feat: serialize subscript and superscript in Markdown export

docling_core/transforms/serializer/markdown.py

@@ -867,6 +867,16 @@ def serialize_strikethrough(self, text: str, **kwargs: Any):
         """Apply Markdown-specific strikethrough serialization."""
         return f"~~{text}~~"
 
+    @override
+    def serialize_subscript(self, text: str, **kwargs: Any):
+        """Apply Markdown-specific subscript serialization."""
+        return f"<sub>{text}</sub>"
+
+    @override
+    def serialize_superscript(self, text: str, **kwargs: Any):
+        """Apply Markdown-specific superscript serialization."""
+        return f"<sup>{text}</sup>"
+
     @override
     def serialize_hyperlink(
         self,

docling_core/transforms/serializer/plain_text.py

@@ -43,8 +43,9 @@ class PlainTextDocSerializer(MarkdownDocSerializer):
     """Document serializer that produces clean plain text.
 
     Strips all Markdown decoration — heading markers, bold/italic/strikethrough
-    markers, and hyperlink syntax — while keeping list bullets (``-``), ordered
-    list numbers, and table-cell separators (``|``) intact.
+    markers, sub/superscript tags, and hyperlink syntax — while keeping list
+    bullets (``-``), ordered list numbers, and table-cell separators (``|``)
+    intact.
     """
 
     text_serializer: BaseTextSerializer = PlainTextTextSerializer()
@@ -65,6 +66,16 @@ def serialize_strikethrough(self, text: str, **kwargs: Any) -> str:
         """Apply plain-text strikethrough serialization."""
         return text
 
+    @override
+    def serialize_subscript(self, text: str, **kwargs: Any) -> str:
+        """Apply plain-text subscript serialization."""
+        return text
+
+    @override
+    def serialize_superscript(self, text: str, **kwargs: Any) -> str:
+        """Apply plain-text superscript serialization."""
+        return text
+
     @override
     def serialize_hyperlink(
         self,

test/data/doc/constructed_doc.embedded.md.gt

@@ -61,7 +61,7 @@ $$E=mc^2$$
 
 <!-- missing-form-item -->
 
-Some formatting chops: **bold** *italic* underline ~~strikethrough~~ subscript superscript [hyperlink](.) &amp; [~~***everything at the same time.***~~](https://github.com/DS4SD/docling)
+Some formatting chops: **bold** *italic* underline ~~strikethrough~~ <sub>subscript</sub> <sup>superscript</sup> [hyperlink](.) &amp; [~~***everything at the same time.***~~](https://github.com/DS4SD/docling)
 
 - (i) Item 1 in A
 - (ii) Item 2 in A

test/data/doc/constructed_doc.placeholder.md.gt

@@ -61,7 +61,7 @@ $$E=mc^2$$
 
 <!-- missing-form-item -->
 
-Some formatting chops: **bold** *italic* underline ~~strikethrough~~ subscript superscript [hyperlink](.) &amp; [~~***everything at the same time.***~~](https://github.com/DS4SD/docling)
+Some formatting chops: **bold** *italic* underline ~~strikethrough~~ <sub>subscript</sub> <sup>superscript</sup> [hyperlink](.) &amp; [~~***everything at the same time.***~~](https://github.com/DS4SD/docling)
 
 - (i) Item 1 in A
 - (ii) Item 2 in A

test/data/doc/constructed_doc.referenced.md.gt

@@ -61,7 +61,7 @@ $$E=mc^2$$
 
 <!-- missing-form-item -->
 
-Some formatting chops: **bold** *italic* underline ~~strikethrough~~ subscript superscript [hyperlink](.) &amp; [~~***everything at the same time.***~~](https://github.com/DS4SD/docling)
+Some formatting chops: **bold** *italic* underline ~~strikethrough~~ <sub>subscript</sub> <sup>superscript</sup> [hyperlink](.) &amp; [~~***everything at the same time.***~~](https://github.com/DS4SD/docling)
 
 - (i) Item 1 in A
 - (ii) Item 2 in A

test/data/doc/constructed_document.yaml.md

@@ -61,7 +61,7 @@ $$E=mc^2$$
 
 <!-- missing-form-item -->
 
-Some formatting chops: **bold** *italic* underline ~~strikethrough~~ subscript superscript [hyperlink](.) &amp; [~~***everything at the same time.***~~](https://github.com/DS4SD/docling)
+Some formatting chops: **bold** *italic* underline ~~strikethrough~~ <sub>subscript</sub> <sup>superscript</sup> [hyperlink](.) &amp; [~~***everything at the same time.***~~](https://github.com/DS4SD/docling)
 
 - (i) Item 1 in A
 - (ii) Item 2 in A

test/data/doc/constructed_legacy_annot_mark_false.gt.md

@@ -63,7 +63,7 @@ $$E=mc^2$$
 
 <!-- missing-form-item -->
 
-Some formatting chops: **bold** *italic* underline ~~strikethrough~~ subscript superscript [hyperlink](.) &amp; [~~***everything at the same time.***~~](https://github.com/DS4SD/docling)
+Some formatting chops: **bold** *italic* underline ~~strikethrough~~ <sub>subscript</sub> <sup>superscript</sup> [hyperlink](.) &amp; [~~***everything at the same time.***~~](https://github.com/DS4SD/docling)
 
 - (i) Item 1 in A
 - (ii) Item 2 in A

test/data/doc/constructed_legacy_annot_mark_true.gt.md

@@ -63,7 +63,7 @@ $$E=mc^2$$
 
 <!-- missing-form-item -->
 
-Some formatting chops: **bold** *italic* underline ~~strikethrough~~ subscript superscript [hyperlink](.) &amp; [~~***everything at the same time.***~~](https://github.com/DS4SD/docling)
+Some formatting chops: **bold** *italic* underline ~~strikethrough~~ <sub>subscript</sub> <sup>superscript</sup> [hyperlink](.) &amp; [~~***everything at the same time.***~~](https://github.com/DS4SD/docling)
 
 - (i) Item 1 in A
 - (ii) Item 2 in A

test/data/doc/constructed_mode_always_valid_false.gt.md

@@ -61,7 +61,7 @@ $$E=mc^2$$
 
 <!-- missing-form-item -->
 
-Some formatting chops: **bold** *italic* underline ~~strikethrough~~ subscript superscript [hyperlink](.) &amp; [~~***everything at the same time.***~~](https://github.com/DS4SD/docling)
+Some formatting chops: **bold** *italic* underline ~~strikethrough~~ <sub>subscript</sub> <sup>superscript</sup> [hyperlink](.) &amp; [~~***everything at the same time.***~~](https://github.com/DS4SD/docling)
 
 (i) Item 1 in A
 (ii) Item 2 in A

test/data/doc/constructed_mode_always_valid_true.gt.md

@@ -61,7 +61,7 @@ $$E=mc^2$$
 
 <!-- missing-form-item -->
 
-Some formatting chops: **bold** *italic* underline ~~strikethrough~~ subscript superscript [hyperlink](.) &amp; [~~***everything at the same time.***~~](https://github.com/DS4SD/docling)
+Some formatting chops: **bold** *italic* underline ~~strikethrough~~ <sub>subscript</sub> <sup>superscript</sup> [hyperlink](.) &amp; [~~***everything at the same time.***~~](https://github.com/DS4SD/docling)
 
 1. (i) Item 1 in A
 2. (ii) Item 2 in A

test/data/doc/constructed_mode_auto_valid_false.gt.md

@@ -61,7 +61,7 @@ $$E=mc^2$$
 
 <!-- missing-form-item -->
 
-Some formatting chops: **bold** *italic* underline ~~strikethrough~~ subscript superscript [hyperlink](.) &amp; [~~***everything at the same time.***~~](https://github.com/DS4SD/docling)
+Some formatting chops: **bold** *italic* underline ~~strikethrough~~ <sub>subscript</sub> <sup>superscript</sup> [hyperlink](.) &amp; [~~***everything at the same time.***~~](https://github.com/DS4SD/docling)
 
 (i) Item 1 in A
 (ii) Item 2 in A

test/data/doc/constructed_mode_auto_valid_true.gt.md

@@ -61,7 +61,7 @@ $$E=mc^2$$
 
 <!-- missing-form-item -->
 
-Some formatting chops: **bold** *italic* underline ~~strikethrough~~ subscript superscript [hyperlink](.) &amp; [~~***everything at the same time.***~~](https://github.com/DS4SD/docling)
+Some formatting chops: **bold** *italic* underline ~~strikethrough~~ <sub>subscript</sub> <sup>superscript</sup> [hyperlink](.) &amp; [~~***everything at the same time.***~~](https://github.com/DS4SD/docling)
 
 - (i) Item 1 in A
 - (ii) Item 2 in A

test/data/doc/constructed_mode_never_valid_false.gt.md

@@ -61,7 +61,7 @@ $$E=mc^2$$
 
 <!-- missing-form-item -->
 
-Some formatting chops: **bold** *italic* underline ~~strikethrough~~ subscript superscript [hyperlink](.) &amp; [~~***everything at the same time.***~~](https://github.com/DS4SD/docling)
+Some formatting chops: **bold** *italic* underline ~~strikethrough~~ <sub>subscript</sub> <sup>superscript</sup> [hyperlink](.) &amp; [~~***everything at the same time.***~~](https://github.com/DS4SD/docling)
 
 Item 1 in A
 Item 2 in A

test/data/doc/constructed_mode_never_valid_true.gt.md

@@ -61,7 +61,7 @@ $$E=mc^2$$
 
 <!-- missing-form-item -->
 
-Some formatting chops: **bold** *italic* underline ~~strikethrough~~ subscript superscript [hyperlink](.) &amp; [~~***everything at the same time.***~~](https://github.com/DS4SD/docling)
+Some formatting chops: **bold** *italic* underline ~~strikethrough~~ <sub>subscript</sub> <sup>superscript</sup> [hyperlink](.) &amp; [~~***everything at the same time.***~~](https://github.com/DS4SD/docling)
 
 1. Item 1 in A
 2. Item 2 in A

test/test_serialization.py

@@ -24,7 +24,7 @@
 )
 from docling_core.transforms.serializer.webvtt import WebVTTDocSerializer, WebVTTParams
 from docling_core.transforms.visualizer.layout_visualizer import LayoutVisualizer
-from docling_core.types.doc import DoclingDocument
+from docling_core.types.doc import DoclingDocument, Formatting, Script
 from docling_core.types.doc.base import ImageRefMode
 from docling_core.types.doc.document import (
     BaseMeta,
@@ -204,6 +204,32 @@ def test_md_inline_and_formatting():
     verify(exp_file=src.with_suffix(".gt.md"), actual=actual)
 
 
+def test_md_subscript_formatting():
+    """Subscript spans are exported as inline <sub> HTML in Markdown."""
+    doc = DoclingDocument(name="test")
+    doc.add_text(
+        label=DocItemLabel.TEXT,
+        text="H2O",
+        formatting=Formatting(script=Script.SUB),
+    )
+
+    actual = MarkdownDocSerializer(doc=doc).serialize().text
+    assert "<sub>H2O</sub>" in actual
+
+
+def test_md_superscript_formatting():
+    """Superscript spans are exported as inline <sup> HTML in Markdown."""
+    doc = DoclingDocument(name="test")
+    doc.add_text(
+        label=DocItemLabel.TEXT,
+        text="2",
+        formatting=Formatting(script=Script.SUPER),
+    )
+
+    actual = MarkdownDocSerializer(doc=doc).serialize().text
+    assert "<sup>2</sup>" in actual
+
+
 def test_md_pb_placeholder_and_page_filter():
     src = Path("./test/data/doc/2408.09869v3_enriched.json")
     doc = DoclingDocument.load_from_json(src)