155 lines
7.1 KiB
Python
155 lines
7.1 KiB
Python
import ast
|
||
import json
|
||
import os
|
||
import re
|
||
|
||
from PyPDF2 import PdfWriter, PdfReader
|
||
|
||
from flask_app.main.通义千问long import upload_file, qianwen_long
|
||
|
||
def extract_and_format_from_paths(json_paths, includes):
|
||
"""
|
||
从多个 JSON 文件路径读取数据，提取包含特定关键词的内容，并按照格式要求合并。
|
||
|
||
参数:
|
||
json_paths (list): 包含多个 JSON 文件路径的列表。
|
||
includes (list): 包含要检查的关键词的列表。
|
||
|
||
返回:
|
||
list: 包含所有文件中满足条件的格式化字符串列表。
|
||
"""
|
||
all_formatted_results = []
|
||
|
||
# 遍历每个文件路径
|
||
for path in json_paths:
|
||
try:
|
||
with open(path, 'r', encoding='utf-8') as file:
|
||
# 加载 JSON 数据
|
||
json_data = json.load(file)
|
||
formatted_results = []
|
||
|
||
# 遍历 JSON 数据的每个键值对
|
||
for key, value in json_data.items():
|
||
if isinstance(value, dict):
|
||
# 如果值是字典，检查嵌套字典的每个键值对
|
||
for sub_key, sub_value in value.items():
|
||
if any(include in sub_value for include in includes):
|
||
# 如果子值包含关键词，格式化并添加到结果列表
|
||
formatted_results.append(f"{sub_key}: {sub_value}")
|
||
elif isinstance(value, str):
|
||
# 如果值是字符串，直接检查是否包含关键词
|
||
if any(include in value for include in includes):
|
||
# 如果值包含关键词，添加到结果列表
|
||
formatted_results.append(value)
|
||
|
||
# 将当前文件的结果添加到总结果列表
|
||
all_formatted_results.extend(formatted_results)
|
||
except FileNotFoundError:
|
||
print(f"Error: The file '{path}' does not exist.")
|
||
except json.JSONDecodeError:
|
||
print(f"Error: The file '{path}' contains invalid JSON.")
|
||
|
||
return all_formatted_results
|
||
|
||
def extract_unique_items_from_texts(texts):
|
||
pattern = re.compile(r'(?:\d+\.|\（\d+\）|\(\d+\)|\d+\))\s*')
|
||
intro_pattern = re.compile(r'^.*[:：]')
|
||
punctuation_pattern = re.compile(r'[；。，、．.,:;!?！？]+$')
|
||
|
||
all_results = []
|
||
seen = set()
|
||
|
||
for text in texts:
|
||
text = intro_pattern.sub('', text)
|
||
items = pattern.split(text)
|
||
|
||
for item in items:
|
||
cleaned_item = item.strip()
|
||
if cleaned_item:
|
||
cleaned_item = pattern.sub('', cleaned_item)
|
||
cleaned_item = punctuation_pattern.sub('', cleaned_item)
|
||
cleaned_item = cleaned_item.strip()
|
||
if cleaned_item and cleaned_item not in seen:
|
||
seen.add(cleaned_item)
|
||
all_results.append(cleaned_item)
|
||
|
||
return all_results
|
||
|
||
def merge_pdfs(paths, output_filename):
|
||
pdf_writer = PdfWriter()
|
||
output_path = None
|
||
|
||
for path in paths:
|
||
pdf_reader = PdfReader(path)
|
||
for page in range(len(pdf_reader.pages)):
|
||
# 将每页添加到写入对象中
|
||
pdf_writer.add_page(pdf_reader.pages[page])
|
||
|
||
if output_path is None:
|
||
# 确定输出文件路径
|
||
output_path = os.path.join(os.path.dirname(path), output_filename)
|
||
|
||
# 写入合并的PDF到文件
|
||
if output_path:
|
||
with open(output_path, 'wb') as out:
|
||
pdf_writer.write(out)
|
||
print(f"Merged PDF saved to {output_path}")
|
||
else:
|
||
print("No files to merge.")
|
||
return output_path
|
||
|
||
def process_string_list(string_list):
|
||
# 使用正则表达式匹配方括号内的内容
|
||
match = re.search(r'\[(.*?)\]', string_list)
|
||
if match:
|
||
# 获取匹配的内容，即方括号内的部分
|
||
content_inside_brackets = match.group(1)
|
||
if content_inside_brackets: # 检查内容是否为空
|
||
# 检查内容是否是数字列表
|
||
if all(item.strip().isdigit() for item in content_inside_brackets.split(',')):
|
||
# 如果是数字，不用加引号，直接保留数字
|
||
formatted_list = '[' + ', '.join(item.strip() for item in content_inside_brackets.split(',') if item.strip()) + ']'
|
||
else:
|
||
# 如果不全是数字，按字符串处理
|
||
formatted_list = '[' + ', '.join(f"'{item.strip()}'" for item in content_inside_brackets.split(',') if item.strip()) + ']'
|
||
else:
|
||
return [] # 直接返回空列表如果内容为空
|
||
|
||
# 使用 ast.literal_eval 来解析格式化后的字符串
|
||
try:
|
||
actual_list = ast.literal_eval(formatted_list)
|
||
return actual_list
|
||
except SyntaxError as e:
|
||
print(f"Error parsing list: {e}")
|
||
return []
|
||
else:
|
||
# 如果没有匹配到内容，返回空列表
|
||
return []
|
||
def find_forbidden(truncate_json_path,clause_path,truncate3): #投标人须知前附表 条款 评分前附表和资格审查表中
|
||
# output_filename="merged.pdf"
|
||
# paths=[truncate1,truncate4]
|
||
# merged_filepath=merge_pdfs(paths,output_filename) #暂时废弃，评分前附表中的在'否决投标'中摘录了。
|
||
|
||
file_id=upload_file(truncate3)
|
||
# user_query_forbidden = "该招标文件规定的投标人不得存在的其他情形有哪些，请按json列表格式给我提供信息，键名为'不得存在的其他情形'，请你不要回答有关\"信誉要求\"的内容,若文件中未说明，请在键值中填'未知'。"
|
||
user_query_forbidden = "该招标文件规定的投标人不得存在的其他情形有哪些，请以列表给我提供信息，形如[xx,xx,...],请你不要回答有关\"信誉要求\"的内容,若原文未提及，返回[]。"
|
||
qianwen_forbidden_str = qianwen_long(file_id, user_query_forbidden)
|
||
actual_list=process_string_list(qianwen_forbidden_str) #提取出字符串列表 ["xxx","xx"]
|
||
|
||
includes = ["不得存在", "禁止投标"]
|
||
forbidden_results = extract_and_format_from_paths([truncate_json_path, clause_path], includes)
|
||
processed_results = extract_unique_items_from_texts(forbidden_results)
|
||
# print(processed_results)
|
||
merged_forbidden_list = list(dict.fromkeys(actual_list + processed_results))
|
||
forbidden_dict={'不得存在的其他情形':merged_forbidden_list}
|
||
|
||
return forbidden_dict
|
||
|
||
|
||
if __name__ == '__main__':
|
||
truncate_json_path = "C:\\Users\\Administrator\\Desktop\\fsdownload\\temp7\\3abb6e16-19db-42ad-9504-53bf1072dfe7\\truncate_output.json"
|
||
clause_path = "C:\\Users\\Administrator\\Desktop\\fsdownload\\temp7\\3abb6e16-19db-42ad-9504-53bf1072dfe7\\clause.json"
|
||
truncate4 = "C:\\Users\\Administrator\\Desktop\\fsdownload\\temp7\\3abb6e16-19db-42ad-9504-53bf1072dfe7\\ztbfile_qualification.pdf"
|
||
output_dir = "C:\\Users\\Administrator\\Desktop\\fsdownload\\temp7\\3abb6e16-19db-42ad-9504-53bf1072dfe7"
|
||
doc_path = 'C:\\Users\\Administrator\\Desktop\\fsdownload\\temp7\\3abb6e16-19db-42ad-9504-53bf1072dfe7\\ztbfile.docx'
|
||
find_forbidden(truncate_json_path,clause_path,truncate4) |