798 lines
40 KiB
Python
798 lines
40 KiB
Python
import glob
|
||
import logging
|
||
|
||
from PyPDF2 import PdfReader, PdfWriter
|
||
import regex # 导入正则表达式库
|
||
import os # 用于文件和文件夹操作
|
||
|
||
from flask_app.general.clean_pdf import clean_page_content, extract_common_header
|
||
from flask_app.general.format_change import docx2pdf
|
||
from flask_app.general.merge_pdfs import merge_and_cleanup, merge_pdfs, merge_selected_pdfs_for_goods
|
||
import concurrent.futures
|
||
|
||
|
||
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
|
||
|
||
def is_pdf_or_doc(filename):
|
||
# 判断文件是否为PDF或Word文档
|
||
return filename.lower().endswith(('.pdf', '.doc', '.docx'))
|
||
|
||
|
||
def convert_to_pdf(file_path):
|
||
# 假设 docx2pdf 函数已经被定义，这里仅根据文件扩展名来决定是否需要转换
|
||
if file_path.lower().endswith(('.doc', '.docx')):
|
||
return docx2pdf(file_path)
|
||
return file_path
|
||
|
||
|
||
def process_files(file_path, output_folder, begin_pattern, begin_page, end_pattern, output_suffix):
|
||
pdf_path = convert_to_pdf(file_path)
|
||
result = extract_pages(pdf_path, output_folder, begin_pattern, begin_page, end_pattern, output_suffix)
|
||
if result:
|
||
if output_suffix == "tobidders_notice":
|
||
# 确保返回的是元组，并将其中的 None 转换为 ""
|
||
path1, path2 = result
|
||
return [path1 or "", path2 or ""]
|
||
elif output_suffix == "qualification1":
|
||
merge_and_cleanup(result, "qualification3")
|
||
return [result or ""]
|
||
return [result or ""]
|
||
return [""] # 返回空字符串
|
||
|
||
def extract_pages_generic(pdf_document, begin_pattern, end_pattern, begin_page, common_header, exclusion_pattern=None,
|
||
output_suffix="normal"):
|
||
start_page = None
|
||
end_page = None
|
||
flag=True
|
||
for i, page in enumerate(pdf_document.pages[begin_page:], start=begin_page):
|
||
text = page.extract_text() or ""
|
||
cleaned_text = clean_page_content(text, common_header)
|
||
if output_suffix == "tobidders_notice":
|
||
if exclusion_pattern and flag and (start_page is not None) and regex.search(exclusion_pattern, cleaned_text):
|
||
flag=False
|
||
continue
|
||
else:
|
||
if exclusion_pattern and flag and regex.search(exclusion_pattern, cleaned_text):
|
||
flag=False
|
||
continue
|
||
if start_page is None and regex.search(begin_pattern, cleaned_text):
|
||
if (output_suffix == "notice" and i >= begin_page) or (output_suffix != "notice" and i > begin_page):
|
||
start_page = i
|
||
continue
|
||
if start_page is not None:
|
||
if output_suffix == "tobidders_notice":
|
||
if regex.search(end_pattern, cleaned_text) and i > start_page:
|
||
end_page = i
|
||
break
|
||
else:
|
||
if regex.search(end_pattern, cleaned_text) and i > start_page and not regex.search(begin_pattern,cleaned_text):
|
||
end_page = i
|
||
break
|
||
return start_page, end_page
|
||
|
||
|
||
def extract_pages(pdf_path, output_folder, begin_pattern, begin_page, end_pattern, output_suffix):
|
||
try:
|
||
common_header = extract_common_header(pdf_path)
|
||
pdf_document = PdfReader(pdf_path)
|
||
exclusion_pattern = None
|
||
total_pages = len(pdf_document.pages) - 1 # 获取总页数
|
||
|
||
if output_suffix == "tobidders_notice":
|
||
exclusion_pattern = regex.compile(
|
||
r'文件的构成|文件的组成|文件构成|文件组成')
|
||
start_page, mid_page, end_page = extract_pages_tobidders_notice(
|
||
pdf_path, begin_pattern, begin_page, common_header, exclusion_pattern
|
||
)
|
||
if not start_page or not mid_page or not end_page:
|
||
print(f"三次提取tobidders_notice均失败!!{pdf_path}")
|
||
return "", ""
|
||
# if start_page is None or end_page is None or mid_page is None:
|
||
# print(f"first: {output_suffix} 未找到起始或结束页在文件 {pdf_path} 中！尝试备用提取策略。")
|
||
# return extract_pages_twice_tobidders_notice(pdf_path, output_folder, output_suffix, common_header,begin_page)
|
||
path1 = save_extracted_pages(pdf_document, start_page, mid_page, pdf_path, output_folder,
|
||
"tobidders_notice_part1")
|
||
if mid_page != end_page:
|
||
path2 = save_extracted_pages(pdf_document, mid_page, end_page, pdf_path, output_folder,
|
||
"tobidders_notice_part2")
|
||
else:
|
||
path2=path1
|
||
return path1, path2
|
||
|
||
else:
|
||
# 原有的处理逻辑保持不变
|
||
if output_suffix == "qualification1" or output_suffix == "procurement" or output_suffix == "evaluation_method":
|
||
exclusion_pattern = regex.compile(
|
||
r'文件的构成|文件的组成|文件构成|文件组成')
|
||
start_page, end_page = extract_pages_generic(pdf_document, begin_pattern, end_pattern, begin_page,
|
||
common_header, exclusion_pattern, output_suffix)
|
||
# 针对 selection = 6 的特殊处理
|
||
if output_suffix == "format":
|
||
if start_page is None:
|
||
print(f"{output_suffix}: 未找到起始页，提取失败！")
|
||
return ""
|
||
if end_page is None:
|
||
# 如果未匹配到结束页，默认截取到文件末尾
|
||
end_page = total_pages
|
||
print(f"{output_suffix}: 未找到结束页，默认截取到文件末尾。")
|
||
return save_extracted_pages(pdf_document, start_page, end_page, pdf_path, output_folder, output_suffix)
|
||
if start_page is None or end_page is None:
|
||
print(f"first: {output_suffix} 未找到起始或结束页在文件 {pdf_path} 中！尝试备用提取策略。")
|
||
return extract_pages_twice(pdf_path, output_folder, output_suffix, common_header, begin_page)
|
||
elif output_suffix == "qualification1":
|
||
truncate_pdf_main(pdf_path, output_folder, 2, "qualification3") # 合并'资格审查'章节和'评标办法'章节
|
||
return save_extracted_pages(pdf_document, start_page, end_page, pdf_path, output_folder, output_suffix)
|
||
except Exception as e:
|
||
print(f"Error processing {pdf_path}: {e}")
|
||
return ""
|
||
|
||
|
||
def get_patterns_for_procurement():
|
||
# begin_pattern = regex.compile(
|
||
# r'^第[一二三四五六七八九十百千]+(?:章|部分).*?(?:服务|项目|商务|技术).*?要求|^第[一二三四五六七八九十1-9]+(?:章|部分).*(?:采购|需求).*',
|
||
# regex.MULTILINE)
|
||
begin_pattern = regex.compile(
|
||
r'(?<!见\s*)(?<!与\s*)(?<!"\s*)(?<!“\s*)(?<!”\s*)'
|
||
r'第[一二三四五六七八九十1-9]+(?:章|部分)\s*' # 匹配“第X章”或“第X部分”
|
||
r'[\u4e00-\u9fff、()（）]*?' # 匹配允许的字符
|
||
r'(?:(?:服务|项目|商务|技术)[\u4e00-\u9fff、()（）]*?要求[\u4e00-\u9fff、()（）]*?\s*$|' # 匹配“服务”、“项目”、“商务”或“技术”后跟“要求”
|
||
r'(?:采购|需求)[\u4e00-\u9fff、()（）]*?)\s*$', # 或者匹配“采购”或“需求”
|
||
regex.MULTILINE
|
||
)
|
||
end_pattern = regex.compile(r'(?<!见\s*)(?<!与\s*)(?<!"\s*)(?<!“\s*)(?<!”\s*)第[一二三四五六七八九1-9]+(?:章|部分)\s*[\u4e00-\u9fff、()（）]+\s*$', regex.MULTILINE)
|
||
return begin_pattern, end_pattern
|
||
|
||
|
||
def get_patterns_for_evaluation_method():
|
||
begin_pattern = regex.compile(
|
||
r'(?:(?<!见\s*)(?<!与\s*)(?<!"\s*)(?<!“\s*)(?<!”\s*)第[一二三四五六七八九十1-9]+(?:章|部分)\s*' # 第一种模式
|
||
r'(?:[\u4e00-\u9fff、()（）]*?)'
|
||
r'(?=.*(?:磋商|谈判|评标|评定|评审))'
|
||
r'(?=.*(?:办法|方法|内容))'
|
||
r'[\u4e00-\u9fff、()（）]*\s*$|'
|
||
r'^\s*(?<!见\s*)(?<!与\s*)(?<!"\s*)(?<!“\s*)(?<!”\s*)评标(方法|办法)前附表\s*$)', # 第二种模式
|
||
regex.MULTILINE
|
||
)
|
||
end_pattern = regex.compile(
|
||
r'(?<!见\s*)(?<!与\s*)(?<!"\s*)(?<!“\s*)(?<!”\s*)第[一二三四五六七八九1-9]+(?:章|部分)\s*[\u4e00-\u9fff、()（）]+\s*$', regex.MULTILINE)
|
||
return begin_pattern, end_pattern
|
||
|
||
|
||
def get_patterns_for_notice():
|
||
begin_pattern = regex.compile(r'^第[一二三四五六七八九十百千]+(?:章|部分).*?(?:公告|邀请书|邀请函).*')
|
||
end_pattern = regex.compile(
|
||
# r'^(?:第[一二三四五六七八九十百千]+(?:章|部分)\s*(?:投标人须知|磋商须知|供应商须知)+|(?:一\s*、\s*)?(?:投标人须知|磋商须知|供应商须知)前附表)',
|
||
r'^第[一二三四五六七八九十百千]+(?:章|部分)\s*[\u4e00-\u9fff]+',
|
||
regex.MULTILINE
|
||
)
|
||
return begin_pattern, end_pattern
|
||
|
||
# def extract_pages_tobidders_notice(pdf_document, begin_pattern, end_pattern, begin_page, common_header,
|
||
# exclusion_pattern):
|
||
# start_page = None
|
||
# mid_page = None
|
||
# end_page = None
|
||
# for i, page in enumerate(pdf_document.pages):
|
||
# text = page.extract_text() or ""
|
||
# cleaned_text = clean_page_content(text, common_header)
|
||
# if exclusion_pattern and regex.search(exclusion_pattern, cleaned_text) and mid_page is not None:
|
||
# continue
|
||
# if start_page is None and regex.search(begin_pattern, cleaned_text) and i > begin_page:
|
||
# start_page = i
|
||
# if start_page is not None and mid_page is None and regex.search(
|
||
# r'^\s*[（(]?\s*[一1]\s*[)）]?\s*[、．.]*\s*(说\s*明|总\s*则)', cleaned_text):
|
||
# mid_page = i
|
||
# if start_page is not None and mid_page is not None and regex.search(end_pattern, cleaned_text) and i > mid_page:
|
||
# end_page = i
|
||
# break
|
||
# return start_page, mid_page, end_page
|
||
|
||
def extract_pages_tobidders_notice(pdf_path, begin_pattern, begin_page, common_header, exclusion_pattern):
|
||
"""
|
||
从PDF文档中提取起始页、中间页和结束页。
|
||
|
||
如果第一次提取失败，则使用新的 begin_pattern 和 end_pattern 重新提取。
|
||
|
||
参数:
|
||
pdf_document (PDFDocument): 要处理的PDF文档对象。
|
||
begin_pattern (str 或 regex.Pattern): 用于识别起始的正则表达式模式。
|
||
begin_page (int): 开始搜索的页码。
|
||
common_header (str): 每页需要清理的公共头部文本。
|
||
exclusion_pattern (str 或 regex.Pattern): 用于排除某些页的模式。
|
||
|
||
返回:
|
||
tuple: (start_page, mid_page, end_page) 如果成功，否则 (None, None, None)
|
||
"""
|
||
|
||
def run_extraction(local_begin_pattern, local_end_pattern=None):
|
||
"""
|
||
使用提供的 begin 和 end 模式运行提取过程。
|
||
|
||
如果未提供 local_end_pattern，则根据匹配的章节类型动态生成 end_pattern。
|
||
|
||
参数:
|
||
local_begin_pattern (str 或 regex.Pattern): 用于识别起始的正则表达式模式。
|
||
local_end_pattern (str 或 regex.Pattern, 可选): 用于识别结束的正则表达式模式。
|
||
|
||
返回:
|
||
tuple: (start_page, mid_page, end_page)
|
||
"""
|
||
start_page = None
|
||
mid_page = None
|
||
end_page = None
|
||
chapter_type = None # 用于存储“章”或“部分”
|
||
combined_mid_pattern = None # 中间页的组合模式
|
||
pdf_document = PdfReader(pdf_path)
|
||
for i, page in enumerate(pdf_document.pages):
|
||
text = page.extract_text() or ""
|
||
cleaned_text = clean_page_content(text, common_header)
|
||
|
||
# 如果已经找到中间页，且当前页匹配排除模式，则跳过
|
||
if exclusion_pattern and regex.search(exclusion_pattern, cleaned_text) and mid_page is not None:
|
||
continue
|
||
|
||
# 识别起始页
|
||
if start_page is None:
|
||
match = regex.search(local_begin_pattern, cleaned_text)
|
||
if match and i > begin_page:
|
||
start_page = i
|
||
matched_text = match.group(0) # 获取整个匹配的文本
|
||
|
||
# 如果未提供固定的 end_pattern，则根据匹配的章节类型动态生成
|
||
if not local_end_pattern:
|
||
if '章' in matched_text:
|
||
chapter_type = '章'
|
||
elif '部分' in matched_text:
|
||
chapter_type = '部分'
|
||
else:
|
||
chapter_type = None # 未匹配到“章”或“部分”
|
||
|
||
if chapter_type:
|
||
# 根据 chapter_type 动态生成 end_pattern
|
||
end_pattern_dynamic = regex.compile(
|
||
rf'^第[一二三四五六七八九十百千]+?(?:{chapter_type})\s*[\u4e00-\u9fff]+|(?<!见\s*)(?<!与\s*)(?<!"\s*)(?<!“\s*)(?<!”\s*)评标(方法|办法)前附表\s*$|(?<!见\s*)(?<!与\s*)(?<!"\s*)(?<!“\s*)(?<!”\s*)评标(方法|办法)正文\s*$',
|
||
regex.MULTILINE
|
||
)
|
||
|
||
# 根据 chapter_type 动态生成 additional_mid_pattern
|
||
if chapter_type == '章':
|
||
additional_mid_pattern = r'^第[一二三四五六七八九十百千]+?(?:部分)'
|
||
elif chapter_type == '部分':
|
||
additional_mid_pattern = r'^第[一二三四五六七八九十百千]+?(?:章)'
|
||
else:
|
||
additional_mid_pattern = ''
|
||
|
||
# 定义基础的 mid_pattern
|
||
base_mid_pattern = r'^\s*(?:[（(]\s*[一二12]?\s*[)）]\s*[、．.]*|' \
|
||
r'[一二12][、．.]+|[、．.]+)\s*(说\s*明|总\s*则|名\s*词\s*解\s*释)' \
|
||
r'|(?<!见\s*)(?<!与\s*)(?<!"\s*)(?<!“\s*)(?<!”\s*)(?:投标人?|磋商|供应商|谈判供应商|磋商供应商)\s*须知正文\s*$'
|
||
|
||
# 合并基础模式和额外模式
|
||
if additional_mid_pattern:
|
||
combined_mid_pattern = regex.compile(
|
||
rf'(?:{base_mid_pattern})|(?:{additional_mid_pattern})',
|
||
regex.MULTILINE
|
||
)
|
||
else:
|
||
combined_mid_pattern = regex.compile(
|
||
rf'{base_mid_pattern}',
|
||
regex.MULTILINE
|
||
)
|
||
else:
|
||
# 如果未匹配到“章”或“部分”，使用默认的 end_pattern 和 mid_pattern
|
||
end_pattern_dynamic = regex.compile(
|
||
r'^第[一二三四五六七八九十百千]+(?:章|部分)\s*[\u4e00-\u9fff]+|(?<!见\s*)(?<!与\s*)(?<!"\s*)(?<!“\s*)(?<!”\s*)评标(方法|办法)前附表\s*$|(?<!见\s*)(?<!与\s*)(?<!"\s*)(?<!“\s*)(?<!”\s*)评标(方法|办法)正文\s*$',
|
||
regex.MULTILINE
|
||
)
|
||
|
||
# 定义基础的 mid_pattern
|
||
base_mid_pattern = r'^\s*(?:[（(]\s*[一二12]?\s*[)）]\s*[、．.]*|' \
|
||
r'[一二12][、．.]+|[、．.]+)\s*(说\s*明|总\s*则|名\s*词\s*解\s*释)'
|
||
combined_mid_pattern = regex.compile(
|
||
rf'{base_mid_pattern}',
|
||
regex.MULTILINE
|
||
)
|
||
else:
|
||
# 如果提供了固定的 end_pattern，则使用默认的 mid_pattern
|
||
base_mid_pattern = r'.*[（(]?\s*[一二12]?[)）]?\s*[、．.]*\s*(说\s*明|总\s*则|名\s*词\s*解\s*释)\s*$' # 可以匹配"东莞市粤隆招标有限公司编制 46一、说明"
|
||
combined_mid_pattern = regex.compile(
|
||
rf'{base_mid_pattern}',
|
||
regex.MULTILINE
|
||
)
|
||
continue
|
||
|
||
# 识别中间页
|
||
if start_page is not None and mid_page is None and combined_mid_pattern:
|
||
if (start_page + 1 == i) and regex.search(local_begin_pattern, cleaned_text):
|
||
continue
|
||
if regex.search(combined_mid_pattern, cleaned_text):
|
||
mid_page = i
|
||
|
||
# 识别结束页
|
||
if start_page is not None and mid_page is not None:
|
||
# 使用提供的 end_pattern 或动态生成的 end_pattern
|
||
current_end_pattern = local_end_pattern if local_end_pattern else end_pattern_dynamic
|
||
if regex.search(current_end_pattern, cleaned_text):
|
||
if i > mid_page:
|
||
end_page = i
|
||
break
|
||
|
||
return start_page, mid_page, end_page
|
||
|
||
# 第一次提取尝试，使用初始的 begin_pattern
|
||
start_page, mid_page, end_page = run_extraction(begin_pattern)
|
||
|
||
# 如果第一次提取失败，则使用新的 begin_pattern 和 end_pattern 重新提取
|
||
if not (start_page and mid_page and end_page):
|
||
print(f"第二次尝试 tobidders_notice!")
|
||
pdf_document = PdfReader(pdf_path)
|
||
start_page, mid_page, end_page = extract_pages_twice_tobidders_notice(pdf_document, common_header, begin_page)
|
||
if start_page and end_page and mid_page:
|
||
return start_page, mid_page, end_page
|
||
else:
|
||
# 定义新的 begin_pattern 和 end_pattern
|
||
new_begin_pattern = regex.compile(
|
||
r'.*(?:投标人|磋商|供应商|谈判供应商|磋商供应商)须知\s*$|'
|
||
r'(?:一\s*、\s*)?(?:投标人?|磋商|供应商)须知前附表',
|
||
regex.MULTILINE
|
||
)
|
||
new_end_pattern = regex.compile(
|
||
r'第[一二三四五六七八九十百千]+(?:章|部分)\s*[\u4e00-\u9fff、()（）]+\s*$',
|
||
regex.MULTILINE
|
||
)
|
||
print("第三次尝试 tobidders_notice! ")
|
||
# 第二次提取尝试，使用新的模式
|
||
start_page, mid_page, end_page = run_extraction(new_begin_pattern, new_end_pattern)
|
||
|
||
return start_page, mid_page, end_page
|
||
|
||
|
||
def extract_pages_twice_tobidders_notice(pdf_document, common_header, begin_page):
|
||
output_suffix = "tobidders_notice"
|
||
begin_pattern = regex.compile(
|
||
r'^第[一二三四五六七八九十百千]+(?:章|部分)\s*(?:(?:投标人?|磋商|供应商|谈判供应商|磋商供应商)须知)+',
|
||
regex.MULTILINE
|
||
)
|
||
end_pattern = regex.compile(
|
||
r'^第[一二三四五六七八九十百千]+(?:章|部分)\s*([\u4e00-\u9fff]+)', regex.MULTILINE # 捕获中文部分
|
||
)
|
||
exclusion_words = ["合同", "评标", "开标", "评审", "采购", "资格"] # 在这里添加需要排除的关键词
|
||
|
||
exclusion_pattern = regex.compile(r'文件的构成|文件的组成|文件构成|文件组成')
|
||
|
||
# 提取第一部分
|
||
start_page1, end_page1 = extract_pages_generic(pdf_document, begin_pattern, end_pattern, begin_page, common_header,exclusion_pattern,output_suffix)
|
||
if start_page1 is None or end_page1 is None:
|
||
return "", "", ""
|
||
|
||
# 提取第二部分
|
||
start_page2 = end_page1
|
||
|
||
# 检查end_page1页面的内容
|
||
text = pdf_document.pages[end_page1].extract_text() or ""
|
||
cleaned_text = clean_page_content(text, common_header)
|
||
match = end_pattern.search(cleaned_text)
|
||
|
||
if match:
|
||
# 获取匹配到的中文部分
|
||
chapter_title = match.group(1)
|
||
# 检查是否包含排除关键词
|
||
if any(word in chapter_title for word in exclusion_words):
|
||
# 如果包含排除关键词，直接返回相同的路径
|
||
return start_page1, end_page1, end_page1
|
||
|
||
# 如果不包含排除关键词，继续提取第二部分
|
||
_, end_page2 = extract_pages_generic(pdf_document, end_pattern, end_pattern, start_page2 - 1, common_header,
|
||
exclusion_pattern, output_suffix)
|
||
|
||
if end_page2 is None:
|
||
return start_page1, end_page1, ""
|
||
|
||
return start_page1, end_page1, end_page2
|
||
|
||
|
||
def extract_pages_qualification(pdf_document, begin_page, common_header):
|
||
# 开始匹配模式，仅匹配“附录”、“附件”或“附表”
|
||
begin_pattern = regex.compile(
|
||
r'^(?:附录(?:一|1)?[：:]?|附件(?:一|1)?[：:]?|附表(?:一|1)?[：:]?)',
|
||
regex.MULTILINE
|
||
)
|
||
|
||
# 优先匹配模式，匹配以“资格性检查”、“资格审查”或“符合性审查”开头
|
||
priority_pattern = regex.compile(
|
||
r'^(资格性检查|资格审查|符合性审查)',
|
||
regex.MULTILINE
|
||
)
|
||
|
||
# 结束匹配模式 - 附录、附件、附表等
|
||
end_pattern_attachment = regex.compile(
|
||
r'^(?:附录.*?[：:]|附件.*?[：:]|附表.*?[：:]|附件\s*\d+).*$',
|
||
regex.MULTILINE
|
||
)
|
||
# 结束匹配模式 - 章节标题
|
||
end_pattern_chapter = regex.compile(
|
||
r'(?<!见\s*)(?<!与\s*)(?<!"\s*)(?<!“\s*)(?<!”\s*)第[一二三四五六七八九十百千]+(?:章|部分)\s*[\u4e00-\u9fff、()（）]+\s*$',
|
||
regex.MULTILINE
|
||
)
|
||
|
||
print("第二次尝试 qualification:匹配附件")
|
||
start_page = None
|
||
end_page = None
|
||
include_keywords = ["资格审查", "资质审查", "符合性审查", "资格性检查", "符合性检查", "资格检查"]
|
||
exclude_keywords = ["声明函", "承诺函"]
|
||
|
||
# 从指定的开始页开始遍历
|
||
for i, page in enumerate(pdf_document.pages[begin_page:], start=begin_page):
|
||
text = page.extract_text()
|
||
if text:
|
||
cleaned_text = clean_page_content(text, common_header)
|
||
|
||
# 优先检查是否匹配优先模式
|
||
priority_match = priority_pattern.search(cleaned_text)
|
||
if priority_match and start_page is None:
|
||
start_page = i
|
||
# print(f"匹配到优先模式，设置起始页为: {start_page}")
|
||
continue
|
||
else:
|
||
# 如果未匹配优先模式，则按照一般模式进行判断
|
||
if (
|
||
any(keyword in cleaned_text for keyword in include_keywords) and
|
||
all(keyword not in cleaned_text for keyword in exclude_keywords) and
|
||
start_page is None
|
||
):
|
||
if begin_pattern.search(cleaned_text):
|
||
start_page = i
|
||
continue
|
||
# print(f"匹配到附录等模式，设置起始页为: {start_page}")
|
||
|
||
# 确定结束页 - 附录、附件、附表等
|
||
if start_page is not None and end_pattern_attachment.search(cleaned_text):
|
||
# 额外检查当前页面是否不包含任何 include_keywords
|
||
if not any(keyword in cleaned_text for keyword in include_keywords):
|
||
if i > start_page:
|
||
end_page = i
|
||
# print(f"找到结束页 (附件类): {end_page}")
|
||
break # 找到结束页后退出循环
|
||
else:
|
||
print(f"当前页面匹配附件结束模式但包含 include_keywords，继续查找。页面: {i}")
|
||
|
||
# 确定结束页 - 章节标题
|
||
elif start_page is not None and end_pattern_chapter.search(cleaned_text):
|
||
if i > start_page:
|
||
end_page = i
|
||
print(f"找到结束页 (章节标题): {end_page}")
|
||
break # 找到结束页后退出循环
|
||
return start_page, end_page
|
||
|
||
|
||
def extract_pages_twice(pdf_path, output_folder, output_suffix, common_header, begin_page):
|
||
try:
|
||
exclusion_pattern = regex.compile(
|
||
r'文件的构成|文件的组成|须对应|需对应|须按照|需按照|须根据|需根据|文件组成|文件构成')
|
||
pdf_document = PdfReader(pdf_path)
|
||
patterns = None
|
||
start_page = None
|
||
end_page = None
|
||
|
||
if output_suffix == "procurement":
|
||
patterns = [get_patterns_for_procurement()]
|
||
elif output_suffix == "evaluation_method" or output_suffix == "qualification2" or output_suffix == "qualification3":
|
||
patterns = [get_patterns_for_evaluation_method()]
|
||
elif output_suffix == "notice":
|
||
patterns = [get_patterns_for_notice()]
|
||
elif output_suffix == "qualification1":
|
||
start_page, end_page = extract_pages_qualification(pdf_document, begin_page, common_header)
|
||
if patterns:
|
||
for pattern_pair in patterns:
|
||
start_page, end_page = extract_pages_generic(pdf_document, pattern_pair[0], pattern_pair[1], begin_page,
|
||
common_header, exclusion_pattern, output_suffix)
|
||
if start_page is not None and end_page is not None:
|
||
break
|
||
|
||
if start_page is None or end_page is None:
|
||
if output_suffix == "qualification1":
|
||
# print(f"second: {output_suffix} 未找到起始或结束页在文件 {pdf_path} 中！")
|
||
print("第三次尝试资格审查:尝试提取评分办法章节...")
|
||
base_file_name = os.path.splitext(os.path.basename(pdf_path))[0]
|
||
evaluation_method_file = os.path.join(output_folder, f"{base_file_name}_evaluation_method.pdf")
|
||
if os.path.isfile(evaluation_method_file):
|
||
print(f"找到评分办法章节文件: {evaluation_method_file}，生成新文件。")
|
||
# 获取文件路径和文件名
|
||
new_file_name = f"{base_file_name}_qualification2.pdf"
|
||
new_file_path = os.path.join(output_folder, new_file_name)
|
||
# 复制文件
|
||
with open(evaluation_method_file, 'rb') as original_file:
|
||
with open(new_file_path, 'wb') as new_file:
|
||
new_file.write(original_file.read())
|
||
return new_file_path
|
||
else:
|
||
temp = truncate_pdf_main(pdf_path, output_folder, 2, "qualification2")
|
||
if len(temp) > 0:
|
||
return temp[0]
|
||
else:
|
||
return ""
|
||
else:
|
||
print(f"second: {output_suffix} 未找到起始或结束页在文件 {pdf_path} 中！")
|
||
return ""
|
||
return save_extracted_pages(pdf_document, start_page, end_page, pdf_path, output_folder, output_suffix)
|
||
except Exception as e:
|
||
print(f"Error in extract_pages_twice: {e}")
|
||
return ""
|
||
|
||
|
||
def save_extracted_pages(pdf_document, start_page, end_page, pdf_path, output_folder, output_suffix):
|
||
try:
|
||
# 检查 start_page 和 end_page 是否为 None
|
||
if start_page is None or end_page is None:
|
||
print("Error: start_page 或 end_page 为 None")
|
||
return ""
|
||
|
||
base_file_name = os.path.splitext(os.path.basename(pdf_path))[0]
|
||
output_pdf_path = os.path.join(output_folder, f"{base_file_name}_{output_suffix}.pdf")
|
||
|
||
if start_page < 0 or end_page >= len(pdf_document.pages) or start_page > end_page:
|
||
print(f"无效的页面范围: {start_page} 到 {end_page}")
|
||
return ""
|
||
|
||
if output_suffix == 'notice' and start_page - 1 >= 0:
|
||
before_pdf_path = os.path.join(output_folder, f"{base_file_name}_before.pdf")
|
||
before_doc = PdfWriter()
|
||
for page_num in range(0, start_page):
|
||
before_doc.add_page(pdf_document.pages[page_num])
|
||
with open(before_pdf_path, 'wb') as f:
|
||
before_doc.write(f)
|
||
# print(f"已保存页面从 0 到 {start_page - 1} 为 {before_pdf_path}")
|
||
|
||
output_doc = PdfWriter()
|
||
for page_num in range(start_page, end_page + 1):
|
||
output_doc.add_page(pdf_document.pages[page_num])
|
||
with open(output_pdf_path, 'wb') as f:
|
||
output_doc.write(f)
|
||
print(f"{output_suffix} 已截取并保存页面从 {start_page} 到 {end_page} 为 {output_pdf_path}")
|
||
return output_pdf_path
|
||
except Exception as e:
|
||
print(f"Error in save_extracted_pages: {e}")
|
||
return "" # 返回空字符串
|
||
|
||
def get_start_and_common_header(input_path):
|
||
common_header = extract_common_header(input_path)
|
||
last_begin_index = 0
|
||
begin_pattern = regex.compile(
|
||
r'.*(?<!见\s*)(?<!与\s*)(?<!"\s*)(?<!“\s*)(?<!”\s*)(?:招标公告|磋商公告|谈判公告|邀请书|邀请函|投标邀请|磋商邀请|谈判邀请)[\)）]?\s*$',
|
||
regex.MULTILINE
|
||
)
|
||
# 新增目录匹配模式
|
||
catalog_pattern = regex.compile(r'\s*目\s*录\s*$', regex.MULTILINE)
|
||
|
||
pdf_document = PdfReader(input_path)
|
||
for i, page in enumerate(pdf_document.pages):
|
||
if i > 10:
|
||
return common_header, 0 # 如果页码大于10，直接返回last_begin_index=0
|
||
text = page.extract_text()
|
||
if text:
|
||
cleaned_text = clean_page_content(text, common_header)
|
||
# 检查是否存在"目录"
|
||
if catalog_pattern.search(cleaned_text):
|
||
continue # 如果存在目录，跳过当前页面
|
||
if begin_pattern.search(cleaned_text):
|
||
last_begin_index = i # 更新第一个匹配的索引，页码从0开始
|
||
return common_header, last_begin_index
|
||
return common_header, last_begin_index
|
||
|
||
|
||
def truncate_pdf_main(input_path, output_folder, selection, output_suffix="default"):
|
||
try:
|
||
# 检查是否为文件夹
|
||
if os.path.isdir(input_path):
|
||
generated_files = []
|
||
for file_name in os.listdir(input_path):
|
||
file_path = os.path.join(input_path, file_name)
|
||
if is_pdf_or_doc(file_path):
|
||
result = process_input(file_path, output_folder, selection, output_suffix)
|
||
if isinstance(result, tuple):
|
||
generated_files.extend([f if f else "" for f in result])
|
||
else:
|
||
generated_files.append(result)
|
||
return generated_files
|
||
|
||
# 单文件情况
|
||
elif os.path.isfile(input_path) and is_pdf_or_doc(input_path):
|
||
return process_input(input_path, output_folder, selection, output_suffix)
|
||
|
||
else:
|
||
print("提供的路径既不是文件夹也不是PDF文件。")
|
||
return ['']
|
||
|
||
except Exception as e:
|
||
print(f"Error in truncate_pdf_main: {e}")
|
||
return [''] # 返回空字符串
|
||
|
||
|
||
def process_input(input_path, output_folder, selection, output_suffix):
|
||
try:
|
||
# 创建输出文件夹
|
||
if not os.path.exists(output_folder):
|
||
os.makedirs(output_folder)
|
||
|
||
# 获取起始和通用页眉
|
||
common_header, last_begin_index = get_start_and_common_header(input_path)
|
||
begin_page = last_begin_index if last_begin_index != 0 else {
|
||
4: 1,
|
||
2: 5,
|
||
3: 5,
|
||
1: 0,
|
||
5: 3
|
||
}.get(selection, 0)
|
||
|
||
# 根据选择设置对应的模式和结束模式
|
||
if selection == 1:
|
||
begin_pattern = regex.compile(
|
||
r'.*(?<!见\s*)(?<!与\s*)(?<!"\s*)(?<!“\s*)(?<!”\s*)(?:招标公告|磋商公告|谈判公告|邀请书|邀请函|投标邀请|磋商邀请|谈判邀请)[\)）]?\s*$',
|
||
regex.MULTILINE
|
||
)
|
||
end_pattern = regex.compile(r'(?<!见\s*)(?<!与\s*)(?<!"\s*)(?<!“\s*)(?<!”\s*)第[一二三四五六七八九十1-9]+(?:章|部分)\s*[\u4e00-\u9fff、()（）]+\s*$|(?<!见\s*)(?<!与\s*)(?<!"\s*)(?<!“\s*)(?<!”\s*)(?:投标人?|磋商|供应商|谈判供应商|磋商供应商)须知前附表\s*$', regex.MULTILINE)
|
||
local_output_suffix = "notice"
|
||
elif selection == 2:
|
||
begin_pattern = regex.compile(
|
||
r'^第[一二三四五六七八九十]+(?:章|部分)\s*(?=.*(?:磋商|谈判|评标|评定|评审))(?=.*(?:办法|方法|内容))')
|
||
end_pattern = regex.compile(r'^第[一二三四五六七八九十百千]+(?:章|部分)\s*[\u4e00-\u9fff]+')
|
||
local_output_suffix = "evaluation_method"
|
||
elif selection == 3:
|
||
begin_pattern = regex.compile(r'^第[一二三四五六七八九十百千]+(?:章|部分).*?(资格审查).*', regex.MULTILINE)
|
||
end_pattern = regex.compile(r'^第[一二三四五六七八九十百千]+(?:章|部分)\s*[\u4e00-\u9fff]+', regex.MULTILINE)
|
||
local_output_suffix = "qualification1"
|
||
elif selection == 4:
|
||
begin_pattern = regex.compile(
|
||
r'^第[一二三四五六七八九十百千]+(?:章|部分)\s*(?:投标人?|磋商|供应商|谈判供应商|磋商供应商)须知+|(?<!见\s*)(?<!与\s*)(?<!"\s*)(?<!“\s*)(?<!”\s*)(?:投标人?|磋商|供应商|谈判供应商|磋商供应商)须知前附表\s*$',
|
||
regex.MULTILINE)
|
||
end_pattern = None
|
||
local_output_suffix = "tobidders_notice"
|
||
elif selection == 5:
|
||
begin_pattern = regex.compile(
|
||
r'^第[一二三四五六七八九十百千]+(?:章|部分).*?(?:服务|项目|商务|技术).*?要求|^第[一二三四五六七八九十百千]+(?:章|部分)(?!.*说明).*(?:采购内容|采购要求|需求).*') # 包头中有一章'采购相关说明'
|
||
end_pattern = regex.compile(r'^第[一二三四五六七八九十百千]+(?:章|部分)\s*[\u4e00-\u9fff]+')
|
||
local_output_suffix = "procurement"
|
||
|
||
# begin_pattern = regex.compile(
|
||
# r'^第[一二三四五六七八九十百千]+(?:章|部分).*?(?:服务|项目|商务).*?要求|'
|
||
# r'^第[一二三四五六七八九十百千]+(?:章|部分).*?(?:采购|技术标准).*|'
|
||
# r'^[一二三四五六七八九十百千]+、\s*采购清单', regex.MULTILINE)
|
||
# end_pattern = regex.compile(
|
||
# r'^第[一二三四五六七八九十百千]+(?:章|部分)\s*[\u4e00-\u9fff]+', regex.MULTILINE)
|
||
|
||
elif selection == 6:
|
||
begin_pattern = regex.compile(r'^第[一二三四五六七八九十百千]+(?:章|部分).*?(?:格式).*')
|
||
end_pattern = regex.compile(r'^第[一二三四五六七八九十百千]+(?:章|部分)\s*[\u4e00-\u9fff]+', regex.MULTILINE)
|
||
local_output_suffix = "format"
|
||
else:
|
||
print("无效的选择:请选择1-5")
|
||
return ['']
|
||
|
||
# 如果传入的 output_suffix 是 'default'，则使用本地生成的 output_suffix
|
||
if output_suffix == "default":
|
||
output_suffix = local_output_suffix
|
||
|
||
# 调用实际处理文件内容的函数
|
||
return process_files(input_path, output_folder, begin_pattern, begin_page, end_pattern, output_suffix)
|
||
|
||
except Exception as e:
|
||
print(f"Error in process_input: {e}")
|
||
return ['']
|
||
|
||
|
||
def truncate_pdf_multiple(pdf_path, output_folder, logger):
|
||
base_file_name = os.path.splitext(os.path.basename(pdf_path))[0]
|
||
truncate_files = []
|
||
|
||
# 定义要处理的选择范围
|
||
selections = range(1, 6)
|
||
|
||
# 使用 ThreadPoolExecutor 进行多线程处理
|
||
with concurrent.futures.ThreadPoolExecutor(max_workers=5) as executor:
|
||
# 提交所有任务并保持 selection 顺序
|
||
future_to_selection = {
|
||
selection: executor.submit(truncate_pdf_main, pdf_path, output_folder, selection, output_suffix="default")
|
||
for selection in selections}
|
||
|
||
# 按 selection 顺序收集结果
|
||
for selection in selections:
|
||
future = future_to_selection.get(selection)
|
||
try:
|
||
files = future.result()
|
||
if files:
|
||
truncate_files.extend(files)
|
||
except Exception as e:
|
||
logger.error(f"Selection {selection} 生成了一个异常: {e}")
|
||
|
||
# 定义合并后的输出路径
|
||
merged_output_path = os.path.join(output_folder, f"{base_file_name}_merged_baseinfo.pdf")
|
||
# 调用 merge_selected_pdfs 并获取返回值
|
||
merged_path = merge_selected_pdfs_for_goods(output_folder, truncate_files, merged_output_path, base_file_name)
|
||
if merged_path:
|
||
# 合并成功，添加合并后的文件路径
|
||
truncate_files.append(merged_path)
|
||
# logger.info(f"已生成合并文件: {merged_output_path}")
|
||
else:
|
||
# 合并失败，添加空字符串
|
||
truncate_files.append("")
|
||
logger.warning(f"合并失败，没有生成合并文件 for {pdf_path}")
|
||
logger.info("已截取文件路径" + str(truncate_files))
|
||
return truncate_files
|
||
|
||
|
||
# 小解析，只需要前三章内容
|
||
def truncate_pdf_specific_goods(pdf_path, output_folder, selections, unique_id="123"):
|
||
"""
|
||
处理 PDF 文件，选择指定的 selections，并合并结果。
|
||
|
||
Args:
|
||
pdf_path (str): 要处理的 PDF 文件路径。
|
||
output_folder (str): 截取后的文件保存文件夹路径。
|
||
selections (iterable): 要处理的 selection 列表。
|
||
|
||
Returns:
|
||
list: 截取的文件路径列表，包括合并后的文件路径（如果有）。
|
||
"""
|
||
logger = get_global_logger(unique_id)
|
||
base_file_name = os.path.splitext(os.path.basename(pdf_path))[0]
|
||
truncate_files = []
|
||
|
||
# 使用 ThreadPoolExecutor 进行多线程处理
|
||
with concurrent.futures.ThreadPoolExecutor(max_workers=len(selections)) as executor:
|
||
# 提交所有任务并保持 selection 顺序
|
||
future_to_selection = {
|
||
selection: executor.submit(truncate_pdf_main, pdf_path, output_folder, selection, output_suffix="default")
|
||
for selection in selections}
|
||
|
||
# 按 selection 顺序收集结果
|
||
for selection in selections:
|
||
future = future_to_selection.get(selection)
|
||
try:
|
||
files = future.result()
|
||
if files:
|
||
if isinstance(files, list):
|
||
truncate_files.extend(files)
|
||
elif isinstance(files, str):
|
||
truncate_files.append(files)
|
||
except Exception as e:
|
||
logger.error(f"Selection {selection} 生成了一个异常: {e}")
|
||
|
||
# 定义合并后的输出路径
|
||
merged_output_path = os.path.join(output_folder, f"{base_file_name}_merged_baseinfo.pdf")
|
||
# 调用 merge_selected_pdfs 并获取返回值
|
||
merged_path = merge_selected_pdfs_for_goods(output_folder, truncate_files, merged_output_path, base_file_name)
|
||
if merged_path:
|
||
# 合并成功，添加合并后的文件路径
|
||
truncate_files.append(merged_path)
|
||
# logger.info(f"已生成合并文件: {merged_output_path}")
|
||
else:
|
||
# 合并失败，添加空字符串
|
||
truncate_files.append("")
|
||
logger.warning(f"合并失败，没有生成merged_baseinfo for {pdf_path}")
|
||
return truncate_files
|
||
|
||
|
||
# TODO:交通智能系统和招标(1)(1)文件有问题 包头 绍兴 工程标中，判断是符合性审查之后，可以将它们设为同一章
|
||
|
||
# ztbfile.pdf少资格评审 包头少符合性评审
|
||
if __name__ == "__main__":
|
||
logger = get_global_logger("123")
|
||
# input_path = r"C:\Users\Administrator\Desktop\new招标文件\货物标"
|
||
# input_path = r"C:\Users\Administrator\Desktop\招标文件-采购类\2024-贵州-贵州医科大学附属医院导视系统零星制作安装项目.pdf"
|
||
# input_path="C:\\Users\\Administrator\\Desktop\\货物标\\zbfiles\\zbtest4_evaluation_method.pdf"
|
||
# input_path = r"C:\Users\Administrator\Desktop\货物标\zbfiles\2-招标文件（广水市教育局封闭管理）.pdf"
|
||
input_path=r"C:\Users\Administrator\Desktop\fsdownload\ff2acdba-3a55-48a7-aa7a-61d9f89b909a\ztbfile.pdf"
|
||
output_folder = r"C:\Users\Administrator\Desktop\fsdownload\ff2acdba-3a55-48a7-aa7a-61d9f89b909a\tmp"
|
||
# output_folder = r"C:\Users\Administrator\Desktop\new招标文件\output2"
|
||
files = truncate_pdf_multiple(input_path, output_folder,logger)
|
||
# selections = [3,5]
|
||
# files=truncate_pdf_specific_goods(input_path,output_folder,selections)
|
||
print(files)
|
||
# selection = 2 # 例如：1 - 公告, 2 - 评标办法, 3 - 资格审查后缀有qualification1或qualification2（与评标办法一致） 4.投标人须知前附表part1 投标人须知正文part2 5-采购需求
|
||
# generated_files = truncate_pdf_main(input_path, output_folder, selection)
|
||
# print(generated_files)
|