Update app.py

app.py

@@ -8,12 +8,11 @@ from huggingface_hub import InferenceClient
 API_KEY = os.getenv("SERPHOUSE_API_KEY")
 hf_client = InferenceClient("CohereForAI/c4ai-command-r-plus-08-2024", token=os.getenv("HF_TOKEN"))
 
-# 국가별 언어 코드 매핑 먼저 정의
+# 국가별 언어 코드 매핑
 COUNTRY_LANGUAGES = {
     "South Korea": "ko",
     "Japan": "ja",
     "China": "zh",
-    "Taiwan": "zh-tw",
     "Russia": "ru",
     "France": "fr",
     "Germany": "de",
@@ -42,153 +41,116 @@ COUNTRY_LANGUAGES = {
     "Slovenia": "sl",
     "Estonia": "et",
     "Latvia": "lv",
-    "Lithuania": "lt",
-    "Ukraine": "uk",
-    "Israel": "he",
-    "Bangladesh": "bn",
-    "Pakistan": "ur",
-    "Finland": "fi",
-    "Denmark": "da",
-    "Norway": "no",
-    "Sweden": "sv",
-    "Iceland": "is",
-    "Philippines": "fil",
-    "Brazil": "pt-br",
-    "Argentina": "es-ar",
-    "Chile": "es-cl",
-    "Colombia": "es-co",
-    "Peru": "es-pe",
-    "Venezuela": "es-ve"
+    "Lithuania": "lt"
 }
 
-COUNTRY_CODES = {
-    "United States": "US",
-    "United Kingdom": "GB",
-    "Canada": "CA",
-    "Australia": "AU",
-    "Germany": "DE",
-    "France": "FR",
-    "Japan": "JP",
-    "South Korea": "KR",
-    "China": "CN",
-    "Taiwan": "TW",
-    "India": "IN",
-    "Brazil": "BR",
-    "Mexico": "MX",
-    "Russia": "RU",
-    "Italy": "IT",
-    "Spain": "ES",
-    "Netherlands": "NL",
-    "Singapore": "SG",
-    "Hong Kong": "HK",
-    "Indonesia": "ID",
-    "Malaysia": "MY",
-    "Philippines": "PH",
-    "Thailand": "TH",
-    "Vietnam": "VN",
-    "Belgium": "BE",
-    "Denmark": "DK",
-    "Finland": "FI",
-    "Ireland": "IE",
-    "Norway": "NO",
-    "Poland": "PL",
-    "Sweden": "SE",
-    "Switzerland": "CH",
-    "Austria": "AT",
-    "Czech Republic": "CZ",
-    "Greece": "GR",
-    "Hungary": "HU",
-    "Portugal": "PT",
-    "Romania": "RO",
-    "Turkey": "TR",
-    "Israel": "IL",
-    "Saudi Arabia": "SA",
-    "United Arab Emirates": "AE",
-    "South Africa": "ZA",
-    "Argentina": "AR",
-    "Chile": "CL",
-    "Colombia": "CO",
-    "Peru": "PE",
-    "Venezuela": "VE",
-    "New Zealand": "NZ",
-    "Bangladesh": "BD",
-    "Pakistan": "PK",
-    "Egypt": "EG",
-    "Morocco": "MA",
-    "Nigeria": "NG",
-    "Kenya": "KE",
-    "Ukraine": "UA",
-    "Croatia": "HR",
-    "Slovakia": "SK",
-    "Bulgaria": "BG",
-    "Serbia": "RS",
-    "Estonia": "EE",
-    "Latvia": "LV",
-    "Lithuania": "LT",
-    "Slovenia": "SI",
-    "Luxembourg": "LU",
-    "Malta": "MT",
-    "Cyprus": "CY",
-    "Iceland": "IS"
+COUNTRY_LOCATIONS = {
+    "United States": "United States",
+    "United Kingdom": "United Kingdom",
+    "Canada": "Canada",
+    "Australia": "Australia",
+    "Germany": "Germany",
+    "France": "France",
+    "Japan": "Japan",
+    "South Korea": "South Korea",
+    "China": "China",
+    "India": "India",
+    "Brazil": "Brazil",
+    "Mexico": "Mexico",
+    "Russia": "Russia",
+    "Italy": "Italy",
+    "Spain": "Spain",
+    "Netherlands": "Netherlands",
+    "Singapore": "Singapore",
+    "Hong Kong": "Hong Kong",
+    "Indonesia": "Indonesia",
+    "Malaysia": "Malaysia",
+    "Philippines": "Philippines",
+    "Thailand": "Thailand",
+    "Vietnam": "Vietnam",
+    "Belgium": "Belgium",
+    "Denmark": "Denmark",
+    "Finland": "Finland",
+    "Ireland": "Ireland",
+    "Norway": "Norway",
+    "Poland": "Poland",
+    "Sweden": "Sweden",
+    "Switzerland": "Switzerland",
+    "Austria": "Austria",
+    "Czech Republic": "Czech Republic",
+    "Greece": "Greece",
+    "Hungary": "Hungary",
+    "Portugal": "Portugal",
+    "Romania": "Romania",
+    "Turkey": "Turkey",
+    "Israel": "Israel",
+    "Saudi Arabia": "Saudi Arabia",
+    "United Arab Emirates": "United Arab Emirates",
+    "South Africa": "South Africa",
+    "Argentina": "Argentina",
+    "Chile": "Chile",
+    "Colombia": "Colombia",
+    "Peru": "Peru",
+    "Venezuela": "Venezuela",
+    "New Zealand": "New Zealand",
+    "Bangladesh": "Bangladesh",
+    "Pakistan": "Pakistan",
+    "Egypt": "Egypt",
+    "Morocco": "Morocco",
+    "Nigeria": "Nigeria",
+    "Kenya": "Kenya",
+    "Ukraine": "Ukraine",
+    "Croatia": "Croatia",
+    "Slovakia": "Slovakia",
+    "Bulgaria": "Bulgaria",
+    "Serbia": "Serbia",
+    "Estonia": "Estonia",
+    "Latvia": "Latvia",
+    "Lithuania": "Lithuania",
+    "Slovenia": "Slovenia",
+    "Luxembourg": "Luxembourg",
+    "Malta": "Malta",
+    "Cyprus": "Cyprus",
+    "Iceland": "Iceland"
 }
 
-MAJOR_COUNTRIES = list(COUNTRY_CODES.keys())
-
-def is_english(text):
-    return all(ord(char) < 128 for char in text.replace(' ', ''))
+MAJOR_COUNTRIES = list(COUNTRY_LOCATIONS.keys())
 
 def translate_query(query, country):
     try:
-        # 영어 입력인 경우 번역하지 않음
-        if is_english(query):
-            print(f"English query detected, using original: {query}")
-            return query[:255]
-            
-        # 한글 입력이고 South Korea가 선택된 경우 번역하지 않음
-        if country == "South Korea":
-            return query[:255]
-            
-        if country in COUNTRY_CODES:
-            query = query[:100]
-            target_lang = COUNTRY_CODES[country]
-            
-            prompt = f"""Translate this text to {target_lang} language. 
-            For Japanese, use Kanji and Kana.
-            For Chinese (China), use Simplified Chinese.
-            For Chinese (Taiwan), use Traditional Chinese.
-            For Korean, use Hangul.
-            Only output the translated text without any explanation.
-            Text to translate: {query}"""
+        if country in COUNTRY_LANGUAGES:
+            target_lang = COUNTRY_LANGUAGES[country]
+            prompt = f"Translate the following English text to {target_lang} language. Only output the translated text without any explanations or quotes: {query}"
             
             translated = hf_client.text_generation(
                 prompt,
-                max_new_tokens=50,
-                temperature=0.1
+                max_new_tokens=100,
+                temperature=0.3
             )
-            translated = translated.strip()[:255]
-            print(f"Original query: {query}")
-            print(f"Translated query: {translated}")
-            return translated
-        return query[:255]
+            return translated.strip()
+        return query
     except Exception as e:
         print(f"Translation error: {str(e)}")
-        return query[:255]
+        return query
 
 def search_serphouse(query, country, page=1, num_result=10):
     url = "https://api.serphouse.com/serp/live"
     
+    # 검색어 번역
+    translated_query = translate_query(query, country)
+    print(f"Original query: {query}")
+    print(f"Translated query: {translated_query}")
+    
     payload = {
         "data": {
             "q": query,
             "domain": "google.com",
-            "loc": country,  # 국가 이름 사용
-            "lang": COUNTRY_LANGUAGES.get(country, "en"),
+            "loc": COUNTRY_LOCATIONS.get(country, "United States"),
+            "lang": "en",
             "device": "desktop",
-            "serp_type": "web",  # web으로 변경
+            "serp_type": "news",
             "page": "1",
-            "verbatim": "0",
-            "gfilter": "0",
-            "num": "10"  # num으로 변경
+            "num": "10"
         }
     }
 
@@ -200,17 +162,13 @@ def search_serphouse(query, country, page=1, num_result=10):
 
     try:
         response = requests.post(url, json=payload, headers=headers)
-        print("Request payload:", json.dumps(payload, indent=2, ensure_ascii=False))
+        print("Request payload:", json.dumps(payload, indent=2))
         print("Response status:", response.status_code)
-        print("Response content:", response.text)
         
         response.raise_for_status()
-        return {"results": response.json(), "translated_query": query}
+        return {"results": response.json(), "translated_query": translated_query}
     except requests.RequestException as e:
-        error_msg = f"Error: {str(e)}"
-        if hasattr(response, 'text'):
-            error_msg += f"\nResponse content: {response.text}"
-        return {"error": error_msg, "translated_query": query}
+        return {"error": f"Error: {str(e)}", "translated_query": query}
 
 def format_results_from_raw(response_data):
     if "error" in response_data:
@@ -220,26 +178,24 @@ def format_results_from_raw(response_data):
         results = response_data["results"]
         translated_query = response_data["translated_query"]
         
-        # 응답 구조 변경
-        organic_results = results.get('results', {}).get('organic', [])
-        if not organic_results:
+        news_results = results.get('results', {}).get('results', {}).get('news', [])
+        if not news_results:
             return "검색 결과가 없습니다.", []
 
         articles = []
-        for idx, result in enumerate(organic_results, 1):
+        for idx, result in enumerate(news_results, 1):
             articles.append({
                 "index": idx,
                 "title": result.get("title", "제목 없음"),
-                "link": result.get("link", "#"),
+                "link": result.get("url", result.get("link", "#")),
                 "snippet": result.get("snippet", "내용 없음"),
-                "channel": result.get("source", "알 수 없음"),
-                "time": result.get("date", "알 수 없는 시간"),
-                "image_url": result.get("thumbnail", ""),
+                "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", "")),
                 "translated_query": translated_query
             })
         return "", articles
     except Exception as e:
-        print(f"Format error: {str(e)}")  # 디버깅용 출력 추가
         return f"결과 처리 중 오류 발생: {str(e)}", []
 
 def serphouse_search(query, country):
@@ -250,16 +206,17 @@ css = """
 footer {visibility: hidden;}
 """
 
+# Gradio 인터페이스 수정
 with gr.Blocks(theme="Nymbo/Nymbo_Theme", css=css, title="NewsAI 서비스") as iface:
-    gr.Markdown("검색어를 입력하고 원하는 국가를 선택하면, 해당 국가의 언어로 번역된 검색어로 뉴스를 검색합니다.")
+    gr.Markdown("검색어를 입력하고 원하는 국가를 선택하면, 검색어와 일치하는 24시간 이내 뉴스를 최대 100개 출력합니다.")
 
     with gr.Column():
         with gr.Row():
             query = gr.Textbox(label="검색어")
             country = gr.Dropdown(MAJOR_COUNTRIES, label="국가", value="South Korea")
             
-        # 번역된 검색어 표시 컴포넌트
-        translated_display = gr.Markdown(visible=True)
+        # 번역 결과를 보여주는 컴포넌트 추가
+        translated_query_display = gr.Markdown(visible=False)
             
         search_button = gr.Button("검색", variant="primary")
 
@@ -287,21 +244,15 @@ with gr.Blocks(theme="Nymbo/Nymbo_Theme", css=css, title="NewsAI 서비스") as
     def search_and_display(query, country, articles_state, progress=gr.Progress()):
         progress(0, desc="검색어 번역 중...")
         
+        # 검색어 번역
         translated_query = translate_query(query, country)
-        if is_english(query):
-            translated_display_text = f"영어 검색어: {query}"
-        elif country == "South Korea":
-            translated_display_text = f"검색어: {query}"
-        elif translated_query != query:
-            translated_display_text = f"원본 검색어: {query}\n번역된 검색어: {translated_query}"
-        else:
-            translated_display_text = f"검색어: {query}"
+        translated_display = f"**원본 검색어:** {query}\n**번역된 검색어:** {translated_query}" if translated_query != query else f"**검색어:** {query}"
         
-        progress(0.2, desc="검색 중...")
-        response_data = search_serphouse(query, country)
-        error_message, articles = format_results_from_raw(response_data)
+        progress(0.2, desc="검색 시작...")
+        error_message, articles = serphouse_search(query, country)
+        progress(0.5, desc="결과 처리 중...")
         
-        outputs = [gr.update(value=translated_display_text, visible=True)]
+        outputs = [gr.update(value=translated_display, visible=True)]  # 번역 결과 표시
         
         if error_message:
             outputs.append(gr.update(value=error_message, visible=True))
@@ -340,14 +291,14 @@ with gr.Blocks(theme="Nymbo/Nymbo_Theme", css=css, title="NewsAI 서비스") as
         outputs.append(gr.update(visible=False))
         return outputs
 
-    search_outputs = [translated_display, gr.Markdown(visible=False)]
+    search_outputs = [translated_query_display, gr.Markdown(visible=False)]  # 번역 결과 컴포넌트 추가
     for comp in article_components:
         search_outputs.extend([comp['group'], comp['title'], comp['image'],
                              comp['snippet'], comp['info']])
     search_outputs.extend([articles_state, status_message])
 
     search_button.click(
-        fn=search_and_display,
+        search_and_display,
         inputs=[query, country, articles_state],
         outputs=search_outputs,
         show_progress=True