app.py

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

API_KEY = "V38CNn4HXpLtynJQyOeoUensTEYoFy8PBUxKpDqAW1pawT1vfJ2BWtPQ98h6"

MAJOR_COUNTRIES = [
    "United States", "United Kingdom", "Canada", "Australia", "Germany",
    "France", "Japan", "South Korea", "China", "India",
    "Brazil", "Mexico", "Russia", "Italy", "Spain",
    "Netherlands", "Sweden", "Switzerland", "Norway", "Denmark",
    "Finland", "Belgium", "Austria", "New Zealand", "Ireland",
    "Singapore", "Hong Kong", "Israel", "United Arab Emirates", "Saudi Arabia",
    "South Africa", "Turkey", "Egypt", "Poland", "Czech Republic",
    "Hungary", "Greece", "Portugal", "Argentina", "Chile",
    "Colombia", "Peru", "Venezuela", "Thailand", "Malaysia",
    "Indonesia", "Philippines", "Vietnam", "Pakistan", "Bangladesh"
]

def search_serphouse(query, country, page, num_result):
    url = "https://api.serphouse.com/serp/live"

    now = datetime.utcnow()
    yesterday = now - timedelta(days=1)
    date_range = f"{yesterday.strftime('%Y-%m-%d')},{now.strftime('%Y-%m-%d')}"

    payload = {
        "data": {
            "q": query,
            "domain": "google.com",
            "loc": country,
            "lang": "en",
            "device": "desktop",
            "serp_type": "news",
            "page": str(page),
            "verbatim": "1",
            "num": str(num_result),
            "date_range": date_range
        }
    }

    headers = {
        "accept": "application/json",
        "content-type": "application/json",
        "authorization": f"Bearer {API_KEY}"
    }

    try:
        response = requests.post(url, json=payload, headers=headers)
        response.raise_for_status()
        return response.json()
    except requests.RequestException as e:
        error_msg = f"Error: {str(e)}"
        if response.text:
            error_msg += f"\nResponse content: {response.text}"
        return {"error": error_msg}

def format_results_from_raw(results):
    try:
        debug_info = f"Raw API Response:\n{json.dumps(results, indent=2, ensure_ascii=False)}"

        if isinstance(results, dict) and "error" in results:
            return "Error: " + results["error"], debug_info

        if not isinstance(results, dict):
            raise ValueError("결과가 사전 형식이 아닙니다.")

        # 'results' 키 내부의 구조 확인
        if 'results' in results:
            results_content = results['results']
            # 'news' 키 확인
            if 'news' in results_content:
                news_results = results_content['news']
            else:
                news_results = []
        else:
            news_results = []

        if not news_results:
            return "검색 결과가 없습니다.", debug_info

        formatted_articles = ""
        for result in news_results:
            title = result.get("title", "제목 없음")
            link = result.get("url", result.get("link", "#"))
            snippet = result.get("snippet", "내용 없음")
            channel = result.get("channel", result.get("source", "알 수 없음"))
            time = result.get("time", result.get("date", "알 수 없는 시간"))
            image_url = result.get("img", result.get("thumbnail", ""))

            # base64로 인코딩된 이미지를 처리하지 않음
            if image_url and not image_url.startswith("data:image"):
                image_html = f'<img src="{image_url}" alt="Image" style="max-width: 100%; height: auto;">'
            else:
                image_html = ''

            article_html = f"""
            <div style="margin-bottom: 20px; border-bottom: 1px solid #ccc; padding-bottom: 20px;">
                <h3><a href="{link}" target="_blank">{title}</a></h3>
                <p><strong>{channel}</strong> - {time}</p>
                {image_html}
                <p>{snippet}</p>
            </div>
            """
            formatted_articles += article_html

        return formatted_articles, debug_info

    except Exception as e:
        error_message = f"결과 처리 중 오류 발생: {str(e)}"
        debug_info = f"Error: {error_message}\n"
        return "Error: " + error_message, debug_info

def serphouse_search(query, country, page, num_result):
    results = search_serphouse(query, country, page, num_result)
    formatted_articles, debug_info = format_results_from_raw(results)
    return formatted_articles, debug_info

css = """
footer {
    visibility: hidden;
}
"""

# Gradio 인터페이스 구성
with gr.Blocks(css=css, title="24시간 이내 뉴스 검색 인터페이스") as iface:
    gr.Markdown("## 24시간 이내 뉴스 검색 인터페이스")
    gr.Markdown("검색어를 입력하고 국가를 선택하여 24시간 이내의 뉴스 결과를 가져옵니다.")

    with gr.Tab("검색"):
        with gr.Row():
            query = gr.Textbox(label="검색어")
            country = gr.Dropdown(MAJOR_COUNTRIES, label="국가", value="South Korea")
        with gr.Row():
            page = gr.Slider(1, 10, 1, label="페이지")
            num_result = gr.Slider(1, 100, 10, label="결과 수")

        search_button = gr.Button("검색")

    with gr.Tab("뉴스 결과"):
        news_output = gr.HTML(label="뉴스 결과")

    with gr.Tab("디버그 정보"):
        debug_output = gr.Textbox(label="디버그 정보", lines=10)

    def search_and_display(query, country, page, num_result):
        articles, debug_info = serphouse_search(query, country, page, num_result)
        return {news_output: articles, debug_output: debug_info}

    search_button.click(
        search_and_display,
        inputs=[query, country, page, num_result],
        outputs=[news_output, debug_output]
    )

iface.launch(auth=("gini", "pick"))