feat: add subscript & superscript formatting (#319)

* feat: add subscript & superscript formatting

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

* switch to enum

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

---------

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

Author: vagenas

docling_core/transforms/serializer/base.py

@@ -202,6 +202,16 @@ def serialize_strikethrough(self, text: str, **kwargs: Any) -> str:
         """Hook for strikethrough formatting serialization."""
         ...
 
+    @abstractmethod
+    def serialize_subscript(self, text: str, **kwargs: Any) -> str:
+        """Hook for subscript formatting serialization."""
+        ...
+
+    @abstractmethod
+    def serialize_superscript(self, text: str, **kwargs: Any) -> str:
+        """Hook for superscript formatting serialization."""
+        ...
+
     @abstractmethod
     def serialize_hyperlink(
         self,

docling_core/transforms/serializer/common.py

@@ -45,6 +45,7 @@
     PictureDataType,
     PictureItem,
     PictureMoleculeData,
+    Script,
     TableAnnotationType,
     TableItem,
     TextItem,
@@ -455,6 +456,10 @@ def post_process(
                 res = self.serialize_underline(text=res)
             if formatting.strikethrough:
                 res = self.serialize_strikethrough(text=res)
+            if formatting.script == Script.SUB:
+                res = self.serialize_subscript(text=res)
+            elif formatting.script == Script.SUPER:
+                res = self.serialize_superscript(text=res)
         if params.include_hyperlinks and hyperlink:
             res = self.serialize_hyperlink(text=res, hyperlink=hyperlink)
         return res
@@ -479,6 +484,16 @@ def serialize_strikethrough(self, text: str, **kwargs: Any) -> str:
         """Hook for strikethrough formatting serialization."""
         return text
 
+    @override
+    def serialize_subscript(self, text: str, **kwargs: Any) -> str:
+        """Hook for subscript formatting serialization."""
+        return text
+
+    @override
+    def serialize_superscript(self, text: str, **kwargs: Any) -> str:
+        """Hook for superscript formatting serialization."""
+        return text
+
     @override
     def serialize_hyperlink(
         self,

docling_core/transforms/serializer/html.py

@@ -847,6 +847,16 @@ def serialize_strikethrough(self, text: str, **kwargs: Any) -> str:
         """Apply HTML-specific strikethrough serialization."""
         return f"<del>{text}</del>"
 
+    @override
+    def serialize_subscript(self, text: str, **kwargs: Any) -> str:
+        """Apply HTML-specific subscript serialization."""
+        return f"<sub>{text}</sub>"
+
+    @override
+    def serialize_superscript(self, text: str, **kwargs: Any) -> str:
+        """Apply HTML-specific superscript serialization."""
+        return f"<sup>{text}</sup>"
+
     @override
     def serialize_hyperlink(
         self,

docling_core/types/doc/document.py

@@ -829,13 +829,22 @@ def get_annotations(self) -> Sequence[BaseAnnotation]:
         return []
 
 
+class Script(str, Enum):
+    """Text script position."""
+
+    BASELINE = "baseline"
+    SUB = "sub"
+    SUPER = "super"
+
+
 class Formatting(BaseModel):
     """Formatting."""
 
     bold: bool = False
     italic: bool = False
     underline: bool = False
     strikethrough: bool = False
+    script: Script = Script.BASELINE
 
 
 class TextItem(DocItem):

docs/DoclingDocument.json

@@ -554,6 +554,10 @@
           "default": false,
           "title": "Strikethrough",
           "type": "boolean"
+        },
+        "script": {
+          "$ref": "#/$defs/Script",
+          "default": "baseline"
         }
       },
       "title": "Formatting",
@@ -1715,6 +1719,16 @@
       "title": "RefItem",
       "type": "object"
     },
+    "Script": {
+      "description": "Text script position.",
+      "enum": [
+        "baseline",
+        "sub",
+        "super"
+      ],
+      "title": "Script",
+      "type": "string"
+    },
     "SectionHeaderItem": {
       "additionalProperties": false,
       "description": "SectionItem.",