Update streamlit_app.py

streamlit_app.py

@@ -30,7 +30,7 @@ UI_TEXT = {
         "menu_search": "🔍 AI 스마트 검색",
         "menu_scanner": "📡 글로벌 시장 스캐너",
         "search_title": "🧠 다국어 AI 주식 비서",
-        "search_placeholder": "예: 베트남 비나밀크, 삼성전자...",
+        "search_placeholder": "예: 베트남 캉디엔, 미국 팔란티어...",
         "btn_analyze": "분석 시작",
         "status_thinking": "🧠 AI가 분석 중입니다...",
         "info_target": "타겟",
@@ -41,11 +41,11 @@ UI_TEXT = {
         "metric_source": "출처",
         "expander_news": "📰 뉴스 원문 보기",
         "insight_title": "🧠 AI 투자 리포트",
-        "scanner_title": "📡 글로벌 시장 전광판",
-        "scanner_caption": "※ 사람처럼 천천히 정밀 검색합니다. (속도: 느림)",
-        "btn_scan": "스캔 시작",
-        "tab_gainers": "🔥 급등 (Top Gainers)",
-        "tab_losers": "💧 급락 (Top Losers)",
+        "scanner_title": "📡 글로벌 시장 전광판 (Max 100)",
+        "scanner_caption": "※ 약 100개 종목을 1차 고속 스캔 후, 실패 종목만 정밀 스캔합니다.",
+        "btn_scan": "광대역 스캔 시작",
+        "tab_gainers": "🔥 급등 (Pure Gainers)",
+        "tab_losers": "💧 급락 (Pure Losers)",
         "col_name": "기업명",
         "col_price": "가격",
         "col_change": "등락률",
@@ -58,7 +58,7 @@ UI_TEXT = {
         "menu_search": "🔍 AI Smart Search",
         "menu_scanner": "📡 Global Market Scanner",
         "search_title": "🧠 AI Stock Assistant",
-        "search_placeholder": "e.g. Vinamilk, Samsung Electronics...",
+        "search_placeholder": "e.g. Vinamilk, Palantir...",
         "btn_analyze": "Analyze",
         "status_thinking": "🧠 AI is analyzing...",
         "info_target": "Target",
@@ -69,11 +69,11 @@ UI_TEXT = {
         "metric_source": "Source",
         "expander_news": "📰 View Source News",
         "insight_title": "🧠 AI Investment Report",
-        "scanner_title": "📡 Global Market Scanner",
-        "scanner_caption": "※ Deep search with human-like delay. (Speed: Slow)",
-        "btn_scan": "Start Scan",
-        "tab_gainers": "🔥 Top Gainers",
-        "tab_losers": "💧 Top Losers",
+        "scanner_title": "📡 Global Market Scanner (Max 100)",
+        "scanner_caption": "※ Hybrid Scanning (Batch High-Speed + Deep Robust).",
+        "btn_scan": "Start Full Scan",
+        "tab_gainers": "🔥 Top Gainers (>0%)",
+        "tab_losers": "💧 Top Losers (<0%)",
         "col_name": "Company",
         "col_price": "Price",
         "col_change": "Change(%)",
@@ -84,38 +84,85 @@ UI_TEXT = {
 }
 
 # ==============================================================================
-# 💾 [데이터] 티커 매핑
+# 💾 [데이터] 대규모 확장 티커 (약 100개 규모)
 # ==============================================================================
 TICKER_NAMES = {
+    # 🇻🇳 베트남
     "VIC.VN": "Vingroup", "VHM.VN": "Vinhomes", "VCB.VN": "Vietcombank", "VNM.VN": "Vinamilk", 
-    "HPG.VN": "Hoa Phat", "MSN.VN": "Masan Group", "GAS.VN": "PV Gas",
-    "BBCA.JK": "BCA Bank", "BBRI.JK": "BRI Bank", "TLKM.JK": "Telkom Indo", "BMRI.JK": "Mandiri", 
-    "ASII.JK": "Astra Intl", "GOTO.JK": "GoTo Group",
+    "HPG.VN": "Hoa Phat", "MSN.VN": "Masan Group", "GAS.VN": "PV Gas", "NVL.VN": "Novaland", 
+    "PDR.VN": "Phat Dat", "DIG.VN": "DIC Corp", "CEO.VN": "CEO Group", "SSI.VN": "SSI Sec", 
+    "VND.VN": "VNDirect", "MWG.VN": "Mobile World", "FRT.VN": "FPT Retail", "FPT.VN": "FPT Corp",
+    "STB.VN": "Sacombank", "MBB.VN": "MB Bank", "TCB.VN": "Techcombank", "VPB.VN": "VPBank",
+    "VRE.VN": "Vincom Retail", "DGC.VN": "Duc Giang Chem", "VHC.VN": "Vinh Hoan", "KBC.VN": "Kinh Bac",
+    "GVR.VN": "Vietnam Rubber", "SAB.VN": "Sabeco", "BID.VN": "BIDV", "CTG.VN": "VietinBank",
+
+    # 🇮🇩 인도네시아
+    "BBCA.JK": "BCA Bank", "BBRI.JK": "BRI Bank", "TLKM.JK": "Telkom", "BMRI.JK": "Mandiri", 
+    "ASII.JK": "Astra Intl", "GOTO.JK": "GoTo", "UNVR.JK": "Unilever", "ADRO.JK": "Adaro Energy", 
+    "MDKA.JK": "Merdeka Copper", "ANTM.JK": "Aneka Tambang", "ICBP.JK": "Indofood CBP", 
+    "KLBF.JK": "Kalbe Farma", "BUKA.JK": "Bukalapak", "PGAS.JK": "Perusahaan Gas",
+
+    # 🇰🇷 한국
     "005930.KS": "Samsung Elec", "000660.KS": "SK Hynix", "035420.KS": "NAVER", "035720.KS": "Kakao", 
-    "005380.KS": "Hyundai Motor", "207940.KS": "Samsung Bio",
+    "005380.KS": "Hyundai Motor", "207940.KS": "Samsung Bio", "068270.KS": "Celltrion",
+    "086520.KQ": "Ecopro", "247540.KQ": "Ecopro BM", "005490.KS": "POSCO Holdings", "010120.KS": "LS ELECTRIC",
+    "012450.KS": "Hanwha Aero", "042700.KS": "Hanmi Semi", "196170.KQ": "Alteogen",
+    "352820.KS": "HYBE", "035900.KQ": "JYP Ent", "000270.KS": "Kia", "010950.KS": "S-Oil",
+    "032830.KQ": "Samsung Life", "329180.KS": "Hyundai Heavy", "015760.KS": "KEPCO",
+
+    # 🇺🇸 미국 (M7 + 성장 + 바이오 + 에너지)
     "AAPL": "Apple", "NVDA": "NVIDIA", "TSLA": "Tesla", "AMZN": "Amazon", "MSFT": "Microsoft", 
-    "GOOGL": "Google", "META": "Meta", "AMD": "AMD",
+    "GOOGL": "Google", "META": "Meta", "AMD": "AMD", "NFLX": "Netflix", "INTC": "Intel",
+    "PLTR": "Palantir", "COIN": "Coinbase", "U": "Unity", "RBLX": "Roblox", "SOFI": "SoFi", 
+    "PYPL": "PayPal", "SQ": "Block", "UBER": "Uber", "ABNB": "Airbnb", "HOOD": "Robinhood",
+    "LLY": "Eli Lilly", "NVO": "Novo Nordisk", "PFE": "Pfizer", "MRK": "Merck",
+    "XOM": "Exxon", "CVX": "Chevron", "KO": "Coca-Cola", "PEP": "Pepsi", "COST": "Costco",
+    "DIS": "Disney", "NKE": "Nike", "SBUX": "Starbucks", "MCD": "McDonalds",
+    
+    # 🇯🇵 일본
     "7203.T": "Toyota", "6758.T": "Sony", "9984.T": "SoftBank", "8035.T": "Tokyo Elec", 
-    "6861.T": "Keyence", "7974.T": "Nintendo",
-    "9988.HK": "Alibaba", "0700.HK": "Tencent", "3690.HK": "Meituan", "1211.HK": "BYD", "1810.HK": "Xiaomi",
+    "6861.T": "Keyence", "7974.T": "Nintendo", "8306.T": "MUFG", "6501.T": "Hitachi",
+    "8058.T": "Mitsubishi", "8001.T": "Itochu", "8031.T": "Mitsui", "6920.T": "Lasertec",
+    "4063.T": "Shin-Etsu", "7741.T": "HOYA", "6146.T": "Disco", "7267.T": "Honda",
+
+    # 🇨🇳 중국/홍콩/대만/인도
+    "9988.HK": "Alibaba", "0700.HK": "Tencent", "3690.HK": "Meituan", "1211.HK": "BYD", 
+    "1810.HK": "Xiaomi", "0941.HK": "China Mobile", "9888.HK": "Baidu", "9618.HK": "JD.com",
+    "2015.HK": "Li Auto", "9868.HK": "Xpeng", "0981.HK": "SMIC",
     "2330.TW": "TSMC", "2454.TW": "MediaTek", "2317.TW": "Foxconn", "2308.TW": "Delta Elec",
-    "RELIANCE.NS": "Reliance Ind", "TCS.NS": "TCS", "HDFCBANK.NS": "HDFC Bank", "INFY.NS": "Infosys"
+    "RELIANCE.NS": "Reliance", "TCS.NS": "TCS", "HDFCBANK.NS": "HDFC Bank", "INFY.NS": "Infosys"
 }
 
 MARKET_SAMPLES = {
-    "🇻🇳 Vietnam": ["VIC.VN", "VHM.VN", "VCB.VN", "VNM.VN", "HPG.VN", "MSN.VN", "GAS.VN"],
-    "🇮🇩 Indonesia": ["BBCA.JK", "BBRI.JK", "TLKM.JK", "BMRI.JK", "ASII.JK", "GOTO.JK"],
-    "🇰🇷 Korea": ["005930.KS", "000660.KS", "035420.KS", "035720.KS", "005380.KS", "207940.KS"],
-    "🇺🇸 USA": ["AAPL", "NVDA", "TSLA", "AMZN", "MSFT", "GOOGL", "META", "AMD"],
-    "🇯🇵 Japan": ["7203.T", "6758.T", "9984.T", "8035.T", "6861.T", "7974.T"],
-    "🇨🇳 China/HK": ["9988.HK", "0700.HK", "3690.HK", "1211.HK", "1810.HK"],
-    "🇹🇼 Taiwan": ["2330.TW", "2454.TW", "2317.TW", "2308.TW"],
-    "🇮🇳 India": ["RELIANCE.NS", "TCS.NS", "HDFCBANK.NS", "INFY.NS"]
+    "🇻🇳 Vietnam (Max 50)": [
+        "VIC.VN", "VHM.VN", "VRE.VN", "VNM.VN", "MSN.VN", "GAS.VN", "HPG.VN", "HSG.VN",
+        "NVL.VN", "PDR.VN", "DIG.VN", "CEO.VN", "DXG.VN", "KBC.VN", "GVR.VN",
+        "VCB.VN", "TCB.VN", "VPB.VN", "MBB.VN", "STB.VN", "ACB.VN", "BID.VN", "CTG.VN",
+        "SSI.VN", "VND.VN", "VCI.VN", "MWG.VN", "FRT.VN", "DGW.VN", "FPT.VN", "DGC.VN", "VHC.VN", "SAB.VN"
+    ],
+    "🇺🇸 USA (Max 50)": [
+        "AAPL", "MSFT", "GOOGL", "AMZN", "NVDA", "META", "TSLA", "AMD", "INTC", "QCOM", 
+        "AVGO", "MU", "PLTR", "COIN", "U", "RBLX", "SOFI", "UBER", "ABNB", "HOOD", "PYPL", "SQ",
+        "LLY", "NVO", "JNJ", "PFE", "MRK", "UNH", "XOM", "CVX", "JPM", "V", "MA", 
+        "KO", "PEP", "COST", "WMT", "DIS", "NKE", "SBUX", "MCD"
+    ],
+    "🇰🇷 Korea (Max 50)": [
+        "005930.KS", "000660.KS", "005380.KS", "000270.KS", "035420.KS", "035720.KS", 
+        "005490.KS", "086520.KQ", "247540.KQ", "010120.KS", "207940.KS", "068270.KS", 
+        "196170.KQ", "012450.KS", "042700.KS", "352820.KS", "010950.KS", "032830.KQ", "329180.KS", "015760.KS"
+    ],
+    "🇮🇩 Indonesia": ["BBCA.JK", "BBRI.JK", "BMRI.JK", "BBNI.JK", "TLKM.JK", "ASII.JK", "GOTO.JK", "ADRO.JK", "MDKA.JK", "ANTM.JK", "UNVR.JK", "ICBP.JK", "KLBF.JK", "BUKA.JK", "PGAS.JK"],
+    "🇯🇵 Japan": ["7203.T", "6758.T", "9984.T", "8035.T", "6861.T", "7974.T", "8306.T", "6501.T", "8058.T", "8001.T", "8031.T", "6920.T", "4063.T", "7741.T", "6146.T", "7267.T"],
+    "🇨🇳 China/HK": ["9988.HK", "0700.HK", "3690.HK", "1211.HK", "1810.HK", "0941.HK", "9888.HK", "9618.HK", "2015.HK", "9868.HK", "0981.HK"],
+    "🇹🇼 Taiwan": ["2330.TW", "2454.TW", "2317.TW", "2308.TW", "2603.TW", "2881.TW", "1301.TW"],
+    "🇮🇳 India": ["RELIANCE.NS", "TCS.NS", "HDFCBANK.NS", "INFY.NS", "ICICIBANK.NS", "TATAMOTORS.NS", "BHARTIARTL.NS"]
 }
 
 # ==============================================================================
-# 🕸️ [엔진] 뉴스 및 데이터 수집
+# 🛠️ [엔진] 데이터 수집 함수들
 # ==============================================================================
+FAKE_HEADERS = {'User-Agent': 'Mozilla/5.0 ...'}
+
 def fetch_news_robust(keyword):
     summary = ""
     seen_urls = set()
@@ -128,38 +175,36 @@ def fetch_news_robust(keyword):
             if not results:
                 time.sleep(random.uniform(0.5, 1.0))
                 results = list(ddgs.text(f"{keyword} latest news", max_results=3))
-            
             for r in results:
                 url = r.get('url') or r.get('href')
                 if url not in seen_urls:
-                    date = r.get('date', 'Unknown')
+                    date = r.get('date', '?')
                     title = r.get('title', '')
                     summary += f"[{date}] {title}\n"
                     seen_urls.add(url)
-    except Exception as e: return f"News Error: {e}"
-    return summary if summary else "No news found."
+    except: return "No news."
+    return summary if summary else "No news."
 
 def scrape_price_from_web(name, ticker):
     try:
-        time.sleep(random.uniform(1.0, 1.5))
+        time.sleep(random.uniform(0.5, 1.0))
         with DDGS() as ddgs:
             query = f"{name} {ticker} stock price quote today"
             results = ddgs.text(query, max_results=2)
             blob = " ".join([r['body'] for r in results])
             match = re.search(r'(\d{1,3}(,\d{3})*(\.\d+)?)', blob)
-            if match:
-                price = float(match.group(0).replace(",", ""))
-                return price, 0.0
-            return None, None
-    except: return None, None
+            if match: return float(match.group(0).replace(",", "")), 0.0
+    except: pass
+    return None, None
 
 def get_price_data_robust(ticker, name):
+    # FDR
     try:
         fdr_symbol = ticker
         if ".VN" in ticker: fdr_symbol = ticker.split('.')[0]
         elif ".JK" in ticker: fdr_symbol = f"IDX:{ticker.split('.')[0]}"
         elif ".KS" in ticker or ".KQ" in ticker: fdr_symbol = ticker.split('.')[0]
-        hist = fdr.DataReader(fdr_symbol, start=(datetime.datetime.now() - datetime.timedelta(days=10)))
+        hist = fdr.DataReader(fdr_symbol, start=(datetime.datetime.now() - datetime.timedelta(days=7)))
         if not hist.empty:
             curr = hist['Close'].iloc[-1]
             prev = hist['Close'].iloc[-2] if len(hist) > 1 else curr
@@ -167,18 +212,16 @@ def get_price_data_robust(ticker, name):
             if prev != 0: pct = ((curr - prev)/prev)*100
             return curr, pct, "FDR"
     except: pass
-    
+    # Yahoo
     try:
         stock = yf.Ticker(ticker)
         price = stock.fast_info.last_price
         prev = stock.fast_info.previous_close
-        if price and prev:
-            pct = ((price - prev)/prev)*100
-            return price, pct, "Yahoo"
+        if price and prev: return price, ((price - prev)/prev)*100, "Yahoo"
     except: pass
-    
+    # Web
     price, pct = scrape_price_from_web(name, ticker)
-    if price is not None: return price, pct, "Web Search"
+    if price is not None: return price, pct, "Web"
     return None, None, "Fail"
 
 def get_chart_data(ticker):
@@ -201,7 +244,6 @@ def identify_target_with_ai(user_query):
     try:
         prompt = f"""
         [ROLE] Entity Resolver. [QUERY] "{user_query}"
-        [TASK] Identify company, ticker, dual keywords.
         [OUTPUT JSON] {{ "name": "Eng Name", "ticker": "TICKER", "eng_key": "Name stock news", "native_key": "Local Name + news keyword" }}
         """
         messages = [{"role": "user", "content": prompt}]
@@ -211,39 +253,26 @@ def identify_target_with_ai(user_query):
     except: return None
 
 def get_polyglot_news(eng_key, native_key):
-    summary = "=== Global ===\n" + fetch_news_robust(eng_key)
-    if eng_key != native_key:
-        summary += "\n\n=== Local ===\n" + fetch_news_robust(native_key)
-    return summary
+    return f"Global:\n{fetch_news_robust(eng_key)}\nLocal:\n{fetch_news_robust(native_key)}"
 
 def parse_stream(stream):
     for chunk in stream:
         if chunk.choices: yield chunk.choices[0].delta.content or ""
 
-# ==============================================================================
-# 📊 [시각화] 차트 함수 (버그 수정됨)
-# ==============================================================================
 def plot_candle_chart(hist, title, source):
     fig = go.Figure(data=[go.Candlestick(x=hist.index, open=hist['Open'], high=hist['High'], low=hist['Low'], close=hist['Close'], name="Price")])
     fig.update_layout(title=f"{title} ({source})", height=350, margin=dict(l=10, r=10, t=30, b=10), template="plotly_dark", paper_bgcolor='rgba(0,0,0,0)')
     return fig
 
 def plot_bar_chart(df, lang_data):
-    # [수정] df['Name'] 대신 현재 언어 설정에 맞는 컬럼(lang_data['col_name']) 사용
     target_col = lang_data['col_name']
-    
     colors = ['#00FF00' if x > 0 else '#FF0000' for x in df['Change(%)']]
-    fig = go.Figure(go.Bar(
-        x=df[target_col],  # <--- 여기서 에러가 났었습니다. 동적 컬럼명 사용!
-        y=df['Change(%)'], 
-        marker_color=colors, 
-        text=df['Change(%)'].apply(lambda x: f"{x:.2f}%")
-    ))
+    fig = go.Figure(go.Bar(x=df[target_col], y=df['Change(%)'], marker_color=colors, text=df['Change(%)'].apply(lambda x: f"{x:.2f}%")))
     fig.update_layout(title="Market Heatmap", height=350, margin=dict(l=10, r=10, t=30, b=10), template="plotly_dark", paper_bgcolor='rgba(0,0,0,0)')
     return fig
 
 # ==============================================================================
-# 🖥️ UI & Sidebar
+# UI
 # ==============================================================================
 with st.sidebar:
     lang_code = st.selectbox("Language / 언어", ["KR", "EN"])
@@ -264,7 +293,6 @@ if menu == T['menu_search']:
             if info:
                 name, ticker = info.get('name'), info.get('ticker')
                 eng_key, native_key = info.get('eng_key'), info.get('native_key')
-                
                 st.info(f"💡 {T['info_target']}: **{name} ({ticker})**")
                 
                 h, src = get_chart_data(ticker)
@@ -301,8 +329,7 @@ if menu == T['menu_search']:
                 prompt = f"""
                 [ROLE] Analyst. [TARGET] {name} ({ticker}). [QUERY] "{query}"
                 [DATA] Price:{curr}, Trend:{pct:.2f}%. [NEWS] {news_data}
-                [TASK] Summarize, Sentiment, Verdict.
-                [OUTPUT LANGUAGE] **{target_lang}** (Strictly output in {target_lang})
+                [TASK] Summarize, Sentiment, Verdict. [LANG] {target_lang}.
                 """
                 msg = [{"role": "user", "content": prompt}]
                 stream = client.chat.completions.create(model="Qwen/Qwen2.5-72B-Instruct", messages=msg, stream=True)
@@ -322,9 +349,38 @@ elif menu == T['menu_scanner']:
         results = []
         bar = st.progress(0)
         
+        # [최적화] 1차: Batch Download (빠름)
+        # 100개씩 나눠서 다운로드 (Yfinance 제한 고려)
+        status_text = st.empty()
+        status_text.text("🚀 Phase 1: Batch Speed Scan...")
+        
+        try:
+            # Batch로 긁어오기
+            batch_data = yf.download(tickers, period="5d", progress=False)['Close']
+        except:
+            batch_data = pd.DataFrame()
+
+        # 2차: 실패한 종목만 Robust Scan (FDR/Web)
+        status_text.text("🕵️ Phase 2: Deep Robust Scan...")
+        
         for i, t in enumerate(tickers):
             name_display = TICKER_NAMES.get(t, t)
-            p, c, s = get_price_data_robust(t, name_display)
+            p, c, s = None, None, None
+            
+            # 1. Batch 결과 확인
+            if not batch_data.empty and t in batch_data.columns:
+                series = batch_data[t].dropna()
+                if len(series) >= 2:
+                    p = series.iloc[-1]
+                    prev = series.iloc[-2]
+                    if prev != 0: 
+                        c = ((p - prev)/prev)*100
+                        s = "Yahoo (Batch)"
+            
+            # 2. Batch 실패 시 Robust 실행
+            if p is None:
+                p, c, s = get_price_data_robust(t, name_display)
+            
             if p is not None:
                 results.append({
                     T['col_name']: name_display, 
@@ -335,19 +391,33 @@ elif menu == T['menu_scanner']:
                 })
             bar.progress((i+1)/len(tickers))
         
+        status_text.empty()
+        
         if results:
             df = pd.DataFrame(results).sort_values('Change(%)', ascending=False)
-            
-            # [수정] 차트 함수에 현재 언어설정(T)을 전달
             st.plotly_chart(plot_bar_chart(df, T), use_container_width=True)
             
             disp_cols = [T['col_name'], 'Ticker', T['col_price'], 'Change(%)', T['col_source']]
             
+            # [필터링 로직 개선]
+            # Gainers: 0보다 큰 것만, 앞에서 10개
+            df_gainers = df[df['Change(%)'] > 0].head(10)
+            
+            # Losers: 0보다 작은 것만, 뒤에서부터 정렬해서 10개 (가장 많이 떨어진 순서)
+            df_losers = df[df['Change(%)'] < 0].sort_values('Change(%)', ascending=True).head(10)
+            
             c_up, c_down = st.columns(2)
             with c_up:
                 st.success(T['tab_gainers'])
-                st.dataframe(df.head(10)[disp_cols].style.format({T['col_price']: "{:,.2f}", "Change(%)": "{:,.2f}%"}), use_container_width=True)
+                if not df_gainers.empty:
+                    st.dataframe(df_gainers[disp_cols].style.format({T['col_price']: "{:,.2f}", "Change(%)": "{:,.2f}%"}), use_container_width=True)
+                else:
+                    st.info("No gainers found.")
+                    
             with c_down:
                 st.error(T['tab_losers'])
-                st.dataframe(df.tail(10)[disp_cols].sort_values('Change(%)').style.format({T['col_price']: "{:,.2f}", "Change(%)": "{:,.2f}%"}), use_container_width=True)
+                if not df_losers.empty:
+                    st.dataframe(df_losers[disp_cols].style.format({T['col_price']: "{:,.2f}", "Change(%)": "{:,.2f}%"}), use_container_width=True)
+                else:
+                    st.info("No losers found.")
         else: st.warning(T['msg_fail'])