Update app.py

app.py

@@ -462,22 +462,77 @@ def get_switzerland_originator(ing_de, page):
 def get_germany_originator(ing_de, page):
     log, brands, companies = [], set(), set()
     try:
-        page.goto(f"https://www.gelbe-liste.de/profi-suche/results?substance={quote(ing_de)}", timeout=30000)
-        page.wait_for_selector('.product-list', timeout=15000)
-        soup = BeautifulSoup(page.content(), 'html.parser')
-        ul = soup.find('ul', class_='product-list')
-        if ul:
-            lis = ul.find_all('li')
-            for li in lis:
-                h5, p_tag = li.find('h5'), li.find('p', class_='small')
-                if h5:
-                    title, comp = h5.get_text(strip=True), (p_tag.get_text(strip=True) if p_tag else "-")
-                    if not is_generic(title, comp, ing_de):
-                        brands.add(clean_brand_name(title))
-                        companies.add(comp)
+        log.append(f"1. 前往 PharmNet.Bund (搜尋: {ing_de})...")
+        page.goto("https://portal.bfarm.de/amguifree/am/search.xhtml", timeout=45000, wait_until="domcontentloaded")
+        
+        log.append("2. 定位並輸入 Stoffname (第二格有效成分)...")
+        # 💡 使用您提供的精準 ID 定位第二個搜尋框
+        search_input = page.locator('input[id="searchForm:searchInputsComponent:searchRows:1:firstSearchTerm"]')
+        search_input.wait_for(state="attached", timeout=15000)
+        search_input.fill(ing_de)
+        
+        # 點擊搜尋按鈕
+        page.locator('input[name="searchForm:searchInputsComponent:suchestarten"]').click()
+        
+        current_page = 1
+        while True:
+            log.append(f"   -> 掃描第 {current_page} 頁資料...")
+            try:
+                # 等待搜尋結果表格載入
+                page.wait_for_selector('table[id="searchResultsForm:searchResultsComponent:titles"]', timeout=15000)
+            except Exception:
+                log.append("❌ 查無資料或等待表格超時。")
+                break
+                
+            soup = BeautifulSoup(page.content(), 'html.parser')
+            table = soup.find('table', id='searchResultsForm:searchResultsComponent:titles')
+            
+            if table and table.find('tbody'):
+                rows = table.find('tbody').find_all('tr')
+                for tr in rows:
+                    tds = tr.find_all('td')
+                    # 依照您提供的 HTML：索引 1 是藥名，索引 3 是藥廠
+                    if len(tds) >= 4:
+                        title = tds[1].get_text(strip=True)
+                        comp = tds[3].get_text(strip=True)
+                        
+                        # 進入黑名單與去劑型化過濾
+                        if not is_generic(title, comp, ing_de):
+                            brands.add(clean_brand_name(title))
+                            if comp != "-": companies.add(comp)
+            else:
+                break
+            
+            # ==========================
+            # 💡 自動翻頁邏輯 (處理 JSF Postback)
+            # ==========================
+            paging_div = soup.find('div', class_='browse')
+            has_next = False
+            
+            if paging_div:
+                buttons = paging_div.find_all('input', class_='submit')
+                for i, btn in enumerate(buttons):
+                    # 找到目前所在的頁碼按鈕 (帶有 disabled class)
+                    if 'disabled' in btn.get('class', []):
+                        # 如果後面還有按鈕，代表有下一頁
+                        if i + 1 < len(buttons):
+                            next_btn_id = buttons[i+1].get('id')
+                            # Playwright 觸發點擊下一頁
+                            page.locator(f'input[id="{next_btn_id}"]').click()
+                            page.wait_for_timeout(2500) # 給予 JSF 更新表格的緩衝時間
+                            current_page += 1
+                            has_next = True
+                            break
+            
+            # 如果沒有找到下一頁按鈕，結束迴圈
+            if not has_next:
+                log.append("✅ 已掃描至最後一頁。")
+                break
+                
         if brands: return ", ".join(brands), ", ".join(companies), "\n".join(log)
-        return "查無原廠", "-", "\n".join(log)
-    except Exception as e: return "執行失敗", "-", "\n".join(log) + f"\n錯誤: {str(e)}"
+        return "查無原廠", "-", "\n".join(log) + "\n❌ 查無資料或全為學名藥"
+    except Exception as e: 
+        return "執行失敗", "-", "\n".join(log) + f"\n錯誤: {str(e)}"
 
 # ==========================================
 # 🚀 主執行中樞