
|
| 69 |
+
Dipanggil dari services/aggregator.py.
|
| 70 |
+
"""
|
| 71 |
+
if not SERPAPI_KEY:
|
| 72 |
+
print("β οΈ SERPAPI_KEY tidak diset β skip Google News")
|
| 73 |
+
return []
|
| 74 |
+
|
| 75 |
+
try:
|
| 76 |
+
from serpapi import GoogleSearch
|
| 77 |
+
except ImportError:
|
| 78 |
+
print("β serpapi package tidak terinstall β tambahkan ke requirements.txt")
|
| 79 |
+
return []
|
| 80 |
+
|
| 81 |
+
try:
|
| 82 |
+
params = {
|
| 83 |
+
"engine": "google_news",
|
| 84 |
+
"q": keyword,
|
| 85 |
+
"hl": "id", # bahasa Indonesia
|
| 86 |
+
"gl": "id", # region Indonesia
|
| 87 |
+
"num": MAX_RESULTS,
|
| 88 |
+
"api_key": SERPAPI_KEY,
|
| 89 |
+
}
|
| 90 |
+
|
| 91 |
+
search = GoogleSearch(params)
|
| 92 |
+
results = search.get_dict()
|
| 93 |
+
|
| 94 |
+
news_results = results.get("news_results", [])
|
| 95 |
+
|
| 96 |
+
if not news_results:
|
| 97 |
+
print(f"β οΈ Google News: tidak ada hasil untuk '{keyword}'")
|
| 98 |
+
return []
|
| 99 |
+
|
| 100 |
+
texts = _extract_texts(news_results)
|
| 101 |
+
print(f"β
Google News: {len(texts)} teks dari {len(news_results)} berita")
|
| 102 |
+
return texts
|
| 103 |
+
|
| 104 |
+
except Exception as e:
|
| 105 |
+
print(f"β Google News error: {e}")
|
| 106 |
+
return []
|