戒酒的李白

Add AI-powered Spider Configuration Assistant.

... ... @@ -103,6 +103,39 @@
</div>
</div>
<!-- AI配置助手 -->
<div class="card mb-4">
<div class="card-header">
<h5 class="mb-0">
<i class="fas fa-robot"></i> AI配置助手
</h5>
</div>
<div class="card-body">
<div class="mb-3">
<label for="aiPrompt" class="form-label">用自然语言描述您的爬虫需求</label>
<textarea class="form-control" id="aiPrompt" rows="3"
placeholder="例如:我想爬取最近一周关于人工智能的热门微博,重点关注转发量超过1000的内容,每个话题爬取前5页内容。"></textarea>
</div>
<div class="d-flex justify-content-between align-items-center">
<button class="btn btn-primary" onclick="generateConfig()">
<i class="fas fa-magic"></i> 生成配置
</button>
<div class="form-check">
<input class="form-check-input" type="checkbox" id="autoApply" checked>
<label class="form-check-label" for="autoApply">
自动应用生成的配置
</label>
</div>
</div>
<div id="aiResponse" class="mt-3" style="display: none;">
<div class="alert alert-info">
<h6 class="alert-heading">AI助手建议:</h6>
<p id="aiSuggestion" class="mb-0"></p>
</div>
</div>
</div>
</div>
<!-- 操作按钮 -->
<div class="d-flex justify-content-between mb-5">
<button class="btn btn-primary" onclick="startCrawling()">
... ... @@ -286,6 +319,63 @@
updateCrawlLog(data.message);
}
};
// AI配置生成
async function generateConfig() {
const prompt = document.getElementById('aiPrompt').value.trim();
if (!prompt) {
alert('请输入您的爬虫需求描述!');
return;
}
const aiResponse = document.getElementById('aiResponse');
const aiSuggestion = document.getElementById('aiSuggestion');
try {
const response = await fetch('/api/spider/ai-config', {
method: 'POST',
headers: {
'Content-Type': 'application/json'
},
body: JSON.stringify({ prompt })
});
const data = await response.json();
if (data.success) {
// 显示AI建议
aiSuggestion.textContent = data.suggestion;
aiResponse.style.display = 'block';
// 如果选择自动应用配置
if (document.getElementById('autoApply').checked) {
// 清除现有选择
selectedTopics.clear();
// 应用新的话题
data.config.topics.forEach(topic => {
selectedTopics.add(topic);
});
// 更新参数
document.getElementById('crawlDepth').value = data.config.parameters.crawlDepth;
document.getElementById('interval').value = data.config.parameters.interval;
document.getElementById('maxRetries').value = data.config.parameters.maxRetries;
document.getElementById('timeout').value = data.config.parameters.timeout;
// 更新UI
updateSelectedTopicsList();
// 添加提示
updateCrawlLog('AI配置已自动应用');
}
} else {
throw new Error(data.message);
}
} catch (error) {
aiSuggestion.textContent = '生成配置时出错:' + error.message;
aiResponse.style.display = 'block';
}
}
</script>
</body>
</html>
\ No newline at end of file
... ...
... ... @@ -8,6 +8,8 @@ import asyncio
import websockets
import logging
from spider.spiderData import SpiderData
from openai import OpenAI
from anthropic import Anthropic
# 创建蓝图
spider_bp = Blueprint('spider', __name__)
... ... @@ -210,4 +212,132 @@ async def spider_status_socket():
finally:
websocket_connections.remove(websocket)
except Exception as e:
logger.error(f"WebSocket连接处理失败: {e}")
\ No newline at end of file
logger.error(f"WebSocket连接处理失败: {e}")
def get_ai_client():
"""获取可用的AI客户端"""
# 按优先级尝试不同的AI服务
if os.getenv('ANTHROPIC_API_KEY'):
return {
'type': 'anthropic',
'client': Anthropic(api_key=os.getenv('ANTHROPIC_API_KEY'))
}
elif os.getenv('OPENAI_API_KEY'):
return {
'type': 'openai',
'client': OpenAI(api_key=os.getenv('OPENAI_API_KEY'))
}
else:
raise ValueError("未找到可用的AI API密钥")
def parse_ai_response(response_text):
"""解析AI响应中的JSON配置"""
try:
# 查找JSON内容
start = response_text.find('{')
end = response_text.rfind('}') + 1
if start == -1 or end == 0:
raise ValueError("未找到有效的JSON配置")
json_str = response_text[start:end]
config = json.loads(json_str)
# 验证配置格式
if not isinstance(config.get('topics'), list):
raise ValueError("配置必须包含话题列表")
parameters = config.get('parameters', {})
if not all(key in parameters for key in ['crawlDepth', 'interval', 'maxRetries', 'timeout']):
raise ValueError("配置缺少必要的参数")
# 提取建议文本(JSON之前的部分)
suggestion = response_text[:start].strip()
return config, suggestion
except Exception as e:
raise ValueError(f"解析AI响应失败: {str(e)}")
@spider_bp.route('/api/spider/ai-config', methods=['POST'])
def generate_ai_config():
"""使用AI生成爬虫配置"""
try:
prompt = request.json.get('prompt', '')
if not prompt:
return jsonify({
'success': False,
'message': '请提供爬虫需求描述'
})
# 构建AI提示
system_prompt = """你是一个专业的爬虫配置助手。请根据用户的自然语言描述,生成合适的微博爬虫配置。
配置应包含以下内容:
1. 要爬取的话题列表
2. 爬虫参数(爬取深度、间隔时间、重试次数、超时时间)
请先用通俗易懂的语言解释你的配置建议,然后在最后提供一个JSON格式的具体配置。
注意:
- 爬取深度(crawlDepth)范围:1-10页
- 间隔时间(interval)范围:3-30秒
- 重试次数(maxRetries)范围:1-5次
- 超时时间(timeout)范围:10-60秒
- 所有参数都必须是整数
示例输出格式:
根据您的需求,我建议...
{
"topics": ["话题1", "话题2"],
"parameters": {
"crawlDepth": 5,
"interval": 5,
"maxRetries": 3,
"timeout": 30
}
}"""
# 获取AI客户端
ai = get_ai_client()
try:
if ai['type'] == 'anthropic':
response = ai['client'].messages.create(
model="claude-3-sonnet-20240229",
max_tokens=1000,
messages=[
{"role": "system", "content": system_prompt},
{"role": "user", "content": prompt}
]
)
response_text = response.content[0].text
else: # OpenAI
response = ai['client'].chat.completions.create(
model="gpt-3.5-turbo",
messages=[
{"role": "system", "content": system_prompt},
{"role": "user", "content": prompt}
]
)
response_text = response.choices[0].message.content
# 解析AI响应
config, suggestion = parse_ai_response(response_text)
return jsonify({
'success': True,
'config': config,
'suggestion': suggestion
})
except Exception as e:
logger.error(f"AI服务调用失败: {e}")
return jsonify({
'success': False,
'message': f"AI配置生成失败: {str(e)}"
})
except Exception as e:
logger.error(f"生成配置失败: {e}")
return jsonify({
'success': False,
'message': str(e)
})
\ No newline at end of file
... ...