app.py

import gradio as gr
import requests
from datetime import datetime, timedelta
import pycountry
import json

# NewsAPI key
API_KEY = "37d83e266422487b8b2e4cb6e1ff0aa6"


COUNTRY_CODES = "ae ar at au be bg br ca ch cn co cu cz de eg fr gb gr hk hu id ie il in it jp kr lt lv ma mx my ng nl no nz ph pl pt ro rs ru sa se sg si sk th tr tw ua us ve za".split()

COUNTRIES = {'all': 'All Countries'}
COUNTRIES.update({code: pycountry.countries.get(alpha_2=code.upper()).name for code in COUNTRY_CODES if pycountry.countries.get(alpha_2=code.upper())})

def get_related_keywords(keyword):
    return [f"{keyword} news", f"{keyword} technology", f"{keyword} company", f"latest {keyword}"]

def get_news(keyword, article_count, country):
    country_code = next((code for code, name in COUNTRIES.items() if name == country), 'all')
    
    if country_code == 'all':
        base_url = "https://newsapi.org/v2/everything"
        params = {
            'apiKey': API_KEY,
            'q': keyword,
            'language': 'en',
            'sortBy': 'publishedAt',
            'pageSize': min(article_count * 2, 100)  # 요청 기사 수를 2배로 늘림 (최대 100)
        }
    else:
        base_url = "https://newsapi.org/v2/top-headlines"
        params = {
            'apiKey': API_KEY,
            'q': keyword,
            'country': country_code,
            'pageSize': min(article_count * 2, 100)  # 요청 기사 수를 2배로 늘림 (최대 100)
        }
    
    two_days_ago = (datetime.utcnow() - timedelta(hours=48)).isoformat()
    params['from'] = two_days_ago

    debug_info = f"API Request URL: {base_url}\n"
    debug_info += f"Parameters: {json.dumps(params, indent=2)}\n\n"

    try:
        response = requests.get(base_url, params=params, timeout=10)
        response.raise_for_status()
        news_data = response.json()

        debug_info += f"API Response Status: {response.status_code}\n"
        debug_info += f"API Response Headers: {json.dumps(dict(response.headers), indent=2)}\n\n"
        debug_info += f"API Response Body: {json.dumps(news_data, indent=2)}\n\n"

    except requests.RequestException as e:
        return f"<p style='color: red;'>Error fetching news: {str(e)}</p><pre>{debug_info}</pre>"

    if news_data['status'] != 'ok':
        return f"<p style='color: red;'>API Error: {news_data.get('message', 'Unknown error occurred')}</p><pre>{debug_info}</pre>"

    articles = news_data['articles']
    
    # removed 기사 필터링
    filtered_articles = [article for article in articles if article.get('title') != '[Removed]' and article.get('description') != '[Removed]']
    
    if not filtered_articles:
        if country_code != 'all':
            # 특정 국가에서 결과가 없을 경우, 전체 국가에서 검색
            return get_news(keyword, article_count, 'All Countries')
        else:
            related_keywords = get_related_keywords(keyword)
            suggestions = (f"<p>No news found for the keyword '<strong>{keyword}</strong>' in {country}.</p>"
                           f"<p>Suggestions:</p>"
                           f"<ul>"
                           f"<li>Try one of these related keywords: {', '.join(related_keywords)}</li>"
                           f"<li>Increase the number of articles</li>"
                           f"<li>Check for any spelling errors in your keyword</li>"
                           f"</ul>")
            return suggestions + f"<pre>{debug_info}</pre>"

    html_output = f"<h2>News results for '{keyword}' in {country}</h2>"
    if country_code == 'all':
        html_output += "<p><em>Showing results from all countries</em></p>"
    else:
        html_output += f"<p><em>Showing results specifically for {country}</em></p>"
    
    html_output += f"<p>Found {len(filtered_articles)} relevant articles (after removing any '[Removed]' articles)</p>"
    
    for article in filtered_articles[:article_count]:  # 요청한 기사 수만큼만 표시
        title = article['title']
        link = article['url']
        pub_date = datetime.strptime(article['publishedAt'], "%Y-%m-%dT%H:%M:%SZ")
        source = article.get('source', {}).get('name', 'Unknown Source')
        description = article.get('description', 'No description available')
        
        html_output += f"""
        <div style='margin-bottom: 20px; padding: 10px; border: 1px solid #ddd; border-radius: 5px;'>
            <h3><a href='{link}' target='_blank' style='text-decoration: none; color: #1a0dab;'>{title}</a></h3>
            <p style='color: #006621;'>{source}</p>
            <p style='color: #545454;'>{pub_date.strftime('%Y-%m-%d %H:%M:%S')}</p>
            <p>{description}</p>
        </div>
        """

    html_output += f"<details><summary>Debug Info</summary><pre>{debug_info}</pre></details>"
    return html_output

iface = gr.Interface(
    fn=get_news,
    inputs=[
        gr.Textbox(label="Enter keyword"),
        gr.Slider(minimum=10, maximum=100, step=10, value=10, label="Number of Articles"),
        gr.Dropdown(choices=list(COUNTRIES.values()), value="All Countries", label="Select Country")
    ],
    outputs=gr.HTML(),
    title="Advanced Visual News Search",
    description="Search for news articles using NewsAPI. Results will be specific to the selected country when applicable.",
    theme=gr.themes.Soft()
)

iface.launch()