zbparse/flask_app/工程标/基础信息整合工程标.py

import copy
import json
import time
import concurrent.futures
from flask_app.general.json_utils import clean_json_string, add_outer_key
from flask_app.general.通用功能函数 import process_judge_questions, aggregate_basic_info
from flask_app.general.投标人须知正文提取指定内容 import extract_from_notice
from flask_app.old_version.判断是否分包等_old import merge_json_to_list
from flask_app.general.llm.多线程提问 import read_questions_from_file, multi_threading
from flask_app.general.llm.通义千问long import upload_file

def update_baseinfo_lists(baseinfo_list1, baseinfo_list2):
    # 创建一个字典，用于存储 baseinfo_list1 中的所有键值对
    combined_dict = {}
    for item in baseinfo_list1:
        combined_dict.update(item)

    # 使用 baseinfo_list2 中的信息更新 combined_dict
    for item in baseinfo_list2:
        for key, value in item.items():
            if key in combined_dict:
                combined_dict[key] = value

    # 重新构建 baseinfo_list1，保持原有的结构
    updated_list = []
    for item in baseinfo_list1:
        updated_item = {}
        for key in item:
            updated_item[key] = combined_dict[key]
        updated_list.append(updated_item)

    return updated_list

#先不带投标人须知正文，如果是未知，再直接问正文，
def process_baseinfo_list(baseinfo_list, merged_baseinfo_path):
    questions_list = []
    for item in baseinfo_list:
        # print(json.dumps(item, ensure_ascii=False, indent=4))
        for key, value in item.items():
            if value == "未知" or (isinstance(value, dict) and all(v == "未知" for v in value.values())):
                question = (
                    f"根据该招标文件中的信息，{key}的内容是怎样的？请按json格式给我提供信息，键名是'{key}'，若存在嵌套信息，嵌套内容键名以文件中对应字段命名（或是你对相应要求的总结），而对应键值需要与原文保持一致。注意：默认情况用普通键值对返回结果即可，外层键名为{key}；若原文中未提及'{key}'相关内容，在键值中填'未知'。"
                )
                questions_list.append(question)

    if questions_list:
        file_id = upload_file(merged_baseinfo_path)
        baseinfo_results = multi_threading(questions_list, "", file_id, 2)
        return [clean_json_string(res) for _, res in baseinfo_results] if baseinfo_results else []
    else:
        return []

def combine_basic_info(merged_baseinfo_path,merged_baseinfo_path_more, clause_path,invalid_path):
    """
    综合和处理基础信息，生成最终的基础信息字典。

    参数：
    - knowledge_name (str): 知识名称。
    - truncate0 (str): 文件路径。
    - output_folder (str): 输出文件夹路径。
    - clause_path (str): 条款路径。

    返回：
    - dict: 综合后的基础信息。
    """
    # baseinfo_prompt_file_path=r'D:\flask_project\flask_app\static\提示词\基本信息工程标.txt'
    baseinfo_prompt_file_path = 'flask_app/static/提示词/基本信息工程标.txt'
    file_id = upload_file(merged_baseinfo_path)
    questions = read_questions_from_file(baseinfo_prompt_file_path)
    baseinfo_results = multi_threading(questions, "", file_id, 2)
    baseinfo_list1 = [clean_json_string(res) for _, res in baseinfo_results] if baseinfo_results else []
    chosen_numbers, merged = merge_json_to_list(baseinfo_list1.pop())
    baseinfo_list1_copy = copy.deepcopy(baseinfo_list1)
    baseinfo_list1.append(merged)
    # judge_file_path=r'D:\flask_project\flask_app\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'

    with concurrent.futures.ThreadPoolExecutor(max_workers=3) as executor:
        # 提交两个任务
        future1 = executor.submit(process_judge_questions, judge_file_path, chosen_numbers, invalid_path, baseinfo_list1)
        future2 = executor.submit(process_baseinfo_list, baseinfo_list1_copy, merged_baseinfo_path)    #未知的内容再问一次
        future3 = executor.submit(extract_from_notice, merged_baseinfo_path_more, clause_path, 3)  # 新增的多线程任务

        # 等待两个任务完成并获取结果
        future1.result()  # process_judge_questions 直接修改 baseinfo_list1，不需要返回值
        baseinfo_list2 = future2.result()
        rebidding_situation = future3.result()  # 获取提取失败的情况

    updated_baseinfo_list = update_baseinfo_lists(baseinfo_list1, baseinfo_list2)
    update_json = add_outer_key(rebidding_situation, "重新招标、不再招标和终止招标")
    updated_baseinfo_list.append(update_json)
    aggregated_baseinfo = aggregate_basic_info(updated_baseinfo_list)

    return {"基础信息": aggregated_baseinfo}


if __name__ == "__main__":
    start_time = time.time()
    merged_baseinfo_path = r"C:\Users\Administrator\Desktop\工程\test\2022-广东-鹏华基金管理有限公司深圳深业上城办公室装修项目.docx"
    more=r"C:\Users\Administrator\Desktop\工程\test\2022-广东-鹏华基金管理有限公司深圳深业上城办公室装修项目.docx"
    # output_folder="C:\\Users\\Administrator\\Desktop\\招标文件\\special_output"
    tobidders_notice_table = r"D:\flask_project\flask_app\static\output\output1\6f2010ee-d7cd-4787-a26a-2db8233d179a\ztbfile_tobidders_notice_table.pdf"
    # tobidders_notice = r"D:\flask_project\flask_app\static\output\output1\6f2010ee-d7cd-4787-a26a-2db8233d179a\ztbfile_tobidders_notice.pdf"
    clause_path = ""
    invalid_path=r"C:\Users\Administrator\Desktop\工程\test\2022-广东-鹏华基金管理有限公司深圳深业上城办公室装修项目.docx"
    res = combine_basic_info(merged_baseinfo_path,more, clause_path,invalid_path)
    print(json.dumps(res, ensure_ascii=False, indent=4))
    end_time = time.time()
    print("elapsed_time:" + str(end_time - start_time))