工具集成自动化
GEO工作流端到端自动化

# 事件驱动架构实现
import json
from datetime import datetime
from typing import Dict, Callable
import requests

# 事件总线（基于Webhook的轻量级实现）
class EventBus:
    def __init__(self):
        self.subscribers: Dict[str, list] = {}

    def subscribe(self, event_type: str, handler: Callable):
        '''订阅事件'''
        if event_type not in self.subscribers:
            self.subscribers[event_type] = []
        self.subscribers[event_type].append(handler)
        print(f'已订阅事件: {event_type}')

    def publish(self, event_type: str, data: dict):
        '''发布事件'''
        event = {
            'event_id': f'{event_type}_{int(datetime.now().timestamp())}',
            'event_type': event_type,
            'source': 'geo_pipeline',
            'timestamp': datetime.now().isoformat(),
            'data': data
        }
        print(f'发布事件: {event_type} | ID: {event["event_id"]}')

        handlers = self.subscribers.get(event_type, [])
        for handler in handlers:
            try:
                handler(event)
            except Exception as e:
                print(f'事件处理失败: {event_type} -> {handler.__name__}: {e}')

# 创建全局事件总线
event_bus = EventBus()

# 事件处理器定义
def on_content_published(event):
    '''内容发布事件处理器'''
    data = event['data']
    print(f'  -> 推送URL到百度: {data["url"]}')
    print(f'  -> 更新sitemap')
    print(f'  -> 通知AI爬虫监控开始追踪: {data["url"]}')

def on_detection_completed(event):
    '''检测完成事件处理器'''
    data = event['data']
    mention_rate = data.get('mention_rate', 0)
    print(f'  -> 更新效果仪表盘: 提及率 {mention_rate:.1%}')
    print(f'  -> 检查是否触发告警')
    if mention_rate < 0.05:
        event_bus.publish('anomaly_detected', {
            'type': 'low_mention_rate',
            'value': mention_rate,
            'keyword': data.get('keyword')
        })

def on_anomaly_detected(event):
    '''异常检测事件处理器'''
    data = event['data']
    print(f'  -> 发送告警通知: {data["type"]}')
    print(f'  -> 启动自动诊断流程')
    # 可以触发Slack通知、邮件告警等

# 注册事件处理器
event_bus.subscribe('content_published', on_content_published)
event_bus.subscribe('detection_completed', on_detection_completed)
event_bus.subscribe('anomaly_detected', on_anomaly_detected)

# 模拟事件触发
event_bus.publish('content_published', {
    'url': 'https://your-domain.com/article/geo-guide',
    'title': 'GEO优化指南',
    'keyword': 'GEO优化'
})

event_bus.publish('detection_completed', {
    'keyword': 'GEO优化',
    'platform': 'chatgpt',
    'mention_rate': 0.03,
    'brand_mentioned': False
})

# 端到端GEO自动化工作流（基于事件驱动）

class GEOWorkflowOrchestrator:
    '''GEO工作流编排器'''

    def __init__(self, event_bus):
        self.event_bus = event_bus
        self.setup_workflow()

    def setup_workflow(self):
        '''注册工作流各环节的事件处理器'''

        # 环节1: 内容需求分析 → 内容生成
        self.event_bus.subscribe('content_needed', self._generate_content)

        # 环节2: 内容生成 → 质量审核
        self.event_bus.subscribe('content_generated', self._review_content)

        # 环节3: 审核通过 → CMS发布
        self.event_bus.subscribe('content_approved', self._publish_content)

        # 环节4: 发布完成 → URL推送+爬虫监控启动
        self.event_bus.subscribe('content_published', self._push_and_monitor)

        # 环节5: AI搜索检测 → 效果分析
        self.event_bus.subscribe('detection_completed', self._analyze_effect)

        # 环节6: 效果分析 → 报告+优化建议
        self.event_bus.subscribe('effect_analyzed', self._generate_report)

    def _generate_content(self, event):
        '''环节1: 调用LLM生成内容'''
        data = event['data']
        print(f'[工作流] 生成内容: {data["keyword"]}')
        # 调用LLM API生成内容
        # content = generate_geo_content(api_key, data['keyword'], ...)
        # 质检通过后发布事件
        self.event_bus.publish('content_generated', {
            'keyword': data['keyword'],
            'content': '生成的内容...',
            'quality_score': 85
        })

    def _review_content(self, event):
        '''环节2: 自动质量审核'''
        data = event['data']
        print(f'[工作流] 审核内容: 质量分 {data["quality_score"]}')
        if data['quality_score'] >= 80:
            self.event_bus.publish('content_approved', {
                'keyword': data['keyword'],
                'content': data['content']
            })
        else:
            print(f'[工作流] 内容质量不达标，退回重新生成')
            self.event_bus.publish('content_needed', {
                'keyword': data['keyword'],
                'retry': True
            })

    def _publish_content(self, event):
        '''环节3: CMS发布'''
        data = event['data']
        print(f'[工作流] 发布到CMS: {data["keyword"]}')
        # 调用CMS API发布
        self.event_bus.publish('content_published', {
            'url': f'https://your-domain.com/article/{data["keyword"]}',
            'keyword': data['keyword']
        })

    def _push_and_monitor(self, event):
        '''环节4: URL推送+启动监控'''
        data = event['data']
        print(f'[工作流] 推送URL: {data["url"]}')
        print(f'[工作流] 注册AI爬虫监控: {data["url"]}')
        # 2周后触发AI搜索检测
        self.event_bus.publish('detection_scheduled', {
            'url': data['url'],
            'keyword': data['keyword'],
            'check_after_days': 14
        })

    def _analyze_effect(self, event):
        '''环节5: 效果分析'''
        data = event['data']
        print(f'[工作流] 分析效果: {data["keyword"]}')
        self.event_bus.publish('effect_analyzed', {
            'keyword': data['keyword'],
            'mention_rate': data.get('mention_rate', 0),
            'recommendation': '增加该主题内容产量' if data.get('mention_rate', 0) > 0.15 else '优化内容结构'
        })

    def _generate_report(self, event):
        '''环节6: 生成报告和优化建议'''
        data = event['data']
        print(f'[工作流] 生成报告: {data["keyword"]}')
        print(f'[工作流] 优化建议: {data["recommendation"]}')

# 启动工作流
# orchestrator = GEOWorkflowOrchestrator(event_bus)
# event_bus.publish('content_needed', {'keyword': '企业CRM选型指南'})

# 持续优化闭环

class ContinuousOptimizationLoop:
    '''GEO持续优化闭环'''

    def __init__(self, event_bus):
        self.event_bus = event_bus
        self.optimization_history = []  # 优化历史记录

    def run_cycle(self, keywords):
        '''执行一轮优化闭环'''
        # Step 1: 检测
        print('=== 优化闭环启动 ===')
        detection_results = self._detect(keywords)

        # Step 2: 分析
        analysis = self._analyze(detection_results)

        # Step 3: 优化建议
        recommendations = self._recommend(analysis)

        # Step 4: 执行优化（需人工确认）
        for rec in recommendations:
            print(f'  优化建议: {rec["action"]} | 关键词: {rec["keyword"]} | 原因: {rec["reason"]}')

        # Step 5: 记录历史
        self.optimization_history.append({
            'date': datetime.now().isoformat(),
            'keywords_checked': len(keywords),
            'issues_found': len(recommendations),
            'analysis': analysis
        })

        return recommendations

    def _detect(self, keywords):
        '''Step 1: 检测AI搜索可见度'''
        print(f'[检测] {len(keywords)}个关键词')
        results = []
        for kw in keywords:
            results.append({
                'keyword': kw,
                'mention_rate': 0.12,  # 实际从检测系统获取
                'sentiment': 'neutral',
                'last_checked': datetime.now().isoformat()
            })
        return results

    def _analyze(self, results):
        '''Step 2: 分析检测结果'''
        issues = []
        for r in results:
            if r['mention_rate'] < 0.05:
                issues.append({
                    'keyword': r['keyword'],
                    'issue': 'mention_rate_too_low',
                    'value': r['mention_rate']
                })
            elif r['sentiment'] == 'negative':
                issues.append({
                    'keyword': r['keyword'],
                    'issue': 'negative_sentiment',
                    'value': r['sentiment']
                })
        print(f'[分析] 发现{len(issues)}个问题')
        return issues

    def _recommend(self, issues):
        '''Step 3: 生成优化建议'''
        recommendations = []
        for issue in issues:
            if issue['issue'] == 'mention_rate_too_low':
                recommendations.append({
                    'keyword': issue['keyword'],
                    'action': '创建新内容或优化现有内容',
                    'reason': f'提及率仅{issue["value"]:.1%}，低于5%阈值',
                    'priority': 'high'
                })
            elif issue['issue'] == 'negative_sentiment':
                recommendations.append({
                    'keyword': issue['keyword'],
                    'action': '审查并修复负面信息来源',
                    'reason': 'AI回答中出现负面情感',
                    'priority': 'critical'
                })
        return recommendations

# 运行优化闭环
# loop = ContinuousOptimizationLoop(event_bus)
# recommendations = loop.run_cycle(['企业CRM', 'SaaS定价', 'API集成'])