2.19 使用脚本额外判断docx文档是否为乱码文件

flask_app/general/读取文件/按页读取pdf.py

@@ -25,7 +25,7 @@ def extract_text_by_page_fitz(file_path):
     return result
 
 
-def extract_text_by_page(file_path):
+def extract_text_by_page_pypdf2(file_path):
     # common_header=""
     common_header = extract_common_header(file_path)
     # print(common_header)
@@ -109,7 +109,7 @@ def save_extracted_text_to_txt(pdf_path, txt_path):
     """
     try:
         # 提取文本内容
-        extracted_text = extract_text_by_page(pdf_path)
+        extracted_text = extract_text_by_page_pypdf2(pdf_path)
 
         # 将提取的文本写入TXT文件
         with open(txt_path, 'w', encoding='utf-8') as txt_file:
@@ -203,7 +203,7 @@ if __name__ == '__main__':
     ress = extract_common_header(pdf_path)
     print(ress)
     print("-----------------")
-    extract_text_by_page(pdf_path)
+    extract_text_by_page_pypdf2(pdf_path)
     # res=extract_text_by_page_fitz(pdf_path)
     # print(res)磋商文件_tobidders_notice_part2.pdf
     # save_extracted_text_to_txt(file_path,"output.txt")

flask_app/old_version/资格评审前判断_old.py

@@ -1,9 +1,9 @@
-from flask_app.general.读取文件.按页读取pdf import extract_text_by_page
+from flask_app.general.读取文件.按页读取pdf import extract_text_by_page_pypdf2
 
 def check_strings_in_pdf(file_path):
     judge_list=['施工机械设备', '企业信息登记']
     # Read text from PDF
-    text = extract_text_by_page(file_path)  # Assuming this returns all text from the PDF
+    text = extract_text_by_page_pypdf2(file_path)  # Assuming this returns all text from the PDF
     full_text = ''.join(text).replace('\n', '').replace(' ', '')  # Clean up the text
 
     # Initialize the questions list

flask_app/routes/判断是否是招标文件.py

@@ -1,7 +1,7 @@
 import time
 from PyPDF2 import PdfReader  # 确保已安装 PyPDF2: pip install PyPDF2
 from docx import Document
-
+import fitz
 from flask_app.general.llm.通义千问long import upload_file, qianwen_long
 
 def judge_zbfile_exec(file_path):
@@ -12,28 +12,38 @@ def judge_zbfile_exec(file_path):
         start_time = time.time()
         # 检查文件是否为PDF格式
         if file_path.lower().endswith('.pdf'):
-            # 使用 with 语句确保文件关闭
-            with open(file_path, 'rb') as f:
-                reader = PdfReader(f)
-                num_pages = len(reader.pages)
-            if num_pages <= 5:
-                return False
-        elif file_path.lower().endswith('.docx'):
-            doc = Document(file_path)
-            accumulated_text = ""
-            chunk_size = 10  # 每次读取10个段落
-            paragraphs = doc.paragraphs
+            try:
+                with open(file_path, 'rb') as f:
+                    reader = PdfReader(f)
+                    num_pages = len(reader.pages)
+            except Exception:
+                try:
+                    doc = fitz.open(file_path)
+                    num_pages = len(doc)
+                except Exception:
+                    print("PDF 文件读取失败")
+                    return False  # 两种解析方式都失败，直接返回 False
 
-            for i in range(0, len(paragraphs), chunk_size):
-                chunk = paragraphs[i:i + chunk_size]
-                for para in chunk:
-                    accumulated_text += para.text
-                # 判断累计字符数是否已经达到1000字,
-                if len(accumulated_text) >= 1000:
-                    break
-            # 若累计内容不足1000字，则直接返回False
-            if len(accumulated_text) < 1000:
-                return False
+            if num_pages <= 5:
+                return False  # 小于等于 5 页的 PDF 直接判定为非招标文件
+        elif file_path.lower().endswith('.docx'):
+            try:
+                doc = Document(file_path)
+                accumulated_text = ""
+                chunk_size = 10  # 每次读取10个段落
+                paragraphs = doc.paragraphs
+
+                for i in range(0, len(paragraphs), chunk_size):
+                    chunk = paragraphs[i:i + chunk_size]
+                    for para in chunk:
+                        accumulated_text += para.text
+                    if len(accumulated_text) >= 1000:
+                        break  # 读取超过1000字后即可停止
+                if len(accumulated_text) < 1000:
+                    return False  # 若累计内容不足1000字，则直接返回 False
+            except Exception:
+                print("DOCX 文件读取失败，可能为乱码文件")
+                return False  # 解析失败直接返回 False
         # 使用大模型进行判断
         user_query = """该文件是否属于招标文件？如果是的话，请返回'是',如果不是的话，返回'否'。请不要返回其他解释或内容。
         以下是常见的招标文件类型：
@@ -58,7 +68,7 @@ def judge_zbfile_exec(file_path):
 
 if __name__ == '__main__':
     start_time = time.time()
-    pdf_path = r"C:\Users\Administrator\Downloads\file1739842556194.docx"
+    pdf_path = r"C:\Users\Administrator\Desktop\fsdownload\19f53a17-ad4c-43b5-a7ed-981958ec3e0fs\ztbfile.docx"
     res = judge_zbfile_exec(pdf_path)
     if res:
         print("yes")