ocr_mk_mm_markdown_with_para和ocr_mk_mm_markdown_with_para_and_pagination逻辑优化

magic_pdf/dict2md/ocr_mkcontent.py

@@ -72,36 +72,27 @@ def ocr_mk_mm_markdown_with_para(pdf_info_dict: dict):
     markdown = []
     for _, page_info in pdf_info_dict.items():
         paras_of_layout = page_info.get("para_blocks")
-        if not paras_of_layout:
-            continue
-        for paras in paras_of_layout:
-            for para in paras:
-                para_text = ''
-                for line in para:
-                    for span in line['spans']:
-                        span_type = span.get('type')
-                        if span_type == ContentType.Text:
-                            content = split_long_words(span['content'])
-                            pass
-                        elif span_type == ContentType.InlineEquation:
-                            content = f" ${span['content']}$ "
-                        elif span_type == ContentType.InterlineEquation:
-                            content = f"\n$$\n{span['content']}\n$$\n"
-                        elif span_type in [ ContentType.Image, ContentType.Table ]:
-                            content = f"\n![]({join_path(s3_image_save_path, span['image_path'])})\n"
-                        para_text += content + ' '
-                markdown.append(para_text.strip() + '  ')
-
+        page_markdown = ocr_mk_mm_markdown_with_para_core(paras_of_layout)
+        markdown.extend(page_markdown)
     return '\n\n'.join(markdown)
 
 
 def ocr_mk_mm_markdown_with_para_and_pagination(pdf_info_dict: dict):
     markdown_with_para_and_pagination = []
     for page_no, page_info in pdf_info_dict.items():
-        page_markdown = []
         paras_of_layout = page_info.get("para_blocks")
         if not paras_of_layout:
             continue
+        page_markdown = ocr_mk_mm_markdown_with_para_core(paras_of_layout)
+        markdown_with_para_and_pagination.append({
+            'page_no': page_no,
+            'md_content': '\n\n'.join(page_markdown)
+        })
+    return markdown_with_para_and_pagination
+
+
+def ocr_mk_mm_markdown_with_para_core(paras_of_layout):
+    page_markdown = []
     for paras in paras_of_layout:
         for para in paras:
             para_text = ''
@@ -119,11 +110,7 @@ def ocr_mk_mm_markdown_with_para_and_pagination(pdf_info_dict: dict):
                         content = f"\n![]({join_path(s3_image_save_path, span['image_path'])})\n"
                     para_text += content + ' '
             page_markdown.append(para_text.strip() + '  ')
-        markdown_with_para_and_pagination.append({
-            'page_no': page_no,
-            'md_content': '\n\n'.join(page_markdown)
-        })
-    return markdown_with_para_and_pagination
+    return page_markdown
 
 
 def make_standard_format_with_para(pdf_info_dict: dict):