Create app.py

app.py

@@ -0,0 +1,136 @@
+import gradio as gr
+import requests
+from bs4 import BeautifulSoup
+import json
+import time
+import logging
+import re
+from urllib.parse import urlparse
+from groq import Groq
+from requests.exceptions import HTTPError, ReadTimeout
+from http.client import RemoteDisconnected
+import os
+from datetime import datetime
+
+# === KONFIGURASI ===
+GAS_URL = os.getenv("GAS_URL", "https://script.google.com/macros/s/AKfycbwstcoUh2CQmuoTgxapW9cUhzQFx6glp25DaCqrvBdwKrb77wqeMN0RzB8UMpiAQ2PtQA/exec")
+GROQ_API_KEY ="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"
+TELEGRAM_BOT_TOKEN = os.getenv("TELEGRAM_BOT_TOKEN", "7166094967:AAHb5S2hN6L527y1-GoXPzBdU4RB8jnYelk")
+TELEGRAM_CHAT_ID = os.getenv("TELEGRAM_CHAT_ID", "6929677613")
+GROQ_MODEL = "gemma2-9b-it"
+REQUEST_TIMEOUT = 10
+GROQ_TIMEOUT = 30
+RETRY_BACKOFF_FACTOR = 2
+MAX_RETRIES = 3
+DELAY_BETWEEN_REQUESTS = 3
+
+log_file = f"scrape_rewrite_{datetime.now().strftime('%Y%m%d_%H%M%S')}.log"
+logging.basicConfig(
+    level=logging.INFO,
+    format="%(asctime)s - %(levelname)s - %(message)s",
+    handlers=[logging.FileHandler(log_file, encoding="utf-8"), logging.StreamHandler()]
+)
+
+client = Groq(api_key=GROQ_API_KEY)
+
+# --- FUNGSI PENDUKUNG ---
+class RateLimitExceeded(Exception):
+    pass
+
+def send_telegram_message(message):
+    try:
+        url = f"https://api.telegram.org/bot{TELEGRAM_BOT_TOKEN}/sendMessage"
+        payload = {"chat_id": TELEGRAM_CHAT_ID, "text": message, "parse_mode": "Markdown"}
+        requests.post(url, json=payload, timeout=5)
+    except: pass
+
+def is_valid_url(url):
+    try:
+        result = urlparse(url)
+        return all([result.scheme in ['http', 'https'], result.netloc])
+    except: return False
+
+def is_valid_html(html):
+    return html and html.strip().startswith('<article') and html.strip().endswith('</article>')
+
+def retry_request(func, *args, **kwargs):
+    for attempt in range(MAX_RETRIES):
+        try:
+            return func(*args, **kwargs)
+        except (HTTPError, RemoteDisconnected, ReadTimeout):
+            time.sleep(RETRY_BACKOFF_FACTOR ** attempt)
+    raise Exception("Max retries exceeded")
+
+def fetch_links(sheet_name="Sheet2"):
+    try:
+        def get_links():
+            resp = requests.get(GAS_URL, params={"sheetName": sheet_name}, timeout=REQUEST_TIMEOUT)
+            return resp.json()
+        data = retry_request(get_links)
+        return [item for item in data if not item.get("judul") and is_valid_url(item.get("link"))]
+    except: return []
+
+def clean_html(soup):
+    for tag in soup(["script", "iframe"]): tag.decompose()
+    return soup
+
+def extract_main_image(soup):
+    try:
+        img = soup.select_one('article img')
+        return img['src'] if img and is_valid_url(img['src']) else ""
+    except: return ""
+
+def scrape_detik(link):
+    try:
+        resp = retry_request(lambda: requests.get(link, timeout=REQUEST_TIMEOUT))
+        soup = BeautifulSoup(resp.text, 'html.parser')
+        content = soup.select_one('article')
+        image_url = extract_main_image(soup)
+        cleaned = clean_html(content)
+        return cleaned.get_text("\n", strip=True), image_url
+    except: return None, None
+
+def rewrite_with_ai(text, image_url):
+    prompt = f"""Tulis ulang artikel berikut menjadi HTML <article>...\n\n{text}"""
+    try:
+        completion = client.chat.completions.create(
+            model=GROQ_MODEL,
+            messages=[{"role": "user", "content": prompt}],
+            stream=True
+        )
+        html = "".join(chunk.choices[0].delta.content or "" for chunk in completion)
+        return html if is_valid_html(html) else None
+    except: return None
+
+def extract_title_from_html(html):
+    try:
+        soup = BeautifulSoup(html, 'html.parser')
+        return soup.find('h2').get_text(strip=True)
+    except: return "Tanpa Judul"
+
+def kirim_ke_sheet(judul, konten_html, link):
+    try:
+        payload = {"method": "updateRowByLink", "link": link, "judul": judul, "konten": konten_html}
+        retry_request(lambda: requests.post(GAS_URL, json=payload, timeout=REQUEST_TIMEOUT))
+    except: pass
+
+# --- FUNGSI UTAMA YANG DIPANGGIL ---
+def jalankan_script():
+    processed = 0
+    rows = fetch_links()
+    for row in rows[:3]:  # BATAS 3 LINK
+        link = row['link']
+        artikel, img = scrape_detik(link)
+        if not artikel: continue
+        hasil = rewrite_with_ai(artikel, img)
+        if not hasil: continue
+        judul = extract_title_from_html(hasil)
+        kirim_ke_sheet(judul, hasil, link)
+        processed += 1
+        time.sleep(DELAY_BETWEEN_REQUESTS)
+    send_telegram_message(f"Selesai! {processed} artikel diproses.")
+    return f"Sukses: {processed} artikel diproses"
+
+# --- GRADIO UNTUK HUGGING FACE ---
+iface = gr.Interface(fn=jalankan_script, inputs=[], outputs="text")
+app = gr.mount_gradio_app(app=None, blocks=iface, path="/run")