Update app.py

app.py

@@ -1,15 +1,16 @@
 import gradio as gr
 import requests
-from bs4 import BeautifulSoup
 from datetime import datetime, timedelta
-import pytz
+
+# NewsAPI key (이것을 실제 API 키로 대체해야 합니다)
+API_KEY = "37d83e266422487b8b2e4cb6e1ff0aa6"
 
 SUPPORTED_COUNTRIES = {
-    'United States': 'US',
-    'United Kingdom': 'GB',
-    'Australia': 'AU',
-    'India': 'IN',
-    'Canada': 'CA'
+    'United States': 'us',
+    'United Kingdom': 'gb',
+    'Australia': 'au',
+    'India': 'in',
+    'Canada': 'ca'
 }
 
 def get_news(country, keyword):
@@ -17,39 +18,44 @@ def get_news(country, keyword):
     if not country_code:
         return "Selected country is not supported."
     
-    base_url = f"https://news.google.com/search?q={keyword}&hl=en-{country_code}&gl={country_code}&ceid={country_code}:en"
+    base_url = "https://newsapi.org/v2/top-headlines"
+    
+    # 48시간 전의 날짜를 ISO 형식으로 얻기
+    two_days_ago = (datetime.utcnow() - timedelta(hours=48)).isoformat()
+
+    params = {
+        'apiKey': API_KEY,
+        'country': country_code,
+        'q': keyword,
+        'from': two_days_ago,
+        'language': 'en'
+    }
     
     try:
-        response = requests.get(base_url, timeout=10)
+        response = requests.get(base_url, params=params, timeout=10)
         response.raise_for_status()
+        news_data = response.json()
     except requests.RequestException as e:
         return f"Error fetching news: {str(e)}"
 
-    soup = BeautifulSoup(response.text, 'html.parser')
-    articles = soup.find_all('article')
-    
-    time_threshold = datetime.now(pytz.utc) - timedelta(hours=48)
-    
-    filtered_news = []
-    for article in articles:
-        title_elem = article.find('h3', class_='ipQwMb ekueJc RD0gLb')
-        link_elem = article.find('a', class_='VDXfz')
-        time_elem = article.find('time', class_='WW6dff uQIVzc Sksgp')
-        
-        if title_elem and link_elem and time_elem:
-            title = title_elem.text
-            link = f"https://news.google.com{link_elem['href'][1:]}"
-            pub_time = datetime.fromtimestamp(int(time_elem['datetime'])/1000, pytz.utc)
-            
-            if pub_time > time_threshold:
-                filtered_news.append(f"Title: {title}\nLink: {link}\nDate: {pub_time}\n")
+    if news_data['status'] != 'ok':
+        return f"API Error: {news_data.get('message', 'Unknown error occurred')}"
+
+    articles = news_data['articles']
     
-    if filtered_news:
-        return "\n".join(filtered_news)
-    else:
+    if not articles:
         return (f"No recent news found for the keyword '{keyword}' in {country} "
                 f"within the last 48 hours.\nTry a different keyword or check back later.")
 
+    filtered_news = []
+    for article in articles:
+        title = article['title']
+        link = article['url']
+        pub_date = datetime.strptime(article['publishedAt'], "%Y-%m-%dT%H:%M:%SZ")
+        filtered_news.append(f"Title: {title}\nLink: {link}\nDate: {pub_date}\n")
+
+    return "\n".join(filtered_news)
+
 iface = gr.Interface(
     fn=get_news,
     inputs=[
@@ -57,8 +63,8 @@ iface = gr.Interface(
         gr.Textbox(label="Enter keyword")
     ],
     outputs="text",
-    title="Google News Search",
-    description="Search for news articles from the last 48 hours using Google News."
+    title="News Search",
+    description="Search for news articles from the last 48 hours using NewsAPI."
 )
 
 iface.launch()