zbparse/flask_app/main/基础信息整合.py

from json_utils import clean_json_string, nest_json_under_key,rename_outer_key, combine_json_results
from flask_app.main.投标人须知正文提取指定内容 import extract_from_notice
from flask_app.main.判断是否分包等 import judge_whether_main, read_questions_from_judge
from flask_app.main.多线程提问 import read_questions_from_file, multi_threading
from flask_app.main.通义千问long import upload_file
def combine_basic_info(baseinfo_list):
    combined_baseinfo_list = []
    key_groups = {
        "招标人/代理信息": ["招标人","招标人联系方式", "招标代理机构","招标代理机构联系方式"],
        "项目信息": ["工程名称", "招标编号","工程概况","招标范围","招标控制价","投标竞争下浮率","是否接受联合体投标"],
        "关键时间/内容":["投标文件递交截止日期","递交方式","投标人要求澄清招标文件的截止时间","投标有效期","评标结果公示媒介"],
        "保证金相关":['质量保证金','退还投标保证金'],
        "其他信息":["重新招标、不再招标和终止招标","是否退还投标文件","费用承担"]
    }
    # 将所有基础信息合并到一个字典中
    combined_data = {}
    relevant_keys_detected = set()

    # 预处理以决定哪些键名将被使用
    for baseinfo in baseinfo_list:
        json_data = clean_json_string(baseinfo)
        combined_data.update(json_data)
        relevant_keys_detected.update(json_data.keys())
        # for key in relevant_keys.keys():
        #     if key in json_data:
        #         relevant_keys[key] = True

        # 根据检测到的键动态调整 key_groups
    dynamic_key_handling(key_groups, relevant_keys_detected)

    # 打印 key_groups 的内容检查它们是否被正确更新
    print("Updated key_groups after dynamic handling:")
    print(key_groups)

        # 使用合并后的字典创建最终输出
    for group_name, keys in key_groups.items():
        group_data = {key: combined_data.get(key, "未提供") for key in keys}
        combined_json = nest_json_under_key(group_data, group_name)
        combined_baseinfo_list.append(combined_json)

    return combined_baseinfo_list

def dynamic_key_handling(key_groups, detected_keys):
    # 检查和调整键组配置
    for key in detected_keys:
        if "投标保证金" in key or "履约保证金" in key:
            key_groups["保证金相关"].append(key)
        elif "联合体投标要求" in key:
            key_groups["项目信息"].append(key)
        elif "分包" in key:
            key_groups["项目信息"].append(key)
        elif "踏勘现场" in key:
            key_groups["其他信息"].append(key)
        elif "投标预备会" in key:
            key_groups["其他信息"].append(key)
        elif "偏离" in key:
            key_groups["其他信息"].append(key)

def judge_consortium_bidding(baseinfo_list):
    for baseinfo in baseinfo_list:
        json_data = clean_json_string(baseinfo)
        # 检查 "是否接受联合体投标" 键是否存在且其值为 "是"
        if json_data.get("是否接受联合体投标") == "是":
            return True
    return False
def project_basic_info(knowledge_name,truncate0,output_folder,clause_path):    #投标人须知前附表
    # 调用大模型回答项目基础信息
    print("starting基础信息...")
    baseinfo_list = []
    # baseinfo_file_path='../static/提示词/前两章提问总结.txt'
    baseinfo_file_path = 'flask_app/static/提示词/前两章提问总结.txt'  # 替换为你的txt文件路径
    questions = read_questions_from_file(baseinfo_file_path)
    res1 = multi_threading(questions, knowledge_name)
    for _, response in res1:  # _占位，代表ques;response[0]也是ques;response[1]是ans
        try:
            if response and len(response) > 1:  # 检查response存在且有至少两个元素
                baseinfo_list.append(response[1])
            else:
                print(f"Warning: Missing or incomplete response data for query index {_}.")
        except Exception as e:
            print(f"Error processing response for query index {_}: {e}")
    print("basic信息done...")
    # 判断是否分包、是否需要递交投标保证金等
    chosen_numbers, merged = judge_whether_main(truncate0,output_folder)
    baseinfo_list.append(merged)
    # judge_file_path = '../static/提示词/是否相关问题.txt'
    judge_file_path ='flask_app/static/提示词/是否相关问题.txt'
    judge_questions = read_questions_from_judge(judge_file_path, chosen_numbers)


    judge_consortium = judge_consortium_bidding(baseinfo_list)     #通过招标公告判断是否接受联合体投标
    if judge_consortium:
        judge_consortium_question = "该招标文件对于联合体投标的要求是怎样的，请按json格式给我提供信息，外层键名为'联合体投标要求'。"
        judge_questions.append(judge_consortium_question)

    file_id=upload_file(truncate0)
    res2 = multi_threading(judge_questions, "",file_id,2)                  #调用千问-long
    if not res2:
        print("errror!")
    else:
        # 打印结果
        for question, response in res2:
            baseinfo_list.append(response)
    # for _, response in res2:  # _占位，代表ques;response[0]也是ques;response[1]是ans     #调用百炼rag
    #     try:
    #         if response and len(response) > 1:  # 检查response存在且有至少两个元素
    #             baseinfo_list.append(response[1])
    #         else:
    #             print(f"Warning: Missing or incomplete response data for query index {_}.")
    #     except Exception as e:
    #         print(f"Error processing response for query index {_}: {e}")

    rebidding_situation = extract_from_notice(clause_path, 3)                     #"重新招标, 不再招标和终止招标"需从投标人须知正文提取

    update_json=rename_outer_key(rebidding_situation,"重新招标、不再招标和终止招标")
    baseinfo_list.append(update_json)

    update_baseinfo_list=combine_basic_info(baseinfo_list)    #整合基础信息核心代码

    baseinfo_combined_res = combine_json_results(update_baseinfo_list)  # 返回值是字典
    print("基础信息done")
    return nest_json_under_key(baseinfo_combined_res, "基础信息")    #返回值是json字符串


if __name__ == "__main__":
    knowledge_name = "ztb"
    output_folder="C:\\Users\\Administrator\\Desktop\\招标文件\\test2"
    truncate0="C:\\Users\\Administrator\\Desktop\\招标文件\\test2\\zbtest10_tobidders_notice_table.pdf"
    clause_path="C:\\Users\\Administrator\\Desktop\\招标文件\\output1\\clause1.json"
    res=project_basic_info(knowledge_name,truncate0,output_folder,clause_path)
    print(res)