623 lines
23 KiB
Python
623 lines
23 KiB
Python
|
# -*- coding: utf-8 -*-
|
|||
|
|
"""
|
|||
|
|
微信群岗位信息提取工具 - GUI版本
|
|||
|
|
支持多群组监听，使用UUID作为岗位ID
|
|||
|
|
"""
|
|||
|
|
import sys
|
|||
|
|
import os
|
|||
|
|
import json
|
|||
|
|
import time
|
|||
|
|
import requests
|
|||
|
|
import threading
|
|||
|
|
import tkinter as tk
|
|||
|
|
from tkinter import ttk, scrolledtext, messagebox
|
|||
|
|
from datetime import datetime
|
|||
|
|
from pathlib import Path
|
|||
|
|
import uuid
|
|||
|
|
|
|||
|
|
# 初始化COM组件（修复打包后的错误）
|
|||
|
|
import pythoncom
|
|||
|
|
|
|||
|
|
_script_dir = os.path.dirname(os.path.abspath(__file__))
|
|||
|
|
if _script_dir not in sys.path:
|
|||
|
|
sys.path.insert(0, _script_dir)
|
|||
|
|
|
|||
|
|
from wxauto import WeChat
|
|||
|
|
|
|||
|
|
|
|||
|
|
class JobExtractorGUI:
|
|||
|
|
def __init__(self, root):
|
|||
|
|
self.root = root
|
|||
|
|
self.root.title("微信群岗位信息提取工具 v1.1")
|
|||
|
|
self.root.geometry("1200x700")
|
|||
|
|
|
|||
|
|
# 变量
|
|||
|
|
self.wx = None
|
|||
|
|
self.is_running = False
|
|||
|
|
self.job_count = 0
|
|||
|
|
self.config_file = "config.json"
|
|||
|
|
self.output_file = "jobs_data.json"
|
|||
|
|
self.active_groups = {} # 存储活跃的群组监听
|
|||
|
|
|
|||
|
|
# 加载配置
|
|||
|
|
self.load_config()
|
|||
|
|
|
|||
|
|
# 创建界面
|
|||
|
|
self.create_widgets()
|
|||
|
|
|
|||
|
|
# 绑定关闭事件
|
|||
|
|
self.root.protocol("WM_DELETE_WINDOW", self.on_closing)
|
|||
|
|
|
|||
|
|
def load_config(self):
|
|||
|
|
"""加载配置"""
|
|||
|
|
if os.path.exists(self.config_file):
|
|||
|
|
try:
|
|||
|
|
with open(self.config_file, "r", encoding="utf-8") as f:
|
|||
|
|
self.config = json.load(f)
|
|||
|
|
except:
|
|||
|
|
self.config = self.get_default_config()
|
|||
|
|
else:
|
|||
|
|
self.config = self.get_default_config()
|
|||
|
|
|
|||
|
|
def get_default_config(self):
|
|||
|
|
"""获取默认配置"""
|
|||
|
|
return {
|
|||
|
|
"target_groups": [],
|
|||
|
|
"bailian_api_url": "https://dashscope.aliyuncs.com/api/v1/services/aigc/text-generation/generation",
|
|||
|
|
"api_key": "",
|
|||
|
|
"output_file": "jobs_data.json"
|
|||
|
|
}
|
|||
|
|
|
|||
|
|
def save_config(self):
|
|||
|
|
"""保存配置"""
|
|||
|
|
try:
|
|||
|
|
with open(self.config_file, "w", encoding="utf-8") as f:
|
|||
|
|
json.dump(self.config, f, ensure_ascii=False, indent=2)
|
|||
|
|
return True
|
|||
|
|
except Exception as e:
|
|||
|
|
messagebox.showerror("错误", f"保存配置失败: {e}")
|
|||
|
|
return False
|
|||
|
|
|
|||
|
|
def create_widgets(self):
|
|||
|
|
"""创建界面组件"""
|
|||
|
|
# 顶部配置区域
|
|||
|
|
config_frame = ttk.LabelFrame(self.root, text="配置", padding=10)
|
|||
|
|
config_frame.pack(fill=tk.X, padx=10, pady=5)
|
|||
|
|
|
|||
|
|
# 目标群组（支持多个，用逗号分隔）
|
|||
|
|
ttk.Label(config_frame, text="目标群组:").grid(row=0, column=0, sticky=tk.W, pady=5)
|
|||
|
|
ttk.Label(config_frame, text="(多个群组用逗号分隔)", font=("", 8), foreground="gray").grid(
|
|||
|
|
row=0, column=1, sticky=tk.W, padx=5, pady=5
|
|||
|
|
)
|
|||
|
|
|
|||
|
|
self.group_entry = ttk.Entry(config_frame, width=60)
|
|||
|
|
self.group_entry.grid(row=1, column=0, columnspan=2, sticky=tk.W, padx=5, pady=5)
|
|||
|
|
|
|||
|
|
# 从配置加载群组
|
|||
|
|
groups = self.config.get("target_groups", [])
|
|||
|
|
if groups:
|
|||
|
|
self.group_entry.insert(0, ", ".join(groups))
|
|||
|
|
|
|||
|
|
# API密钥（只读显示，打包时配置）
|
|||
|
|
ttk.Label(config_frame, text="API密钥:").grid(row=2, column=0, sticky=tk.W, pady=5)
|
|||
|
|
api_key = self.config.get("api_key", "")
|
|||
|
|
if api_key:
|
|||
|
|
masked_key = api_key[:10] + "..." + api_key[-10:] if len(api_key) > 20 else api_key
|
|||
|
|
ttk.Label(config_frame, text=masked_key, foreground="green").grid(
|
|||
|
|
row=2, column=1, sticky=tk.W, padx=5, pady=5
|
|||
|
|
)
|
|||
|
|
else:
|
|||
|
|
ttk.Label(config_frame, text="未配置", foreground="red").grid(
|
|||
|
|
row=2, column=1, sticky=tk.W, padx=5, pady=5
|
|||
|
|
)
|
|||
|
|
|
|||
|
|
# 保存配置按钮
|
|||
|
|
ttk.Button(config_frame, text="保存群组配置", command=self.save_config_click).grid(
|
|||
|
|
row=1, column=2, padx=10
|
|||
|
|
)
|
|||
|
|
|
|||
|
|
# 控制区域
|
|||
|
|
control_frame = ttk.Frame(self.root, padding=10)
|
|||
|
|
control_frame.pack(fill=tk.X, padx=10)
|
|||
|
|
|
|||
|
|
self.start_btn = ttk.Button(control_frame, text="开始任务", command=self.start_task, width=15)
|
|||
|
|
self.start_btn.pack(side=tk.LEFT, padx=5)
|
|||
|
|
|
|||
|
|
self.stop_btn = ttk.Button(control_frame, text="停止任务", command=self.stop_task,
|
|||
|
|
width=15, state=tk.DISABLED)
|
|||
|
|
self.stop_btn.pack(side=tk.LEFT, padx=5)
|
|||
|
|
|
|||
|
|
ttk.Button(control_frame, text="清空列表", command=self.clear_jobs, width=15).pack(
|
|||
|
|
side=tk.LEFT, padx=5
|
|||
|
|
)
|
|||
|
|
|
|||
|
|
ttk.Button(control_frame, text="导出数据", command=self.export_data, width=15).pack(
|
|||
|
|
side=tk.LEFT, padx=5
|
|||
|
|
)
|
|||
|
|
|
|||
|
|
# 状态栏
|
|||
|
|
status_frame = ttk.Frame(self.root)
|
|||
|
|
status_frame.pack(fill=tk.X, padx=10, pady=5)
|
|||
|
|
|
|||
|
|
ttk.Label(status_frame, text="状态:").pack(side=tk.LEFT)
|
|||
|
|
self.status_label = ttk.Label(status_frame, text="未启动", foreground="gray")
|
|||
|
|
self.status_label.pack(side=tk.LEFT, padx=5)
|
|||
|
|
|
|||
|
|
ttk.Label(status_frame, text="已提取岗位:").pack(side=tk.LEFT, padx=(20, 0))
|
|||
|
|
self.count_label = ttk.Label(status_frame, text="0", foreground="blue")
|
|||
|
|
self.count_label.pack(side=tk.LEFT, padx=5)
|
|||
|
|
|
|||
|
|
# 岗位列表区域
|
|||
|
|
list_frame = ttk.LabelFrame(self.root, text="提取的岗位信息", padding=10)
|
|||
|
|
list_frame.pack(fill=tk.BOTH, expand=True, padx=10, pady=5)
|
|||
|
|
|
|||
|
|
# 创建Treeview
|
|||
|
|
columns = ("group", "job_name", "company", "location", "salary", "contact", "time")
|
|||
|
|
self.tree = ttk.Treeview(list_frame, columns=columns, show="tree headings", height=15)
|
|||
|
|
|
|||
|
|
# 设置列
|
|||
|
|
self.tree.heading("#0", text="序号")
|
|||
|
|
self.tree.heading("group", text="来源群组")
|
|||
|
|
self.tree.heading("job_name", text="岗位名称")
|
|||
|
|
self.tree.heading("company", text="公司")
|
|||
|
|
self.tree.heading("location", text="地点")
|
|||
|
|
self.tree.heading("salary", text="薪资")
|
|||
|
|
self.tree.heading("contact", text="联系方式")
|
|||
|
|
self.tree.heading("time", text="提取时间")
|
|||
|
|
|
|||
|
|
# 设置列宽
|
|||
|
|
self.tree.column("#0", width=50)
|
|||
|
|
self.tree.column("group", width=120)
|
|||
|
|
self.tree.column("job_name", width=150)
|
|||
|
|
self.tree.column("company", width=130)
|
|||
|
|
self.tree.column("location", width=100)
|
|||
|
|
self.tree.column("salary", width=100)
|
|||
|
|
self.tree.column("contact", width=130)
|
|||
|
|
self.tree.column("time", width=140)
|
|||
|
|
|
|||
|
|
# 滚动条
|
|||
|
|
scrollbar = ttk.Scrollbar(list_frame, orient=tk.VERTICAL, command=self.tree.yview)
|
|||
|
|
self.tree.configure(yscrollcommand=scrollbar.set)
|
|||
|
|
|
|||
|
|
self.tree.pack(side=tk.LEFT, fill=tk.BOTH, expand=True)
|
|||
|
|
scrollbar.pack(side=tk.RIGHT, fill=tk.Y)
|
|||
|
|
|
|||
|
|
# 双击查看详情
|
|||
|
|
self.tree.bind("<Double-1>", self.show_job_detail)
|
|||
|
|
|
|||
|
|
# 日志区域
|
|||
|
|
log_frame = ttk.LabelFrame(self.root, text="运行日志", padding=10)
|
|||
|
|
log_frame.pack(fill=tk.BOTH, expand=True, padx=10, pady=5)
|
|||
|
|
|
|||
|
|
self.log_text = scrolledtext.ScrolledText(log_frame, height=8, wrap=tk.WORD)
|
|||
|
|
self.log_text.pack(fill=tk.BOTH, expand=True)
|
|||
|
|
|
|||
|
|
# 加载已有数据
|
|||
|
|
self.load_existing_jobs()
|
|||
|
|
|
|||
|
|
def log(self, message):
|
|||
|
|
"""添加日志"""
|
|||
|
|
timestamp = datetime.now().strftime("%H:%M:%S")
|
|||
|
|
log_msg = f"[{timestamp}] {message}\n"
|
|||
|
|
self.log_text.insert(tk.END, log_msg)
|
|||
|
|
self.log_text.see(tk.END)
|
|||
|
|
self.root.update()
|
|||
|
|
|
|||
|
|
def save_config_click(self):
|
|||
|
|
"""保存配置按钮点击"""
|
|||
|
|
groups_text = self.group_entry.get().strip()
|
|||
|
|
|
|||
|
|
if not groups_text:
|
|||
|
|
messagebox.showwarning("警告", "请输入至少一个目标群组名称")
|
|||
|
|
return
|
|||
|
|
|
|||
|
|
# 解析群组列表（支持逗号、分号、换行分隔）
|
|||
|
|
import re
|
|||
|
|
groups = re.split(r'[,，;；\n]+', groups_text)
|
|||
|
|
groups = [g.strip() for g in groups if g.strip()]
|
|||
|
|
|
|||
|
|
if not groups:
|
|||
|
|
messagebox.showwarning("警告", "请输入至少一个有效的群组名称")
|
|||
|
|
return
|
|||
|
|
|
|||
|
|
self.config["target_groups"] = groups
|
|||
|
|
|
|||
|
|
if self.save_config():
|
|||
|
|
messagebox.showinfo("成功", f"已保存 {len(groups)} 个群组配置:\n" + "\n".join(f"- {g}" for g in groups))
|
|||
|
|
self.log(f"配置已保存: {len(groups)} 个群组")
|
|||
|
|
|
|||
|
|
def start_task(self):
|
|||
|
|
"""开始任务"""
|
|||
|
|
# 验证配置
|
|||
|
|
groups_text = self.group_entry.get().strip()
|
|||
|
|
api_key = self.config.get("api_key", "")
|
|||
|
|
|
|||
|
|
if not groups_text:
|
|||
|
|
messagebox.showwarning("警告", "请输入目标群组名称")
|
|||
|
|
return
|
|||
|
|
|
|||
|
|
if not api_key:
|
|||
|
|
messagebox.showerror("错误", "API密钥未配置，请在config.json中配置后重新打包")
|
|||
|
|
return
|
|||
|
|
|
|||
|
|
# 解析群组列表
|
|||
|
|
import re
|
|||
|
|
groups = re.split(r'[,，;；\n]+', groups_text)
|
|||
|
|
groups = [g.strip() for g in groups if g.strip()]
|
|||
|
|
|
|||
|
|
if not groups:
|
|||
|
|
messagebox.showwarning("警告", "请输入至少一个有效的群组名称")
|
|||
|
|
return
|
|||
|
|
|
|||
|
|
# 更新配置
|
|||
|
|
self.config["target_groups"] = groups
|
|||
|
|
self.save_config()
|
|||
|
|
|
|||
|
|
# 启动监听线程
|
|||
|
|
self.is_running = True
|
|||
|
|
self.start_btn.config(state=tk.DISABLED)
|
|||
|
|
self.stop_btn.config(state=tk.NORMAL)
|
|||
|
|
self.status_label.config(text="正在启动...", foreground="orange")
|
|||
|
|
|
|||
|
|
threading.Thread(target=self.run_task, daemon=True).start()
|
|||
|
|
|
|||
|
|
def run_task(self):
|
|||
|
|
"""运行任务（在线程中）"""
|
|||
|
|
# 初始化COM组件（每个线程都需要）
|
|||
|
|
pythoncom.CoInitialize()
|
|||
|
|
|
|||
|
|
try:
|
|||
|
|
self.log("正在连接微信...")
|
|||
|
|
self.wx = WeChat()
|
|||
|
|
self.log(f"✓ 已连接微信，当前用户: {self.wx.nickname}")
|
|||
|
|
|
|||
|
|
# 获取要监听的群组列表
|
|||
|
|
groups = self.config.get("target_groups", [])
|
|||
|
|
if not groups:
|
|||
|
|
self.log("× 错误: 未配置目标群组")
|
|||
|
|
self.root.after(0, lambda: messagebox.showerror("错误", "未配置目标群组"))
|
|||
|
|
self.root.after(0, self.stop_task)
|
|||
|
|
return
|
|||
|
|
|
|||
|
|
# 为每个群组添加监听
|
|||
|
|
success_count = 0
|
|||
|
|
for group_name in groups:
|
|||
|
|
self.log(f"正在添加监听: {group_name}")
|
|||
|
|
|
|||
|
|
# 创建带群组名称的回调函数
|
|||
|
|
def make_callback(gname):
|
|||
|
|
return lambda msg, chat: self.on_message(msg, chat, gname)
|
|||
|
|
|
|||
|
|
result = self.wx.AddListenChat(
|
|||
|
|
nickname=group_name,
|
|||
|
|
callback=make_callback(group_name)
|
|||
|
|
)
|
|||
|
|
|
|||
|
|
if isinstance(result, str) and "失败" in result:
|
|||
|
|
self.log(f"× 添加监听失败: {group_name} - {result}")
|
|||
|
|
else:
|
|||
|
|
self.log(f"✓ 成功监听群组: {group_name}")
|
|||
|
|
self.active_groups[group_name] = result
|
|||
|
|
success_count += 1
|
|||
|
|
|
|||
|
|
if success_count == 0:
|
|||
|
|
self.log("× 错误: 所有群组监听都失败")
|
|||
|
|
self.root.after(0, lambda: messagebox.showerror("错误", "所有群组监听都失败，请检查群组名称"))
|
|||
|
|
self.root.after(0, self.stop_task)
|
|||
|
|
return
|
|||
|
|
|
|||
|
|
self.log(f"✓ 成功监听 {success_count}/{len(groups)} 个群组")
|
|||
|
|
self.root.after(0, lambda: self.status_label.config(
|
|||
|
|
text=f"运行中 ({success_count}个群组)", foreground="green"
|
|||
|
|
))
|
|||
|
|
|
|||
|
|
# 保持运行
|
|||
|
|
while self.is_running:
|
|||
|
|
time.sleep(1)
|
|||
|
|
|
|||
|
|
except Exception as e:
|
|||
|
|
self.log(f"× 错误: {e}")
|
|||
|
|
self.root.after(0, lambda: messagebox.showerror("错误", f"任务执行失败: {e}"))
|
|||
|
|
self.root.after(0, self.stop_task)
|
|||
|
|
finally:
|
|||
|
|
# 清理COM组件
|
|||
|
|
pythoncom.CoUninitialize()
|
|||
|
|
|
|||
|
|
def on_message(self, msg, chat, group_name):
|
|||
|
|
"""消息回调"""
|
|||
|
|
try:
|
|||
|
|
self.log(f"[{group_name}] 收到消息 - 发送者: {msg.sender}, 类型: {msg.type}")
|
|||
|
|
|
|||
|
|
# 只处理文本消息
|
|||
|
|
if msg.type != "text" or not msg.content:
|
|||
|
|
return
|
|||
|
|
|
|||
|
|
self.log(f"[{group_name}] 正在分析消息内容...")
|
|||
|
|
job_info = self.extract_job_info(msg.content)
|
|||
|
|
|
|||
|
|
if job_info and any(job_info.values()):
|
|||
|
|
self.log(f"[{group_name}] ✓ 提取到岗位信息")
|
|||
|
|
|
|||
|
|
# 生成UUID作为岗位ID
|
|||
|
|
job_id = str(uuid.uuid4())
|
|||
|
|
|
|||
|
|
# 添加元数据
|
|||
|
|
job_info["_id"] = job_id
|
|||
|
|
job_info["_metadata"] = {
|
|||
|
|
"source": "wechat_group",
|
|||
|
|
"group_name": group_name,
|
|||
|
|
"sender": msg.sender,
|
|||
|
|
"extract_time": datetime.now().strftime('%Y-%m-%d %H:%M:%S'),
|
|||
|
|
"original_message": msg.content
|
|||
|
|
}
|
|||
|
|
|
|||
|
|
# 保存并显示
|
|||
|
|
self.root.after(0, lambda: self.add_job_to_list(job_info))
|
|||
|
|
self.save_job_data(job_info)
|
|||
|
|
else:
|
|||
|
|
self.log(f"[{group_name}] × 未提取到有效岗位信息")
|
|||
|
|
|
|||
|
|
except Exception as e:
|
|||
|
|
self.log(f"[{group_name}] × 处理消息时出错: {e}")
|
|||
|
|
|
|||
|
|
def extract_job_info(self, message_content):
|
|||
|
|
"""提取岗位信息"""
|
|||
|
|
prompt = f"""请从以下消息中提取招聘岗位信息，并以JSON格式返回。如果消息不包含招聘信息，返回空对象。
|
|||
|
|
|
|||
|
|
要提取的字段：
|
|||
|
|
- job_name: 工作名称
|
|||
|
|
- job_description: 工作描述
|
|||
|
|
- job_location: 工作地点
|
|||
|
|
- salary_min: 月薪最低（数字，单位：元）
|
|||
|
|
- salary_max: 月薪最高（数字，单位：元）
|
|||
|
|
- company_name: 公司名称
|
|||
|
|
- contact_person: 联系人
|
|||
|
|
- contact_info: 联系方式（电话/微信等）
|
|||
|
|
|
|||
|
|
消息内容：
|
|||
|
|
{message_content}
|
|||
|
|
|
|||
|
|
请直接返回JSON格式，不要包含其他说明文字。"""
|
|||
|
|
|
|||
|
|
headers = {
|
|||
|
|
"Content-Type": "application/json",
|
|||
|
|
"Authorization": f"Bearer {self.config['api_key']}"
|
|||
|
|
}
|
|||
|
|
|
|||
|
|
payload = {
|
|||
|
|
"model": "qwen-plus",
|
|||
|
|
"input": {
|
|||
|
|
"messages": [
|
|||
|
|
{
|
|||
|
|
"role": "system",
|
|||
|
|
"content": "你是一个专业的招聘信息提取助手，擅长从文本中提取结构化的岗位信息。"
|
|||
|
|
},
|
|||
|
|
{
|
|||
|
|
"role": "user",
|
|||
|
|
"content": prompt
|
|||
|
|
}
|
|||
|
|
]
|
|||
|
|
},
|
|||
|
|
"parameters": {
|
|||
|
|
"result_format": "message"
|
|||
|
|
}
|
|||
|
|
}
|
|||
|
|
|
|||
|
|
try:
|
|||
|
|
response = requests.post(
|
|||
|
|
self.config["bailian_api_url"],
|
|||
|
|
headers=headers,
|
|||
|
|
json=payload,
|
|||
|
|
timeout=30
|
|||
|
|
)
|
|||
|
|
response.raise_for_status()
|
|||
|
|
result = response.json()
|
|||
|
|
|
|||
|
|
if "output" in result and "choices" in result["output"]:
|
|||
|
|
content = result["output"]["choices"][0]["message"]["content"]
|
|||
|
|
try:
|
|||
|
|
job_data = json.loads(content)
|
|||
|
|
return job_data if job_data else None
|
|||
|
|
except json.JSONDecodeError:
|
|||
|
|
return None
|
|||
|
|
return None
|
|||
|
|
|
|||
|
|
except Exception as e:
|
|||
|
|
self.log(f"API调用失败: {e}")
|
|||
|
|
return None
|
|||
|
|
|
|||
|
|
def add_job_to_list(self, job_info):
|
|||
|
|
"""添加岗位到列表"""
|
|||
|
|
self.job_count += 1
|
|||
|
|
|
|||
|
|
# 格式化薪资
|
|||
|
|
salary_min = job_info.get("salary_min", "")
|
|||
|
|
salary_max = job_info.get("salary_max", "")
|
|||
|
|
if salary_min and salary_max:
|
|||
|
|
salary = f"{salary_min}-{salary_max}"
|
|||
|
|
elif salary_min:
|
|||
|
|
salary = f"{salary_min}+"
|
|||
|
|
elif salary_max:
|
|||
|
|
salary = f"<{salary_max}"
|
|||
|
|
else:
|
|||
|
|
salary = "面议"
|
|||
|
|
|
|||
|
|
# 联系方式
|
|||
|
|
contact = job_info.get("contact_person", "")
|
|||
|
|
if job_info.get("contact_info"):
|
|||
|
|
contact += f" {job_info['contact_info']}" if contact else job_info['contact_info']
|
|||
|
|
|
|||
|
|
# 来源群组
|
|||
|
|
group_name = job_info.get("_metadata", {}).get("group_name", "未知")
|
|||
|
|
|
|||
|
|
# 插入到树形视图
|
|||
|
|
self.tree.insert("", 0, text=str(self.job_count), values=(
|
|||
|
|
group_name,
|
|||
|
|
job_info.get("job_name", "未知"),
|
|||
|
|
job_info.get("company_name", "未知"),
|
|||
|
|
job_info.get("job_location", "未知"),
|
|||
|
|
salary,
|
|||
|
|
contact,
|
|||
|
|
job_info.get("_metadata", {}).get("extract_time", "")
|
|||
|
|
), tags=(json.dumps(job_info, ensure_ascii=False),))
|
|||
|
|
|
|||
|
|
# 更新计数
|
|||
|
|
self.count_label.config(text=str(self.job_count))
|
|||
|
|
|
|||
|
|
def save_job_data(self, job_data):
|
|||
|
|
"""保存岗位数据"""
|
|||
|
|
try:
|
|||
|
|
existing_data = []
|
|||
|
|
if os.path.exists(self.output_file):
|
|||
|
|
with open(self.output_file, "r", encoding="utf-8") as f:
|
|||
|
|
existing_data = json.load(f)
|
|||
|
|
|
|||
|
|
existing_data.append(job_data)
|
|||
|
|
|
|||
|
|
with open(self.output_file, "w", encoding="utf-8") as f:
|
|||
|
|
json.dump(existing_data, f, ensure_ascii=False, indent=2)
|
|||
|
|
except Exception as e:
|
|||
|
|
self.log(f"保存数据失败: {e}")
|
|||
|
|
|
|||
|
|
def load_existing_jobs(self):
|
|||
|
|
"""加载已有岗位数据"""
|
|||
|
|
if not os.path.exists(self.output_file):
|
|||
|
|
return
|
|||
|
|
|
|||
|
|
try:
|
|||
|
|
with open(self.output_file, "r", encoding="utf-8") as f:
|
|||
|
|
jobs = json.load(f)
|
|||
|
|
|
|||
|
|
for job in jobs:
|
|||
|
|
self.add_job_to_list(job)
|
|||
|
|
|
|||
|
|
if jobs:
|
|||
|
|
self.log(f"已加载 {len(jobs)} 条历史岗位数据")
|
|||
|
|
except Exception as e:
|
|||
|
|
self.log(f"加载历史数据失败: {e}")
|
|||
|
|
|
|||
|
|
def show_job_detail(self, event):
|
|||
|
|
"""显示岗位详情"""
|
|||
|
|
selection = self.tree.selection()
|
|||
|
|
if not selection:
|
|||
|
|
return
|
|||
|
|
|
|||
|
|
item = self.tree.item(selection[0])
|
|||
|
|
tags = item.get("tags", ())
|
|||
|
|
if not tags:
|
|||
|
|
return
|
|||
|
|
|
|||
|
|
try:
|
|||
|
|
job_info = json.loads(tags[0])
|
|||
|
|
|
|||
|
|
# 创建详情窗口
|
|||
|
|
detail_win = tk.Toplevel(self.root)
|
|||
|
|
detail_win.title("岗位详情")
|
|||
|
|
detail_win.geometry("600x550")
|
|||
|
|
|
|||
|
|
text = scrolledtext.ScrolledText(detail_win, wrap=tk.WORD, padx=10, pady=10)
|
|||
|
|
text.pack(fill=tk.BOTH, expand=True)
|
|||
|
|
|
|||
|
|
# 显示详情
|
|||
|
|
text.insert(tk.END, f"岗位ID: {job_info.get('_id', '未知')}\n\n")
|
|||
|
|
text.insert(tk.END, f"岗位名称: {job_info.get('job_name', '未知')}\n\n")
|
|||
|
|
text.insert(tk.END, f"公司名称: {job_info.get('company_name', '未知')}\n\n")
|
|||
|
|
text.insert(tk.END, f"工作地点: {job_info.get('job_location', '未知')}\n\n")
|
|||
|
|
|
|||
|
|
salary_min = job_info.get("salary_min", "")
|
|||
|
|
salary_max = job_info.get("salary_max", "")
|
|||
|
|
if salary_min or salary_max:
|
|||
|
|
text.insert(tk.END, f"薪资范围: {salary_min}-{salary_max} 元\n\n")
|
|||
|
|
|
|||
|
|
if job_info.get("job_description"):
|
|||
|
|
text.insert(tk.END, f"工作描述:\n{job_info['job_description']}\n\n")
|
|||
|
|
|
|||
|
|
if job_info.get("contact_person"):
|
|||
|
|
text.insert(tk.END, f"联系人: {job_info['contact_person']}\n")
|
|||
|
|
if job_info.get("contact_info"):
|
|||
|
|
text.insert(tk.END, f"联系方式: {job_info['contact_info']}\n\n")
|
|||
|
|
|
|||
|
|
if "_metadata" in job_info:
|
|||
|
|
meta = job_info["_metadata"]
|
|||
|
|
text.insert(tk.END, "=" * 50 + "\n")
|
|||
|
|
text.insert(tk.END, f"来源群组: {meta.get('group_name', '未知')}\n")
|
|||
|
|
text.insert(tk.END, f"发送者: {meta.get('sender', '未知')}\n")
|
|||
|
|
text.insert(tk.END, f"提取时间: {meta.get('extract_time', '未知')}\n\n")
|
|||
|
|
|
|||
|
|
if meta.get("original_message"):
|
|||
|
|
text.insert(tk.END, "原始消息:\n")
|
|||
|
|
text.insert(tk.END, meta["original_message"])
|
|||
|
|
|
|||
|
|
text.config(state=tk.DISABLED)
|
|||
|
|
|
|||
|
|
except Exception as e:
|
|||
|
|
messagebox.showerror("错误", f"显示详情失败: {e}")
|
|||
|
|
|
|||
|
|
def stop_task(self):
|
|||
|
|
"""停止任务"""
|
|||
|
|
if self.wx and self.is_running:
|
|||
|
|
self.log("正在停止监听...")
|
|||
|
|
self.is_running = False
|
|||
|
|
try:
|
|||
|
|
self.wx.StopListening()
|
|||
|
|
except:
|
|||
|
|
pass
|
|||
|
|
self.wx = None
|
|||
|
|
|
|||
|
|
self.start_btn.config(state=tk.NORMAL)
|
|||
|
|
self.stop_btn.config(state=tk.DISABLED)
|
|||
|
|
self.status_label.config(text="已停止", foreground="gray")
|
|||
|
|
self.log("任务已停止")
|
|||
|
|
|
|||
|
|
def clear_jobs(self):
|
|||
|
|
"""清空岗位列表"""
|
|||
|
|
if messagebox.askyesno("确认", "确定要清空所有岗位数据吗？"):
|
|||
|
|
# 清空树形视图
|
|||
|
|
for item in self.tree.get_children():
|
|||
|
|
self.tree.delete(item)
|
|||
|
|
|
|||
|
|
# 删除数据文件
|
|||
|
|
if os.path.exists(self.output_file):
|
|||
|
|
os.remove(self.output_file)
|
|||
|
|
|
|||
|
|
self.job_count = 0
|
|||
|
|
self.count_label.config(text="0")
|
|||
|
|
self.log("已清空所有岗位数据")
|
|||
|
|
|
|||
|
|
def export_data(self):
|
|||
|
|
"""导出数据"""
|
|||
|
|
if not os.path.exists(self.output_file):
|
|||
|
|
messagebox.showinfo("提示", "暂无数据可导出")
|
|||
|
|
return
|
|||
|
|
|
|||
|
|
try:
|
|||
|
|
export_file = f"jobs_export_{datetime.now().strftime('%Y%m%d_%H%M%S')}.json"
|
|||
|
|
with open(self.output_file, "r", encoding="utf-8") as f:
|
|||
|
|
data = f.read()
|
|||
|
|
with open(export_file, "w", encoding="utf-8") as f:
|
|||
|
|
f.write(data)
|
|||
|
|
|
|||
|
|
messagebox.showinfo("成功", f"数据已导出到: {export_file}")
|
|||
|
|
self.log(f"数据已导出到: {export_file}")
|
|||
|
|
except Exception as e:
|
|||
|
|
messagebox.showerror("错误", f"导出失败: {e}")
|
|||
|
|
|
|||
|
|
def on_closing(self):
|
|||
|
|
"""关闭窗口"""
|
|||
|
|
if self.is_running:
|
|||
|
|
if messagebox.askyesno("确认", "任务正在运行，确定要退出吗？"):
|
|||
|
|
self.stop_task()
|
|||
|
|
self.root.destroy()
|
|||
|
|
else:
|
|||
|
|
self.root.destroy()
|
|||
|
|
|
|||
|
|
|
|||
|
|
def main():
|
|||
|
|
root = tk.Tk()
|
|||
|
|
app = JobExtractorGUI(root)
|
|||
|
|
root.mainloop()
|
|||
|
|
|
|||
|
|
|
|||
|
|
if __name__ == "__main__":
|
|||
|
|
main()
|