first commit

.gitignore

@@ -0,0 +1,3 @@
+build/*
+dist/*
+outputs/*

README.md

@@ -0,0 +1,29 @@
+# cwExcel - Python Desktop Application
+
+一个基于 Python Tkinter 的桌面应用程序。
+
+## 环境要求
+
+- Python 3.8+
+- Tkinter（Python 内置）
+
+## 安装依赖
+
+```bash
+pip install -r requirements.txt
+```
+
+## 运行
+
+```bash
+python main.py
+```
+
+## 打包
+
+```bash
+pip install pyinstaller
+pyinstaller --noconfirm cwExcel.spec
+```
+
+输出在 `dist/cwExcel/cwExcel.exe`。

app/__init__.py

@@ -0,0 +1,6 @@
+"""
+app 包 - 应用程序核心模块
+"""
+from app.application import MainApplication
+
+__all__ = ["MainApplication"]

app/application.py

@@ -0,0 +1,22 @@
+"""
+主应用程序类
+"""
+import tkinter as tk
+from app.ui import MainWindow
+
+
+class MainApplication:
+    """应用程序主控制器"""
+
+    def __init__(self, root: tk.Tk):
+        self.root = root
+        self.root.title("cwExcel")
+        self.root.geometry("1024x680")
+        self.root.minsize(800, 500)
+
+        # 构建主窗口
+        self.main_window = MainWindow(root)
+
+    def run(self):
+        """启动应用程序主循环"""
+        self.root.mainloop()

app/finance_parse.py

@@ -0,0 +1,633 @@
+# spacy_training/scripts/predict.py
+import re
+import sys
+import cn2an
+import os
+import os.path
+from tqdm import tqdm
+from pathlib import Path
+
+# PyInstaller --noconsole 兼容：无控制台时重定向输出到空设备
+if sys.stdout is None:
+    sys.stdout = open(os.devnull, 'w')
+if sys.stderr is None:
+    sys.stderr = open(os.devnull, 'w')
+
+
+# 单位与连接符定义
+MAX_ITEM_OFFSET = 10
+UNIT_PATTERN = "个|张|片|块|条|幅|根"
+SIZE_UNITS_PATTERN = r'mm|cm|dm|m'
+SIZE_JOIN_PATTERN = "xX×"
+NUMS_PATTERN = r'\d+(?:\.\d+)'  # 修复：原缺少 ? 导致整数不匹配
+CNT_PATTERN = rf'{NUMS_PATTERN}?[^克+\s\d-]*'
+EXCEPT_TEXT = "解析失败"
+INDEX_COL = '序号'
+
+base_colume = ['解析备注', '尺寸备注', '描述', '异常信息']
+mapping_size = {
+        "width_mm": "长",
+        "height_mm": "宽",
+        # "original_size": "原始尺寸",
+        # "total_quantity": "总数量",
+        "style_count": "款数",
+        "quantity_per_style": "数量",
+        "unit": "单位",
+        "描述": "描述",
+        "解析备注": "解析备注",
+        "size_unit": "尺寸单位",
+        "exception_msg": '异常信息'
+    }
+
+import spacy
+import pandas as pd
+
+# 将 doc 转换为字典
+def doc_to_dict(doc):
+    return {
+        "text": doc.text,
+        "entities": [
+            {
+                "text": ent.text,
+                "label": ent.label_,
+                "start": ent.start_char,
+                "end": ent.end_char
+            }
+            for ent in doc.ents
+        ],
+    }
+
+def predict_file(df, model_path="./spacy_training/model", key='备注'):
+    print(model_path)
+    nlp = spacy.load(model_path)
+    if "sentencizer" not in nlp.pipe_names:
+        nlp.add_pipe("sentencizer")
+
+    # 统一表头空格
+    df.columns = df.columns.str.strip()
+    key = key.strip()
+
+    # 流式批处理：用 as_tuples 边读边处理，batch_size=32 兼顾速度与内存
+    stream = ((str(row[key]).strip(), {k: v for k, v in row.items()}) for _, row in df.iterrows())
+    for doc, base_row in tqdm(nlp.pipe(stream, as_tuples=True, batch_size=32), total=len(df)):
+        yield doc.text, base_row, doc_to_dict(doc)
+
+from dataclasses import dataclass
+
+@dataclass
+class StructuredOrder:
+    key: str
+    width_mm: float
+    height_mm: float
+    original_size: str
+    style_count: int
+    quantity_per_style: float
+    unit: str
+    total_quantity: float
+    original_qty: str
+    size_position: list
+    qty_position: list
+    exception_msg: str
+    calc_type: str
+    size_unit: str
+    meta: dict
+
+def expand_to_structured(raw_pairs, size_unit=''):
+    unit_map = {
+        '百': 100, '千': 1000, '万': 10000,
+        '百万': 1000000, '千万': 10000000, '亿': 100000000,
+    }
+
+    UNIT_REGEX = re.compile(rf'({UNIT_PATTERN})$')
+    exception_msg = []
+
+    def extract_multiplier(text):
+        """
+        从字符串中提取数值乘数，支持：
+        - 阿拉伯数字 + 中文单位：3.5万 → 35000
+        - 纯中文数字 + 单位：三万五千 → 35000
+        - 纯中文：两千万 → 20000000
+        """
+        # 去除末尾单位（保留前面的部分）
+        clean = UNIT_REGEX.sub('', text.strip())
+
+        # 如果去单位后为空，原字符串可能是纯单位（如“万”），默认系数为1
+        if not clean:
+            matched_unit = UNIT_REGEX.search(text)
+            if matched_unit:
+                unit = matched_unit.group(1)
+                return unit_map.get(unit, 1.0)
+            return 1.0
+
+        # 尝试直接转阿拉伯数字
+        try:
+            return float(clean)
+        except ValueError:
+            pass
+
+        # 尝试转中文数字（如“三万五千”虽然不合理，但“三万”可以）
+        try:
+            # 注意：cn2an 可以直接处理“三万五千”这种
+            num = cn2an.cn2an(clean, "smart")  # smart 模式支持混合写法
+            if isinstance(num, (int, float)):
+                # 检查原字符串是否有单位后缀（比如“三万”中的“万”已被去除，需补回）
+                matched_unit = UNIT_REGEX.search(text)
+                if matched_unit:
+                    unit = matched_unit.group(1)
+                    num *= unit_map[unit]
+                return float(num)
+        except Exception:
+            exception_msg.append(f"尺寸解析异常：{text}")
+            pass
+
+        return 1.0
+
+    def parse_size(size_str):
+        join_str = ''.join(re.findall(f"[{SIZE_JOIN_PATTERN}]+", size_str))
+        if len(join_str) == 0 or re.search('比例|等高|等比|等宽', size_str):
+            result = f'异常值：{size_str}'
+            return result, result, size_unit
+        size_str = size_str.replace(join_str, "x")
+        match = re.match(rf'({NUMS_PATTERN}?)([a-z]+)?[{SIZE_JOIN_PATTERN}]({NUMS_PATTERN}?)([a-z]+)?', size_str, re.IGNORECASE)
+        if not match:
+            return None, None, size_unit
+        w_val, w_unit, h_val, h_unit = match.groups()
+        unit = (w_unit or h_unit or size_unit).lower()
+        conv = {'m': 1000, 'dm': 100, 'cm': 10, 'mm': 1}
+        if unit not in conv:
+            print(f'未知单位: {unit} size_str:{size_str} size_unit:{size_unit}')
+            exception_msg.append(f"尺寸解析异常：{size_str}, 未知单位: {unit}")
+            return None, None, unit
+        w_mm = conv[unit] * float(w_val)
+        h_mm = conv[unit] * float(h_val)
+        return round(w_mm, 3), round(h_mm, 3), unit
+
+    def parse_quantity(qty_str):
+        qty_str = "".join(qty_str.split())
+        multi_match = re.match((
+            rf'[共计]?(.+?)款\w*?([共计]?[\d.一二两三四五六七八九十百千万亿]+)({UNIT_PATTERN})'
+        ), qty_str, re.IGNORECASE)
+        if multi_match:
+            style_part, qty_part = multi_match.group(1), multi_match.group(2)
+            if qty_part[0] in '共计':
+                qty_part = qty_part[1:]
+                calc_type = 'total'
+            else:
+                calc_type = 'single'
+            style_count = int(extract_multiplier(style_part))
+            quantity_per_style = int(extract_multiplier(qty_part))
+            unit = multi_match.group(3)
+
+            return {"style_count": style_count, "quantity_per_style": quantity_per_style, "unit": unit, "calc_type": calc_type}
+        
+        single_match = re.match(rf'(.+?)({UNIT_PATTERN})', qty_str, re.IGNORECASE)
+        if single_match:
+            qty_part = single_match.group(1)
+            if qty_part[0] in '共计':
+                qty_part = qty_part[1:]
+                calc_type = 'total'
+            else:
+                calc_type = 'single'
+            quantity_per_style = int(extract_multiplier(qty_part))
+            unit = single_match.group(2)
+            return {"style_count": 1, "quantity_per_style": quantity_per_style, "unit": unit, "calc_type": calc_type}
+
+        return {"style_count": 0, "quantity_per_style": 0, "unit": "个"}
+
+    structured = []
+
+    for size_info, qty_info in raw_pairs:
+        s_start, s_end, s_text, *_ = size_info
+        q_start, q_end, q_text, *_ = qty_info
+
+        if s_text == EXCEPT_TEXT:
+            width_mm, height_mm = EXCEPT_TEXT, EXCEPT_TEXT
+        else:
+            size_parsed = parse_size(s_text)
+            if size_parsed:
+                width_mm, height_mm, size_unit = size_parsed
+            else:
+                width_mm, height_mm, size_unit = None, None, None
+        
+        if q_text == EXCEPT_TEXT:
+            style_count, quantity_per_style, unit = 0, EXCEPT_TEXT, EXCEPT_TEXT
+        else:
+            qty_parsed = parse_quantity(q_text)
+            style_count, quantity_per_style, unit = qty_parsed["style_count"], qty_parsed["quantity_per_style"], qty_parsed["unit"]
+
+        structured.append(StructuredOrder(
+            key=f"{s_start}{s_end}{s_text}",
+            width_mm=width_mm,
+            height_mm=height_mm,
+            original_size=s_text,
+            style_count=style_count,
+            quantity_per_style=quantity_per_style,
+            unit=unit,
+            total_quantity=style_count * quantity_per_style if style_count and style_count else None,
+            original_qty=q_text,
+            size_position=[s_start, s_end],
+            qty_position=[q_start, q_end],
+            exception_msg="; ".join(exception_msg),
+            calc_type=qty_parsed.get('calc_type', None),
+            size_unit=size_unit,
+            meta={
+                "size_text_except": s_text if s_text == EXCEPT_TEXT or (not width_mm and not height_mm) else None,
+                "qty_text_except": q_text if q_text == EXCEPT_TEXT else None,
+            }
+        ))
+
+    return structured
+
+
+def decode(doc_dict_list, is_horizontal=False):
+    ent_group = {
+        # "单号": ["单号"],
+        "刮刮膜尺寸": ["刮刮膜尺寸"],
+        # "产品": ["产品"],
+        # "材质": ["材质"],
+        # "工艺": ["工艺"],
+        # "用户": ["用户信息"],
+        # "加急": ["加急"],
+        "数量": ["尺寸", "数量"],
+    }
+    columns = list(ent_group.keys())
+    columns.remove('数量')
+    columns.extend(base_colume)
+    group_reflect = {v: k for k, vs in ent_group.items() for v in vs}
+    result_items = []
+    exception_list = []
+    
+    for doc_dict in doc_dict_list:
+        sentence = doc_dict["text"]
+        entities = doc_dict["entities"]
+
+        is_except = False
+
+        # 强关联实体需要进行组合处理
+        # ======================================== 实体分组 ========================================
+        ent_group_dict = {k: [] for k in ent_group.keys()}
+        base_ent_group_dict = {}
+
+        for ent in entities:
+            start, end, label, text = ent["start"], ent["end"], ent["label"], ent["text"]
+            if label not in group_reflect:
+                continue
+            group_name = group_reflect[label]
+            ent_group_dict[group_name].append((start, end, text, label))
+            base_ent_group_dict[group_name] = text
+    
+        result = []
+
+        # 解析出尺寸和数量的组合
+        # ======================================== 解析尺寸和数量 ========================================
+        group_list = []
+        size_and_qty_list = [[], []]
+
+        # 数量需要和尺寸进行组合处理
+        base_ent_group_dict.pop("数量", None)
+        for cnt_item in ent_group_dict["数量"]:
+            size_info, qty_info = size_and_qty_list
+            label = cnt_item[-1]
+            if label == '数量':
+                qty_info.append(cnt_item)
+            elif label == '尺寸':
+                start, end, text, _label = cnt_item
+                if f'-({text[:2]}' in sentence:
+                    continue
+                if qty_info:
+                    group_list.append(size_and_qty_list)
+                    size_and_qty_list = [[], []]
+                    size_info, qty_info = size_and_qty_list
+                size_info.append(cnt_item)
+        
+        # 存在数量和尺寸
+        if size_and_qty_list[0] or size_and_qty_list[1]:
+            group_list.append(size_and_qty_list)
+        
+        # 如果只有一对尺寸，有可能反过来描述
+        if len(group_list) == 2:
+            _1, qty_info = group_list[0]
+            size_info, _2 = group_list[1]
+
+            if not _1 and not _2:
+                group_list[:] = [[qty_info, size_info]]
+        
+        # print(f'订单中尺寸和数量组合：{group_list}')
+
+        # 解析出尺寸和数量的组合
+        # ======================================== 解析长、宽、款数、数量 ========================================
+        size_unit = ''
+        for size_info, qty_info in group_list:
+            units = set(''.join(re.findall(SIZE_UNITS_PATTERN, ent[2])) for ent in size_info)
+            units = [unit for unit in units if unit]
+            if len(units) == 1:
+                size_unit = units[0]
+
+        size_and_qty_parsed_result = []
+        for size_info, qty_info in group_list:
+            size_len, qty_len = len(size_info), len(qty_info)
+            if size_len == 0:
+                print(f'解析异常，尺寸为空，数量为{qty_info}，原文：{sentence}', entities)
+                is_except = True
+                # 起始位置、结束位置、尺寸文本、尺寸标签
+                qty_info = [(0, 0, EXCEPT_TEXT, '数量')]
+            if qty_len == 0:
+                print(f'解析异常，数量为空，尺寸为{size_info}，原文：{sentence}', entities)
+                is_except = True
+                # 起始位置、结束位置、尺寸文本、尺寸标签
+                size_info = [(0, 0, EXCEPT_TEXT, '尺寸')]
+                # continue
+
+            # print('size_info, qty_info', size_info, qty_info)
+            if size_len == qty_len:
+                decode_desc = "单尺寸、单款式描述的订单" if size_len == 1 else "多尺寸、多款式描述一一匹配的订单"
+                
+                for size, qty in zip(size_info, qty_info):
+                    expand = expand_to_structured([(size, qty)], size_unit) or []
+                    exceptions = [item.meta for item in expand if item.meta.get('size_text_except') or item.meta.get('qty_text_except')]
+                    if exceptions: print("解析异常", exceptions, '原文', sentence, entities)
+                    for item in expand:
+                        parsed_item = item.__dict__.copy()
+                        parsed_item['描述'] = f"{size[2]}|{qty[2]}"
+                        parsed_item['解析备注'] = decode_desc
+                        parsed_item['size_len'] = size_len
+                        parsed_item['qty_len'] = qty_len
+                        size_and_qty_parsed_result.append(parsed_item)
+            elif size_len>1 and qty_len>1:
+                
+                if size_len > qty_len:
+                    total_qty = []
+                    for size, qty in zip(size_info, qty_info):
+                        expand = expand_to_structured([(size, qty)], size_unit) or []
+                        for item in expand:
+                            total_qty.append(item.style_count)
+
+                    if size_len == sum(total_qty):
+                        decode_desc = "多尺寸、多款式描述:多款式描述和尺寸相等的订单"
+                        item = []
+                        for idx, qty in enumerate(total_qty):
+                            item.extend(qty_info[idx] for _ in range(qty))
+                        print('解析：', size_info, qty_info, item)
+                        # breakpoint()
+                        qty_info = item
+                        qty_len = len(qty_info)
+                    else:
+                        # qty_info = []
+                        decode_desc = f"匹配异常：多尺寸、多款式描述的订单: {size_info} {qty_info}"
+                else:
+                    # qty_info = []
+                    decode_desc = f"匹配异常：多尺寸、多款式描述的订单: {size_info} {qty_info}"
+
+                for size, qty in zip(size_info, qty_info):
+                    expand = expand_to_structured([(size, qty)], size_unit) or []
+                    exceptions = [item.meta for item in expand if item.meta.get('size_text_except') or item.meta.get('qty_text_except')]
+                    if exceptions: print("解析异常", exceptions, '原文', sentence, entities)
+                    for item in expand:
+                        parsed_item = item.__dict__.copy()
+                        parsed_item['style_count'] = 1
+                        parsed_item['描述'] = f"{size[2]}|{qty[2]}"
+                        parsed_item['解析备注'] = decode_desc
+                        parsed_item['size_len'] = size_len
+                        parsed_item['qty_len'] = qty_len
+                        size_and_qty_parsed_result.append(parsed_item)
+            else:
+                # 多尺寸-单数量、单尺寸-多数量 告警多尺寸-多数量
+                iter_item = ((size, qty) for size in size_info for qty in qty_info)
+                if size_len > 1 and qty_len == 1:
+                    decode_desc = "多尺寸、单款式描述的订单"
+                elif size_len == 1 and qty_len > 1:
+                    decode_desc = "单尺寸、多款式描述的订单"
+                else:
+                    decode_desc = f"异常：多尺寸、多款式描述的订单: {size_info} {qty_info}"
+
+                for item in iter_item:
+                    size, qty = item
+                    expand = expand_to_structured([(size, qty)], size_unit) or []
+                    exceptions = [item.meta for item in expand if item.meta.get('size_text_except') or item.meta.get('qty_text_except')]
+                    if exceptions: print("解析异常", exceptions, '原文', sentence, entities)
+                    for item in expand:
+                        parsed_item = item.__dict__.copy()
+                        parsed_item['描述'] = f"{size[2]}|{qty[2]}"
+                        parsed_item['解析备注'] = decode_desc
+                        parsed_item['size_len'] = size_len
+                        parsed_item['qty_len'] = qty_len
+                        size_and_qty_parsed_result.append(parsed_item)
+
+        result.extend(size_and_qty_parsed_result)
+    
+        for item in result:
+            # print('item', item)
+            new_item = base_ent_group_dict.copy()
+            for key in mapping_size:
+                col_name = mapping_size[key]
+                new_item[col_name] = item[key]
+                columns.append(col_name)
+            size_len = item.pop('size_len', None)
+            qty_len = item.pop('qty_len', None)
+            calc_type = item.pop('calc_type', None)
+
+            if size_len > 1 and calc_type == 'total':
+                new_item[f'数量'] = f"警告：异常值(总计值:{new_item[f'数量']})"
+
+            if size_len == 1 or size_len == qty_len:
+                pass
+                # new_item[f'解析备注'] = '单尺寸的订单'
+            else:
+                if size_len == new_item['款数']:
+                    new_item[f'款数'] = 1
+                    # new_item[f'解析备注'] = '尺寸数量和款数相同的订单'
+                elif new_item['款数'] == 1:
+                    new_item[f'款数'] = 1
+                    # new_item[f'解析备注'] = '单款的订单'
+                else:
+                    new_item[f'款数'] = "异常值：款数和尺寸数不一致"
+                    new_item[f'数量'] = "异常值：款数和尺寸数不一致"
+                    # new_item[f'解析备注'] = '多尺寸并且和款数不同的订单'
+                    
+            result_items.append(new_item)
+            # print(json.dumps(new_item, indent=4, ensure_ascii=False))
+
+        if not result:
+            if base_ent_group_dict:
+                base_ent_group_dict['是否异常'] = is_except
+                base_ent_group_dict[f'解析备注'] = '未解析到实际尺寸的订单1'
+                result_items.append(base_ent_group_dict.copy())
+            else:
+                result_items.append({
+                    "是否异常": is_except,
+                    "解析备注": "未解析到任何实体"
+                })
+
+    return exception_list, columns, result_items
+
+
+def get_model_path():
+    """获取模型路径，兼容开发环境和PyInstaller打包"""
+    if getattr(sys, '_MEIPASS', None):
+        # PyInstaller 打包后
+        return str(Path(sys._MEIPASS) / 'app' / 'resources' / 'models' / '订单尺寸识别' / 'best_model')
+    # 开发环境
+    return str(Path(__file__).parent / 'resources' / 'models' / '订单尺寸识别' / 'best_model')
+
+
+def parse_finance_data(file_path, target_index, is_horizontal, sheet_name="Sheet1"):
+    
+    df = pd.read_excel(file_path, dtype=str, sheet_name=sheet_name)
+    total_amount = len(df)
+
+    model_path = get_model_path()
+    print(f"使用模型路径: {model_path}")
+    items = predict_file(df, model_path, target_index)
+    print(f"解析完成，共{items}行")
+    # items = predict_file(file_path, './resources/models/lintao/best_model', target_index)
+    exception_list, results, outputs_columns = [], [], []
+    origin_item = None
+
+    parsed_dict = {}
+    mapping_size_values = mapping_size.values()
+
+    for text, origin_item, item in items:
+
+        decode_result = []
+
+        base_info = {}
+        result_index = parsed_dict.get(f"{text}_index")
+        if result_index:
+            exceptions, columns = [], []
+            try:
+                decode_item = parsed_dict[text][result_index]
+                for key in origin_item:
+                    if key in mapping_size_values:
+                        continue
+                    decode_item[key] = origin_item[key]
+                    decode_item['尺寸备注'] = '多组尺寸解析'
+                parsed_dict[f"{text}_index"] += 1
+                continue
+            except IndexError:
+                print(f'解析到的尺寸不足：{text}，{result_index}， {parsed_dict[text]}')
+                base_item = parsed_dict[text][0]
+                for col in base_colume:
+                    base_info[col] = base_item.get(col, '')
+                base_info["长"] = '异常值：没有解析到那么多尺寸'
+                
+                decode_items = []
+                parsed_dict[f"{text}_index"] += 1
+        else:
+            exceptions, columns, decode_items = decode([item], is_horizontal) or []
+        
+        # if '(250731202428388557)' in text:
+        #     print(result_index, text, exceptions, columns, decode_items)
+        #     breakpoint()
+
+        if len(columns) > len(outputs_columns):
+            outputs_columns = columns
+
+        if len(decode_items) > 1:
+            for item in decode_items:
+                item['尺寸备注'] = '多组尺寸解析'
+        
+        for idx, decode_item in enumerate(decode_items):
+            base_row = origin_item.copy()
+            if idx and INDEX_COL in base_row: base_row[INDEX_COL] = ''
+            # base_row = origin_item.copy() if idx == 0 else {}
+            base_row.update(decode_item)
+            is_except = base_row.pop('是否异常', False)
+            if is_except:
+                for col in base_colume:
+                    origin_item[col] = base_row.get(col, '')
+                decode_result.append(origin_item)
+                exception_list.append(base_row)
+            else:
+                decode_result.append(base_row)
+        if not decode_items:
+            if base_info:
+                origin_item.update(base_info)
+
+            decode_result.append(origin_item)
+        else:
+            if is_horizontal:
+                new_decode_result = decode_result[0].copy()
+                for idx, item in enumerate(decode_result[1:], 1):
+                    for key in mapping_size_values:
+                        new_key = f"{key}{idx}"
+                        new_decode_result[new_key] = item[key]
+                        if new_key not in outputs_columns:
+                            outputs_columns.append(new_key)
+                decode_result = [new_decode_result]
+
+            if text not in parsed_dict:
+                parsed_dict[text] = decode_result
+                parsed_dict[f"{text}_index"] = 0
+            parsed_dict[f"{text}_index"] += 1
+        
+        # if decode_result: 
+        #     results.append(decode_result[0])
+        if len(decode_items) > 1:
+            for item in decode_result[1:]:
+                item['尺寸备注'] = '复制新增：多组尺寸解析'
+        results.extend(decode_result)
+
+        # break
+    
+    print('横向解析', outputs_columns)
+    if origin_item:
+        # outputs_columns.
+        _outputs_columns = list(origin_item.keys()) + outputs_columns
+        col_set = set()
+        outputs_columns = []
+
+        for col in _outputs_columns:
+            if col not in col_set:
+                col_set.add(col)
+                outputs_columns.append(col)
+    
+    df = pd.DataFrame(results, columns=outputs_columns)
+
+    file_name = os.path.basename(file_path).replace(".xlsx", '')
+    filename = file_name + '_解析' + ('_横向排列' if is_horizontal else '_纵向排列')
+
+    upload_dir = Path(file_path).parent
+    upload_dir.mkdir(parents=True, exist_ok=True)
+    save_file_path = upload_dir / f'{filename}_1.xlsx'
+
+    # 使用 ExcelWriter 同时写入多个 sheet
+    with pd.ExcelWriter(save_file_path, engine='openpyxl') as writer:
+        df.to_excel(writer, sheet_name='正常解析', index=False)
+        if exception_list:
+            except_df = pd.DataFrame(exception_list, columns=outputs_columns)
+            except_df.to_excel(writer, sheet_name='异常解析', index=False)
+
+    print(f'解析结果保存在：{save_file_path}')
+    return save_file_path, total_amount
+
+
+if __name__ == "__main__":
+
+    # print(expand_to_structured([[(0,1,'600x50cm','r'), (0,1,'3款各1张','e')]]))
+    # exit()
+    # predict("Apple is opening a new office in Tokyo.")
+    # predict("Google hired Sarah Connor from Berlin last year.")
+    file_info = ['d:/会计组/新领图8月.xlsx', '备注']
+    # file_info = ['d:/会计组/即客2025年8月尺寸整理.xlsx', '系统文件名']
+    file_info = ['d:/会计组/8月订单明细9.6.xlsx', '文件名']
+    file_info = ['d:/会计组/ZHX-8月订单明细.xlsx', '文件名']
+    file_info = ['d:/会计组/国税数据源/智韬2025年8月尺寸整理(1).xlsx', '系统文件名']
+    file_info = ['d:/会计组/彩印通8月数码9.23.xlsx', 'ERP系统文件名']
+    file_info = ['d:/会计组/CYT8月明细9.24.xlsx', 'ERP系统文件名']
+    file_info = ['d:/会计组/9.1-9.26.xlsx', '备注']
+    file_info = ['d:/会计组/JD.xlsx', '文件名']
+    file_info = ['d:/会计组/泰州即客2025年9月尺寸整理.xlsx', '系统文件名']
+    file_info = ['d:/会计组/ZHX需拆明细9月.xlsx', '文件名', False]
+    file_info = ['d:/会计组/七彩2024年9月账单尺寸整理.xlsx', '系统文件名', True]
+    file_info = ['d:/会计组/艾印图文2024年9月账单尺寸整理.xlsx', '系统文件名', True]
+    file_info = ['d:/会计组/9月转印.xlsx', 'erp', True]
+    file_info = ['d:/会计组/9月CYT.xlsx', '文件名', False]
+    file_info = ['d:/会计组/智韬2025年9月尺寸整理.xlsx', '系统文件名', True]
+    file_info = ['d:/会计组/9月UV转印贴.xlsx', '文件名', True]
+    file_info = ['d:/会计组/彩印通2025年9月（数码）.xlsx', '文件名', True]
+    file_info = ['d:/会计组/9月名片.xlsx', '文件名', True]
+    file_info = ['d:/会计组/9月不干胶.xlsx', '文件名', True]
+    file_path, target_index, is_horizontal = file_info
+    parse_finance_data(file_path, target_index, is_horizontal)
+
+    

app/resources/models/订单尺寸识别/best_model/config.cfg

@@ -0,0 +1,135 @@
+[paths]
+train = null
+dev = null
+vectors = null
+init_tok2vec = null
+
+[system]
+seed = 0
+gpu_allocator = null
+
+[nlp]
+lang = "zh"
+pipeline = ["ner"]
+disabled = []
+before_creation = null
+after_creation = null
+after_pipeline_creation = null
+batch_size = 1000
+vectors = {"@vectors":"spacy.Vectors.v1"}
+
+[nlp.tokenizer]
+@tokenizers = "spacy.zh.ChineseTokenizer"
+segmenter = "char"
+
+[components]
+
+[components.ner]
+factory = "ner"
+incorrect_spans_key = null
+moves = null
+scorer = {"@scorers":"spacy.ner_scorer.v1"}
+update_with_oracle_cut_size = 100
+
+[components.ner.model]
+@architectures = "spacy.TransitionBasedParser.v2"
+state_type = "ner"
+extra_state_tokens = false
+hidden_width = 64
+maxout_pieces = 2
+use_upper = true
+nO = null
+
+[components.ner.model.tok2vec]
+@architectures = "spacy.HashEmbedCNN.v2"
+pretrained_vectors = null
+width = 96
+depth = 4
+embed_size = 2000
+window_size = 1
+maxout_pieces = 3
+subword_features = true
+
+[corpora]
+
+[corpora.dev]
+@readers = "spacy.Corpus.v1"
+path = ${paths.dev}
+gold_preproc = false
+max_length = 0
+limit = 0
+augmenter = null
+
+[corpora.train]
+@readers = "spacy.Corpus.v1"
+path = ${paths.train}
+gold_preproc = false
+max_length = 0
+limit = 0
+augmenter = null
+
+[training]
+seed = ${system.seed}
+gpu_allocator = ${system.gpu_allocator}
+dropout = 0.1
+accumulate_gradient = 1
+patience = 1600
+max_epochs = 0
+max_steps = 20000
+eval_frequency = 200
+frozen_components = []
+annotating_components = []
+dev_corpus = "corpora.dev"
+train_corpus = "corpora.train"
+before_to_disk = null
+before_update = null
+
+[training.batcher]
+@batchers = "spacy.batch_by_words.v1"
+discard_oversize = false
+tolerance = 0.2
+get_length = null
+
+[training.batcher.size]
+@schedules = "compounding.v1"
+start = 100
+stop = 1000
+compound = 1.001
+t = 0.0
+
+[training.logger]
+@loggers = "spacy.ConsoleLogger.v1"
+progress_bar = false
+
+[training.optimizer]
+@optimizers = "Adam.v1"
+beta1 = 0.9
+beta2 = 0.999
+L2_is_weight_decay = true
+L2 = 0.01
+grad_clip = 1.0
+use_averages = false
+eps = 0.00000001
+learn_rate = 0.001
+
+[training.score_weights]
+ents_f = 1.0
+ents_p = 0.0
+ents_r = 0.0
+ents_per_type = null
+
+[pretraining]
+
+[initialize]
+vectors = ${paths.vectors}
+init_tok2vec = ${paths.init_tok2vec}
+vocab_data = null
+lookups = null
+before_init = null
+after_init = null
+
+[initialize.components]
+
+[initialize.tokenizer]
+pkuseg_model = null
+pkuseg_user_dict = "default"

app/resources/models/订单尺寸识别/best_model/meta.json

@@ -0,0 +1,41 @@
+{
+  "lang":"zh",
+  "name":"pipeline",
+  "version":"0.0.0",
+  "spacy_version":">=3.8.7,<3.9.0",
+  "description":"",
+  "author":"",
+  "email":"",
+  "url":"",
+  "license":"",
+  "spacy_git_version":"4b65aa7",
+  "vectors":{
+    "width":0,
+    "vectors":0,
+    "keys":0,
+    "name":null,
+    "mode":"default"
+  },
+  "labels":{
+    "ner":[
+      "\u4ea7\u54c1",
+      "\u522e\u522e\u819c\u5c3a\u5bf8",
+      "\u52a0\u6025",
+      "\u5355\u53f7",
+      "\u5c3a\u5bf8",
+      "\u5de5\u827a",
+      "\u6570\u91cf",
+      "\u6750\u8d28",
+      "\u7528\u6237\u4fe1\u606f"
+    ]
+  },
+  "pipeline":[
+    "ner"
+  ],
+  "components":[
+    "ner"
+  ],
+  "disabled":[
+
+  ]
+}

app/resources/models/订单尺寸识别/best_model/ner/cfg

@@ -0,0 +1,13 @@
+{
+  "moves":null,
+  "update_with_oracle_cut_size":100,
+  "multitasks":[
+
+  ],
+  "min_action_freq":1,
+  "learn_tokens":false,
+  "beam_width":1,
+  "beam_density":0.0,
+  "beam_update_prob":0.0,
+  "incorrect_spans_key":null
+}

app/resources/models/订单尺寸识别/best_model/ner/moves

@@ -0,0 +1 @@
+‚¥movesÚ0{"0":{},"1":{"\u522e\u522e\u819c\u5c3a\u5bf8":-1,"\u6570\u91cf":-2,"\u5c3a\u5bf8":-3,"\u5355\u53f7":-4,"\u4ea7\u54c1":-5,"\u6750\u8d28":-6,"\u5de5\u827a":-7,"\u52a0\u6025":-8,"\u7528\u6237\u4fe1\u606f":-9},"2":{"\u522e\u522e\u819c\u5c3a\u5bf8":-1,"\u6570\u91cf":-2,"\u5c3a\u5bf8":-3,"\u5355\u53f7":-4,"\u4ea7\u54c1":-5,"\u6750\u8d28":-6,"\u5de5\u827a":-7,"\u52a0\u6025":-8,"\u7528\u6237\u4fe1\u606f":-9},"3":{"\u522e\u522e\u819c\u5c3a\u5bf8":-1,"\u6570\u91cf":-2,"\u5c3a\u5bf8":-3,"\u5355\u53f7":-4,"\u4ea7\u54c1":-5,"\u6750\u8d28":-6,"\u5de5\u827a":-7,"\u52a0\u6025":-8,"\u7528\u6237\u4fe1\u606f":-9},"4":{"":1,"\u522e\u522e\u819c\u5c3a\u5bf8":-1,"\u6570\u91cf":-2,"\u5c3a\u5bf8":-3,"\u5355\u53f7":-4,"\u4ea7\u54c1":-5,"\u6750\u8d28":-6,"\u5de5\u827a":-7,"\u52a0\u6025":-8,"\u7528\u6237\u4fe1\u606f":-9},"5":{"":1}}£cfg�§neg_keyÀ

app/resources/models/订单尺寸识别/best_model/tokenizer/cfg

@@ -0,0 +1,3 @@
+{
+  "segmenter":"char"
+}

app/resources/models/订单尺寸识别/best_model/vocab/key2row

@@ -0,0 +1 @@
+�

app/resources/models/订单尺寸识别/best_model/vocab/lookups.bin

@@ -0,0 +1 @@
+�

app/resources/models/订单尺寸识别/best_model/vocab/vectors.cfg

@@ -0,0 +1,3 @@
+{
+  "mode":"default"
+}

app/ui/__init__.py

@@ -0,0 +1,6 @@
+"""
+ui 包 - 用户界面组件
+"""
+from app.ui.main_window import MainWindow
+
+__all__ = ["MainWindow"]

app/ui/main_window.py

@@ -0,0 +1,676 @@
+"""
+主窗口 UI 组件
+"""
+import os
+import sys
+import threading
+import tkinter as tk
+from tkinter import ttk, filedialog, messagebox
+import openpyxl
+
+# ---- 配色 ----
+C = {
+    "bg":            "#f0f2f5",
+    "white":         "#ffffff",
+    "text":          "#1a1a2e",
+    "text_muted":    "#6b7280",
+    "accent":        "#4f46e5",
+    "accent_hover":  "#4338ca",
+    "accent_light":  "#eef2ff",
+    "border":        "#e5e7eb",
+    "success":       "#059669",
+}
+
+# 列宽
+COL_W = [220, 1, 220, 1, 220, 1, 200]
+COL_X = [0]
+for w in COL_W[:-1]:
+    COL_X.append(COL_X[-1] + w)
+
+ROW_H = 44
+CELL_PAD = 8
+
+
+# ======================================================================
+# 自定义下拉组件
+# ======================================================================
+class Dropdown(tk.Frame):
+    """纯自定义下拉选择器 — 不用任何系统 Combobox"""
+
+    def __init__(self, parent, values=None, default="", on_select=None,
+                 font=("Microsoft YaHei", 10), width=180, **kw):
+        super().__init__(parent, bg=C["white"], highlightbackground=C["border"],
+                         highlightthickness=1, cursor="hand2", **kw)
+        self.values = values or []
+        self._selected = default or ""
+        self._on_select = on_select
+        self._font = font
+        self._popup = None
+        self._hover = False
+
+        # 文字标签
+        self._label = tk.Label(
+            self, text=self._selected or "请选择",
+            font=font, bg=C["white"], fg=C["text"] if self._selected else C["text_muted"],
+            anchor=tk.W, padx=8,
+        )
+        self._label.pack(side=tk.LEFT, fill=tk.BOTH, expand=True)
+
+        # 箭头
+        self._arrow = tk.Label(
+            self, text="▾", font=("Microsoft YaHei", 8),
+            bg=C["white"], fg="#9ca3af", padx=6,
+        )
+        self._arrow.pack(side=tk.RIGHT, fill=tk.Y)
+
+        # 事件绑定
+        for w in (self, self._label, self._arrow):
+            w.bind("<Button-1>", self._toggle)
+            w.bind("<Enter>", self._on_enter)
+            w.bind("<Leave>", self._on_leave)
+
+    def _on_enter(self, _):
+        self._hover = True
+        self.configure(highlightbackground=C["accent"])
+        self._label.configure(bg="#f9fafb")
+        self._arrow.configure(bg="#f9fafb", fg=C["accent"])
+
+    def _on_leave(self, _):
+        self._hover = False
+        if not self._popup:
+            self.configure(highlightbackground=C["border"])
+        self._label.configure(bg=C["white"])
+        self._arrow.configure(bg=C["white"], fg="#9ca3af")
+
+    def _toggle(self, event=None):
+        if self._popup:
+            self._close()
+        else:
+            self._open()
+
+    def _open(self):
+        self.configure(highlightbackground=C["accent"])
+        self._label.configure(bg="#f9fafb")
+        self._arrow.configure(bg="#f9fafb", fg=C["accent"])
+
+        self._popup = tk.Toplevel(self)
+        self._popup.overrideredirect(True)
+        self._popup.configure(bg=C["white"])
+
+        # 计算位置
+        x = self.winfo_rootx()
+        y = self.winfo_rooty() + self.winfo_height() + 1
+        w = self.winfo_width()
+        item_h = 32
+        max_h = min(len(self.values) * item_h, 240)
+        list_h = max(item_h, max_h)
+
+        self._popup.geometry(f"{w}x{list_h}+{x}+{y}")
+
+        # 外边框
+        border = tk.Frame(self._popup, bg=C["border"])
+        border.pack(fill=tk.BOTH, expand=True)
+
+        # 列表
+        self._listbox = tk.Listbox(
+            border, font=self._font,
+            bg=C["white"], fg=C["text"],
+            selectbackground=C["accent_light"],
+            selectforeground=C["accent"],
+            selectborderwidth=0,
+            activestyle="none",
+            highlightthickness=0,
+            borderwidth=0,
+            relief=tk.FLAT,
+            height=len(self.values),
+        )
+        self._listbox.pack(fill=tk.BOTH, expand=True, padx=1, pady=1)
+
+        for i, v in enumerate(self.values):
+            self._listbox.insert(tk.END, f"  {v}")
+            if v == self._selected:
+                self._listbox.selection_set(i)
+                self._listbox.see(i)
+
+        self._listbox.bind("<ButtonRelease-1>", self._on_pick)
+        self._listbox.bind("<Return>", self._on_pick)
+        self._listbox.bind("<Escape>", lambda e: self._close())
+        self._listbox.focus_set()
+
+        # 点击外部关闭
+        self._popup.bind("<FocusOut>", lambda e: self.after(100, self._check_focus))
+
+        # 绑定全局点击关闭（鼠标点击 popup 外）
+        self._grab_bind = self.winfo_toplevel().bind(
+            "<Button-1>", lambda e: self._on_global_click(e), add="+"
+        )
+
+    def _on_global_click(self, event):
+        if self._popup:
+            px, py = event.x_root, event.y_root
+            px2 = self._popup.winfo_rootx()
+            py2 = self._popup.winfo_rooty()
+            pw = self._popup.winfo_width()
+            ph = self._popup.winfo_height()
+            if not (px2 <= px <= px2 + pw and py2 <= py <= py2 + ph):
+                self._close()
+
+    def _check_focus(self):
+        if self._popup and not self._popup.focus_displayof():
+            self._close()
+
+    def _close(self, *_):
+        if self._popup:
+            self._popup.destroy()
+            self._popup = None
+            self._listbox = None
+            self.configure(highlightbackground=C["border"])
+            self._label.configure(bg=C["white"])
+            self._arrow.configure(bg=C["white"], fg="#9ca3af")
+            # 移除全局绑定
+            try:
+                self.winfo_toplevel().unbind("<Button-1>", self._grab_bind)
+            except Exception:
+                pass
+
+    def _on_pick(self, event):
+        sel = self._listbox.curselection()
+        if sel:
+            idx = sel[0]
+            self._selected = self.values[idx]
+            self._label.configure(text=self._selected, fg=C["text"])
+            if self._on_select:
+                self._on_select(self._selected)
+        self._close()
+
+    # ── 公开 API ──
+    def get(self):
+        return self._selected
+
+    def set_values(self, values, default=None):
+        self.values = values
+        if default is not None:
+            self._selected = default
+            self._label.configure(text=default, fg=C["text"] if default else C["text_muted"])
+        # 如果已打开则重建
+        if self._popup:
+            self._close()
+            self._open()
+
+
+# ======================================================================
+# 主窗口
+# ======================================================================
+class MainWindow:
+    """应用程序主窗口"""
+
+    def __init__(self, root: tk.Tk):
+        self.root = root
+        self.root.configure(bg=C["bg"])
+        self._files_data = []
+        self._build_ui()
+
+    def _build_ui(self):
+        # ── 选择卡片 ──
+        card = tk.Frame(self.root, bg=C["white"])
+        card.pack(fill=tk.X, padx=24, pady=(20, 0))
+
+        card_border = tk.Frame(card, bg=C["white"],
+                               highlightbackground=C["border"],
+                               highlightthickness=1)
+        card_border.pack(fill=tk.X)
+
+        tk.Label(
+            card_border, text="📂  选择 Excel 文件",
+            font=("Microsoft YaHei", 12, "bold"),
+            bg=C["white"], fg=C["text"], anchor=tk.W,
+        ).pack(fill=tk.X, padx=20, pady=(16, 10))
+
+        input_row = tk.Frame(card_border, bg=C["white"])
+        input_row.pack(fill=tk.X, padx=20, pady=(0, 10))
+
+        entry_border = tk.Frame(
+            input_row, bg=C["white"],
+            highlightbackground=C["border"], highlightthickness=1,
+        )
+        entry_border.pack(side=tk.LEFT, fill=tk.X, expand=True)
+
+        self._path_entry = tk.Entry(
+            entry_border,
+            font=("Microsoft YaHei", 10),
+            bg="#f9fafb", fg=C["text_muted"],
+            relief=tk.FLAT, borderwidth=0,
+            readonlybackground="#f9fafb",
+        )
+        self._path_entry.pack(fill=tk.X, padx=10, pady=8)
+        self._path_entry.insert(0, "请选择 Excel 文件（.xlsx / .xls，可多选）")
+        self._path_entry.configure(state="readonly")
+
+        self._browse_btn = tk.Button(
+            input_row, text="浏览...",
+            font=("Microsoft YaHei", 10, "bold"),
+            bg=C["accent"], fg="white", relief=tk.FLAT,
+            activebackground=C["accent_hover"], activeforeground="white",
+            cursor="hand2", padx=20, pady=7, borderwidth=0,
+            command=self._on_select_files,
+        )
+        self._browse_btn.pack(side=tk.LEFT, padx=(8, 0))
+
+        status_bar = tk.Frame(card_border, bg="#f9fafb")
+        status_bar.pack(fill=tk.X)
+
+        self._status_label = tk.Label(
+            status_bar, text="  未选择任何文件",
+            font=("Microsoft YaHei", 9),
+            bg="#f9fafb", fg=C["text_muted"], anchor=tk.W,
+        )
+        self._status_label.pack(side=tk.LEFT, padx=16, pady=(6, 6))
+
+        # ── 表格区域 ──
+        self._table_area = tk.Frame(self.root, bg=C["bg"])
+        self._table_area.pack(fill=tk.BOTH, expand=True, padx=24, pady=(12, 0))
+
+        # ── 日志面板（底部固定高度，内部独立滚动） ──
+        self._log_frame = tk.Frame(self.root, bg="white", height=130)
+        self._log_frame.pack(fill=tk.X, padx=24, pady=(8, 8))
+        self._log_frame.pack_propagate(False)
+
+        log_header = tk.Frame(self._log_frame, bg="white")
+        log_header.pack(fill=tk.X)
+        tk.Label(log_header, text="📋  处理日志",
+                 font=("Microsoft YaHei", 9, "bold"),
+                 bg="white", fg=C["text"], anchor=tk.W,
+        ).pack(side=tk.LEFT, padx=10, pady=(6, 2))
+
+        log_text_frame = tk.Frame(self._log_frame, bg="white")
+        log_text_frame.pack(fill=tk.BOTH, expand=True, padx=10, pady=(0, 8))
+
+        self._log_text = tk.Text(
+            log_text_frame,
+            font=("Consolas", 9),
+            bg="#f9fafb", fg=C["text"],
+            wrap=tk.WORD, state=tk.DISABLED,
+            relief=tk.FLAT, borderwidth=0,
+            height=6,
+        )
+        log_scrollbar = tk.Scrollbar(log_text_frame, orient=tk.VERTICAL,
+                                     command=self._log_text.yview)
+        self._log_text.configure(yscrollcommand=log_scrollbar.set)
+        log_scrollbar.pack(side=tk.RIGHT, fill=tk.Y)
+        self._log_text.pack(side=tk.LEFT, fill=tk.BOTH, expand=True)
+
+        self._show_empty()
+
+    # ==================================================================
+    # 空状态
+    # ==================================================================
+    def _show_empty(self):
+        for w in self._table_area.winfo_children():
+            w.destroy()
+
+        drop = tk.Frame(self._table_area, bg="#fafafe",
+                        highlightbackground="#c7d2fe",
+                        highlightthickness=2)
+        drop.pack(fill=tk.BOTH, expand=True)
+
+        inner = tk.Frame(drop, bg="#fafafe")
+        inner.place(relx=0.5, rely=0.5, anchor=tk.CENTER)
+
+        tk.Label(inner, text="📂", font=("Segoe UI Emoji", 40),
+                 bg="#fafafe", fg="#c7d2fe").pack(pady=(0, 4))
+        tk.Label(inner, text="点击上方「浏览...」按钮选择文件",
+                 font=("Microsoft YaHei", 13, "bold"),
+                 bg="#fafafe", fg=C["text_muted"]).pack()
+        tk.Label(inner, text="支持 .xlsx、.xls 格式，可同时选择多个文件",
+                 font=("Microsoft YaHei", 9),
+                 bg="#fafafe", fg="#c0c4cc").pack(pady=(2, 0))
+
+    # ==================================================================
+    # 读取 xlsx 信息（用内置 zipfile，避免 openpyxl 打包问题）
+    # ==================================================================
+    def _read_xlsx_info(self, filepath):
+        import pandas as pd
+        sheet_names = []
+        columns = []
+        for engine in ["openpyxl", "xlrd"]:
+            try:
+                xls = pd.ExcelFile(filepath, engine=engine)
+                sheet_names = xls.sheet_names
+                if sheet_names:
+                    df = pd.read_excel(xls, sheet_name=sheet_names[0], nrows=1)
+                    columns = [str(c) for c in df.columns.tolist()]
+                xls.close()
+                break
+            except Exception:
+                continue
+        return sheet_names, columns
+
+    # ==================================================================
+    # 选择文件
+    # ==================================================================
+    def _on_select_files(self):
+        paths = filedialog.askopenfilenames(
+            title="选择 Excel 文件",
+            filetypes=[
+                ("Excel 文件", "*.xlsx *.xls"),
+                ("Excel 2007+", "*.xlsx"),
+                ("Excel 97-2003", "*.xls"),
+            ],
+        )
+        if not paths:
+            return
+
+        for p in paths:
+            if any(d["path"] == p for d in self._files_data):
+                continue
+            try:
+                sheet_names, columns = self._read_xlsx_info(p)
+                first_sheet = sheet_names[0] if sheet_names else None
+                self._files_data.append({
+                    "path": p,
+                    "filename": os.path.basename(p),
+                    "sheet_names": sheet_names,
+                    "columns": columns,
+                    "selected_sheet": first_sheet,
+                    "selected_col": columns[0] if columns else None,
+                    "parse_mode": "横向排列",
+                })
+            except Exception as e:
+                messagebox.showerror("读取失败", f"无法读取文件:\n{os.path.basename(p)}\n\n{e}")
+                continue
+
+        n = len(self._files_data)
+        if n == 0:
+            self._path_entry.configure(state="normal")
+            self._path_entry.delete(0, tk.END)
+            self._path_entry.insert(0, "请选择 Excel 文件（.xlsx / .xls，可多选）")
+            self._path_entry.configure(state="readonly", fg=C["text_muted"])
+            self._status_label.configure(text="  未选择任何文件")
+        elif n == 1:
+            self._path_entry.configure(state="normal")
+            self._path_entry.delete(0, tk.END)
+            self._path_entry.insert(0, self._files_data[0]["path"])
+            self._path_entry.configure(state="readonly", fg=C["text"])
+            self._status_label.configure(
+                text=f"  已选择 1 个文件  |  点击「浏览...」可继续添加",
+                fg=C["success"],
+            )
+        else:
+            self._path_entry.configure(state="normal")
+            self._path_entry.delete(0, tk.END)
+            self._path_entry.insert(0, f"已选择 {n} 个文件")
+            self._path_entry.configure(state="readonly", fg=C["text"])
+            self._status_label.configure(
+                text=f"  已选择 {n} 个文件  |  点击「浏览...」可继续添加",
+                fg=C["success"],
+            )
+
+        if self._files_data:
+            self._build_table()
+        else:
+            self._show_empty()
+
+    # ==================================================================
+    # 构建表格
+    # ==================================================================
+    def _build_table(self):
+        for w in self._table_area.winfo_children():
+            w.destroy()
+
+        self._table_frame = tk.Frame(self._table_area, bg=C["bg"])
+        self._table_frame.pack(fill=tk.BOTH, expand=True)
+
+        self._draw_header()
+        for fi, fd in enumerate(self._files_data):
+            self._draw_row(fi, fd)
+
+        # ── 确认按钮 ──
+        btn_bar = tk.Frame(self._table_frame, bg=C["bg"])
+        btn_bar.pack(fill=tk.X, pady=(16, 0))
+
+        self._confirm_btn = tk.Button(
+            btn_bar, text="✅  开始解析",
+            font=("Microsoft YaHei", 12, "bold"),
+            bg=C["accent"], fg="white", relief=tk.FLAT,
+            activebackground=C["accent_hover"], activeforeground="white",
+            cursor="hand2", padx=28, pady=10, borderwidth=0,
+            command=self._on_confirm,
+        )
+        self._confirm_btn.pack(side=tk.RIGHT)
+
+        self._progress_label = tk.Label(
+            btn_bar, text="",
+            font=("Microsoft YaHei", 9),
+            bg=C["bg"], fg=C["text_muted"], anchor=tk.E,
+        )
+        self._progress_label.pack(side=tk.RIGHT, padx=(0, 12))
+
+    def _draw_header(self):
+        hdr = tk.Frame(self._table_frame, bg=C["white"], height=38)
+        hdr.pack(fill=tk.X)
+        hdr.pack_propagate(False)
+
+        headers = ["文件名", "工作表", "解析列", "解析方式"]
+        for i, title in enumerate(headers):
+            col_idx = i * 2
+            x, w = COL_X[col_idx], COL_W[col_idx]
+            tk.Label(
+                hdr, text=title, font=("Microsoft YaHei", 10, "bold"),
+                bg=C["white"], fg=C["text"], anchor=tk.W,
+            ).place(x=x + CELL_PAD, y=0, width=w - CELL_PAD * 2, height=38)
+            if i < 3:
+                tk.Frame(hdr, bg=C["border"], width=1).place(
+                    x=COL_X[col_idx + 1], y=6, height=26
+                )
+
+        tk.Frame(self._table_frame, bg=C["border"], height=1).pack(fill=tk.X)
+
+    def _draw_row(self, fi: int, fd: dict):
+        row = tk.Frame(self._table_frame, bg=C["white"], height=ROW_H)
+        row.pack(fill=tk.X)
+        row.pack_propagate(False)
+
+        # ── 文件名 ──
+        x, w = COL_X[0], COL_W[0]
+        tk.Label(
+            row, text=f"📄  {fd['filename']}",
+            font=("Microsoft YaHei", 10), bg=C["white"], fg=C["text"], anchor=tk.W,
+        ).place(x=x + CELL_PAD, y=0, width=w - CELL_PAD * 2, height=ROW_H)
+
+        tk.Frame(row, bg=C["border"], width=1).place(x=COL_X[1], y=6, height=ROW_H - 12)
+
+        # ── 工作表（自定义 Dropdown） ──
+        x, w = COL_X[2], COL_W[2]
+        dd_sheet = Dropdown(
+            row,
+            values=fd["sheet_names"],
+            default=fd["selected_sheet"] or "",
+            on_select=lambda val, i=fi: self._on_sheet_change(i, val),
+        )
+        dd_sheet.place(x=x + CELL_PAD, y=6, width=w - CELL_PAD * 2, height=ROW_H - 12)
+        fd["sheet_dd"] = dd_sheet
+
+        tk.Frame(row, bg=C["border"], width=1).place(x=COL_X[3], y=6, height=ROW_H - 12)
+
+        # ── 解析列（自定义 Dropdown） ──
+        x, w = COL_X[4], COL_W[4]
+        dd_col = Dropdown(
+            row,
+            values=fd["columns"],
+            default=fd["selected_col"] or "",
+            on_select=lambda val, i=fi: self._on_col_change(i, val),
+        )
+        dd_col.place(x=x + CELL_PAD, y=6, width=w - CELL_PAD * 2, height=ROW_H - 12)
+        fd["col_dd"] = dd_col
+
+        tk.Frame(row, bg=C["border"], width=1).place(x=COL_X[5], y=6, height=ROW_H - 12)
+
+        # ── 解析方式（自定义 Dropdown） ──
+        x, w = COL_X[6], COL_W[6]
+        dd_mode = Dropdown(
+            row,
+            values=["横向排列", "纵向排序"],
+            default=fd["parse_mode"],
+            on_select=lambda val, i=fi: self._on_mode_change(i, val),
+        )
+        dd_mode.place(x=x + CELL_PAD, y=6, width=w - CELL_PAD * 2, height=ROW_H - 12)
+        fd["mode_dd"] = dd_mode
+
+        fd["row_frame"] = row
+        tk.Frame(self._table_frame, bg=C["border"], height=1).pack(fill=tk.X)
+
+    # ==================================================================
+    # 切换工作表 → 更新解析列
+    # ==================================================================
+    def _on_sheet_change(self, idx: int, new_sheet: str):
+        fd = self._files_data[idx]
+        fd["selected_sheet"] = new_sheet
+
+        try:
+            import pandas as pd
+            for engine in ["openpyxl", "xlrd"]:
+                try:
+                    df = pd.read_excel(fd["path"], sheet_name=new_sheet, nrows=1, engine=engine)
+                    columns = [str(c) for c in df.columns.tolist()]
+                    break
+                except Exception:
+                    continue
+
+            fd["columns"] = columns
+            first_col = columns[0] if columns else None
+            fd["selected_col"] = first_col
+            fd["col_dd"].set_values(columns, default=first_col)
+        except Exception as e:
+            messagebox.showerror("错误", f"读取工作表失败: {e}")
+
+    # ==================================================================
+    # 下拉选择回调
+    # ==================================================================
+    def _on_col_change(self, idx: int, val: str):
+        self._files_data[idx]["selected_col"] = val
+
+    def _on_mode_change(self, idx: int, val: str):
+        self._files_data[idx]["parse_mode"] = val
+
+    # ==================================================================
+    # 日志输出
+    # ==================================================================
+    def _log_write(self, text: str):
+        """向底部日志面板追加一行"""
+        self._log_text.configure(state=tk.NORMAL)
+        self._log_text.insert(tk.END, text + "\n")
+        self._log_text.see(tk.END)
+        self._log_text.configure(state=tk.DISABLED)
+
+    def _log_clear(self):
+        """清空日志面板"""
+        self._log_text.configure(state=tk.NORMAL)
+        self._log_text.delete("1.0", tk.END)
+        self._log_text.configure(state=tk.DISABLED)
+
+    # ==================================================================
+    # 确认解析
+    # ==================================================================
+    def _on_confirm(self):
+        if not self._files_data:
+            return
+
+        self._confirm_btn.configure(state=tk.DISABLED, text="⏳  解析中...")
+        self._log_clear()
+        self._log_write("🚀 开始解析...")
+
+        # 捕获 print 输出到日志面板
+        class _LogWriter:
+            def __init__(self, root, log_func):
+                self._root = root
+                self._log = log_func
+                self._buf = ""
+            def write(self, s):
+                self._buf += s
+                if "\n" in self._buf:
+                    lines = self._buf.split("\n")
+                    self._buf = lines[-1]
+                    for line in lines[:-1]:
+                        stripped = line.strip()
+                        if stripped:
+                            self._root.after(0, lambda t=stripped: self._log(t))
+            def flush(self):
+                if self._buf.strip():
+                    self._root.after(0, lambda t=self._buf.strip(): self._log(t))
+                    self._buf = ""
+
+        def _run():
+            import sys
+            from app import finance_parse as fp
+            total = len(self._files_data)
+            success = 0
+            old_stdout = sys.stdout
+            old_stderr = sys.stderr
+            log_writer = _LogWriter(self.root, self._log_write)
+            sys.stdout = log_writer
+            sys.stderr = open(os.devnull, "w")
+
+            try:
+                for i, fd in enumerate(self._files_data):
+                    fname = fd["filename"]
+                    self.root.after(0, lambda n=i+1, fn=fname:
+                        self._log_write(f"⏳ [{n}/{total}] 正在处理: {fn}"))
+                    try:
+                        save_path, amount = fp.parse_finance_data(
+                            file_path=fd["path"],
+                            target_index=fd["col_dd"].get(),
+                            is_horizontal=(fd["mode_dd"].get() == "横向排列"),
+                            sheet_name=fd["sheet_dd"].get(),
+                        )
+                        success += 1
+                        fd["result"] = save_path
+                        fd["amount"] = amount
+                        self.root.after(0, lambda fn=fname, sp=save_path, amt=amount:
+                            self._log_write(f"✅ [{success}/{total}] 完成: {fn}  共 {amt} 条"))
+                    except Exception as e:
+                        fd["error"] = str(e)
+                        self.root.after(0, lambda fn=fname, err=str(e):
+                            self._log_write(f"❌ [{i+1}/{total}] 失败: {fn}  — {err}"))
+            finally:
+                sys.stdout = old_stdout
+                sys.stderr = old_stderr
+
+            self.root.after(0, lambda: self._on_parse_done(success, total))
+
+        threading.Thread(target=_run, daemon=True).start()
+
+    def _on_parse_done(self, success: int, total: int):
+        self._confirm_btn.configure(state=tk.NORMAL, text="✅  开始解析")
+
+        # 从列表中移除已成功解析的文件，保留失败的
+        self._files_data = [fd for fd in self._files_data if "error" in fd]
+
+        if success == total:
+            self._log_write(f"🎉 全部 {total} 个文件解析完成！结果已保存到原文件所在目录。")
+            messagebox.showinfo("解析完成",
+                f"全部 {total} 个文件解析完成！\n\n结果已保存到原文件所在目录。")
+        else:
+            failed_count = len(self._files_data)
+            self._log_write(f"⚠️ 完成 {success}/{total}，有 {failed_count} 个文件失败。")
+            errors = []
+            for fd in self._files_data:
+                if "error" in fd:
+                    errors.append(f"• {fd['filename']}: {fd['error']}")
+            msg = f"完成 {success}/{total}\n\n失败:\n" + "\n".join(errors[:5])
+            messagebox.showwarning("解析完成（部分失败）", msg)
+
+        # 刷新 UI：清除成功项后重建表格或显示空状态
+        if self._files_data:
+            self._build_table()
+            n = len(self._files_data)
+            self._status_label.configure(
+                text=f"  剩余 {n} 个失败文件  |  可调整配置后重新解析",
+                fg="#dc2626",
+            )
+        else:
+            self._show_empty()
+            self._path_entry.configure(state="normal")
+            self._path_entry.delete(0, tk.END)
+            self._path_entry.insert(0, "请选择 Excel 文件（.xlsx / .xls，可多选）")
+            self._path_entry.configure(state="readonly", fg=C["text_muted"])
+            self._status_label.configure(text="  未选择任何文件")

cwExcel.spec

@@ -0,0 +1,52 @@
+# -*- mode: python ; coding: utf-8 -*-
+from PyInstaller.utils.hooks import collect_all
+
+datas = [('app/resources/models/订单尺寸识别/best_model', 'app/resources/models/订单尺寸识别/best_model')]
+binaries = []
+hiddenimports = ['spacy', 'cn2an', 'tqdm', 'pandas', 'scipy']
+tmp_ret = collect_all('spacy')
+datas += tmp_ret[0]; binaries += tmp_ret[1]; hiddenimports += tmp_ret[2]
+tmp_ret = collect_all('openpyxl')
+datas += tmp_ret[0]; binaries += tmp_ret[1]; hiddenimports += tmp_ret[2]
+
+
+a = Analysis(
+    ['main.py'],
+    pathex=[],
+    binaries=binaries,
+    datas=datas,
+    hiddenimports=hiddenimports,
+    hookspath=[],
+    hooksconfig={},
+    runtime_hooks=[],
+    excludes=[],
+    noarchive=False,
+)
+pyz = PYZ(a.pure)
+
+exe = EXE(
+    pyz,
+    a.scripts,
+    [],
+    exclude_binaries=True,
+    name='cwExcel',
+    debug=False,
+    bootloader_ignore_signals=False,
+    strip=False,
+    upx=True,
+    console=False,
+    disable_windowed_traceback=False,
+    argv_emulation=False,
+    target_arch=None,
+    codesign_identity=None,
+    entitlements_file=None,
+)
+coll = COLLECT(
+    exe,
+    a.binaries,
+    a.datas,
+    strip=False,
+    upx=True,
+    upx_exclude=[],
+    name='cwExcel',
+)

main.py

@@ -0,0 +1,16 @@
+"""
+cwExcel - Python Desktop Application
+入口模块
+"""
+import tkinter as tk
+from app import MainApplication
+
+
+def main():
+    root = tk.Tk()
+    app = MainApplication(root)
+    app.run()
+
+
+if __name__ == "__main__":
+    main()

requirements.txt

@@ -0,0 +1,5 @@
+openpyxl==3.1.2
+pandas==2.0.3
+spacy==3.7.5
+cn2an==0.5.24
+tqdm==4.68.3