Introduce ListItem API, with marker and enumerated properties

Signed-off-by: Christoph Auer <cau@zurich.ibm.com>

docling_core/types/experimental/document.py

@@ -143,6 +143,14 @@ class DocumentOrigin(BaseModel):
         # from any file handler protocol (e.g. https://, file://, s3://)
     )
 
+    _extra_mimetypes = [
+        "application/vnd.openxmlformats-officedocument.wordprocessingml.document",
+        "application/vnd.openxmlformats-officedocument.wordprocessingml.template",
+        "application/vnd.openxmlformats-officedocument.presentationml.template",
+        "application/vnd.openxmlformats-officedocument.presentationml.slideshow",
+        "application/vnd.openxmlformats-officedocument.presentationml.presentation",
+    ]
+
     @field_validator("binary_hash", mode="before")
     @classmethod
     def parse_hex_string(cls, value):
@@ -164,7 +172,7 @@ def parse_hex_string(cls, value):
     def validate_mimetype(cls, v):
         """validate_mimetype."""
         # Check if the provided MIME type is valid using mimetypes module
-        if v not in mimetypes.types_map.values():
+        if v not in mimetypes.types_map.values() and v not in cls._extra_mimetypes:
             raise ValueError(f"'{v}' is not a valid MIME type")
         return v
 
@@ -347,6 +355,14 @@ class SectionHeaderItem(TextItem):
     level: LevelNumber
 
 
+class ListItem(TextItem):
+    """SectionItem."""
+
+    label: typing.Literal[DocItemLabel.LIST_ITEM] = DocItemLabel.LIST_ITEM
+    enumerated: bool = False
+    marker: str  # The bullet or number symbol that prefixes this list item
+
+
 class FloatingItem(DocItem):
     """FloatingItem."""
 
@@ -689,7 +705,7 @@ class DoclingDocument(BaseModel):
     body: GroupItem = GroupItem(name="_root_", self_ref="#/body")  # List[RefItem] = []
 
     groups: List[GroupItem] = []
-    texts: List[Union[SectionHeaderItem, TextItem]] = []
+    texts: List[Union[SectionHeaderItem, ListItem, TextItem]] = []
     pictures: List[PictureItem] = []
     tables: List[TableItem] = []
     key_value_items: List[KeyValueItem] = []
@@ -726,6 +742,50 @@ def add_group(
 
         return group
 
+    def add_list_item(
+        self,
+        text: str,
+        enumerated: bool = False,
+        marker: Optional[str] = None,
+        orig: Optional[str] = None,
+        prov: Optional[ProvenanceItem] = None,
+        parent: Optional[GroupItem] = None,
+    ):
+        """add_paragraph.
+
+        :param label: str:
+        :param text: str:
+        :param orig: Optional[str]:  (Default value = None)
+        :param prov: Optional[ProvenanceItem]:  (Default value = None)
+        :param parent: Optional[GroupItem]:  (Default value = None)
+
+        """
+        if not parent:
+            parent = self.body
+
+        if not orig:
+            orig = text
+
+        marker = marker or "-"
+
+        text_index = len(self.texts)
+        cref = f"#/texts/{text_index}"
+        list_item = ListItem(
+            text=text,
+            orig=orig,
+            self_ref=cref,
+            parent=parent.get_ref(),
+            enumerated=enumerated,
+            marker=marker,
+        )
+        if prov:
+            list_item.prov.append(prov)
+
+        self.texts.append(list_item)
+        parent.children.append(RefItem(cref=cref))
+
+        return list_item
+
     def add_text(
         self,
         label: str,
@@ -1061,6 +1121,15 @@ def export_to_markdown(  # noqa: C901
                         else:
                             markdown_text = f"## {text}"
 
+                    # secondary titles
+                    elif isinstance(item, ListItem):
+                        if item.enumerated:
+                            marker = item.marker
+                        else:
+                            marker = "-"
+
+                        markdown_text = f"{marker} {text}"
+
                     # normal text
                     else:
                         markdown_text = text

docling_core/types/experimental/labels.py

@@ -34,7 +34,10 @@ class GroupLabel(str, Enum):
     """GroupLabel."""
 
     UNSPECIFIED = "unspecified"
-    LIST = "list"  # group label for list container (not the list-items)
+    LIST = (
+        "list"  # group label for list container (not the list-items) (e.g. HTML <ul/>)
+    )
+    ORDERED_LIST = "ordered_list"  # List with enumeration (e.g. HTML <ol/>)
     CHAPTER = "chapter"
     SECTION = "section"
     SHEET = "sheet"

test/data/docling_document/unit/ListItem.yaml

@@ -0,0 +1,9 @@
+children: []
+enumerated: true
+label: list_item
+marker: (1)
+orig: whatever
+parent: null
+prov: []
+self_ref: '#'
+text: whatever

test/test_docling_doc.py

@@ -13,6 +13,7 @@
     DoclingDocument,
     FloatingItem,
     KeyValueItem,
+    ListItem,
     PictureItem,
     SectionHeaderItem,
     TableCell,
@@ -52,9 +53,9 @@ def read(name: str):
         return gold
 
     def verify(dc, obj):
-        pred = serialise(obj)
+        pred = serialise(obj).strip()
         # print(f"\t{dc.__name__}:\n {pred}")
-        gold = read(dc.__name__)
+        gold = read(dc.__name__).strip()
 
         assert pred == gold, f"pred!=gold for {dc.__name__}"
 
@@ -70,7 +71,15 @@ def verify(dc, obj):
                 self_ref="#",
             )
             verify(dc, obj)
-
+        elif dc is ListItem:
+            obj = dc(
+                text="whatever",
+                orig="whatever",
+                marker="(1)",
+                enumerated=True,
+                self_ref="#",
+            )
+            verify(dc, obj)
         elif dc is FloatingItem:
             obj = dc(
                 label=DocItemLabel.TEXT,