37 lines
1.9 KiB
Python
37 lines
1.9 KiB
Python
|
import os
|
|||
|
|
import sys
|
|||
|
|
from 货物标截取pdf import truncate_pdf_main
|
|||
|
|
sys.path.append(os.path.abspath(os.path.join(os.path.dirname(__file__), '..', '..')))
|
|||
|
|
# from ..main.format_change import docx2pdf
|
|||
|
|
# from ..main.多线程提问 import multi_threading
|
|||
|
|
from ..main.通义千问long import upload_file,qianwen_long
|
|||
|
|
from ..main.json_utils import clean_json_string
|
|||
|
|
|
|||
|
|
def generate_key_paths(data, parent_key=''):
|
|||
|
|
key_paths = []
|
|||
|
|
for key, value in data.items():
|
|||
|
|
current_key = f"{parent_key}.{key}" if parent_key else key
|
|||
|
|
if isinstance(value, dict):
|
|||
|
|
# 如果值是字典，递归调用
|
|||
|
|
key_paths.extend(generate_key_paths(value, current_key))
|
|||
|
|
else:
|
|||
|
|
# 如果到达了末端，添加当前键路径
|
|||
|
|
key_paths.append(current_key)
|
|||
|
|
return key_paths
|
|||
|
|
|
|||
|
|
#获取采购清单
|
|||
|
|
def fetch_purchasing_list(file_path):
|
|||
|
|
output_folder="C:\\Users\\Administrator\\Desktop\\货物标\\output"
|
|||
|
|
# file_path = docx2pdf(file_path)
|
|||
|
|
truncate_path=truncate_pdf_main(file_path,output_folder,1)
|
|||
|
|
user_query="这是一份货物标中采购要求部分的内容，你需要摘取出需要采购的系统（货物），一个大系统（大项）中可能包含多个小系统（小项），你需要保留这种层次关系，给出货物名称，请以json格式返回，外层键名为\"采购需求\"，嵌套键名为对应的系统名称或货物名称，无需给出采购数量和单位，如有未知内容，在对应键值处填\"未知\"。"
|
|||
|
|
file_id=upload_file(truncate_path)
|
|||
|
|
res=qianwen_long(file_id,user_query)
|
|||
|
|
cleaned_res=clean_json_string(res)
|
|||
|
|
keys_list=generate_key_paths(cleaned_res['采购需求'])
|
|||
|
|
print(keys_list)
|
|||
|
|
|
|||
|
|
if __name__ == "__main__":
|
|||
|
|
file_path="C:\\Users\\Administrator\\Desktop\\货物标\\output1\\招招招标文件（一中多媒体报告厅教学设备）_20240829101650_tobidders_notice_table.pdf"
|
|||
|
|
fetch_purchasing_list(file_path)
|