fix: The image uploaded from the workflow knowledge base zip file cannot be parsed (#4505)

Author: shaohuzhang1

apps/common/handle/impl/text/text_split_handle.py

@@ -28,11 +28,13 @@
 
 class TextSplitHandle(BaseSplitHandle):
     def support(self, file, get_buffer):
-        buffer = get_buffer(file)
         file_name: str = file.name.lower()
         if file_name.endswith(".md") or file_name.endswith('.txt') or file_name.endswith('.TXT') or file_name.endswith(
                 '.MD'):
             return True
+        if file_name.index('.') > 0:
+            return False
+        buffer = get_buffer(file)
         result = detect(buffer)
         if result['encoding'] is not None and result['confidence'] is not None and result['encoding'] != 'ascii' and \
                 result['confidence'] > 0.5:

apps/common/handle/impl/text/zip_split_handle.py

@@ -216,20 +216,18 @@ def get_content(self, file, save_image):
                         real_name = get_file_name(zf.name)
                     except Exception:
                         real_name = zf.name
-
                     # 为 split_handle 提供可重复读取的 file-like 对象
                     zf.name = real_name
+                    get_buffer = FileBufferHandle().get_buffer
                     for split_handle in split_handles:
-                        # 准备一个简单的 get_buffer 回调，返回当前 raw
-                        get_buffer = FileBufferHandle().get_buffer
                         if split_handle.support(zf, get_buffer):
                             row = get_buffer(zf)
                             md_text = split_handle.get_content(io.BytesIO(row), save_image)
                             file_content_list.append({'content': md_text, 'name': real_name})
                             break
             for file_content in file_content_list:
                 _image_list, content = get_image_list_by_content(file_content.get('name'), file_content.get("content"),
-                                                                files)
+                                                                 files)
                 content_parts.append(content)
                 for image in _image_list:
                     image_list.append(image)