refactor: optimize material deletion with exact code matching and database queries

- Change from fuzzy material name matching to exact material code matching
- Migrate from MaterialsTypeToBeDeleted to MaterialsToBeDeleted table
- Add should_delete_material() for on-demand database queries
- Remove in-memory material list loading to reduce memory footprint
- Update documentation to reflect new matching logic
- Remove main scripts from git tracking (contain user-specific config)

Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
This commit is contained in:
Misaka_Company
2026-02-10 10:53:53 +08:00
parent 6ec7484036
commit cff5e78c1d
7 changed files with 405 additions and 149 deletions

6
.gitignore vendored
View File

@@ -23,6 +23,12 @@ tests/
config/user_settings.json config/user_settings.json
nul nul
# 主脚本(包含用户特定配置)
main.py
main_clean.py
main_ui.py
record.py
# 环境变量 # 环境变量
.env .env
.env.local .env.local

View File

@@ -1,6 +1,6 @@
""" """
待删除物料查询组件 待删除物料查询组件
从数据库查询指定负责人需要删除的物料名称 从数据库查询指定负责人需要删除的物料编码
""" """
from typing import List, Dict, Any from typing import List, Dict, Any
@@ -9,25 +9,25 @@ from db.connection import get_connection
def get_materials_to_delete(manager_name): def get_materials_to_delete(manager_name):
""" """
根据负责人名称查询待删除物料名称列表 根据负责人名称查询待删除物料编码列表
Args: Args:
manager_name: 负责人姓名 manager_name: 负责人姓名
Returns: Returns:
物料名称列表(关键字) 物料编码列表
""" """
query = """ query = """
SELECT [MaterialName] SELECT [MaterialCode]
FROM [dbo].[MaterialsTypeToBeDeleted] FROM [dbo].[MaterialsToBeDeleted]
WHERE [ManagerName] = ? WHERE [ManagerName] = ?
""" """
with get_connection() as conn: with get_connection() as conn:
results = conn.execute_query(query, (manager_name,)) results = conn.execute_query(query, (manager_name,))
# 提取物料名称并去除空值 # 提取物料编码并去除空值
material_names = [row["MaterialName"] for row in results if row["MaterialName"]] material_codes = [row["MaterialCode"] for row in results if row["MaterialCode"]]
return material_names return material_codes
def get_all_materials_to_delete() -> List[Dict[str, Any]]: def get_all_materials_to_delete() -> List[Dict[str, Any]]:
@@ -35,15 +35,40 @@ def get_all_materials_to_delete() -> List[Dict[str, Any]]:
获取所有待删除物料记录 获取所有待删除物料记录
Returns: Returns:
List[Dict[str, Any]]: 包含MaterialName和ManagerName的记录列表 List[Dict[str, Any]]: 包含MaterialCode和ManagerName的记录列表
""" """
query = """ query = """
SELECT [MaterialName], [ManagerName] SELECT [MaterialCode], [ManagerName]
FROM [dbo].[MaterialsTypeToBeDeleted] FROM [dbo].[MaterialsToBeDeleted]
WHERE [MaterialName] IS NOT NULL WHERE [MaterialCode] IS NOT NULL
ORDER BY [ManagerName], [MaterialName] ORDER BY [ManagerName], [MaterialCode]
""" """
with get_connection() as conn: with get_connection() as conn:
results = conn.execute_query(query) results = conn.execute_query(query)
return results return results
def should_delete_material(manager_name: str, material_code: str) -> bool:
"""
检查指定物料编码是否需要删除
直接在数据库层面查询,避免将大量数据加载到内存
Args:
manager_name: 负责人姓名
material_code: 物料编码
Returns:
bool: 如果物料需要删除返回 True否则返回 False
"""
query = """
SELECT COUNT(*) as count
FROM [dbo].[MaterialsToBeDeleted]
WHERE [ManagerName] = ?
AND [MaterialCode] = ?
"""
with get_connection() as conn:
results = conn.execute_query(query, (manager_name, material_code))
return results[0]["count"] > 0 if results else False

345
docs/MAIN_CLEAN_FLOW.md Normal file
View File

@@ -0,0 +1,345 @@
# main_clean.py 流程说明
## 概述
`main_clean.py` 是离散备料计划维护数据清理工具的主程序入口。该工具通过 Playwright 自动化登录用友 BIP 系统,读取生产订单号,然后逐个清理订单中不需要的物料数据。
---
## 系统架构图
```mermaid
flowchart TB
Start([开始]) --> Init[初始化 DiscreteMaterialPlanCleaner]
Init --> QueryDB[(查询数据库<br/>获取待删除物料列表)]
QueryDB --> Login[登录用友BIP系统]
Login --> OpenMenu[点击功能菜单]
OpenMenu --> OpenPage[打开离散生产订单维护页面]
OpenPage --> SetupQuery[设置查询界面]
SetupQuery --> ReadFile[(读取ProductionID.txt)]
ReadFile --> GetOrders[(查询数据库<br/>获取生产订单号)]
GetOrders --> LoopStart{遍历订单}
LoopStart --> ProcessOrder[处理单个订单]
ProcessOrder --> QueryOrder[查询订单详情]
QueryOrder --> OpenPlanPage[打开备料计划页面]
OpenPlanPage --> WaitLoad[等待页面加载完成]
WaitLoad --> CheckData{检查数据状态}
CheckData -->|detail_count = 0| SkipOrder[跳过该订单]
CheckData -->|status != 审批通过| SkipOrder
CheckData -->|detail_count > 0<br/>status = 审批通过| ClickModify[点击修改按钮]
ClickModify --> ExpandForm[展开详细信息表单]
ExpandForm --> LoopMaterials{遍历物料}
LoopMaterials --> GetMaterial[获取物料信息]
GetMaterial --> CheckDelete{需要删除?}
CheckDelete -->|是| MarkDelete[标记待删除]
CheckDelete -->|否| KeepMaterial[保留物料]
MarkDelete --> NextMaterial{下一个物料?}
KeepMaterial --> NextMaterial
NextMaterial -->|是| LoopMaterials
NextMaterial -->|否| ClosePlanPage[关闭备料计划页面]
SkipOrder --> ClosePlanPage
ClosePlanPage --> NextOrder{下一个订单?}
NextOrder -->|是| LoopStart
NextOrder -->|否| Logout[注销账号]
Logout --> CloseBrowser[关闭浏览器]
CloseBrowser --> End([结束])
style Init fill:#e1f5ff
style QueryDB fill:#fff4e1
style Login fill:#e1f5ff
style End fill:#ffe1e1
style CheckDelete fill:#fff4e1
style CheckData fill:#fff4e1
```
---
## 详细流程说明
### 1. 初始化阶段
```mermaid
flowchart LR
A[main函数] --> B[创建 DiscreteMaterialPlanCleaner 实例]
B --> C[配置参数:<br/>username, password,<br/>manager_name, headless, verbose]
style A fill:#e1f5ff
style B fill:#e1f5ff
style C fill:#e1f5ff
```
**代码位置**: `main_clean.py:9-24`
```python
cleaner = DiscreteMaterialPlanCleaner(
username="BLDpengqiangqiang",
password="Cqbld123456.",
manager_name="彭羽",
headless=False,
verbose=True,
)
```
---
### 2. 数据准备阶段
```mermaid
flowchart TB
A[clean.clean方法] --> B[查询数据库<br/>MaterialsToBeDeleted表]
B --> C[获取指定负责人的<br/>待删除物料编码列表]
style A fill:#fff4e1
style B fill:#fff4e1
style C fill:#fff4e1
```
**代码位置**: `utils/离散备料计划维护数据清理.py:292-304`
| 模块 | 功能 | 文件 |
|------|------|------|
| `get_materials_to_delete()` | 查询待删除物料编码 | `db/materials_to_delete.py` |
---
### 3. 登录与导航
```mermaid
sequenceDiagram
participant Main as main_clean.py
participant Auth as auth.py
participant Browser as Chromium Browser
participant ERP as 用友BIP系统
Main->>Auth: login(playwright, username, password)
Auth->>Browser: launch(headless=False)
Auth->>Browser: new_context()
Auth->>Browser: new_page()
Auth->>ERP: goto(登录页面URL)
Auth->>ERP: 填写用户名和密码
Auth->>ERP: 点击登录按钮
alt 有强制登录弹窗
ERP-->>Auth: 显示弹窗
Auth->>ERP: 点击确定
end
Auth-->>Main: 返回 (browser, context, page, main_frame)
```
**代码位置**: `utils/auth.py:8-72`
---
### 4. 设置查询界面
```mermaid
flowchart TB
A[点击图标按钮] --> B[选择"订单号查询"]
B --> C[点击"全部"标签]
C --> D{填充显示条数}
D --> E{成功?}
E -->|否| F[重试<br/>最多3次]
F --> D
E -->|是| G[继续]
style D fill:#fff4e1
style E fill:#fff4e1
style F fill:#ffe1e1
```
**代码位置**: `utils/离散备料计划维护数据清理.py:263-290`
---
### 5. 获取生产订单号
```mermaid
flowchart LR
A[read_production_ids] --> B[读取ProductionID.txt]
B --> C[解析总排号列表]
C --> D[query_production_order_numbers]
D --> E[查询数据库<br/>获取生产订单号]
style A fill:#fff4e1
style D fill:#fff4e1
```
**代码位置**: `utils/离散备料计划维护数据清理.py:41-59`
| 函数 | 功能 | 数据库表 |
|------|------|----------|
| `read_production_ids()` | 读取总排号文件 | 文件输入 |
| `query_production_order_numbers()` | 查询生产订单号 | ProductionOrder |
---
### 6. 订单处理流程
```mermaid
flowchart TB
Start([开始处理订单]) --> FillInput[填充生产订单号]
FillInput --> ClickQuery[点击查询按钮]
ClickQuery --> WaitLoad[等待加载完成]
WaitLoad --> ClickMore[点击"更多"按钮]
ClickMore --> OpenPlan[打开"备料计划"页面]
OpenPlan --> GetIframe[获取嵌套iframe]
GetIframe --> WaitPlanLoad[等待备料计划页面加载]
WaitPlanLoad --> ExtractInfo[提取信息:<br/>- 编码<br/>- 详细信息数量<br/>- 备料状态]
ExtractInfo --> CheckStatus{状态检查}
style Start fill:#e1f5ff
style CheckStatus fill:#fff4e1
```
**代码位置**: `utils/离散备料计划维护数据清理.py:61-167`
---
### 7. 物料清理流程
```mermaid
flowchart TB
Check{detail_count > 0<br/>&<br/>status = 审批通过} -->|否| Skip[跳过处理]
Check -->|是| ClickModify[点击修改按钮]
ClickModify --> WaitSave[等待保存按钮可见]
WaitSave --> Expand[点击展开按钮]
Expand --> GetForm[获取表单容器]
GetForm --> LoopStart{遍历物料}
LoopStart --> GetSerial[定位序号]
GetSerial --> GetCode[获取材料编码]
GetCode --> GetName[获取材料名称]
GetName --> GetQty1[获取累计待发数量]
GetQty1 --> GetQty2[获取累计出库数量]
GetQty2 --> Match{物料编码<br/>在待删除列表中?}
Match -->|精确匹配| Delete[标记待删除]
Match -->|未匹配| Keep[保留]
Delete --> Next{下一个?}
Keep --> Next
Next -->|是| LoopStart
Next -->|否| Done([完成])
style Check fill:#fff4e1
style Match fill:#fff4e1
style Done fill:#e1f5ff
```
**代码位置**: `utils/离散备料计划维护数据清理.py:169-261`
---
## 关键数据结构
### DiscreteMaterialPlanCleaner 类
```mermaid
classDiagram
class DiscreteMaterialPlanCleaner {
+str username
+str password
+str manager_name
+bool headless
+bool verbose
+_print(args)
+get_production_order_numbers(file)
+process_order(frame, order_id, ...)
+setup_query_interface(frame)
+clean(file, debug_mode, ...)
}
class auth {
+login(playwright, username, password)
+logout(main_frame, verbose)
+close_session(browser, context)
}
class materials_to_delete {
+get_materials_to_delete(manager_name)
+get_all_materials_to_delete()
}
class production_order_query {
+read_production_ids(file)
+query_production_order_numbers(ids)
}
DiscreteMaterialPlanCleaner --> auth
DiscreteMaterialPlanCleaner --> materials_to_delete
DiscreteMaterialPlanCleaner --> production_order_query
```
---
## 配置参数说明
| 参数 | 类型 | 默认值 | 说明 |
|------|------|--------|------|
| `username` | str | - | 登录用户名 |
| `password` | str | - | 登录密码 |
| `manager_name` | str | - | 负责人姓名(用于查询待删除物料) |
| `headless` | bool | False | 是否无头模式运行浏览器 |
| `verbose` | bool | True | 是否打印详细日志 |
---
## 文件依赖关系
```mermaid
graph TD
A[main_clean.py] --> B[utils/离散备料计划维护数据清理.py]
B --> C[utils/auth.py]
B --> D[db/production_order_query.py]
B --> E[db/materials_to_delete.py]
B --> F[db/connection.py]
style A fill:#e1f5ff
style B fill:#e1f5ff
```
---
## 执行入口
```bash
# 直接运行
python main_clean.py
# 或在虚拟环境中运行
.venv\Scripts\activate
python main_clean.py
```
---
## 注意事项
1. **浏览器模式**: 默认使用非无头模式 (`headless=False`),便于调试观察
2. **日志输出**: 启用详细日志 (`verbose=True`),打印每一步操作
3. **数据库连接**: 使用上下文管理器自动管理数据库连接
4. **异常处理**: 加载超时会自动重试或跳过
5. **调试模式**: 支持在指定订单处暂停调试

27
main.py
View File

@@ -1,27 +0,0 @@
"""
主程序 - 使用离散备料计划维护数据提取工具
"""
import os
from utils.离散备料计划维护数据提取 import DiscreteMaterialPlanExtractor
def main():
# 创建提取器
extractor = DiscreteMaterialPlanExtractor(
username="BLDpengqiangqiang",
password="Cqbld123456.",
headless=True,
verbose=True,
)
# 设置文件路径
order_id_file = os.path.join(os.path.dirname(__file__), "ProductionID.txt")
output_file = r"D:/python/playwrite/data/离散备料计划维护_合并.xlsx"
# 执行提取
extractor.extract(order_id_file, output_file)
if __name__ == "__main__":
main()

View File

@@ -1,27 +0,0 @@
"""
主程序 - 使用离散备料计划维护数据清理工具
"""
import os
from utils.离散备料计划维护数据清理 import DiscreteMaterialPlanCleaner
def main():
# 创建清理器
cleaner = DiscreteMaterialPlanCleaner(
username="BLDpengqiangqiang",
password="Cqbld123456.",
manager_name="彭羽",
headless=True,
verbose=True,
)
# 设置文件路径
order_id_file = os.path.join(os.path.dirname(__file__), "ProductionID.txt")
# 执行清理
cleaner.clean(order_id_file)
if __name__ == "__main__":
main()

View File

@@ -1,59 +0,0 @@
#!/usr/bin/env python
# -*- coding: utf-8 -*-
"""
ERP 自动化工具 - GUI 版本
这是一个基于 Tkinter 的图形用户界面应用程序,
用于自动化 ERP 系统(用友 YonBIP的数据提取、校验和查询任务。
运行方式:
python main_ui.py
功能:
- 数据提取:从 ERP 系统提取备料计划数据
- 物料校验:校验物料状态并匹配待删除物料
- 数据查询:查询生产订单号等信息
- 设置管理:管理系统配置
- 用户认证:基于用户权限的数据访问控制
"""
import tkinter as tk
from tkinter import messagebox
from gui.main_window import MainWindow
from gui.login_dialog import LoginDialog
from auth.session_manager import SessionManager
def main():
"""主函数"""
# 创建根窗口
root = tk.Tk()
# 显示登录对话框
login_dialog = LoginDialog(root)
credentials = login_dialog.get_credentials()
if not credentials:
# 用户取消登录或关闭对话框
root.destroy()
return
username, password = credentials
# 认证并初始化会话
session_manager = SessionManager.get_instance()
if not session_manager.login(username, password):
messagebox.showerror("登录失败", "用户名或密码错误")
root.destroy()
return
# 创建主窗口(传入 session_manager
app = MainWindow(root, session_manager)
# 启动主事件循环
root.mainloop()
if __name__ == "__main__":
main()

View File

@@ -64,7 +64,7 @@ class DiscreteMaterialPlanCleaner:
order_id, order_id,
order_index, order_index,
page1, page1,
materials_to_delete=None, manager_name=None,
debug_mode=False, debug_mode=False,
debug_order=None, debug_order=None,
): ):
@@ -75,12 +75,12 @@ class DiscreteMaterialPlanCleaner:
order_id: 订单号 order_id: 订单号
order_index: 订单索引 order_index: 订单索引
page1: 页面对象 page1: 页面对象
materials_to_delete: 待删除物料关键字列表 manager_name: 负责人姓名(用于查询待删除物料)
debug_mode: 是否启用调试模式 debug_mode: 是否启用调试模式
debug_order: 调试订单号 debug_order: 调试订单号
""" """
if materials_to_delete is None: if manager_name is None:
materials_to_delete = [] manager_name = self.manager_name
from playwright.sync_api import TimeoutError from playwright.sync_api import TimeoutError
import re import re
import time import time
@@ -180,22 +180,23 @@ class DiscreteMaterialPlanCleaner:
child_form.wait_for(state="visible", timeout=5000) child_form.wait_for(state="visible", timeout=5000)
self._print(f"父容器 .card-table-side-box 已找到") self._print(f"父容器 .card-table-side-box 已找到")
page2.pause() # page2.pause()
for id in range(detail_count): for id in range(detail_count):
id_lable_locator = child_form.get_by_text("序号 " + str(id + 1)) id_lable_locator = child_form.get_by_text("序号 " + str(id + 1))
id_lable_locator.wait_for(state="visible", timeout=10000) id_lable_locator.wait_for(state="visible", timeout=10000)
self._print(f"处理 {id_lable_locator.inner_text()} ") self._print(f"处理 {id_lable_locator.inner_text()} ")
# 获取材料编码(通过文本定位取第一个input # 获取材料编码(用于精确匹配
input_box = ( input_box = (
child_form.locator("div") child_form.locator("div")
.filter(has_text=re.compile(r"^材料编码\d{11}$", re.MULTILINE)) .filter(has_text=re.compile(r"^材料编码\d{11}$", re.MULTILINE))
.locator("input") .locator("input")
.first .first
) )
self._print(f"材料编码:{input_box.input_value()}") material_code = input_box.input_value()
self._print(f"材料编码:{material_code}")
# 获取材料名称 # 获取材料名称(仅用于日志)
input_box = ( input_box = (
child_form.locator("div") child_form.locator("div")
.filter(has_text=re.compile(r"^材料名称$")) .filter(has_text=re.compile(r"^材料名称$"))
@@ -220,22 +221,17 @@ class DiscreteMaterialPlanCleaner:
) )
self._print(f"累计出库数量:{input_box.input_value()}") self._print(f"累计出库数量:{input_box.input_value()}")
# 检查是否需要清理该物料 # 检查是否需要清理该物料(直接数据库查询)
should_delete = False from db.materials_to_delete import should_delete_material
matched_keyword = None should_delete = should_delete_material(manager_name, material_code)
for keyword in materials_to_delete:
if keyword in material_name:
should_delete = True
matched_keyword = keyword
break
if should_delete: if should_delete:
self._print( self._print(
f">>> 需要清理:材料名称【{material_name}匹配关键字{matched_keyword}" f">>> 需要清理:材料名称【{material_name}材料编码{material_code}"
) )
# TODO: 执行删除操作 # TODO: 执行删除操作
else: else:
self._print(f"保留:材料名称【{material_name}】无需清理") self._print(f"保留:材料名称【{material_name}材料编码【{material_code}无需清理")
if id != detail_count - 1: if id != detail_count - 1:
child_form.get_by_role("button").filter( child_form.get_by_role("button").filter(
@@ -298,10 +294,7 @@ class DiscreteMaterialPlanCleaner:
debug_mode: 是否启用调试模式 debug_mode: 是否启用调试模式
debug_order: 调试订单索引 debug_order: 调试订单索引
""" """
# 获取待删除物料列表 self._print(f"使用负责人 [{self.manager_name}] 进行数据清理")
self._print(f"正在查询负责人 [{self.manager_name}] 的待删除物料列表...")
materials_to_delete = get_materials_to_delete(self.manager_name)
self._print(f"查询到 {len(materials_to_delete)} 个待删除物料关键字")
with sync_playwright() as playwright: with sync_playwright() as playwright:
# 调用登录模块 # 调用登录模块
@@ -348,7 +341,7 @@ class DiscreteMaterialPlanCleaner:
order_id, order_id,
order_index, order_index,
page1, page1,
materials_to_delete, self.manager_name,
debug_mode=debug_mode, debug_mode=debug_mode,
debug_order=debug_order, debug_order=debug_order,
) )