fix: HTML serialization for single image documents (#261)

* fix for HTML serialization for single image documents

Signed-off-by: Peter Staar <[email protected]>

* minor refactor, add test

Signed-off-by: Panos Vagenas <[email protected]>

---------

Signed-off-by: Peter Staar <[email protected]>
Signed-off-by: Panos Vagenas <[email protected]>
Co-authored-by: Panos Vagenas <[email protected]>
Co-authored-by: Panos Vagenas <[email protected]>

Author: PeterStaar-IBM

docling_core/experimental/serializer/html.py

@@ -802,6 +802,8 @@ def serialize_doc(
         ]
 
         if self.params.output_style == HTMLOutputStyle.SPLIT_PAGE:
+            applicable_pages = self._get_applicable_pages()
+
             html_content = "\n".join([p.text for p in parts if p.text])
             next_page: Optional[int] = None
             prev_full_match_end = 0
@@ -814,11 +816,12 @@ def serialize_doc(
             # capture last page
             if next_page is not None:
                 pages[next_page] = html_content[prev_full_match_end:]
+            elif applicable_pages is not None and len(applicable_pages) == 1:
+                pages[applicable_pages[0]] = html_content
 
             html_parts.append("<table>")
             html_parts.append("<tbody>")
 
-            applicable_pages = self._get_applicable_pages()
             for page_no, page in pages.items():
 
                 if isinstance(page_no, int):