knowledge_search/run_enhanced_search.py

#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
增强搜索系统V2 - 快速启动脚本
"""

import os
import sys
import logging
from enhanced_search_v2 import EnhancedSearchV2

# 配置日志
logging.basicConfig(
    level=logging.INFO,
    format='%(asctime)s - %(name)s - %(levelname)s - %(message)s',
    handlers=[
        logging.FileHandler('search_execution.log', encoding='utf-8'),
        logging.StreamHandler()
    ]
)

logger = logging.getLogger(__name__)


def check_files_exist():
    """检查必需文件是否存在"""
    required_files = [
        '69114f150000000007001f30_how copy.json',
        'dimension_associations_analysis.json',
        'optimized_clustered_data_gemini-3-pro-preview.json'
    ]

    missing_files = []
    for file in required_files:
        if not os.path.exists(file):
            missing_files.append(file)

    if missing_files:
        logger.error("缺少以下必需文件：")
        for file in missing_files:
            logger.error(f"  - {file}")
        return False

    return True


def check_api_key():
    """检查API密钥"""
    api_key = os.getenv('OPENROUTER_API_KEY')
    if not api_key:
        logger.error("未找到OPENROUTER_API_KEY环境变量")
        logger.error("请设置环境变量: export OPENROUTER_API_KEY='your_key'")
        return False

    logger.info(f"已找到API密钥: {api_key[:10]}...")
    return True


def main():
    """主函数"""
    print("=" * 80)
    print("增强搜索系统 V2")
    print("=" * 80)
    print()

    # 检查环境
    logger.info("检查运行环境...")

    if not check_api_key():
        logger.error("环境检查失败：API密钥未设置")
        sys.exit(1)

    if not check_files_exist():
        logger.error("环境检查失败：缺少必需文件")
        sys.exit(1)

    logger.info("环境检查通过 ✓")
    print()

    # 初始化系统
    logger.info("初始化搜索系统...")
    try:
        system = EnhancedSearchV2(
            how_json_path='69114f150000000007001f30_how copy.json',
            dimension_associations_path='dimension_associations_analysis.json',
            optimized_clustered_data_path='optimized_clustered_data_gemini-3-pro-preview.json',
            output_dir='output_v2'
        )
        logger.info("系统初始化成功 ✓")
        print()
    except Exception as e:
        logger.error(f"系统初始化失败: {e}")
        sys.exit(1)

    # 运行完整流程
    logger.info("=" * 80)
    logger.info("开始执行7阶段搜索流程")
    logger.info("=" * 80)
    print()

    try:
        final_results = system.run_full_pipeline()

        print()
        logger.info("=" * 80)
        logger.info("搜索流程执行完成！")
        logger.info("=" * 80)
        print()

        # 显示结果摘要
        logger.info("输出文件位置: output_v2/")
        logger.info("- stage1_filtered_features.json")
        logger.info("- stage2_associations.json")
        logger.info("- stage3_features.json")
        logger.info("- stage4_with_llm_scores.json")
        logger.info("- stage5_with_search_results.json")
        logger.info("- stage6_with_evaluations.json")
        logger.info("- stage7_final_results.json")
        print()

        # 统计信息
        logger.info("执行统计：")
        stage1_count = len(final_results.get('结果', []))
        logger.info(f"- 筛选出的特征数: {stage1_count}")

        # 统计搜索词数量
        total_search_words = 0
        total_extensions = 0
        for result in final_results.get('结果', []):
            for assoc in result.get('找到的关联', []):
                for feature in assoc.get('特征列表', []):
                    if feature.get('search_word'):
                        total_search_words += 1
                    total_extensions += len(feature.get('extended_searches', []))

        logger.info(f"- 生成的搜索词数: {total_search_words}")
        logger.info(f"- 扩展搜索数: {total_extensions}")

        return 0

    except KeyboardInterrupt:
        logger.warning("\n用户中断执行")
        return 1
    except Exception as e:
        logger.error(f"\n执行过程中出错: {e}", exc_info=True)
        return 1


if __name__ == '__main__':
    sys.exit(main())