docling-core 2.45.0__py3-none-any.whl → 2.47.0__py3-none-any.whl

docling_core/transforms/serializer/common.py

@@ -359,6 +359,7 @@ class DocSerializer(BaseModel, BaseDocSerializer):
                 item=item,
                 doc_serializer=self,
                 doc=self.doc,
+                visited=my_visited,
                 **my_kwargs,
             )
         elif isinstance(item, PictureItem):

docling_core/transforms/serializer/doctags.py

@@ -157,6 +157,7 @@ class DocTagsTableSerializer(BaseTableSerializer):
         item: TableItem,
         doc_serializer: BaseDocSerializer,
         doc: DoclingDocument,
+        visited: Optional[set[str]] = None,
         **kwargs: Any,
     ) -> SerializationResult:
         """Serializes the passed item."""
@@ -179,6 +180,7 @@ class DocTagsTableSerializer(BaseTableSerializer):
                 add_cell_text=params.add_table_cell_text,
                 xsize=params.xsize,
                 ysize=params.ysize,
+                visited=visited,
             )
             res_parts.append(create_ser_result(text=otsl_text, span_source=item))
 

docling_core/transforms/serializer/html.py

@@ -65,8 +65,8 @@ from docling_core.types.doc.document import (
     PictureItem,
     PictureMoleculeData,
     PictureTabularChartData,
+    RichTableCell,
     SectionHeaderItem,
-    TableCell,
     TableItem,
     TextItem,
     TitleItem,
@@ -346,9 +346,6 @@ class HTMLTableSerializer(BaseTableSerializer):
         **kwargs: Any,
     ) -> SerializationResult:
         """Serializes the passed table item to HTML."""
-        nrows = item.data.num_rows
-        ncols = item.data.num_cols
-
         res_parts: list[SerializationResult] = []
         cap_res = doc_serializer.serialize_captions(item=item, tag="caption", **kwargs)
         if cap_res.text:
@@ -356,11 +353,11 @@ class HTMLTableSerializer(BaseTableSerializer):
 
         if item.self_ref not in doc_serializer.get_excluded_refs(**kwargs):
             body = ""
+            span_source: Union[DocItem, list[SerializationResult]] = []
 
-            for i in range(nrows):
+            for i, row in enumerate(item.data.grid):
                 body += "<tr>"
-                for j in range(ncols):
-                    cell: TableCell = item.data.grid[i][j]
+                for j, cell in enumerate(row):
 
                     rowspan, rowstart = (
                         cell.row_span,
@@ -376,7 +373,16 @@ class HTMLTableSerializer(BaseTableSerializer):
                     if colstart != j:
                         continue
 
-                    content = html.escape(cell.text.strip())
+                    if isinstance(cell, RichTableCell):
+                        ser_res = doc_serializer.serialize(
+                            item=cell.ref.resolve(doc=doc), **kwargs
+                        )
+                        content = ser_res.text
+                        span_source = [ser_res]
+                    else:
+                        content = html.escape(cell.text.strip())
+                        span_source = item
+
                     celltag = "td"
                     if cell.column_header or cell.row_header or cell.row_section:
                         celltag = "th"
@@ -389,14 +395,14 @@ class HTMLTableSerializer(BaseTableSerializer):
 
                     text_dir = get_text_direction(content)
                     if text_dir == "rtl":
-                        opening_tag += f' dir="{dir}"'
+                        opening_tag += f' dir="{text_dir}"'
 
                     body += f"<{opening_tag}>{content}</{celltag}>"
                 body += "</tr>"
 
             if body:
                 body = f"<tbody>{body}</tbody>"
-                res_parts.append(create_ser_result(text=body, span_source=item))
+                res_parts.append(create_ser_result(text=body, span_source=span_source))
 
         text_res = "".join([r.text for r in res_parts])
         text_res = f"<table>{text_res}</table>" if text_res else ""

docling_core/transforms/serializer/markdown.py

@@ -55,6 +55,7 @@ from docling_core.types.doc.document import (
     PictureItem,
     PictureMoleculeData,
     PictureTabularChartData,
+    RichTableCell,
     SectionHeaderItem,
     TableItem,
     TextItem,
@@ -320,7 +321,13 @@ class MarkdownTableSerializer(BaseTableSerializer):
                 [
                     # make sure that md tables are not broken
                     # due to newline chars in the text
-                    col.text.replace("\n", " ")
+                    (
+                        doc_serializer.serialize(
+                            item=col.ref.resolve(doc=doc), **kwargs
+                        ).text
+                        if isinstance(col, RichTableCell)
+                        else col.text
+                    ).replace("\n", " ")
                     for col in row
                 ]
                 for row in item.data.grid

docling_core/types/doc/__init__.py

@@ -7,6 +7,7 @@
 
 from .base import BoundingBox, CoordOrigin, ImageRefMode, Size
 from .document import (
+    AnyTableCell,
     BaseAnnotation,
     ChartBar,
     ChartLine,
@@ -52,6 +53,7 @@ from .document import (
     PictureTabularChartData,
     ProvenanceItem,
     RefItem,
+    RichTableCell,
     Script,
     SectionHeaderItem,
     TableCell,

docling_core/types/doc/document.py

@@ -34,7 +34,7 @@ from pydantic import (
     validate_call,
 )
 from tabulate import tabulate
-from typing_extensions import Annotated, Self, deprecated
+from typing_extensions import Annotated, Self, deprecated, override
 
 from docling_core.search.package import VERSION_PATTERN
 from docling_core.types.base import _JSON_POINTER_REGEX
@@ -60,7 +60,7 @@ _logger = logging.getLogger(__name__)
 
 Uint64 = typing.Annotated[int, Field(ge=0, le=(2**64 - 1))]
 LevelNumber = typing.Annotated[int, Field(ge=1, le=100)]
-CURRENT_VERSION: Final = "1.5.0"
+CURRENT_VERSION: Final = "1.6.0"
 
 DEFAULT_EXPORT_LABELS = {
     DocItemLabel.TITLE,
@@ -325,7 +325,7 @@ class TableCell(BaseModel):
                 in data
             ):
                 return data
-            text = data["bbox"].get("token", "")
+            text = data.get("bbox", {}).get("token", "")
             if not len(text):
                 text_cells = data.pop("text_cell_bboxes", None)
                 if text_cells:
@@ -337,11 +337,39 @@ class TableCell(BaseModel):
 
         return data
 
+    def _get_text(self, doc: Optional["DoclingDocument"] = None, **kwargs: Any) -> str:
+        return self.text
+
+
+class RichTableCell(TableCell):
+    """RichTableCell."""
+
+    ref: "RefItem"
+
+    @override
+    def _get_text(self, doc: Optional["DoclingDocument"] = None, **kwargs: Any) -> str:
+        from docling_core.transforms.serializer.markdown import MarkdownDocSerializer
+
+        if doc is not None:
+            doc_serializer = kwargs.pop(
+                "doc_serializer", MarkdownDocSerializer(doc=doc)
+            )
+            ser_res = doc_serializer.serialize(item=self.ref.resolve(doc=doc), **kwargs)
+            return ser_res.text
+        else:
+            return "<!-- rich cell -->"
+
+
+AnyTableCell = Annotated[
+    Union[RichTableCell, TableCell],
+    Field(union_mode="left_to_right"),
+]
+
 
 class TableData(BaseModel):  # TBD
     """BaseTableData."""
 
-    table_cells: List[TableCell] = []
+    table_cells: List[AnyTableCell] = []
     num_rows: int = 0
     num_cols: int = 0
 
@@ -380,7 +408,9 @@ class TableData(BaseModel):  # TBD
 
         return table_data
 
-    def remove_rows(self, indices: List[int]) -> List[List[TableCell]]:
+    def remove_rows(
+        self, indices: List[int], doc: Optional["DoclingDocument"] = None
+    ) -> List[List[TableCell]]:
         """Remove rows from the table by their indices.
 
         :param indices: List[int]: A list of indices of the rows to remove. (Starting from 0)
@@ -392,6 +422,7 @@ class TableData(BaseModel):  # TBD
 
         indices = sorted(indices, reverse=True)
 
+        refs_to_remove = []
         all_removed_cells = []
         for row_index in indices:
             if row_index < 0 or row_index >= self.num_rows:
@@ -403,6 +434,10 @@ class TableData(BaseModel):  # TBD
             end_idx = start_idx + self.num_cols
             removed_cells = self.table_cells[start_idx:end_idx]
 
+            for cell in removed_cells:
+                if isinstance(cell, RichTableCell):
+                    refs_to_remove.append(cell.ref)
+
             # Remove the cells from the table
             self.table_cells = self.table_cells[:start_idx] + self.table_cells[end_idx:]
 
@@ -417,9 +452,18 @@ class TableData(BaseModel):  # TBD
 
             all_removed_cells.append(removed_cells)
 
+        if refs_to_remove:
+            if doc is None:
+                _logger.warning(
+                    "When table contains rich cells, `doc` argument must be provided, "
+                    "otherwise rich cell content will be left dangling."
+                )
+            else:
+                doc._delete_items(refs_to_remove)
+
         return all_removed_cells
 
-    def pop_row(self) -> List[TableCell]:
+    def pop_row(self, doc: Optional["DoclingDocument"] = None) -> List[TableCell]:
         """Remove and return the last row from the table.
 
         :returns: List[TableCell]: A list of TableCell objects representing the popped row.
@@ -427,16 +471,18 @@ class TableData(BaseModel):  # TBD
         if self.num_rows == 0:
             raise IndexError("Cannot pop from an empty table.")
 
-        return self.remove_row(self.num_rows - 1)
+        return self.remove_row(self.num_rows - 1, doc=doc)
 
-    def remove_row(self, row_index: int) -> List[TableCell]:
+    def remove_row(
+        self, row_index: int, doc: Optional["DoclingDocument"] = None
+    ) -> List[TableCell]:
         """Remove a row from the table by its index.
 
         :param row_index: int: The index of the row to remove. (Starting from 0)
 
         :returns: List[TableCell]: A list of TableCell objects representing the removed row.
         """
-        return self.remove_rows([row_index])[0]
+        return self.remove_rows([row_index], doc=doc)[0]
 
     def insert_rows(
         self, row_index: int, rows: List[List[str]], after: bool = False
@@ -1509,8 +1555,15 @@ class TableItem(FloatingItem):
 
     annotations: List[TableAnnotationType] = []
 
-    def export_to_dataframe(self) -> pd.DataFrame:
+    def export_to_dataframe(
+        self, doc: Optional["DoclingDocument"] = None
+    ) -> pd.DataFrame:
         """Export the table as a Pandas DataFrame."""
+        if doc is None:
+            _logger.warning(
+                "Usage of TableItem.export_to_dataframe() without `doc` argument is deprecated."
+            )
+
         if self.data.num_rows == 0 or self.data.num_cols == 0:
             return pd.DataFrame()
 
@@ -1539,14 +1592,15 @@ class TableItem(FloatingItem):
             columns = ["" for _ in range(self.data.num_cols)]
             for i in range(num_headers):
                 for j, cell in enumerate(self.data.grid[i]):
-                    col_name = cell.text
+                    col_name = cell._get_text(doc=doc)
                     if columns[j] != "":
                         col_name = f".{col_name}"
                     columns[j] += col_name
 
         # Create table data
         table_data = [
-            [cell.text for cell in row] for row in self.data.grid[num_headers:]
+            [cell._get_text(doc=doc) for cell in row]
+            for row in self.data.grid[num_headers:]
         ]
 
         # Create DataFrame
@@ -1577,7 +1631,7 @@ class TableItem(FloatingItem):
 
                     # make sure that md tables are not broken
                     # due to newline chars in the text
-                    text = col.text
+                    text = col._get_text(doc=doc)
                     text = text.replace("\n", " ")
                     tmp.append(text)
 
@@ -1623,6 +1677,7 @@ class TableItem(FloatingItem):
         add_cell_text: bool = True,
         xsize: int = 500,
         ysize: int = 500,
+        **kwargs: Any,
     ) -> str:
         """Export the table as OTSL."""
         # Possible OTSL tokens...
@@ -1639,6 +1694,9 @@ class TableItem(FloatingItem):
         # Headers (column, row, section row):
         # "ched", "rhed", "srow"
 
+        from docling_core.transforms.serializer.doctags import DocTagsDocSerializer
+
+        doc_serializer = DocTagsDocSerializer(doc=doc)
         body = []
         nrows = self.data.num_rows
         ncols = self.data.num_cols
@@ -1652,7 +1710,9 @@ class TableItem(FloatingItem):
         for i in range(nrows):
             for j in range(ncols):
                 cell: TableCell = self.data.grid[i][j]
-                content = cell.text.strip()
+                content = cell._get_text(
+                    doc=doc, doc_serializer=doc_serializer, **kwargs
+                ).strip()
                 rowspan, rowstart = (
                     cell.row_span,
                     cell.start_row_offset_idx,
@@ -2304,6 +2364,15 @@ class DoclingDocument(BaseModel):
                 refs_to_be_deleted=refs_to_be_deleted,
                 lookup=lookup,
             )
+            if isinstance(node, TableItem):
+                for cell in node.data.table_cells:
+                    if isinstance(cell, RichTableCell):
+                        path = cell.ref._split_ref_to_path()
+                        cell.ref = self._update_ref_with_lookup(
+                            item_label=path[1],
+                            item_index=int(path[2]),
+                            lookup=lookup,
+                        )
 
         # Update the self_ref reference
         if node.parent is not None:
@@ -3945,16 +4014,22 @@ class DoclingDocument(BaseModel):
         """num_pages."""
         return len(self.pages.values())
 
-    def validate_tree(self, root) -> bool:
+    def validate_tree(self, root: NodeItem) -> bool:
         """validate_tree."""
-        res = []
         for child_ref in root.children:
             child = child_ref.resolve(self)
-            if child.parent.resolve(self) != root:
+            if child.parent.resolve(self) != root or not self.validate_tree(child):
                 return False
-            res.append(self.validate_tree(child))
 
-        return all(res) or len(res) == 0
+        if isinstance(root, TableItem):
+            for cell in root.data.table_cells:
+                if isinstance(cell, RichTableCell) and (
+                    (par_ref := cell.ref.resolve(self).parent) is None
+                    or par_ref.resolve(self) != root
+                ):
+                    return False
+
+        return True
 
     def iterate_items(
         self,
@@ -3963,14 +4038,14 @@ class DoclingDocument(BaseModel):
         traverse_pictures: bool = False,
         page_no: Optional[int] = None,
         included_content_layers: Optional[set[ContentLayer]] = None,
-        _level: int = 0,  # fixed parameter, carries through the node nesting level
+        _level: int = 0,  # deprecated
     ) -> typing.Iterable[Tuple[NodeItem, int]]:  # tuple of node and level
         """Iterate elements with level."""
         for item, stack in self._iterate_items_with_stack(
             root=root,
             with_groups=with_groups,
             traverse_pictures=traverse_pictures,
-            page_no=page_no,
+            page_nrs={page_no} if page_no is not None else None,
             included_content_layers=included_content_layers,
         ):
             yield item, len(stack)
@@ -3980,7 +4055,7 @@ class DoclingDocument(BaseModel):
         root: Optional[NodeItem] = None,
         with_groups: bool = False,
         traverse_pictures: bool = False,
-        page_no: Optional[int] = None,
+        page_nrs: Optional[set[int]] = None,
         included_content_layers: Optional[set[ContentLayer]] = None,
         _stack: Optional[list[int]] = None,
     ) -> typing.Iterable[Tuple[NodeItem, list[int]]]:  # tuple of node and level
@@ -4003,8 +4078,8 @@ class DoclingDocument(BaseModel):
             and (
                 not isinstance(root, DocItem)
                 or (
-                    page_no is None
-                    or any(prov.page_no == page_no for prov in root.prov)
+                    page_nrs is None
+                    or any(prov.page_no in page_nrs for prov in root.prov)
                 )
             )
             and root.content_layer in my_layers
@@ -4038,7 +4113,7 @@ class DoclingDocument(BaseModel):
                     child,
                     with_groups=with_groups,
                     traverse_pictures=traverse_pictures,
-                    page_no=page_no,
+                    page_nrs=page_nrs,
                     _stack=my_stack,
                     included_content_layers=my_layers,
                 )
@@ -5324,7 +5399,9 @@ class DoclingDocument(BaseModel):
                         grid.append([])
                         for j, cell in enumerate(row):
                             if j < 10:
-                                text = get_text(text=cell.text, max_text_len=16)
+                                text = get_text(
+                                    cell._get_text(doc=self), max_text_len=16
+                                )
                                 grid[-1].append(text)
 
                     result.append("\n" + tabulate(grid) + "\n")
@@ -5526,7 +5603,9 @@ class DoclingDocument(BaseModel):
         def get_item_list(self, key: str) -> list[NodeItem]:
             return getattr(self, key)
 
-        def index(self, doc: "DoclingDocument") -> None:
+        def index(
+            self, doc: "DoclingDocument", page_nrs: Optional[set[int]] = None
+        ) -> None:
 
             orig_ref_to_new_ref: dict[str, str] = {}
             page_delta = self._max_page - min(doc.pages.keys()) + 1 if doc.pages else 0
@@ -5537,10 +5616,11 @@ class DoclingDocument(BaseModel):
             self._names.append(doc.name)
 
             # collect items in traversal order
-            for item, _ in doc.iterate_items(
+            for item, _ in doc._iterate_items_with_stack(
                 with_groups=True,
                 traverse_pictures=True,
                 included_content_layers={c for c in ContentLayer},
+                page_nrs=page_nrs,
             ):
                 key = item.self_ref.split("/")[1]
                 is_body = key == "body"
@@ -5588,6 +5668,16 @@ class DoclingDocument(BaseModel):
                                         )
                                         break
 
+                            # update rich table cells references:
+                            if isinstance(parent_item, TableItem):
+                                for cell in parent_item.data.table_cells:
+                                    if (
+                                        isinstance(cell, RichTableCell)
+                                        and cell.ref.cref == item.self_ref
+                                    ):
+                                        cell.ref.cref = new_cref
+                                        break
+
                         elif num_components == 2 and path_components[1] == "body":
                             parent_item = self._body
                         else:
@@ -5599,12 +5689,13 @@ class DoclingDocument(BaseModel):
             # update pages
             new_max_page = None
             for page_nr in doc.pages:
-                new_page = copy.deepcopy(doc.pages[page_nr])
-                new_page_nr = page_nr + page_delta
-                new_page.page_no = new_page_nr
-                self.pages[new_page_nr] = new_page
-                if new_max_page is None or new_page_nr > new_max_page:
-                    new_max_page = new_page_nr
+                if page_nrs is None or page_nr in page_nrs:
+                    new_page = copy.deepcopy(doc.pages[page_nr])
+                    new_page_nr = page_nr + page_delta
+                    new_page.page_no = new_page_nr
+                    self.pages[new_page_nr] = new_page
+                    if new_max_page is None or new_page_nr > new_max_page:
+                        new_max_page = new_page_nr
             if new_max_page is not None:
                 self._max_page = new_max_page
 
@@ -5628,6 +5719,14 @@ class DoclingDocument(BaseModel):
         doc_index.index(doc=self)
         self._update_from_index(doc_index)
 
+    def filter(self, page_nrs: Optional[set[int]] = None) -> "DoclingDocument":
+        """Create a new document based on the provided filter parameters."""
+        doc_index = DoclingDocument._DocIndex()
+        doc_index.index(doc=self, page_nrs=page_nrs)
+        res_doc = DoclingDocument(name=self.name)
+        res_doc._update_from_index(doc_index)
+        return res_doc
+
     @classmethod
     def concatenate(cls, docs: Sequence["DoclingDocument"]) -> "DoclingDocument":
         """Concatenate multiple documents into a single document."""
@@ -5676,6 +5775,18 @@ class DoclingDocument(BaseModel):
             elif isinstance(item, ListItem):
                 validate_list_item(self, item)
 
+    def add_table_cell(self, table_item: TableItem, cell: TableCell) -> None:
+        """Add a table cell to the table."""
+        if isinstance(cell, RichTableCell):
+            item = cell.ref.resolve(doc=self)
+            if isinstance(item, NodeItem) and (
+                (not item.parent) or item.parent.cref != table_item.self_ref
+            ):
+                raise ValueError(
+                    f"Trying to add cell with another parent {item.parent} to {table_item.self_ref}"
+                )
+        table_item.data.table_cells.append(cell)
+
 
 # deprecated aliases (kept for backwards compatibility):
 BasePictureData = BaseAnnotation

docling_core/utils/legacy.py

@@ -252,7 +252,7 @@ def docling_document_to_legacy(doc: DoclingDocument, fallback_filaname: str = "f
 
                             spans = list(_make_spans(cell, item))
                             table_data[i][j] = GlmTableCell(
-                                text=cell.text,
+                                text=cell._get_text(doc=doc),
                                 bbox=(
                                     cell.bbox.as_tuple()
                                     if cell.bbox is not None

{docling_core-2.45.0.dist-info → docling_core-2.47.0.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: docling-core
-Version: 2.45.0
+Version: 2.47.0
 Summary: A python library to define and validate data types in Docling.
 Author-email: Cesar Berrospi Ramis <ceb@zurich.ibm.com>, Panos Vagenas <pva@zurich.ibm.com>, Michele Dolfi <dol@zurich.ibm.com>, Christoph Auer <cau@zurich.ibm.com>, Peter Staar <taa@zurich.ibm.com>
 Maintainer-email: Panos Vagenas <pva@zurich.ibm.com>, Michele Dolfi <dol@zurich.ibm.com>, Christoph Auer <cau@zurich.ibm.com>, Peter Staar <taa@zurich.ibm.com>, Cesar Berrospi Ramis <ceb@zurich.ibm.com>

{docling_core-2.45.0.dist-info → docling_core-2.47.0.dist-info}/RECORD

@@ -28,11 +28,11 @@ docling_core/transforms/chunker/tokenizer/huggingface.py,sha256=aZ_RNQIzcNkAHGHZ
 docling_core/transforms/chunker/tokenizer/openai.py,sha256=zt2kwcC-r8MafeEG0CESab8E4RIC9aaFXxxnxOGyTMA,918
 docling_core/transforms/serializer/__init__.py,sha256=CECQlMoCDUxkg4RAUdC3itA3I3qFhKhe2HcYghN6_xw,105
 docling_core/transforms/serializer/base.py,sha256=TI8Epj7gyxdTet9j-Rs4o5U09gfACfAIVoirlschviM,7266
-docling_core/transforms/serializer/common.py,sha256=0TNEGoA_rJ-qkVYp-X8SMUr3jTrbf6TRzPzwufYh5JM,19114
-docling_core/transforms/serializer/doctags.py,sha256=TD0yAm1qSVy-GsE6svpUAI-Yqjcf2rrTZ3ac9YU3gbE,19858
-docling_core/transforms/serializer/html.py,sha256=PPlHVu3_wnc0cD-n6n8v9clCmeY_LPqII7euVYqi6Kk,38089
+docling_core/transforms/serializer/common.py,sha256=RwfdzZ9FRSHQjKM0vskg1CVqar0Z_ms38arSlLAgITc,19150
+docling_core/transforms/serializer/doctags.py,sha256=VXPjAZPhBur7LaEeuqH9k31TgZWSN32lK8z8rJXzFwY,19935
+docling_core/transforms/serializer/html.py,sha256=GRfRaqFIb4FXRMplB4Agl4fSNa5jsHV7P4tBtFMro9I,38453
 docling_core/transforms/serializer/html_styles.py,sha256=-jBwS4EU7yfKoz0GSoxhwx90OmIKieO6TwPw57IuxcA,4692
-docling_core/transforms/serializer/markdown.py,sha256=VwonuAkuOPmQM7ibDIGvQBHOqhTcTJ_t187fLQQiNPo,23951
+docling_core/transforms/serializer/markdown.py,sha256=hilGM1yWpbbRTjuEjfBRrhavspD5vFF_6SDvlKx8BrM,24230
 docling_core/transforms/visualizer/__init__.py,sha256=gUfF25yiJ_KO46ZIUNqZQOZGy2PLx6gnnr6AZYxKHXI,35
 docling_core/transforms/visualizer/base.py,sha256=aEF7b3rHq6DVdX8zDYEPoq55BHDYe4Hh_97lBdcW4lY,555
 docling_core/transforms/visualizer/key_value_visualizer.py,sha256=fp7nFLy4flOSiavdRgg5y1Mu7WVLIDGh1zEHsq8kgVM,8979
@@ -41,9 +41,9 @@ docling_core/transforms/visualizer/reading_order_visualizer.py,sha256=muqmaxOBao
 docling_core/transforms/visualizer/table_visualizer.py,sha256=iJPjk-XQSSCH3oujcjPMz-redAwNNHseZ41lFyd-u3k,8097
 docling_core/types/__init__.py,sha256=MVRSgsk5focwGyAplh_TRR3dEecIXpd98g_u3zZ5HXo,260
 docling_core/types/base.py,sha256=PusJskRVL19y-hq0BgXr5e8--QEqSqLnFNJ8UbOqW88,8318
-docling_core/types/doc/__init__.py,sha256=8hOhm5W9mArf3zwgfoMxDs1pHizhLFSAZlLu1tPBBRk,1641
+docling_core/types/doc/__init__.py,sha256=Vsl3oJV3_BLpS7rIwvahhcWOwmEBvj7ZbQzQCCl-IQk,1678
 docling_core/types/doc/base.py,sha256=i98y4IF250adR-8BSS374K90fwfwG-vBfWh14tLC5Cs,15906
-docling_core/types/doc/document.py,sha256=gMruWRH1ELYepSMaGA5b8_l9bIjIz7JU4Yh78W_sb00,198193
+docling_core/types/doc/document.py,sha256=jyMcK1oiu8X8juNa9DuI3S1imn4hXwjOS7iTLQ1HykU,202707
 docling_core/types/doc/labels.py,sha256=-W1-LW6z0J9F9ExJqR0Wd1WeqWTaY3Unm-j1UkQGlC4,7330
 docling_core/types/doc/page.py,sha256=35h1xdtCM3-AaN8Dim9jDseZIiw-3GxpB-ofF-H2rQQ,41878
 docling_core/types/doc/tokens.py,sha256=z22l9J81_sg9CYMvOuLmPuLsNT7h_s7wao2UT89DvI8,9278
@@ -73,12 +73,12 @@ docling_core/utils/alias.py,sha256=B6Lqvss8CbaNARHLR4qSmNh9OkB6LvqTpxfsFmkLAFo,8
 docling_core/utils/file.py,sha256=CSNclJGL2OwLIc8DQFdoLxr22FUc4_UC7zS6pNrFfkQ,6858
 docling_core/utils/generate_docs.py,sha256=BdKAoduWXOc7YMvcmlhjoJOFlUxij1ybxglj6LZDtC8,2290
 docling_core/utils/generate_jsonschema.py,sha256=uNX1O5XnjyB5nA66XqZXTt3YbGuR2tyi_OhHepHYtZg,1654
-docling_core/utils/legacy.py,sha256=5lghO48OEcV9V51tRnH3YSKgLtdqhr-Q5C_OcJZ8TOs,24392
+docling_core/utils/legacy.py,sha256=G7ed8fkBpIO8hG3DKEY83cHsrKJHyvDst_1jSdgBXMI,24406
 docling_core/utils/validate.py,sha256=aQ11UbFyl8iD_N7yTTZmm_VVeXz8KcCyn3GLXgkfYRM,2049
 docling_core/utils/validators.py,sha256=azcrndLzhNkTWnbFSu9shJ5D3j_znnLrIFA5R8hzmGU,2798
-docling_core-2.45.0.dist-info/licenses/LICENSE,sha256=2M9-6EoQ1sxFztTOkXGAtwUDJvnWaAHdB9BYWVwGkIw,1087
-docling_core-2.45.0.dist-info/METADATA,sha256=VX5jfhqswstEumhmLi6VrGD9crC8RKy52z835nCvORw,6453
-docling_core-2.45.0.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-docling_core-2.45.0.dist-info/entry_points.txt,sha256=ER4zROQWkFMHIrY-oqY5E4HeCcCIg8dLkNztYGxdb7c,59
-docling_core-2.45.0.dist-info/top_level.txt,sha256=O-tcXpGiurlud-1ZxMq1b-OmrfAVA4sajcgWU32RtfA,13
-docling_core-2.45.0.dist-info/RECORD,,
+docling_core-2.47.0.dist-info/licenses/LICENSE,sha256=2M9-6EoQ1sxFztTOkXGAtwUDJvnWaAHdB9BYWVwGkIw,1087
+docling_core-2.47.0.dist-info/METADATA,sha256=jW4Zdx0WwStnLDifSsvYyGLw-5C2IYiEeK4IQRGQi-I,6453
+docling_core-2.47.0.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+docling_core-2.47.0.dist-info/entry_points.txt,sha256=ER4zROQWkFMHIrY-oqY5E4HeCcCIg8dLkNztYGxdb7c,59
+docling_core-2.47.0.dist-info/top_level.txt,sha256=O-tcXpGiurlud-1ZxMq1b-OmrfAVA4sajcgWU32RtfA,13
+docling_core-2.47.0.dist-info/RECORD,,