zbparse/flask_app/general/little_zbparse.py

# -*- encoding:utf-8 -*-
import json
import logging
import os
import time

from flask_app.main.format_change import docx2pdf, pdf2docx
from flask_app.main.json_utils import clean_json_string
from flask_app.main.判断是否分包等 import merge_json_to_list, read_questions_from_judge
from flask_app.main.基础信息整合 import judge_consortium_bidding
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
from flask_app.货物标.基础信息解析main import aggregate_basic_info_goods
from flask_app.货物标.货物标截取pdf import truncate_pdf_specific
from flask_app.main.post_processing import inner_post_processing

def get_global_logger(unique_id):
    if unique_id is None:
        return logging.getLogger()  # 获取默认的日志器
    logger = logging.getLogger(unique_id)
    return logger


logger = None
def get_base_info(baseinfo_file_path):
    file_id = upload_file(baseinfo_file_path)
    # baseinfo_file_path='flask_app/static/提示词/基本信息货物标.txt'
    baseinfo_file_path = 'D:\\flask_project\\flask_app\\static\\提示词\\基本信息货物标.txt'
    questions = read_questions_from_file(baseinfo_file_path)
    more_query = "请你根据招标文件信息，回答以下问题：是否组织踏勘现场？是否召开投标预备会（或投标答疑会）？是否退还投标文件？是否允许分包? 是否需要递交投标保证金（或磋商保证金）？是否需要提交履约保证金（或履约担保）？是否有招标代理服务费（或中标、成交服务费）？请按json格式给我提供信息，键名分别为'是否组织踏勘现场','是否召开投标预备会'（或'是否召开投标答疑会'）,'是否退还投标文件',是否允许分包','是否递交投标保证金'（或'是否递交磋商保证金'）,'是否提交履约保证金','是否有招标代理服务费',键值仅限于'是','否','未知',若存在矛盾信息，请回答'未知'。"
    questions.append(more_query)
    baseinfo_results = multi_threading(questions, "", file_id, 2)  # 1代表使用百炼rag 2代表使用qianwen-long
    baseinfo_list = [clean_json_string(res) for _, res in baseinfo_results] if baseinfo_results else []
    chosen_numbers, merged = merge_json_to_list(baseinfo_list.pop())
    baseinfo_list.append(merged)

    # judge_file_path = 'flask_app/static/提示词/是否相关问题货物标.txt'
    judge_file_path = 'D:\\flask_project\\flask_app\\static\\提示词\\是否相关问题货物标.txt'
    judge_questions = read_questions_from_judge(judge_file_path, chosen_numbers)

    res2 = multi_threading(judge_questions, "", file_id, 2)  # 调用千问-long
    if not res2:
        print("基础信息整合： multi_threading error!")
    else:
        for question, response in res2:
            baseinfo_list.append(clean_json_string(response))
    return baseinfo_list

#货物标
def little_parse_goods(output_folder,file_path):
    files=truncate_pdf_specific(file_path,output_folder)
    baseinfo_list=get_base_info(files[-1])
    aggregated_baseinfo = aggregate_basic_info_goods(baseinfo_list)
    return {"基础信息": aggregated_baseinfo}
def little_parse_engineering(output_folder,downloaded_filepath):
    return True

def little_parse_main(output_folder, file_path, file_type,zb_type,unique_id):
    """
        主解析函数，根据文件类型和招标类型处理文件，并保存结果为JSON文件。

        参数:
            output_folder (str): 输出文件夹路径。
            file_path (str): 待处理文件的路径。
            file_type (int): 文件类型（1: docx, 2: pdf, 3: doc）。
            zb_type (int): 招标类型（2: 货物标, 其他: 工程标）。

        返回:
            str: 保存的JSON文件的路径。
        """
    global logger
    logger = get_global_logger(unique_id)
    # 根据文件类型处理文件路径
    if file_type == 1:  # docx
        docx_path = file_path
        pdf_path = docx2pdf(docx_path)  # 将docx转换为pdf以供后续处理
    elif file_type == 2:  # pdf
        pdf_path = file_path
        # docx_path = pdf2docx(pdf_path)  # 将pdf转换为docx以供上传到知识库
    elif file_type == 3:  # doc
        pdf_path = docx2pdf(file_path)
        # docx_path = pdf2docx(pdf_path)
    else:
        logger.error("Unsupported file type provided. Preprocessing halted.")
        return None
 # 根据招标类型调用相应的解析函数
    if zb_type == 2:  # 货物标
        combined_data = little_parse_goods(output_folder, pdf_path)
        logger.info("Parsed goods data: %s", json.dumps(combined_data, ensure_ascii=False, indent=4))
        res = inner_post_processing(combined_data["基础信息"])
    else:
        combined_data = little_parse_engineering(output_folder, pdf_path)
        logger.info("Parsed engineering data: %s", json.dumps(combined_data, ensure_ascii=False, indent=4))
        res = inner_post_processing(combined_data["基础信息"])
        # 定义保存JSON文件的路径
    json_path = os.path.join(output_folder, "final_result.json")
    # 将结果保存为JSON文件
    with open(json_path, 'w', encoding='utf-8') as json_file:
        json.dump(res, json_file, ensure_ascii=False, indent=4)
        logger.info(f"Extraction results saved to {json_path}")

        # 返回JSON文件的路径
    return json_path

if __name__ == "__main__":
    output_folder = "C:\\Users\\Administrator\\Desktop\\货物标\\little_parse_output"
    start_time = time.time()
    file_type = 2  # 1:docx 2:pdf 3:其他
    zb_type=2   #1:工程标 2：货物标
    input_file = "C:\\Users\\Administrator\\Desktop\\货物标\\zbfiles\\6.2定版视频会议磋商文件.pdf"
    res=little_parse_main(output_folder, input_file, file_type, zb_type)
    print(json.dumps(res, ensure_ascii=False, indent=4))
    end_time = time.time()
    elapsed_time = end_time - start_time  # 计算耗时
    print(f"Function execution took {elapsed_time} seconds.")