feat: 京东账单支持 LOW 复核等级

analyzer/cleaners/__init__.py

@@ -4,6 +4,7 @@
 from .base import BaseCleaner
 from .alipay import AlipayCleaner
 from .wechat import WechatCleaner
+from .jd import JDCleaner
 
-__all__ = ['BaseCleaner', 'AlipayCleaner', 'WechatCleaner']
+__all__ = ['BaseCleaner', 'AlipayCleaner', 'WechatCleaner', 'JDCleaner']
 

analyzer/cleaners/jd.py

@@ -0,0 +1,370 @@
+"""
+京东白条账单清理模块
+"""
+import csv
+import re
+from decimal import Decimal
+from pathlib import Path
+
+import yaml
+
+from .base import (
+    BaseCleaner, parse_amount, format_amount,
+    is_in_date_range, create_arg_parser
+)
+from category import infer_category
+
+
+# 加载京东专属分类配置
+JD_CONFIG_FILE = Path(__file__).parent.parent / "config" / "category_jd.yaml"
+
+def load_jd_config():
+    """加载京东分类配置"""
+    with open(JD_CONFIG_FILE, "r", encoding="utf-8") as f:
+        return yaml.safe_load(f)
+
+_jd_config = load_jd_config()
+
+
+def infer_jd_category(merchant: str, product: str, original_category: str) -> tuple[str, bool, int]:
+    """
+    根据京东账单的商户名称、商品说明和原分类推断统一分类
+    
+    Args:
+        merchant: 商户名称（如"京东外卖"、"京东平台商户"）
+        product: 交易说明/商品说明
+        original_category: 京东原始分类（如"食品酒饮"、"数码电器"）
+    
+    Returns:
+        (分类名称, 是否确定, 复核等级)
+        
+        复核等级:
+            0 = 无需复核（商户映射或原分类映射成功，高置信度）
+            1 = 低优先级复核（通用关键词匹配成功，需确认）
+            2 = 高优先级复核（全部匹配失败或未知分类，需人工分类）
+    """
+    # 1. 先检查商户名称直接映射（如"京东外卖" -> "餐饮美食"）
+    merchant_mapping = _jd_config.get("商户映射", {})
+    for merchant_key, category in merchant_mapping.items():
+        if merchant_key in merchant:
+            return category, True, 0  # 商户映射，无需复核
+    
+    # 2. 尝试直接映射京东原分类
+    category_mapping = _jd_config.get("分类映射", {})
+    
+    # 处理多分类情况（如"食品酒饮 其他网购"）
+    original_cats = original_category.split() if original_category else []
+    for orig_cat in original_cats:
+        if orig_cat in category_mapping:
+            mapped = category_mapping[orig_cat]
+            if mapped:  # 非空映射 → 使用映射结果
+                return mapped, True, 0  # 原分类映射，无需复核
+            # 空映射（如"其他"→""）→ 继续检查下一个原分类或进入关键词匹配
+        else:
+            # 未知分类（不在映射表中）→ 保留原分类，HIGH 复核
+            return orig_cat, True, 2
+    
+    # 3. 使用通用分类推断（已包含京东平台商户关键词）
+    category, is_certain = infer_category(merchant, product, "支出")
+    if is_certain:
+        return category, True, 1  # 关键词匹配，低优先级复核
+    
+    # 4. 返回默认分类
+    return _jd_config.get("默认分类", "其他支出"), False, 2  # 全部失败，高优先级复核
+
+
+# 与支付宝/微信对齐的表头（包含"复核等级"字段）
+ALIGNED_HEADER = [
+    "交易时间", "交易分类", "交易对方", "对方账号", "商品说明",
+    "收/支", "金额", "收/付款方式", "交易状态", "交易订单号", "商家订单号", "备注", "复核等级"
+]
+
+
+class JDCleaner(BaseCleaner):
+    """京东白条账单清理器"""
+    
+    def clean(self) -> None:
+        """执行清理"""
+        self.print_header()
+        
+        # 读取数据，跳过京东导出文件的头部信息
+        with open(self.input_file, "r", encoding="utf-8") as f:
+            reader = csv.reader(f)
+            header = None
+            rows = []
+            
+            for row in reader:
+                # 跳过空行
+                if not row or not row[0].strip():
+                    continue
+                
+                # 清理每个字段的 tab 字符
+                row = [cell.strip().replace('\t', '') for cell in row]
+                
+                # 查找实际的CSV头部行（包含"交易时间"和"商户名称"）
+                if header is None:
+                    if len(row) >= 2 and "交易时间" in row[0] and "商户名称" in row[1]:
+                        header = row
+                        continue
+                    # 跳过头部信息行
+                    continue
+                
+                # 收集数据行
+                rows.append(row)
+        
+        # 确保找到了有效的头部
+        if header is None:
+            raise ValueError("无法找到有效的京东账单表头（需包含'交易时间'和'商户名称'列）")
+        
+        self.stats["original_count"] = len(rows)
+        print(f"原始数据行数: {len(rows)}")
+        
+        # 第一步：按日期范围筛选
+        rows_filtered = [
+            row for row in rows
+            if row and is_in_date_range(row[0], self.start_date, self.end_date)
+        ]
+        self.stats["filtered_count"] = len(rows_filtered)
+        
+        date_desc = f"{self.start_date} ~ {self.end_date}" if self.start_date or self.end_date else "全部"
+        print(f"筛选后数据行数: {len(rows_filtered)} ({date_desc})")
+        
+        # 第二步：分离退款和支出条目（过滤掉"不计收支"）
+        refund_rows = []
+        expense_rows = []
+        skipped_count = 0  # 不计收支（还款、冻结等）
+        
+        for row in rows_filtered:
+            if len(row) < 7:
+                continue
+            
+            income_expense = row[6].strip()  # 收/支 列
+            transaction_desc = row[2].strip()  # 交易说明
+            status = row[5].strip()  # 交易状态
+            
+            # 过滤掉"不计收支"记录（还款、冻结、预授权等）
+            if income_expense == "不计收支":
+                skipped_count += 1
+                continue
+            
+            # 退款判断：交易说明以"退款-"开头 或 状态包含"退款成功"
+            if transaction_desc.startswith("退款-") or "退款" in status:
+                refund_rows.append(row)
+            elif income_expense == "支出":
+                expense_rows.append(row)
+        
+        print(f"退款条目数: {len(refund_rows)}")
+        print(f"支出条目数: {len(expense_rows)}")
+        print(f"不计收支过滤: {skipped_count} 条（还款/冻结等）")
+        
+        # 第三步：处理退款
+        # 京东账单特点：已全额退款的记录金额会显示为 "179.00(已全额退款)"
+        final_expense_rows = self._process_expenses(expense_rows, refund_rows)
+        
+        print(f"\n处理结果:")
+        print(f"  全额退款删除: {self.stats['fully_refunded']} 条")
+        print(f"  部分退款调整: {self.stats['partially_refunded']} 条")
+        if self.stats.get("zero_amount", 0) > 0:
+            print(f"  0元记录过滤: {self.stats['zero_amount']} 条")
+        print(f"  最终保留行数: {len(final_expense_rows)}")
+        
+        # 第四步：转换为对齐格式并重新分类
+        aligned_rows = [self._convert_and_reclassify(row_data) for row_data in final_expense_rows]
+        
+        # 按时间排序（最新在前）
+        aligned_rows.sort(key=lambda x: x[0], reverse=True)
+        
+        # 统计复核数量
+        review_high_count = sum(1 for row in aligned_rows if row[-1] == "HIGH")
+        
+        self.stats["final_count"] = len(aligned_rows)
+        if review_high_count > 0:
+            print(f"  高优先级复核: {review_high_count} 条（无法判断）")
+        
+        # 写入文件
+        self.write_output(ALIGNED_HEADER, aligned_rows)
+        
+        print(f"\n清理后的数据已保存到: {self.output_file}")
+        
+        # 统计支出
+        self._print_expense_summary(aligned_rows)
+    
+    def _parse_jd_amount(self, amount_str: str) -> tuple[Decimal, bool]:
+        """
+        解析京东账单金额
+        
+        京东金额格式特点：
+        - 普通金额: "179.00"
+        - 全额退款: "179.00(已全额退款)"
+        
+        Returns:
+            (金额, 是否已全额退款)
+        """
+        amount_str = amount_str.strip()
+        
+        # 检查是否包含"已全额退款"
+        if "(已全额退款)" in amount_str or "（已全额退款）" in amount_str:
+            # 提取金额部分
+            amount_part = re.sub(r'[（(]已全额退款[)）]', '', amount_str)
+            return parse_amount(amount_part), True
+        
+        return parse_amount(amount_str), False
+    
+    def _process_expenses(self, expense_rows: list, refund_rows: list) -> list:
+        """
+        处理支出记录
+        
+        京东账单特点：
+        1. 已全额退款的记录金额显示为 "金额(已全额退款)"
+        2. 部分退款可能有单独的退款记录
+        """
+        # 构建退款索引（按订单号）
+        order_refunds = {}
+        for row in refund_rows:
+            if len(row) >= 9:
+                order_no = row[8].strip()  # 交易订单号
+                amount = parse_amount(row[3])  # 金额
+                if order_no:
+                    if order_no not in order_refunds:
+                        order_refunds[order_no] = Decimal("0")
+                    order_refunds[order_no] += amount
+                    print(f"  退款记录: {row[0]} | {row[1]} | {amount}元")
+        
+        final_rows = []
+        
+        for row in expense_rows:
+            if len(row) < 9:
+                continue
+            
+            order_no = row[8].strip()  # 交易订单号
+            amount, is_fully_refunded = self._parse_jd_amount(row[3])
+            
+            # 情况1：金额已标注"已全额退款"
+            if is_fully_refunded:
+                self.stats["fully_refunded"] += 1
+                desc = row[2][:25] if len(row[2]) > 25 else row[2]
+                print(f"  全额退款删除: {row[0]} | {row[1]} | {desc}... | {row[3]}")
+                continue
+            
+            # 情况2：检查是否有对应的退款记录
+            refund_amount = order_refunds.get(order_no, Decimal("0"))
+            if refund_amount > 0:
+                if refund_amount >= amount:
+                    # 全额退款
+                    self.stats["fully_refunded"] += 1
+                    desc = row[2][:25] if len(row[2]) > 25 else row[2]
+                    print(f"  全额退款删除: {row[0]} | {row[1]} | {desc}... | 原{amount}元")
+                else:
+                    # 部分退款
+                    remaining = amount - refund_amount
+                    new_row = row.copy()
+                    new_row[3] = format_amount(remaining)
+                    remark = f"原金额{amount}元,退款{refund_amount}元"
+                    
+                    final_rows.append((new_row, remark))
+                    self.stats["partially_refunded"] += 1
+                    print(f"  部分退款: {row[0]} | {row[1]} | 原{amount}元 -> {format_amount(remaining)}元")
+            else:
+                # 无退款，正常记录
+                if amount > 0:
+                    final_rows.append((row, None))
+                else:
+                    self.stats["zero_amount"] = self.stats.get("zero_amount", 0) + 1
+        
+        return final_rows
+    
+    def _convert_and_reclassify(self, row_tuple: tuple) -> list:
+        """
+        转换为对齐格式并重新分类
+        
+        京东原始字段:
+            0: 交易时间, 1: 商户名称, 2: 交易说明, 3: 金额,
+            4: 收/付款方式, 5: 交易状态, 6: 收/支, 7: 交易分类,
+            8: 交易订单号, 9: 商家订单号, 10: 备注
+        
+        对齐后字段:
+            交易时间, 交易分类, 交易对方, 对方账号, 商品说明,
+            收/支, 金额, 收/付款方式, 交易状态, 交易订单号, 商家订单号, 备注, 复核等级
+        """
+        if isinstance(row_tuple, tuple):
+            row, remark = row_tuple
+        else:
+            row, remark = row_tuple, None
+        
+        transaction_time = row[0]
+        merchant = row[1]  # 商户名称
+        product = row[2]  # 交易说明
+        amount, _ = self._parse_jd_amount(row[3])
+        payment_method = row[4] if len(row) > 4 else ""
+        status = row[5] if len(row) > 5 else ""
+        income_expense = row[6] if len(row) > 6 else "支出"
+        original_category = row[7] if len(row) > 7 else ""
+        order_no = row[8] if len(row) > 8 else ""
+        merchant_order_no = row[9] if len(row) > 9 else ""
+        final_remark = remark if remark else (row[10] if len(row) > 10 else "/")
+        
+        # 使用京东专属分类推断
+        category, is_certain, review_level = infer_jd_category(merchant, product, original_category)
+        
+        # 复核等级映射: 0=空, 1=LOW, 2=HIGH
+        review_marks = {0: "", 1: "LOW", 2: "HIGH"}
+        review_mark = review_marks.get(review_level, "")
+        
+        return [
+            transaction_time,
+            category,
+            merchant,
+            "/",  # 对方账号（京东无此字段）
+            product,
+            income_expense,
+            format_amount(amount),
+            payment_method,
+            status,
+            order_no,
+            merchant_order_no,
+            final_remark,
+            review_mark
+        ]
+    
+    def reclassify(self, rows: list) -> list:
+        """
+        重新分类京东账单
+        
+        京东账单在 _convert_and_reclassify 中已完成分类
+        此方法为接口兼容保留
+        """
+        return rows
+    
+    def _print_expense_summary(self, expense_rows: list):
+        """打印支出统计"""
+        total = Decimal("0")
+        categories = {}
+        
+        for row in expense_rows:
+            if row[5] == "支出":
+                amt = Decimal(row[6])
+                total += amt
+                cat = row[1]
+                categories[cat] = categories.get(cat, Decimal("0")) + amt
+        
+        print(f"清理后支出总额: ￥{total}")
+        print("\n=== 按分类统计 ===")
+        for cat, amt in sorted(categories.items(), key=lambda x: -x[1]):
+            print(f"  {cat}: ￥{amt}")
+
+
+def main():
+    """命令行入口"""
+    parser = create_arg_parser("清理京东白条账单数据")
+    args = parser.parse_args()
+    
+    from .base import compute_date_range
+    
+    cleaner = JDCleaner(args.input_file, args.output_file)
+    start_date, end_date = compute_date_range(args)
+    cleaner.set_date_range(start_date, end_date)
+    cleaner.clean()
+
+
+if __name__ == "__main__":
+    main()

analyzer/config/category.yaml

@@ -84,6 +84,40 @@
     - 供暖
     - 暖气
 
+  # 宠物用品
+  宠物用品:
+    - 宠物
+    - 猫咪
+    - 狗
+    - 猫粮
+    - 狗粮
+    - 猫砂
+    - 喂水
+    - 猫零食
+    - 犬猫
+
+  # 数码电器
+  数码电器:
+    - 饮水机
+    - 净水
+    - 制冰
+    - nas
+    - 存储
+    - 硬盘
+    - 电脑
+    - 手机
+    - 平板
+    - 电器
+    - 小家电
+    - 充电
+    - 数据线
+    - 路由器
+    - 音箱
+    - 耳机
+    - 键盘
+    - 鼠标
+    - 显示器
+
   # 运动健身
   运动健身:
     - 健身
@@ -113,6 +147,9 @@
     - 电影
     - 游戏
     - 娱乐
+    - 书
+    - 图书
+    - 文娱
     - 旅游
     - 景区
     - 门票
@@ -157,6 +194,15 @@
     - 妍丽      # AFIONA妍丽美妆店
     - 屈臣氏
     - 丝芙兰
+    - 保鲜盒
+    - 收纳
+    - 厨房
+    - 清洁
+    - 洗衣
+    - 纸巾
+    - 毛巾
+    - 床品
+    - 家居
 
   # 餐饮美食
   餐饮美食:

analyzer/config/category_jd.yaml

@@ -0,0 +1,48 @@
+# =============================================================================
+# 京东账单分类映射配置
+# 将京东原始分类转换为统一分类
+# =============================================================================
+
+# =============================================================================
+# 京东原始分类 -> 统一分类映射
+# 京东账单中的"交易分类"字段可能包含以下值：
+#   - 余额、小金库、白条：财务操作（已在清洗时过滤）
+#   - 其他、其他网购、网购：需要根据商品说明进一步判断
+#   - 食品酒饮：餐饮美食
+#   - 数码电器、电脑办公：数码电器
+#   - 日用百货：日用百货
+#   - 图书文娱：文化休闲
+# =============================================================================
+
+分类映射:
+  # 直接映射（京东分类 -> 统一分类）
+  食品酒饮: 餐饮美食
+  数码电器: 数码电器
+  电脑办公: 数码电器
+  日用百货: 日用百货
+  图书文娱: 文化休闲
+  
+  # 需要进一步判断的分类（返回空字符串，由关键词推断）
+  其他: ""
+  其他网购: ""
+  网购: ""
+  
+  # 财务类（通常已被过滤，但以防万一）
+  余额: ""
+  小金库: ""
+  白条: ""
+
+# =============================================================================
+# 商户名称 -> 统一分类映射
+# 根据商户名称直接映射分类，无需关键词匹配
+# =============================================================================
+
+商户映射:
+  京东外卖: 餐饮美食
+
+# =============================================================================
+# 默认分类
+# 当无法匹配任何规则时使用
+# =============================================================================
+
+默认分类: 其他支出

analyzer/converter.py

@@ -49,7 +49,7 @@ def detect_bill_type_from_content(content: str, filename: str = "") -> str:
     从内容和文件名检测账单类型
     
     Returns:
-        'alipay', 'wechat', 或 ''
+        'alipay', 'wechat', 'jd', 或 ''
     """
     # 从文件名检测
     filename_lower = filename.lower()
@@ -57,6 +57,8 @@ def detect_bill_type_from_content(content: str, filename: str = "") -> str:
         return 'alipay'
     if '微信' in filename or 'wechat' in filename_lower:
         return 'wechat'
+    if '京东' in filename or 'jd' in filename_lower:
+        return 'jd'
     
     # 从内容检测
     # 支付宝特征: 有 "交易分类" 和 "对方账号" 列
@@ -67,6 +69,12 @@ def detect_bill_type_from_content(content: str, filename: str = "") -> str:
     if '交易类型' in content and '金额(元)' in content:
         return 'wechat'
     
+    # 京东特征: 有 "商户名称" 和 "交易说明" 列，或头部包含 "京东账号名"
+    if '商户名称' in content and '交易说明' in content:
+        return 'jd'
+    if '京东账号名' in content:
+        return 'jd'
+    
     return ''
 
 

analyzer/server.py

@@ -22,7 +22,7 @@ if sys.stdout.encoding != 'utf-8':
     sys.stdout = io.TextIOWrapper(sys.stdout.buffer, encoding='utf-8', errors='replace')
 
 from cleaners.base import compute_date_range_from_values
-from cleaners import AlipayCleaner, WechatCleaner
+from cleaners import AlipayCleaner, WechatCleaner, JDCleaner
 from category import infer_category, get_all_categories, get_all_income_categories
 from converter import convert_bill_file
 
@@ -43,7 +43,7 @@ class CleanRequest(BaseModel):
     start: Optional[str] = None
     end: Optional[str] = None
     format: Optional[str] = "csv"
-    bill_type: Optional[str] = "auto"  # auto, alipay, wechat
+    bill_type: Optional[str] = "auto"  # auto, alipay, wechat, jd
 
 
 class CleanResponse(BaseModel):
@@ -90,7 +90,7 @@ def detect_bill_type(filepath: str) -> str | None:
     检测账单类型
     
     Returns:
-        'alipay' | 'wechat' | None
+        'alipay' | 'wechat' | 'jd' | None
     """
     try:
         with open(filepath, "r", encoding="utf-8") as f:
@@ -107,6 +107,14 @@ def detect_bill_type(filepath: str) -> str | None:
                 if "交易类型" in line and "金额(元)" in line:
                     return "wechat"
                 
+                # 京东特征：表头包含 "商户名称" 和 "交易说明"
+                if "商户名称" in line and "交易说明" in line:
+                    return "jd"
+                
+                # 京东特征：头部信息包含 "京东账号名"
+                if "京东账号名" in line:
+                    return "jd"
+                
                 # 数据行特征
                 if line.startswith("202"):
                     if "￥" in line:
@@ -155,14 +163,16 @@ def do_clean(
     try:
         if bill_type == "alipay":
             cleaner = AlipayCleaner(input_path, output_path, output_format)
+        elif bill_type == "jd":
+            cleaner = JDCleaner(input_path, output_path, output_format)
         else:
             cleaner = WechatCleaner(input_path, output_path, output_format)
         
         cleaner.set_date_range(start_date, end_date)
         cleaner.clean()
         
-        type_names = {"alipay": "支付宝", "wechat": "微信"}
+        type_names = {"alipay": "支付宝", "wechat": "微信", "jd": "京东白条"}
-        return True, bill_type, f"✅ {type_names[bill_type]}账单清洗完成"
+        return True, bill_type, f"✅ {type_names.get(bill_type, bill_type)}账单清洗完成"
     
     except Exception as e:
         return False, bill_type, f"清洗失败: {str(e)}"
@@ -324,7 +334,7 @@ async def detect_bill_type_api(file: UploadFile = File(...)):
     """
     检测账单类型
     
-    上传文件后自动检测是支付宝还是微信账单
+    上传文件后自动检测是支付宝、微信还是京东账单
     """
     suffix = Path(file.filename).suffix or ".csv"
     with tempfile.NamedTemporaryFile(delete=False, suffix=suffix) as tmp:
@@ -336,10 +346,10 @@ async def detect_bill_type_api(file: UploadFile = File(...)):
         if bill_type is None:
             raise HTTPException(status_code=400, detail="无法识别账单类型")
         
-        type_names = {"alipay": "支付宝", "wechat": "微信"}
+        type_names = {"alipay": "支付宝", "wechat": "微信", "jd": "京东白条"}
         return {
             "bill_type": bill_type,
-            "display_name": type_names[bill_type]
+            "display_name": type_names.get(bill_type, bill_type)
         }
     finally:
         if os.path.exists(tmp_path):

server/adapter/adapter.go

@@ -13,14 +13,14 @@ type CleanOptions struct {
 
 // CleanResult 清洗结果
 type CleanResult struct {
-	BillType string // 检测到的账单类型: alipay/wechat
+	BillType string // 检测到的账单类型: alipay/wechat/jd
 	Output   string // 脚本输出信息
 }
 
 // ConvertResult 格式转换结果
 type ConvertResult struct {
 	OutputPath string // 转换后的文件路径
-	BillType   string // 检测到的账单类型: alipay/wechat
+	BillType   string // 检测到的账单类型: alipay/wechat/jd
 }
 
 // Cleaner 账单清洗器接口

server/adapter/python/cleaner.go

@@ -87,6 +87,9 @@ func detectBillTypeFromOutput(output string) string {
 	if strings.Contains(output, "微信") {
 		return "wechat"
 	}
+	if strings.Contains(output, "京东") {
+		return "jd"
+	}
 	return ""
 }
 

server/handler/bills.go

@@ -18,7 +18,7 @@ type ListBillsRequest struct {
 	StartDate     string `form:"start_date"`     // 开始日期 YYYY-MM-DD
 	EndDate       string `form:"end_date"`       // 结束日期 YYYY-MM-DD
 	Category      string `form:"category"`       // 分类筛选
-	Type          string `form:"type"`           // 账单类型 alipay/wechat
+	Type          string `form:"type"`           // 账单类型 alipay/wechat/jd
 	IncomeExpense string `form:"income_expense"` // 收支类型 收入/支出
 }
 

server/handler/upload.go

@@ -15,6 +15,7 @@ import (
 
 	"billai-server/config"
 	"billai-server/model"
+	"billai-server/repository"
 	"billai-server/service"
 )
 
@@ -145,6 +146,8 @@ func Upload(c *gin.Context) {
 			billType = "alipay"
 		} else if strings.Contains(fileName, "微信") || strings.Contains(fileName, "wechat") {
 			billType = "wechat"
+		} else if strings.Contains(fileName, "京东") || strings.Contains(fileName, "jd") {
+			billType = "jd"
 		}
 	}
 	if billType == "" {
@@ -152,15 +155,15 @@ func Upload(c *gin.Context) {
 		service.CleanupExtractedFiles(extractedFiles)
 		c.JSON(http.StatusBadRequest, model.UploadResponse{
 			Result:  false,
-			Message: "无法识别账单类型，请指定 type 参数 (alipay 或 wechat)",
+			Message: "无法识别账单类型，请指定 type 参数 (alipay/wechat/jd)",
 		})
 		return
 	}
-	if billType != "alipay" && billType != "wechat" {
+	if billType != "alipay" && billType != "wechat" && billType != "jd" {
 		service.CleanupExtractedFiles(extractedFiles)
 		c.JSON(http.StatusBadRequest, model.UploadResponse{
 			Result:  false,
-			Message: "账单类型无效，仅支持 alipay 或 wechat",
+			Message: "账单类型无效，仅支持 alipay/wechat/jd",
 		})
 		return
 	}
@@ -252,22 +255,41 @@ func Upload(c *gin.Context) {
 	}
 	service.CleanupExtractedFiles(extractedFiles)
 
-	// 13. 返回成功响应
+	// 13. 如果是京东账单，软删除其他来源中包含"京东-订单编号"的记录
+	var jdRelatedDeleted int64
+	if billType == "jd" {
+		repo := repository.GetRepository()
+		if repo != nil {
+			deleted, err := repo.SoftDeleteJDRelatedBills()
+			if err != nil {
+				fmt.Printf("⚠️  软删除京东关联记录失败: %v\n", err)
+			} else if deleted > 0 {
+				jdRelatedDeleted = deleted
+				fmt.Printf("🗑️  已软删除 %d 条其他来源中的京东关联记录\n", deleted)
+			}
+		}
+	}
+
+	// 14. 返回成功响应
 	message := fmt.Sprintf("处理成功，新增 %d 条记录", cleanedCount)
 	if dedupResult.DuplicateCount > 0 {
 		message = fmt.Sprintf("处理成功，新增 %d 条，跳过 %d 条重复记录", cleanedCount, dedupResult.DuplicateCount)
 	}
+	if jdRelatedDeleted > 0 {
+		message = fmt.Sprintf("%s，标记删除 %d 条重复的京东订单", message, jdRelatedDeleted)
+	}
 
 	c.JSON(http.StatusOK, model.UploadResponse{
 		Result:  true,
 		Message: message,
 		Data: &model.UploadData{
-			BillType:       billType,
+			BillType:         billType,
-			FileURL:        fmt.Sprintf("/download/%s", outputFileName),
+			FileURL:          fmt.Sprintf("/download/%s", outputFileName),
-			FileName:       outputFileName,
+			FileName:         outputFileName,
-			RawCount:       rawCount,
+			RawCount:         rawCount,
-			CleanedCount:   cleanedCount,
+			CleanedCount:     cleanedCount,
-			DuplicateCount: dedupResult.DuplicateCount,
+			DuplicateCount:   dedupResult.DuplicateCount,
+			JDRelatedDeleted: jdRelatedDeleted,
 		},
 	})
 }

server/model/bill.go

@@ -70,7 +70,7 @@ func (t LocalTime) Time() time.Time {
 // RawBill 原始账单记录（存储上传的原始数据）
 type RawBill struct {
 	ID          primitive.ObjectID     `bson:"_id,omitempty" json:"id,omitempty"`
-	BillType    string                 `bson:"bill_type" json:"bill_type"`       // 账单类型: alipay/wechat
+	BillType    string                 `bson:"bill_type" json:"bill_type"`       // 账单类型: alipay/wechat/jd
 	SourceFile  string                 `bson:"source_file" json:"source_file"`   // 来源文件名
 	UploadBatch string                 `bson:"upload_batch" json:"upload_batch"` // 上传批次（时间戳）
 	RowIndex    int                    `bson:"row_index" json:"row_index"`       // 原始行号
@@ -81,7 +81,7 @@ type RawBill struct {
 // CleanedBill 清洗后账单记录（标准化后的数据）
 type CleanedBill struct {
 	ID              primitive.ObjectID `bson:"_id,omitempty" json:"id,omitempty"`
-	BillType        string             `bson:"bill_type" json:"bill_type"`                 // 账单类型: alipay/wechat
+	BillType        string             `bson:"bill_type" json:"bill_type"`                 // 账单类型: alipay/wechat/jd
 	TransactionID   string             `bson:"transaction_id" json:"transaction_id"`       // 交易订单号（用于去重）
 	MerchantOrderNo string             `bson:"merchant_order_no" json:"merchant_order_no"` // 商家订单号（用于去重）
 	Time            LocalTime          `bson:"time" json:"time"`                           // 交易时间（本地时间格式）

server/model/request.go

@@ -2,7 +2,7 @@ package model
 
 // UploadRequest 上传请求参数
 type UploadRequest struct {
-	Type     string `form:"type"`     // 账单类型: alipay/wechat（可选，会自动检测）
+	Type     string `form:"type"`     // 账单类型: alipay/wechat/jd（可选，会自动检测）
 	Password string `form:"password"` // ZIP 文件密码（可选）
 	Year     string `form:"year"`     // 年份筛选
 	Month    string `form:"month"`    // 月份筛选

server/model/response.go

@@ -2,12 +2,13 @@ package model
 
 // UploadData 上传响应数据
 type UploadData struct {
-	BillType       string `json:"bill_type,omitempty"`       // alipay/wechat
+	BillType         string `json:"bill_type,omitempty"`          // alipay/wechat/jd
-	FileURL        string `json:"file_url,omitempty"`        // 下载链接
+	FileURL          string `json:"file_url,omitempty"`           // 下载链接
-	FileName       string `json:"file_name,omitempty"`       // 文件名
+	FileName         string `json:"file_name,omitempty"`          // 文件名
-	RawCount       int    `json:"raw_count,omitempty"`       // 存储到原始数据集合的记录数
+	RawCount         int    `json:"raw_count,omitempty"`          // 存储到原始数据集合的记录数
-	CleanedCount   int    `json:"cleaned_count,omitempty"`   // 存储到清洗后数据集合的记录数
+	CleanedCount     int    `json:"cleaned_count,omitempty"`      // 存储到清洗后数据集合的记录数
-	DuplicateCount int    `json:"duplicate_count,omitempty"` // 重复跳过的记录数
+	DuplicateCount   int    `json:"duplicate_count,omitempty"`    // 重复跳过的记录数
+	JDRelatedDeleted int64  `json:"jd_related_deleted,omitempty"` // 软删除的京东关联记录数（其他来源中描述包含京东订单号的记录）
 }
 
 // UploadResponse 上传响应

server/repository/mongo/repository.go

@@ -458,6 +458,41 @@ func (r *Repository) DeleteCleanedBillByID(id string) error {
 	return nil
 }
 
+// SoftDeleteJDRelatedBills 软删除描述中包含"京东-订单编号"的非京东账单
+// 用于避免京东账单与其他来源（微信、支付宝）账单重复计算
+func (r *Repository) SoftDeleteJDRelatedBills() (int64, error) {
+	if r.cleanedCollection == nil {
+		return 0, fmt.Errorf("cleaned collection not initialized")
+	}
+
+	ctx, cancel := context.WithTimeout(context.Background(), 30*time.Second)
+	defer cancel()
+
+	// 筛选条件：
+	// 1. 账单类型不是 jd（只处理微信、支付宝等其他来源）
+	// 2. 描述中包含"京东-订单编号"
+	// 3. 尚未被删除
+	filter := bson.M{
+		"bill_type":   bson.M{"$ne": "jd"},
+		"description": bson.M{"$regex": "京东-订单编号", "$options": ""},
+		"is_deleted":  bson.M{"$ne": true},
+	}
+
+	update := bson.M{
+		"$set": bson.M{
+			"is_deleted": true,
+			"updated_at": time.Now(),
+		},
+	}
+
+	result, err := r.cleanedCollection.UpdateMany(ctx, filter, update)
+	if err != nil {
+		return 0, fmt.Errorf("soft delete JD related bills failed: %w", err)
+	}
+
+	return result.ModifiedCount, nil
+}
+
 // GetClient 获取 MongoDB 客户端（用于兼容旧代码）
 func (r *Repository) GetClient() *mongo.Client {
 	return r.client

server/repository/repository.go

@@ -51,4 +51,9 @@ type BillRepository interface {
 
 	// CountRawByField 按字段统计原始数据数量
 	CountRawByField(fieldName, value string) (int64, error)
+
+	// SoftDeleteJDRelatedBills 软删除描述中包含"京东-订单编号"的非京东账单
+	// 用于避免京东账单与其他来源（微信、支付宝）账单重复计算
+	// 返回: 删除数量、错误
+	SoftDeleteJDRelatedBills() (int64, error)
 }

server/service/archive.go

@@ -105,6 +105,8 @@ func ExtractZip(zipPath, destDir, password string) (*ExtractResult, error) {
 				result.BillType = "alipay"
 			} else if strings.Contains(fileName, "微信") || strings.Contains(strings.ToLower(fileName), "wechat") {
 				result.BillType = "wechat"
+			} else if strings.Contains(fileName, "京东") || strings.Contains(strings.ToLower(fileName), "jd") {
+				result.BillType = "jd"
 			}
 		}
 	}

server/service/bill.go

@@ -150,6 +150,10 @@ func detectBillTypeAndIdField(header []string) (billType string, idFieldIdx int)
 		if col == "交易类型" || col == "金额(元)" {
 			billType = "wechat"
 		}
+		// 京东特征
+		if col == "商户名称" || col == "交易说明" {
+			billType = "jd"
+		}
 
 		// 查找去重字段（优先使用交易订单号/交易号）
 		if col == "交易订单号" || col == "交易号" || col == "交易单号" {

server/service/cleaner.go

@@ -36,6 +36,9 @@ func DetectBillTypeFromOutput(output string) string {
 	if containsSubstring(output, "微信") {
 		return "wechat"
 	}
+	if containsSubstring(output, "京东") {
+		return "jd"
+	}
 	return ""
 }
 

web/src/lib/api.ts

@@ -42,7 +42,7 @@ export async function checkHealth(): Promise<boolean> {
 }
 
 // 类型定义
-export type BillType = 'alipay' | 'wechat';
+export type BillType = 'alipay' | 'wechat' | 'jd';
 
 export interface UploadData {
   bill_type: BillType;

web/src/routes/+page.svelte

@@ -213,6 +213,8 @@
       selectedType = 'alipay';
     } else if (fileName.includes('微信') || fileName.includes('wechat')) {
       selectedType = 'wechat';
+    } else if (fileName.includes('京东') || fileName.includes('jd')) {
+      selectedType = 'jd';
     }
   }
   
@@ -263,7 +265,7 @@
   <!-- 页面标题 -->
   <div>
     <h1 class="text-2xl font-bold tracking-tight">账单管理</h1>
-    <p class="text-muted-foreground">上传并分析您的支付宝、微信账单</p>
+    <p class="text-muted-foreground">上传并分析您的支付宝、微信、京东账单</p>
   </div>
 
   <!-- 统计卡片 -->
@@ -297,7 +299,7 @@
       <Card.Header class="flex flex-row items-center justify-between space-y-0">
         <div>
           <Card.Title>上传账单</Card.Title>
-          <Card.Description>支持支付宝、微信账单 CSV、XLSX 或 ZIP 文件</Card.Description>
+          <Card.Description>支持支付宝、微信、京东账单 CSV、XLSX 或 ZIP 文件</Card.Description>
         </div>
         <Button variant="outline" size="sm" onclick={() => goto('/bills?tab=manual')}>
           <Plus class="mr-2 h-4 w-4" />
@@ -397,6 +399,13 @@
             >
               微信
             </Button>
+            <Button 
+              variant={selectedType === 'jd' ? 'default' : 'outline'}
+              size="sm"
+              onclick={() => selectedType = 'jd'}
+            >
+              京东
+            </Button>
           </div>
         </div>
 
@@ -438,7 +447,7 @@
               <div class="flex items-center justify-between">
                 <span class="text-sm text-muted-foreground">账单类型</span>
                 <Badge variant="secondary">
-                  {uploadResult.data?.bill_type === 'alipay' ? '支付宝' : '微信'}
+                  {uploadResult.data?.bill_type === 'alipay' ? '支付宝' : uploadResult.data?.bill_type === 'wechat' ? '微信' : '京东'}
                 </Badge>
               </div>
               <div class="flex items-center justify-between">

web/src/routes/bills/+page.svelte

@@ -380,13 +380,14 @@
             <Label class="text-xs">来源</Label>
             <Select.Root type="single" value={filterBillType || undefined} onValueChange={handleBillTypeChange}>
               <Select.Trigger class="h-9 w-full">
-                <span class="text-sm">{filterBillType === 'alipay' ? '支付宝' : filterBillType === 'wechat' ? '微信' : filterBillType === 'manual' ? '手动' : '全部'}</span>
+                <span class="text-sm">{filterBillType === 'alipay' ? '支付宝' : filterBillType === 'wechat' ? '微信' : filterBillType === 'jd' ? '京东' : filterBillType === 'manual' ? '手动' : '全部'}</span>
               </Select.Trigger>
               <Select.Portal>
                 <Select.Content>
                   <Select.Item value="">全部</Select.Item>
                   <Select.Item value="alipay">支付宝</Select.Item>
                   <Select.Item value="wechat">微信</Select.Item>
+                  <Select.Item value="jd">京东</Select.Item>
                   <Select.Item value="manual">手动</Select.Item>
                 </Select.Content>
               </Select.Portal>
@@ -438,8 +439,8 @@
                     {formatDateTime(record.time)}
                   </Table.Cell>
                   <Table.Cell class="hidden xl:table-cell">
-                    <Badge variant={record.bill_type === 'manual' ? 'outline' : (record.bill_type === 'alipay' ? 'default' : 'secondary')}>
+<Badge variant={record.bill_type === 'manual' ? 'outline' : (record.bill_type === 'alipay' ? 'default' : (record.bill_type === 'jd' ? 'destructive' : 'secondary'))}>
-                      {record.bill_type === 'manual' ? '手动输入' : (record.bill_type === 'alipay' ? '支付宝' : '微信')}
+                      {record.bill_type === 'manual' ? '手动输入' : (record.bill_type === 'alipay' ? '支付宝' : (record.bill_type === 'jd' ? '京东' : '微信'))}
                     </Badge>
                   </Table.Cell>
                   <Table.Cell>