fix(pdf_parse): Fixed the issue where some headings were missing in certain complex layouts.

magic_pdf/pdf_parse_union_core_v2.py

@@ -436,7 +436,8 @@ def cal_block_index(fix_blocks, sorted_bboxes):
             block_bboxes.append(block['bbox'])
 
             # 删除图表body block中的虚拟line信息, 并用real_lines信息回填
-            if block['type'] in [BlockType.ImageBody, BlockType.TableBody]:
+            if block['type'] in [BlockType.ImageBody, BlockType.TableBody, BlockType.Title, BlockType.InterlineEquation]:
+                if 'real_lines' in block:
                     block['virtual_lines'] = copy.deepcopy(block['lines'])
                     block['lines'] = copy.deepcopy(block['real_lines'])
                     del block['real_lines']