diff --git a/带公式版本.py b/带公式版本.py
new file mode 100644
index 0000000..61025a3
--- /dev/null
+++ b/带公式版本.py
@@ -0,0 +1,1160 @@
+import pandas as pd
+import tkinter as tk
+from tkinter import ttk, filedialog, messagebox, simpledialog
+import os
+import re
+import numpy as np
+from datetime import datetime
+import traceback
+from openpyxl.utils import get_column_letter
+
+
+# ==========================================
+# 第一部分：业务逻辑核心
+# ==========================================
+
+class DataProcessor:
+    def __init__(self):
+        # ==================== 表头定义 ====================
+        self.cols_asd_foreign_general = [
+            "合同编号", "签署公司", "外贸合同号", "收款情况", "合同签订日期", "销售员",
+            "最终用户单位", "最终用户信息\n联系人、电话、邮箱", "最终用户所在地",
+            "厂家", "型号/货号", "合同标的", "数量", "单位", "币种", "折扣率",
+            "合同额", "总合同额", "外购", "已收款", "未收款", "收款日期",
+            "最晚发货期", "付款方式", "发货港", "目的港", "发货日期",
+            "买方单位", "买方信息\n联系人、电话、邮箱", "收货人信息",
+            "转为美元净合同额", "转为美元总合同额"
+        ]
+
+        self.cols_nonasd_foreign_general = [
+            "合同编号", "签署公司", "外贸合同号", "收款情况", "合同签订日期", "销售员",
+            "最终用户单位", "最终用户信息\n联系人、电话、邮箱", "最终用户所在地",
+            "厂家", "型号/货号", "合同标的", "数量", "单位", "币种", "折扣率",
+            "合同额", "总合同额", "外购", "已收款", "未收款", "收款日期",
+            "最晚发货期", "付款方式", "发货港", "目的港", "发货日期",
+            "买方单位", "买方信息\n联系人、电话、邮箱", "收货人信息",
+            "合同币种/美元", "转为美元净合同额", "转为美元总合同额"
+        ]
+
+        self.cols_domestic_general = [
+            "合同编号", "签署公司", "内贸合同号", "收款情况", "签订日期", "销售员",
+            "最终用户单位", "最终用户信息\n联系人、电话、邮箱", "最终用户所在地",
+            "买方单位", "买方信息\n联系人、电话、邮箱",
+            "厂家", "型号", "合同标的", "数量", "单位", "折扣率（％）",
+            "合同额", "合同总额", "外购", "付款方式", "最晚发货期",
+            "已收款", "未收款", "收款日期",
+            "转为美元净合同额", "转为美元总合同额"
+        ]
+
+        self.cols_foreign_detail = [
+            "合同编号", "销售员", "合同标的", "厂家", "货号", "产品描述", "数量", "单位",
+            "币种", "报价单价", "报价总价", "销售单价", "销售总价", "折扣率",
+            "外购", "合同币种/美元", "外购转美元", "报价总价美元", "净合同额美元"
+        ]
+
+        self.cols_domestic_detail = [
+            "合同编号", "销售员", "合同标的", "厂家", "货号", "产品描述", "数量", "单位",
+            "外币币种", "外币报价单价", "报价RMB单价", "报价RMB总价",
+            "售价RMB单价", "售价RMB总价", "折扣率（％）", "外购",
+            "计算汇率", "外购转美元", "报价总价美元", "净合同额美元"
+        ]
+
+        self.cols_om = [
+            "合同编号", "签署公司", "内贸合同号", "收款情况", "签订日期", "销售员",
+            "最终用户单位", "最终用户信息\n联系人、电话、邮箱", "最终用户所在地",
+            "买方单位", "买方信息\n联系人、电话、邮箱", "合同标的",
+            "合同总额", "已收款", "未收款", "收款日期"
+        ]
+
+        # 辅助集合：金额列
+        self.money_cols = set([
+            "合同额", "总合同额", "合同总额", "外购", "已收款", "未收款",
+            "净合同额美元", "外购转美元", "报价总价美元",
+            "外币报价单价", "报价RMB单价", "报价RMB总价",
+            "售价RMB单价", "售价RMB总价", "外购产品金额",
+            "转为美元净合同额", "转为美元总合同额", "报价单价", "报价总价", "销售单价", "销售总价"
+        ])
+
+        # [重要] 这里只保留真正的折扣率列，绝不包含汇率列
+        self.percent_cols = set(["折扣率", "折扣率（％）"])
+
+        self.date_cols = set(["合同签订日期", "签订日期", "收款日期", "最晚发货期", "发货日期"])
+
+        self.legacy_map = {
+            "外币币种": "币种", "汇率": "计算汇率", "折扣率(%)": "折扣率",
+            "折扣率（%）": "折扣率（％）", "合同": "合同额"
+        }
+
+        self.standard_col_map = {}
+        all_lists = [
+            self.cols_asd_foreign_general, self.cols_nonasd_foreign_general,
+            self.cols_domestic_general, self.cols_foreign_detail,
+            self.cols_domestic_detail, self.cols_om
+        ]
+        for lst in all_lists:
+            for col in lst:
+                clean_key = self.clean_header_key(col)
+                self.standard_col_map[clean_key] = col
+
+    def clean_header_key(self, text):
+        if not isinstance(text, str): return str(text)
+        return re.sub(r'[\s\n\r]+', '', text)
+
+    def safe_float(self, val):
+        try:
+            if isinstance(val, (int, float)): return float(val)
+            if isinstance(val, str):
+                val = val.replace(',', '').replace('¥', '').replace('$', '').strip()
+                if val == '': return 0.0
+            if pd.isna(val): return 0.0
+            return float(val)
+        except:
+            return 0.0
+
+    def format_money_str(self, val):
+        if pd.isna(val) or str(val).strip() == "": return ""
+        try:
+            f_val = self.safe_float(val)
+            return "{:.2f}".format(f_val)
+        except:
+            return str(val)
+
+    def format_percent_str(self, val):
+        if pd.isna(val) or str(val).strip() == "": return ""
+        try:
+            s_val = str(val).strip()
+            if '%' in s_val: return s_val
+            f_val = self.safe_float(val)
+            return "{:.2f}%".format(f_val * 100)
+        except:
+            return str(val)
+
+    def format_date_str(self, val):
+        if pd.isna(val) or str(val).strip() == "": return ""
+        try:
+            s_val = str(val).strip()
+            if len(s_val) == 10 and s_val[4] == '-' and s_val[7] == '-': return s_val
+            dt = pd.to_datetime(val, errors='coerce')
+            if pd.isnull(dt): return s_val
+            return dt.strftime('%Y-%m-%d')
+        except:
+            return str(val)
+
+    def load_multiple_csvs(self, file_paths):
+        if isinstance(file_paths, str):
+            paths = [p.strip() for p in file_paths.split(';') if p.strip()]
+        else:
+            paths = list(file_paths)
+
+        if not paths: return None, "未选择文件"
+
+        all_dfs = []
+        error_msgs = []
+        col_factory_general = '厂家'
+        col_factory_detail = '厂家'
+
+        for path in paths:
+            if not os.path.exists(path): continue
+            df = None
+            for enc in ['utf-8', 'gbk', 'gb18030']:
+                try:
+                    df = pd.read_csv(path, encoding=enc)
+                    break
+                except UnicodeDecodeError:
+                    continue
+            if df is None:
+                try:
+                    df = pd.read_csv(path, encoding='gb18030', encoding_errors='replace')
+                except:
+                    error_msgs.append(f"无法读取: {os.path.basename(path)}")
+                    continue
+
+            df['合同类型'] = df['合同类型'].fillna('').astype(str)
+            if '厂家.1' in df.columns: col_factory_detail = '厂家.1'
+            all_dfs.append(df)
+
+        if not all_dfs: return None, "\n".join(error_msgs) if error_msgs else "没有有效的数据文件"
+
+        try:
+            final_df = pd.concat(all_dfs, ignore_index=True)
+        except Exception as e:
+            return None, f"文件合并失败: {str(e)}"
+
+        col_factory_detail = '厂家.1' if '厂家.1' in final_df.columns else '厂家'
+        final_df[col_factory_general] = final_df[col_factory_general].fillna('').astype(str)
+
+        if '合同订单编号' in final_df.columns:
+            final_df['原始_合同订单编号'] = final_df['合同订单编号'].astype(str).str.strip()
+            final_df['Clean_ID'] = final_df['原始_合同订单编号'].apply(lambda x: x.split()[0] if x else "")
+
+        return final_df, (col_factory_general, col_factory_detail)
+
+    def parse_buyer_info(self, text):
+        info = {'name': '', 'contact_full': ''}
+        if not isinstance(text, str) or not text.strip(): return info
+        lines = [l.strip() for l in text.split('\n') if l.strip()]
+        if not lines: return info
+        info['name'] = lines[0]
+        info['contact_full'] = " ".join(lines[1:])
+        return info
+
+    def parse_single_line_subject(self, text):
+        res = {'name': '', 'model': '', 'qty': '', 'unit': '', 'price': '', 'sort_price': 0.0}
+        if not isinstance(text, str) or not text.strip(): return res
+        text = text.strip()
+
+        if re.match(r'^Model[:：]', text, re.IGNORECASE):
+            m = re.match(
+                r'Model[:：]\s*(.+?)\s+([a-zA-Z\u4e00-\u9fa5]+)\s+(\d+(?:\.\d+)?)\s+(\d+(?:\.\d+)?)\s+(\d+(?:\.\d+)?)',
+                text, re.IGNORECASE)
+            if m:
+                res['name'] = m.group(1).strip()
+                res['model'] = m.group(1).strip()
+                res['unit'] = m.group(2)
+                res['qty'] = m.group(3)
+                res['price'] = m.group(4)
+                res['sort_price'] = self.safe_float(m.group(4))
+                return res
+            else:
+                m2 = re.match(r'Model[:：]\s*(.+?)\s+(\d+(?:\.\d+)?)\s+(\d+(?:\.\d+)?)\s+(\d+(?:\.\d+)?)', text,
+                              re.IGNORECASE)
+                if m2:
+                    res['name'] = m2.group(1).strip()
+                    res['model'] = m2.group(1).strip()
+                    res['qty'] = m2.group(2)
+                    res['price'] = m2.group(3)
+                    res['sort_price'] = self.safe_float(m2.group(3))
+                    return res
+
+        if '/' in text:
+            parts = [p.strip() for p in text.split('/')]
+            if len(parts) >= 1: res['name'] = parts[0]
+            if len(parts) >= 2: res['model'] = parts[1]
+            if len(parts) >= 3:
+                m_qty = re.match(r'^(\d+(\.\d+)?)\s*([\u4e00-\u9fa5a-zA-Z]+)?$', parts[2])
+                if m_qty:
+                    res['qty'] = m_qty.group(1)
+                    res['unit'] = m_qty.group(3) if m_qty.group(3) else ""
+                else:
+                    res['qty'] = parts[2]
+            if len(parts) >= 4:
+                res['price'] = parts[3]
+                res['sort_price'] = self.safe_float(parts[3])
+            return res
+
+        name_match = re.search(r'(?:中文品名|中文名称|名称|Name)[:：]\s*(.*?)(?:\n|$)', text, re.IGNORECASE)
+        if name_match:
+            res['name'] = name_match.group(1).strip()
+        else:
+            res['name'] = text.split('\n')[0]
+
+        nums = re.findall(r'\d+(?:\.\d+)?', text.replace(',', ''))
+        if nums:
+            res['sort_price'] = self.safe_float(nums[-1])
+            res['price'] = nums[-1]
+        return res
+
+    def prepare_new_data_general(self, csv_df, trade_type, target_cols, col_factory):
+        if csv_df.empty: return pd.DataFrame(columns=target_cols)
+
+        def extract_items(row):
+            target_raw = str(row.get('合同标的（品名/型号/数量/单价/总价）', ''))
+            lines = [line.strip() for line in target_raw.split('\n') if line.strip()]
+            items = []
+            if not lines:
+                items.append(self.parse_single_line_subject(""))
+            else:
+                for line in lines: items.append(self.parse_single_line_subject(line))
+            return items
+
+        parsed_series = csv_df.apply(extract_items, axis=1)
+
+        expanded_data = []
+        for idx, row in enumerate(csv_df.itertuples(index=False)):
+            items = parsed_series.iloc[idx]
+            r_dict = csv_df.iloc[idx].to_dict()
+
+            for item in items:
+                row_base = {
+                    'Clean_ID': r_dict.get('Clean_ID', ''),
+                    '原始_合同订单编号': r_dict.get('原始_合同订单编号', ''),
+                    '收款账户': r_dict.get('收款账户', ''),
+                    '签约日期': r_dict.get('签约日期', ''),
+                    '负责人': r_dict.get('负责人', ''),
+                    '客户名称': r_dict.get('客户名称', ''),
+                    '联系人姓名': r_dict.get('联系人姓名', ''),
+                    '合同买方_Raw': r_dict.get('合同买方（名称/联系人/电话/邮箱）', ''),
+                    '进口代理_Raw': r_dict.get('进口代理（名称/USCI/地址/联系人/电话/邮箱）', ''),
+                    '厂家_Val': r_dict.get(col_factory, ''),
+                    '货币': r_dict.get('货币(选完产品再改)', ''),
+                    '发货地': r_dict.get('发货地', ''),
+                    '目的港': r_dict.get('目的港', ''),
+                    '折扣率': r_dict.get('折扣率', ''),
+                    '合同总额': r_dict.get('合同总额', ''),
+                    '收款状态': r_dict.get('收款状态', ''),
+                    '外购产品金额': r_dict.get('外购产品金额', ''),
+                    '最新收款日期': r_dict.get('最新收款日期', ''),
+                    '最晚发货期': r_dict.get('最晚发货期', ''),
+                    '付款比例及期限': r_dict.get('付款比例及期限', ''),
+                    '合同币种/美元': r_dict.get('合同币种/美元', ''),
+                    '_item_name': item['name'],
+                    '_item_model': item['model'],
+                    '_item_qty': item['qty'],
+                    '_item_unit': item['unit'],
+                    '_item_price': item['price'],
+                    '_sort_price': item['sort_price']
+                }
+                expanded_data.append(row_base)
+
+        df_expanded = pd.DataFrame(expanded_data)
+        if df_expanded.empty: return pd.DataFrame(columns=target_cols)
+
+        # 排序去重：取金额最大
+        df_expanded.sort_values(by=['Clean_ID', '_sort_price'], ascending=[True, False], inplace=True)
+        df_unique = df_expanded.drop_duplicates(subset=['Clean_ID'], keep='first').copy()
+
+        result = pd.DataFrame(index=df_unique.index)
+
+        parts = df_unique['原始_合同订单编号'].str.split(n=1, expand=True)
+        result['合同编号'] = parts[0]
+        contract_no_col = '外贸合同号' if trade_type == '外贸' else '内贸合同号'
+        result[contract_no_col] = parts[1] if parts.shape[1] > 1 else ""
+
+        result['签署公司'] = df_unique['收款账户']
+        result['合同签订日期'] = df_unique['签约日期'].apply(self.format_date_str)
+        if '签订日期' in target_cols: result['签订日期'] = result['合同签订日期']
+
+        result['销售员'] = df_unique['负责人']
+        result['最终用户单位'] = df_unique['客户名称']
+
+        if any("最终用户信息" in c for c in target_cols):
+            col_name = next(c for c in target_cols if "最终用户信息" in c)
+            result[col_name] = df_unique['联系人姓名']
+
+        def get_buyer_info(row):
+            raw = row['进口代理_Raw'] if trade_type == '外贸' and str(row['进口代理_Raw']) not in ['', 'nan'] else row[
+                '合同买方_Raw']
+            return self.parse_buyer_info(str(raw))
+
+        buyer_infos = df_unique.apply(get_buyer_info, axis=1)
+        result['买方单位'] = [x['name'] for x in buyer_infos]
+
+        if any("买方信息" in c for c in target_cols):
+            col_name = next(c for c in target_cols if "买方信息" in c)
+            result[col_name] = [x['contact_full'] for x in buyer_infos]
+
+        if '收货人信息' in target_cols: result['收货人信息'] = result['买方单位']
+
+        result['厂家'] = df_unique['厂家_Val']
+        if '币种' in target_cols: result['币种'] = df_unique['货币']
+        if '发货港' in target_cols: result['发货港'] = df_unique['发货地']
+        if '目的港' in target_cols: result['目的港'] = df_unique['目的港']
+
+        discount_col = '折扣率' if '折扣率' in target_cols else '折扣率（％）'
+        result[discount_col] = df_unique['折扣率'].apply(self.format_percent_str)
+
+        result['合同标的'] = df_unique['_item_name']
+        if '型号/货号' in target_cols: result['型号/货号'] = df_unique['_item_model']
+        if '型号' in target_cols: result['型号'] = df_unique['_item_model']
+        result['数量'] = df_unique['_item_qty']
+        result['单位'] = df_unique['_item_unit']
+
+        if '合同额' in target_cols: result['合同额'] = df_unique['_item_price'].apply(self.format_money_str)
+
+        total_col_name = '总合同额' if '总合同额' in target_cols else '合同总额'
+        result[total_col_name] = df_unique['合同总额'].apply(self.format_money_str)
+
+        result['收款情况'] = df_unique['收款状态'].fillna('').astype(str).str.strip()
+
+        is_received = result['收款情况'].str.contains('已收')
+        result['已收款'] = ""
+        result['未收款'] = ""
+        result.loc[is_received, '已收款'] = result.loc[is_received, total_col_name]
+        result.loc[is_received, '未收款'] = "0.00"
+
+        result['外购'] = df_unique['外购产品金额'].apply(self.format_money_str)
+        result['收款日期'] = df_unique['最新收款日期'].apply(self.format_date_str)
+
+        if '最晚发货期' in target_cols: result['最晚发货期'] = df_unique['最晚发货期'].apply(self.format_date_str)
+        if '付款方式' in target_cols: result['付款方式'] = df_unique['付款比例及期限']
+        if '发货日期' in target_cols: result['发货日期'] = ""
+
+        # [修改] 汇率原样读取
+        if '合同币种/美元' in target_cols:
+            result['合同币种/美元'] = df_unique['合同币种/美元'].fillna("").astype(str)
+
+        result['_sort_price'] = df_unique['_sort_price']
+
+        for col in target_cols:
+            if col not in result.columns: result[col] = ""
+
+        return result[target_cols + ['_sort_price']]
+
+    def prepare_new_data_detail(self, csv_df, trade_type, target_cols, col_factory):
+        if csv_df.empty: return pd.DataFrame(columns=target_cols)
+
+        new_rows = csv_df.apply(lambda r: self.process_row_detail_single(r, col_factory, trade_type), axis=1)
+        if isinstance(new_rows, pd.Series): new_rows = pd.DataFrame([new_rows])
+
+        for col in target_cols:
+            if col not in new_rows.columns: new_rows[col] = ""
+
+        # 分摊计算逻辑
+        new_rows['_temp_id'] = csv_df['Clean_ID'].values
+
+        if '合同总额' in csv_df.columns:
+            contract_totals = csv_df.groupby('Clean_ID')['合同总额'].first().apply(self.safe_float)
+        else:
+            contract_totals = pd.Series(0, index=new_rows['_temp_id'].unique())
+
+        if '外购产品金额' in csv_df.columns:
+            outsourced_totals = csv_df.groupby('Clean_ID')['外购产品金额'].first().apply(self.safe_float)
+        else:
+            outsourced_totals = pd.Series(0, index=new_rows['_temp_id'].unique())
+
+        quote_col = '报价总价' if trade_type == '外贸' else '报价RMB总价'
+        new_rows['_temp_quote_val'] = new_rows[quote_col].apply(self.safe_float)
+
+        sum_quotes = new_rows.groupby('_temp_id')['_temp_quote_val'].transform('sum')
+
+        mapped_contract_total = new_rows['_temp_id'].map(contract_totals).fillna(0)
+        mapped_outsourced = new_rows['_temp_id'].map(outsourced_totals).fillna(0)
+
+        net_contract_amt = mapped_contract_total - mapped_outsourced
+
+        ratio = pd.Series(0.0, index=new_rows.index)
+        valid_sum = sum_quotes != 0
+        ratio[valid_sum] = net_contract_amt[valid_sum] / sum_quotes[valid_sum]
+
+        sales_total_val = new_rows['_temp_quote_val'] * ratio
+
+        qty_val = new_rows['数量'].apply(self.safe_float)
+        sales_unit_val = pd.Series(0.0, index=new_rows.index)
+        valid_qty = qty_val != 0
+        sales_unit_val[valid_qty] = sales_total_val[valid_qty] / qty_val[valid_qty]
+
+        price_col = '销售单价' if trade_type == '外贸' else '售价RMB单价'
+        total_col = '销售总价' if trade_type == '外贸' else '售价RMB总价'
+
+        new_rows[price_col] = sales_unit_val.apply(lambda x: "{:.2f}".format(x))
+        new_rows[total_col] = sales_total_val.apply(lambda x: "{:.2f}".format(x))
+
+        new_rows.drop(columns=['_temp_id', '_temp_quote_val'], inplace=True, errors='ignore')
+
+        return new_rows[target_cols + ['_is_outsourced']]
+
+    def process_row_detail_single(self, row, col_factory, trade_type):
+        if trade_type == '外贸':
+            target_cols = self.cols_foreign_detail
+        else:
+            target_cols = self.cols_domestic_detail
+        new_row = {col: "" for col in target_cols}
+
+        detail_manuf_val = str(row.get(col_factory, ''))
+        raw_order_no = str(row.get('原始_合同订单编号', row.get('合同订单编号', ''))).strip()
+        parts_no = raw_order_no.split()
+        new_row['合同编号'] = parts_no[0] if len(parts_no) > 0 else raw_order_no
+
+        new_row['销售员'] = row.get('负责人', '')
+        new_row['厂家'] = detail_manuf_val
+        new_row['货号'] = row.get('产品编码', '')
+
+        # 标记外购
+        is_outsourced = 1 if '外购' in detail_manuf_val else 0
+        new_row['_is_outsourced'] = is_outsourced
+
+        if trade_type == '外贸':
+            new_row['币种'] = row.get('原币种', '')
+        else:
+            new_row['外币币种'] = row.get('原币种', '')
+
+        target_raw = str(row.get('合同标的（品名/型号/数量/单价/总价）', ''))
+        if '/' in target_raw:
+            new_row['合同标的'] = target_raw.split('/')[0].strip()
+        else:
+            new_row['合同标的'] = target_raw.split('\n')[0].strip()
+
+        csv_qty = str(row.get('数量', '')).strip()
+        if csv_qty and csv_qty.lower() != 'nan': new_row['数量'] = csv_qty
+
+        val_product_subtotal = self.safe_float(row.get('产品小计', 0))
+        if '外购' in detail_manuf_val:
+            new_row['外购'] = self.format_money_str(val_product_subtotal)
+            remark = str(row.get('备注', '')).strip()
+            if not remark or remark.lower() == 'nan':
+                outsourced = str(row.get('外购产品明细', '')).strip()
+                new_row['产品描述'] = outsourced if outsourced.lower() != 'nan' else ""
+            else:
+                new_row['产品描述'] = remark
+        else:
+            new_row['外购'] = ""
+            new_row['产品描述'] = row.get('产品名称', '')
+            if '外币报价单价' in new_row: new_row['外币报价单价'] = self.format_money_str(row.get('美元报价', ''))
+            if '报价单价' in new_row: new_row['报价单价'] = self.format_money_str(row.get('美元报价', ''))
+            if '报价RMB总价' in new_row: new_row['报价RMB总价'] = self.format_money_str(row.get('产品小计', ''))
+            if '报价总价' in new_row: new_row['报价总价'] = self.format_money_str(row.get('产品小计', ''))
+
+        # [修改] 汇率原样读取，不做处理
+        rate_val = str(row.get('汇率', '')).strip()
+        if rate_val.lower() == 'nan': rate_val = ""
+
+        if '计算汇率' in new_row: new_row['计算汇率'] = rate_val
+        if '合同币种/美元' in new_row: new_row['合同币种/美元'] = rate_val
+
+        discount_col = '折扣率' if '折扣率' in new_row else '折扣率（％）'
+        if discount_col in new_row: new_row[discount_col] = self.format_percent_str(row.get('折扣率', ''))
+
+        new_row['外购转美元'] = self.format_money_str(row.get('外购转美元', ''))
+        new_row['报价总价美元'] = self.format_money_str(row.get('报价总价美元', ''))
+        new_row['净合同额美元'] = self.format_money_str(row.get('净合同额美元', ''))
+        if '报价RMB单价' in new_row: new_row['报价RMB单价'] = self.format_money_str(row.get('报价RMB单价', ''))
+
+        return pd.Series(new_row)
+
+    def prepare_new_data_om(self, csv_df, target_cols):
+        if csv_df.empty: return pd.DataFrame(columns=target_cols)
+
+        def extract_items(row):
+            target_raw = str(row.get('合同标的（品名/型号/数量/单价/总价）', ''))
+            lines = [line.strip() for line in target_raw.split('\n') if line.strip()]
+            items = []
+            if not lines:
+                items.append(self.parse_single_line_subject(""))
+            else:
+                for line in lines: items.append(self.parse_single_line_subject(line))
+            return items
+
+        parsed_series = csv_df.apply(extract_items, axis=1)
+        expanded_data = []
+        for idx, row in enumerate(csv_df.itertuples(index=False)):
+            r_dict = csv_df.iloc[idx].to_dict()
+            items = parsed_series.iloc[idx]
+            for item in items:
+                row_base = {
+                    'Clean_ID': r_dict.get('Clean_ID', ''),
+                    '原始_合同订单编号': r_dict.get('原始_合同订单编号', ''),
+                    '收款账户': r_dict.get('收款账户', ''),
+                    '签约日期': r_dict.get('签约日期', ''),
+                    '负责人': r_dict.get('负责人', ''),
+                    '客户名称': r_dict.get('客户名称', ''),
+                    '联系人姓名': r_dict.get('联系人姓名', ''),
+                    '合同买方_Raw': r_dict.get('合同买方（名称/联系人/电话/邮箱）', ''),
+                    '合同总额': r_dict.get('合同总额', ''),
+                    '收款状态': r_dict.get('收款状态', ''),
+                    '最新收款日期': r_dict.get('最新收款日期', ''),
+                    '_item_name': item['name'],
+                    '_sort_price': item['sort_price']
+                }
+                expanded_data.append(row_base)
+
+        df_expanded = pd.DataFrame(expanded_data)
+        if df_expanded.empty: return pd.DataFrame(columns=target_cols)
+
+        df_expanded.sort_values(by=['Clean_ID', '_sort_price'], ascending=[True, False], inplace=True)
+        df_unique = df_expanded.drop_duplicates(subset=['Clean_ID'], keep='first').copy()
+
+        result = pd.DataFrame(index=df_unique.index)
+        parts = df_unique['原始_合同订单编号'].str.split(n=1, expand=True)
+        result['合同编号'] = parts[0]
+        result['内贸合同号'] = parts[1] if parts.shape[1] > 1 else ""
+
+        result['合同总额'] = df_unique['合同总额'].apply(self.format_money_str)
+        result['收款情况'] = df_unique['收款状态'].fillna('').astype(str).str.strip()
+        is_received = result['收款情况'].str.contains('已收')
+        result['已收款'] = ""
+        result['未收款'] = ""
+        result.loc[is_received, '已收款'] = result.loc[is_received, '合同总额']
+        result.loc[is_received, '未收款'] = "0.00"
+
+        result['签署公司'] = df_unique['收款账户']
+        result['签订日期'] = df_unique['签约日期'].apply(self.format_date_str)
+        result['销售员'] = df_unique['负责人']
+        result['最终用户单位'] = df_unique['客户名称']
+
+        if '最终用户信息\n联系人、电话、邮箱' in target_cols:
+            result['最终用户信息\n联系人、电话、邮箱'] = df_unique['联系人姓名']
+
+        buyer_infos = df_unique['合同买方_Raw'].astype(str).apply(self.parse_buyer_info)
+        result['买方单位'] = [x['name'] for x in buyer_infos]
+        if '买方信息\n联系人、电话、邮箱' in target_cols:
+            result['买方信息\n联系人、电话、邮箱'] = [x['contact_full'] for x in buyer_infos]
+
+        result['收款日期'] = df_unique['最新收款日期'].apply(self.format_date_str)
+        result['合同标的'] = df_unique['_item_name']
+        result['_sort_price'] = df_unique['_sort_price']
+
+        return result[target_cols + ['_sort_price']]
+
+    def merge_datasets(self, old_dfs, csv_df, is_asd):
+        col_gen = '厂家'
+        col_det = '厂家.1' if '厂家.1' in csv_df.columns else '厂家'
+
+        if is_asd:
+            df_subset = csv_df[csv_df[col_gen].str.contains('ASD', case=False, na=False)]
+        else:
+            df_subset = csv_df[~csv_df[col_gen].str.contains('ASD', case=False, na=False)]
+
+        csv_foreign = df_subset[df_subset['合同类型'] == '外贸'].copy()
+        csv_domestic = df_subset[df_subset['合同类型'] == '内贸'].copy()
+        csv_om = df_subset[~df_subset['合同类型'].isin(['外贸', '内贸'])].copy()
+
+        result_dfs = {}
+
+        def vectorized_merge(old_df, new_df, unique_col, target_columns):
+            if new_df.empty:
+                if old_df is None or old_df.empty: return pd.DataFrame(columns=target_columns + ['_status'])
+                old_df['_status'] = ''
+                return old_df
+
+            for col in target_columns:
+                if col not in new_df.columns: new_df[col] = ""
+            if '_sort_price' not in new_df.columns: new_df['_sort_price'] = 0.0
+            if '_is_outsourced' not in new_df.columns: new_df['_is_outsourced'] = 0
+
+            if old_df is None or old_df.empty:
+                combined = new_df.copy()
+                combined['_status'] = 'new'
+                return combined
+
+            old_df = old_df.copy()
+            if unique_col not in old_df.columns: old_df[unique_col] = ""
+            if '_status' not in old_df.columns: old_df['_status'] = ''
+            if '_is_outsourced' not in old_df.columns: old_df['_is_outsourced'] = 0
+
+            for col in target_columns:
+                if col in old_df.columns:
+                    old_df[col] = old_df[col].astype(object)
+
+            is_unique_index = (old_df[unique_col].duplicated().sum() == 0) and (
+                        new_df[unique_col].duplicated().sum() == 0)
+
+            if is_unique_index:
+                old_df.set_index(unique_col, inplace=True, drop=False)
+                new_df.set_index(unique_col, inplace=True, drop=False)
+
+                new_ids = new_df.index.difference(old_df.index)
+                rows_new = new_df.loc[new_ids].copy()
+                rows_new['_status'] = 'new'
+
+                common_ids = new_df.index.intersection(old_df.index)
+                if not common_ids.empty:
+                    for col in target_columns:
+                        new_vals = new_df.loc[common_ids, col].astype(str).str.strip()
+                        old_vals = old_df.loc[common_ids, col].fillna("").astype(str).str.strip()
+
+                        valid_new_mask = (new_vals != "") & (new_vals != "nan") & (new_vals != "None")
+                        diff_mask = valid_new_mask & (new_vals != old_vals)
+
+                        ids_to_update = diff_mask[diff_mask].index
+
+                        if not ids_to_update.empty:
+                            old_df.loc[ids_to_update, col] = new_df.loc[ids_to_update, col]
+                            old_df.loc[ids_to_update, '_status'] = 'modified'
+
+                    old_df.loc[common_ids, '_sort_price'] = new_df.loc[common_ids, '_sort_price']
+                    old_df.loc[common_ids, '_is_outsourced'] = new_df.loc[common_ids, '_is_outsourced']
+
+                old_df.reset_index(drop=True, inplace=True)
+                rows_new.reset_index(drop=True, inplace=True)
+                final_df = pd.concat([old_df, rows_new], ignore_index=True)
+                return final_df
+
+            else:
+                new_ids = set(new_df[unique_col]) - set(old_df[unique_col])
+                rows_to_add = new_df[new_df[unique_col].isin(new_ids)].copy()
+                rows_to_add['_status'] = 'new'
+                final_df = pd.concat([old_df, rows_to_add], ignore_index=True)
+                return final_df
+
+        target_cols_foreign = self.cols_asd_foreign_general if is_asd else self.cols_nonasd_foreign_general
+        old_gen = old_dfs.get('外贸', pd.DataFrame(columns=target_cols_foreign))
+        new_gen_df = self.prepare_new_data_general(csv_foreign, '外贸', target_cols_foreign, col_gen)
+        result_dfs['外贸'] = vectorized_merge(old_gen, new_gen_df, '合同编号', target_cols_foreign)
+
+        old_det = old_dfs.get('外贸明细', pd.DataFrame(columns=self.cols_foreign_detail))
+        new_det_df = self.prepare_new_data_detail(csv_foreign, '外贸', self.cols_foreign_detail, col_det)
+        result_dfs['外贸明细'] = vectorized_merge(old_det, new_det_df, '合同编号', self.cols_foreign_detail)
+
+        old_dom_gen = old_dfs.get('内贸', pd.DataFrame(columns=self.cols_domestic_general))
+        new_dom_df = self.prepare_new_data_general(csv_domestic, '内贸', self.cols_domestic_general, col_gen)
+        result_dfs['内贸'] = vectorized_merge(old_dom_gen, new_dom_df, '合同编号', self.cols_domestic_general)
+
+        old_dom_det = old_dfs.get('内贸明细', pd.DataFrame(columns=self.cols_domestic_detail))
+        new_dom_det_df = self.prepare_new_data_detail(csv_domestic, '内贸', self.cols_domestic_detail, col_det)
+        result_dfs['内贸明细'] = vectorized_merge(old_dom_det, new_dom_det_df, '合同编号', self.cols_domestic_detail)
+
+        old_om = old_dfs.get('OM合同', pd.DataFrame(columns=self.cols_om))
+        new_om_df = self.prepare_new_data_om(csv_om, self.cols_om)
+        result_dfs['OM合同'] = vectorized_merge(old_om, new_om_df, '合同编号', self.cols_om)
+
+        return result_dfs
+
+    def apply_formatting_to_all(self, data_dict):
+        for sheet_name, df in data_dict.items():
+            if df.empty: continue
+
+            # [关键] 强制跳过汇率列的格式化
+            skip_cols = ["计算汇率", "合同币种/美元", "汇率"]
+
+            for col in df.columns:
+                if col in skip_cols: continue
+
+                if col in self.money_cols:
+                    df[col] = df[col].apply(self.format_money_str)
+                elif col in self.percent_cols:
+                    df[col] = df[col].apply(self.format_percent_str)
+                elif col in self.date_cols:
+                    df[col] = df[col].apply(self.format_date_str)
+        return data_dict
+
+
+# ==========================================
+# 第二部分：GUI 界面
+# ==========================================
+
+class ContractApp:
+    def __init__(self, root):
+        self.root = root
+        self.root.title("合同数据处理系统 V4.9 (排序优化&汇率原样版)")
+        self.root.geometry("1300x850")
+        self.root.minsize(1000, 700)
+
+        self.colors = {
+            'bg': '#F0F2F5',
+            'panel': '#FFFFFF',
+            'primary': '#1890FF',
+            'primary_hover': '#40A9FF',
+            'success': '#52C41A',
+            'success_hover': '#73D13D',
+            'text_main': '#262626',
+            'text_sub': '#8C8C8C',
+            'border': '#D9D9D9',
+            'tag_new': '#FFFBE6',
+            'tag_mod': '#E6F7FF'
+        }
+
+        self.setup_styles()
+        self.processor = DataProcessor()
+
+        self.csv_paths = tk.StringVar()
+        self.asd_path = tk.StringVar()
+        self.non_asd_path = tk.StringVar()
+        self.status_var = tk.StringVar(value="准备就绪")
+
+        self.final_data = {}
+        self.create_widgets()
+
+    def setup_styles(self):
+        self.style = ttk.Style()
+        self.style.theme_use('clam')
+
+        self.style.configure("TFrame", background=self.colors['bg'])
+        self.style.configure("Panel.TFrame", background=self.colors['panel'], relief="flat")
+        self.style.configure("TLabel", background=self.colors['panel'], foreground=self.colors['text_main'],
+                             font=("Microsoft YaHei UI", 10))
+        self.style.configure("Header.TLabel", font=("Microsoft YaHei UI", 18, "bold"), background=self.colors['bg'],
+                             foreground=self.colors['text_main'])
+        self.style.configure("SubHeader.TLabel", font=("Microsoft YaHei UI", 12, "bold"),
+                             background=self.colors['panel'], foreground=self.colors['text_main'])
+        self.style.configure("Status.TLabel", background=self.colors['bg'], foreground=self.colors['text_sub'],
+                             font=("Microsoft YaHei UI", 9))
+        self.style.configure("TButton", font=("Microsoft YaHei UI", 10), borderwidth=0, padding=8)
+        self.style.map("TButton", background=[('active', '#E0E0E0')], relief=[('pressed', 'sunken')])
+        self.style.configure("Primary.TButton", background=self.colors['primary'], foreground='white')
+        self.style.map("Primary.TButton", background=[('active', self.colors['primary_hover'])])
+        self.style.configure("Success.TButton", background=self.colors['success'], foreground='white')
+        self.style.map("Success.TButton", background=[('active', self.colors['success_hover'])])
+        self.style.configure("TEntry", fieldbackground="white", padding=5)
+        self.style.configure("Treeview", background="white", foreground=self.colors['text_main'], rowheight=30,
+                             font=("Microsoft YaHei UI", 9), fieldbackground="white")
+        self.style.configure("Treeview.Heading", font=("Microsoft YaHei UI", 10, "bold"), background="#FAFAFA",
+                             foreground=self.colors['text_main'], relief="flat")
+        self.style.map("Treeview", background=[('selected', self.colors['primary_hover'])],
+                       foreground=[('selected', 'white')])
+
+    def create_widgets(self):
+        header_frame = ttk.Frame(self.root)
+        header_frame.pack(side="top", fill="x", padx=25, pady=(25, 10))
+        ttk.Label(header_frame, text="🚀 合同数据智能处理系统", style="Header.TLabel").pack(side="left")
+
+        bottom_bar = ttk.Frame(self.root, style="Panel.TFrame", padding=15)
+        bottom_bar.pack(side="bottom", fill="x", padx=25, pady=(0, 25))
+
+        legend_frame = ttk.Frame(bottom_bar, style="Panel.TFrame")
+        legend_frame.pack(side="left")
+        self.create_legend(legend_frame, "● 新增数据", self.colors['tag_new'], "#D48806")
+        self.create_legend(legend_frame, "● 发生变更", self.colors['tag_mod'], self.colors['primary'])
+
+        ttk.Label(bottom_bar, textvariable=self.status_var, style="Status.TLabel").pack(side="left", padx=20)
+        ttk.Button(bottom_bar, text="💾  确认无误，保存写入", style="Success.TButton", command=self.save_files).pack(
+            side="right")
+
+        input_panel = ttk.Frame(self.root, style="Panel.TFrame", padding=25)
+        input_panel.pack(side="top", fill="x", padx=25, pady=5)
+
+        ttk.Label(input_panel, text="文件配置与导入", style="SubHeader.TLabel").grid(row=0, column=0, columnspan=3,
+                                                                                     sticky="w", pady=(0, 20))
+
+        self.create_file_row(input_panel, "📂  导入 CSV 源文件 (支持多选):", self.csv_paths, 1, is_multiple=True)
+        self.create_file_row(input_panel, "📘  旧 ASD Excel 文件:", self.asd_path, 2)
+        self.create_file_row(input_panel, "📗  旧 非ASD Excel 文件:", self.non_asd_path, 3)
+
+        btn_frame = ttk.Frame(input_panel, style="Panel.TFrame")
+        btn_frame.grid(row=4, column=0, columnspan=3, pady=(20, 0), sticky="e")
+
+        ttk.Button(btn_frame, text="▶  开始极速处理 (仅预览)", style="Primary.TButton",
+                   command=self.process_files).pack(side="right")
+
+        self.notebook = ttk.Notebook(self.root)
+        self.notebook.pack(side="top", fill="both", expand=True, padx=25, pady=15)
+
+    def create_file_row(self, parent, label_text, var, row_idx, is_multiple=False):
+        lbl = ttk.Label(parent, text=label_text, width=28)
+        lbl.grid(row=row_idx, column=0, sticky="w", pady=8)
+
+        entry = ttk.Entry(parent, textvariable=var, font=("Microsoft YaHei UI", 9))
+        entry.grid(row=row_idx, column=1, sticky="ew", padx=10, pady=8)
+
+        btn = ttk.Button(parent, text="浏览...", width=8, command=lambda: self.browse_file(var, is_multiple))
+        btn.grid(row=row_idx, column=2, padx=5)
+
+        parent.columnconfigure(1, weight=1)
+
+    def create_legend(self, parent, text, bg_color, fg_color):
+        f = tk.Frame(parent, bg=bg_color, padx=10, pady=4)
+        f.pack(side="left", padx=5)
+        tk.Label(f, text=text, bg=bg_color, fg=fg_color, font=("Microsoft YaHei UI", 9, "bold")).pack()
+
+    def browse_file(self, variable, is_multiple=False):
+        if is_multiple:
+            files = filedialog.askopenfilenames(filetypes=[("CSV Files", "*.csv")])
+            if files: variable.set("; ".join(files))
+        else:
+            f = filedialog.askopenfilename(filetypes=[("Excel/CSV Files", "*.csv;*.xlsx")])
+            if f: variable.set(f)
+
+    def load_excel_safe(self, path):
+        if not path or not os.path.exists(path): return {}
+        try:
+            dfs = pd.read_excel(path, sheet_name=None)
+            clean_dfs = {}
+            for k, v in dfs.items():
+                new_columns = []
+                for col in v.columns:
+                    clean_col = self.processor.clean_header_key(str(col))
+                    if clean_col in self.processor.standard_col_map:
+                        new_columns.append(self.processor.standard_col_map[clean_col])
+                    elif col in self.processor.legacy_map:
+                        new_columns.append(self.processor.legacy_map[col])
+                    else:
+                        new_columns.append(col)
+                v.columns = new_columns
+                v = v.loc[:, ~v.columns.duplicated()]
+                if '合同编号' in v.columns: v['合同编号'] = v['合同编号'].astype(str)
+                clean_dfs[k.strip()] = v
+            return clean_dfs
+        except Exception as e:
+            messagebox.showwarning("读取错误", f"读取旧文件失败: {path}\n错误: {str(e)}")
+            return {}
+
+    def process_files(self):
+        csv_paths_str = self.csv_paths.get()
+        if not csv_paths_str:
+            messagebox.showerror("提示", "请先选择 CSV 源文件！")
+            return
+
+        self.status_var.set("⏳ 正在读取多个数据源...")
+        self.root.update()
+
+        try:
+            csv_df, headers_or_msg = self.processor.load_multiple_csvs(csv_paths_str)
+            if csv_df is None:
+                messagebox.showerror("读取错误", headers_or_msg)
+                return
+
+            self.status_var.set("🚀 正在极速合并数据...")
+            self.root.update()
+
+            self.final_data = {}
+
+            path_asd = self.asd_path.get()
+            asd_old = self.load_excel_safe(path_asd)
+            self.final_data['ASD'] = self.processor.merge_datasets(asd_old, csv_df, True)
+
+            path_non = self.non_asd_path.get()
+            non_old = self.load_excel_safe(path_non)
+            self.final_data['NonASD'] = self.processor.merge_datasets(non_old, csv_df, False)
+
+            self.final_data['ASD'] = self.processor.apply_formatting_to_all(self.final_data['ASD'])
+            self.final_data['NonASD'] = self.processor.apply_formatting_to_all(self.final_data['NonASD'])
+
+            self.refresh_preview()
+            self.status_var.set("✅ 预览已生成。确认无误后请点击右下角保存！")
+            messagebox.showinfo("完成",
+                                "数据预览已生成！\n\n注意：此时尚未写入文件。\n请在下方检查数据，确认无误后点击 [保存] 按钮。")
+
+        except Exception as e:
+            self.status_var.set("❌ 发生错误")
+            traceback.print_exc()
+            messagebox.showerror("运行错误", str(e))
+
+    def refresh_preview(self):
+        for tab in self.notebook.tabs():
+            self.notebook.forget(tab)
+
+        for file_type in ['ASD', 'NonASD']:
+            if file_type not in self.final_data: continue
+
+            data_dict = self.final_data[file_type]
+            main_frame = ttk.Frame(self.notebook, style="Panel.TFrame", padding=10)
+            self.notebook.add(main_frame, text=f"  {file_type} 预览  ")
+
+            inner_notebook = ttk.Notebook(main_frame)
+            inner_notebook.pack(fill="both", expand=True)
+
+            sheet_order = ['外贸', '外贸明细', '内贸', '内贸明细', 'OM合同']
+            for sheet_name in sheet_order:
+                if sheet_name in data_dict:
+                    df = data_dict[sheet_name]
+                    if not df.empty:
+                        # [预览阶段也应用排序逻辑]
+                        if '合同编号' in df.columns:
+                            df['合同编号'] = df['合同编号'].astype(str)
+
+                            sort_cols = ['合同编号']
+                            asc_order = [True]
+
+                            if '_is_outsourced' in df.columns:
+                                sort_cols.append('_is_outsourced')
+                                asc_order.append(True)
+
+                            if '_sort_price' in df.columns:
+                                sort_cols.append('_sort_price')
+                                asc_order.append(False)
+
+                            df = df.sort_values(by=sort_cols, ascending=asc_order)
+
+                            if '明细' in sheet_name:
+                                mask = df.duplicated(subset=['合同编号'], keep='first')
+                                df.loc[mask, '合同标的'] = ""
+
+                        standard_cols = []
+                        is_asd = (file_type == 'ASD')
+                        if sheet_name == '外贸':
+                            standard_cols = self.processor.cols_asd_foreign_general if is_asd else self.processor.cols_nonasd_foreign_general
+                        elif sheet_name == '内贸':
+                            standard_cols = self.processor.cols_domestic_general
+                        elif sheet_name == 'OM合同':
+                            standard_cols = self.processor.cols_om
+                        elif sheet_name == '外贸明细':
+                            standard_cols = self.processor.cols_foreign_detail
+                        elif sheet_name == '内贸明细':
+                            standard_cols = self.processor.cols_domestic_detail
+
+                        self.create_treeview(inner_notebook, df, sheet_name, standard_cols)
+
+    def create_treeview(self, parent, df, title, target_cols):
+        frame = ttk.Frame(parent)
+        parent.add(frame, text=f" {title} ")
+
+        scroll_y = ttk.Scrollbar(frame, orient="vertical")
+        scroll_x = ttk.Scrollbar(frame, orient="horizontal")
+
+        tree = ttk.Treeview(frame, columns=target_cols, show='headings',
+                            yscrollcommand=scroll_y.set, xscrollcommand=scroll_x.set)
+
+        scroll_y.config(command=tree.yview)
+        scroll_x.config(command=tree.xview)
+
+        scroll_y.pack(side="right", fill="y")
+        scroll_x.pack(side="bottom", fill="x")
+        tree.pack(fill="both", expand=True)
+
+        for col in target_cols:
+            clean_header = col.replace('\n', ' ')
+            tree.heading(col, text=clean_header)
+            tree.column(col, width=130, anchor="center")
+
+        tree.tag_configure('new', background=self.colors['tag_new'])
+        tree.tag_configure('modified', background=self.colors['tag_mod'], foreground=self.colors['primary'])
+        tree.tag_configure('odd', background='white')
+        tree.tag_configure('even', background='#FAFAFA')
+
+        if not df.empty:
+            df_display = df.fillna("")
+            last_contract_id = None
+
+            count = 0
+            for _, row in df_display.iterrows():
+                values = []
+                for col in target_cols:
+                    val = row.get(col, "")
+                    if '明细' in title and col == '合同标的':
+                        current_id = row.get('合同编号', '')
+                        if current_id == last_contract_id: val = ""
+                    values.append(val)
+
+                if '明细' in title: last_contract_id = row.get('合同编号', '')
+
+                status = row.get('_status', '')
+                tags = [status] if status else [('even' if count % 2 == 0 else 'odd')]
+                tree.insert("", "end", values=values, tags=tags)
+                count += 1
+
+        tree.bind("<Double-1>", lambda event: self.on_double_click(event, tree))
+
+    def on_double_click(self, event, tree):
+        region = tree.identify("region", event.x, event.y)
+        if region != "cell": return
+        column = tree.identify_column(event.x)
+        row_id = tree.identify_row(event.y)
+        col_idx = int(column.replace('#', '')) - 1
+        col_name = tree['columns'][col_idx]
+        current_val = tree.item(row_id, "values")[col_idx]
+
+        new_val = simpledialog.askstring("快速编辑", f"修改 [{col_name}]:", initialvalue=current_val, parent=self.root)
+        if new_val is not None:
+            current_values = list(tree.item(row_id, "values"))
+            current_values[col_idx] = new_val
+            tree.item(row_id, values=current_values)
+
+    def apply_excel_formulas(self, df, sheet_name):
+        if df.empty: return df
+        col_map = {col: get_column_letter(i + 1) for i, col in enumerate(df.columns)}
+
+        if '明细' in sheet_name:
+            rate_col = col_map.get('计算汇率')
+            if rate_col:
+                if '外购' in col_map and '外购转美元' in col_map:
+                    col_target = '外购转美元'
+                    col_src = col_map['外购']
+                    for idx in df.index:
+                        excel_row = idx + 2
+                        formula = f'=IF(AND(ISNUMBER({col_src}{excel_row}), ISNUMBER({rate_col}{excel_row}), {rate_col}{excel_row}<>0), {col_src}{excel_row}/{rate_col}{excel_row}, "")'
+                        df.at[idx, col_target] = formula
+
+                src_quote = '报价总价' if '报价总价' in col_map else '报价RMB总价'
+                if src_quote in col_map and '报价总价美元' in col_map:
+                    col_target = '报价总价美元'
+                    col_src = col_map[src_quote]
+                    for idx in df.index:
+                        excel_row = idx + 2
+                        formula = f'=IF(AND(ISNUMBER({col_src}{excel_row}), ISNUMBER({rate_col}{excel_row}), {rate_col}{excel_row}<>0), {col_src}{excel_row}/{rate_col}{excel_row}, "")'
+                        df.at[idx, col_target] = formula
+
+                src_sales = '销售总价' if '销售总价' in col_map else '售价RMB总价'
+                src_out = '外购'
+                if src_sales in col_map and src_out in col_map and '净合同额美元' in col_map:
+                    col_target = '净合同额美元'
+                    col_sales = col_map[src_sales]
+                    col_out = col_map[src_out]
+                    for idx in df.index:
+                        excel_row = idx + 2
+                        formula = f'=IF(AND(ISNUMBER({col_sales}{excel_row}), ISNUMBER({rate_col}{excel_row}), {rate_col}{excel_row}<>0), ({col_sales}{excel_row}-{col_out}{excel_row})/{rate_col}{excel_row}, "")'
+                        df.at[idx, col_target] = formula
+
+        elif '外贸' in sheet_name or '内贸' in sheet_name:
+            rate_col = col_map.get('合同币种/美元')
+            if rate_col:
+                total_src = '总合同额' if '总合同额' in col_map else '合同总额'
+                if total_src in col_map and '转为美元总合同额' in col_map:
+                    col_target = '转为美元总合同额'
+                    col_src = col_map[total_src]
+                    for idx in df.index:
+                        excel_row = idx + 2
+                        formula = f'=IF(AND(ISNUMBER({col_src}{excel_row}), ISNUMBER({rate_col}{excel_row}), {rate_col}{excel_row}<>0), {col_src}{excel_row}/{rate_col}{excel_row}, "")'
+                        df.at[idx, col_target] = formula
+
+                if total_src in col_map and '外购' in col_map and '转为美元净合同额' in col_map:
+                    col_target = '转为美元净合同额'
+                    col_total = col_map[total_src]
+                    col_out = col_map['外购']
+                    for idx in df.index:
+                        excel_row = idx + 2
+                        formula = f'=IF(AND(ISNUMBER({col_total}{excel_row}), ISNUMBER({rate_col}{excel_row}), {rate_col}{excel_row}<>0), ({col_total}{excel_row}-{col_out}{excel_row})/{rate_col}{excel_row}, "")'
+                        df.at[idx, col_target] = formula
+
+        return df
+
+    def save_files(self):
+        if not self.final_data:
+            messagebox.showwarning("提示", "没有可保存的数据，请先处理文件！")
+            return
+
+        csv_path_str = self.csv_paths.get()
+        first_path = csv_path_str.split(';')[0].strip() if csv_path_str else ""
+        base_dir = os.path.dirname(first_path) if first_path else ""
+
+        confirm = messagebox.askyesno("确认写入",
+                                      "您确定要将预览的数据写入到 Excel 文件吗？\n\n注意：程序将自动注入Excel公式以支持动态汇率计算。")
+        if not confirm:
+            self.status_var.set("已取消写入")
+            return
+
+        self.status_var.set("💾 正在写入文件...")
+        self.root.update()
+
+        try:
+            for file_type, sheets in self.final_data.items():
+                target_path = ""
+                if file_type == 'ASD':
+                    target_path = self.asd_path.get()
+                    if not target_path: target_path = os.path.join(base_dir, "ASD_Combined.xlsx")
+                elif file_type == 'NonASD':
+                    target_path = self.non_asd_path.get()
+                    if not target_path: target_path = os.path.join(base_dir, "NonASD_Combined.xlsx")
+
+                with pd.ExcelWriter(target_path, engine='openpyxl') as writer:
+                    valid_sheets = ['外贸', '外贸明细', '内贸', '内贸明细', 'OM合同']
+                    for sheet_name in valid_sheets:
+                        if sheet_name in sheets:
+                            df = sheets[sheet_name].copy()
+                            df.reset_index(drop=True, inplace=True)
+                            df = self.apply_excel_formulas(df, sheet_name)
+
+                            # [核心修改] 写入时应用严格排序逻辑
+                            if '合同编号' in df.columns:
+                                sort_cols = ['合同编号']
+                                asc_order = [True]
+
+                                # 优先级1: 外购 (0在前, 1在后)
+                                if '_is_outsourced' in df.columns:
+                                    sort_cols.append('_is_outsourced')
+                                    asc_order.append(True)
+
+                                # 优先级2: 价格降序
+                                if '_sort_price' in df.columns:
+                                    sort_cols.append('_sort_price')
+                                    asc_order.append(False)
+
+                                df = df.sort_values(by=sort_cols, ascending=asc_order)
+
+                            save_df = df.drop(columns=['_status', '_sort_price', '_is_outsourced'], errors='ignore')
+
+                            if not save_df.empty:
+                                if '明细' in sheet_name:
+                                    mask = save_df.duplicated(subset=['合同编号'], keep='first')
+                                    save_df.loc[mask, '合同标的'] = ""
+                                save_df.to_excel(writer, sheet_name=sheet_name, index=False)
+
+            self.status_var.set("✅ 写入成功！")
+            messagebox.showinfo("成功", f"文件已成功写入！\n位置: {base_dir or '当前目录'}")
+        except PermissionError:
+            messagebox.showerror("保存失败", "文件被占用！\n请先关闭 Excel 文件后再点击保存。")
+        except Exception as e:
+            messagebox.showerror("保存失败", str(e))
+        finally:
+            if self.status_var.get() != "已取消写入":
+                self.status_var.set("准备就绪")
+
+
+if __name__ == "__main__":
+    root = tk.Tk()
+    try:
+        from ctypes import windll
+
+        windll.shcore.SetProcessDpiAwareness(1)
+    except:
+        pass
+    app = ContractApp(root)
+    root.mainloop()
\ No newline at end of file