docling-core 2.19.1__py3-none-any.whl → 2.21.0__py3-none-any.whl

docling_core/types/doc/base.py

@@ -1,7 +1,7 @@
 """Models for the base data types."""
 
 from enum import Enum
-from typing import Tuple
+from typing import List, Tuple
 
 from pydantic import BaseModel
 
@@ -365,3 +365,30 @@ class BoundingBox(BaseModel):
             raise ValueError("BoundingBoxes have different CoordOrigin")
 
         return False
+
+    @classmethod
+    def enclosing_bbox(cls, boxes: List["BoundingBox"]) -> "BoundingBox":
+        """Create a bounding box that covers all of the given boxes."""
+        if not boxes:
+            raise ValueError("No bounding boxes provided for union.")
+
+        origin = boxes[0].coord_origin
+        if any(box.coord_origin != origin for box in boxes):
+            raise ValueError(
+                "All bounding boxes must have the same \
+                CoordOrigin to compute their union."
+            )
+
+        left = min(box.l for box in boxes)
+        right = max(box.r for box in boxes)
+
+        if origin == CoordOrigin.TOPLEFT:
+            top = min(box.t for box in boxes)
+            bottom = max(box.b for box in boxes)
+        elif origin == CoordOrigin.BOTTOMLEFT:
+            top = max(box.t for box in boxes)
+            bottom = min(box.b for box in boxes)
+        else:
+            raise ValueError("BoundingBoxes have different CoordOrigin")
+
+        return cls(l=left, t=top, r=right, b=bottom, coord_origin=origin)

docling_core/types/doc/document.py

@@ -43,7 +43,13 @@ from docling_core.search.package import VERSION_PATTERN
 from docling_core.types.base import _JSON_POINTER_REGEX
 from docling_core.types.doc import BoundingBox, Size
 from docling_core.types.doc.base import ImageRefMode
-from docling_core.types.doc.labels import CodeLanguageLabel, DocItemLabel, GroupLabel
+from docling_core.types.doc.labels import (
+    CodeLanguageLabel,
+    DocItemLabel,
+    GraphCellLabel,
+    GraphLinkLabel,
+    GroupLabel,
+)
 from docling_core.types.doc.tokens import DocumentToken, TableToken
 from docling_core.types.doc.utils import (
     get_html_tag_with_text_direction,
@@ -55,7 +61,7 @@ _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 = "1.1.0"
+CURRENT_VERSION: Final = "1.2.0"
 
 DEFAULT_EXPORT_LABELS = {
     DocItemLabel.TITLE,
@@ -855,11 +861,11 @@ class PictureItem(FloatingItem):
         image_placeholder: str = "<!-- image -->",
     ) -> str:
         """Export picture to Markdown format."""
-        default_response = "\n" + image_placeholder + "\n"
+        default_response = image_placeholder
         error_response = (
-            "\n<!-- 🖼️❌ Image not available. "
+            "<!-- 🖼️❌ Image not available. "
             "Please use `PdfPipelineOptions(generate_picture_images=True)`"
-            " --> \n"
+            " -->"
         )
 
         if image_mode == ImageRefMode.PLACEHOLDER:
@@ -873,7 +879,7 @@ class PictureItem(FloatingItem):
                 and isinstance(self.image.uri, AnyUrl)
                 and self.image.uri.scheme == "data"
             ):
-                text = f"\n![Image]({self.image.uri})\n"
+                text = f"![Image]({self.image.uri})"
                 return text
 
             # get the self.image._pil or crop it out of the page-image
@@ -881,7 +887,7 @@ class PictureItem(FloatingItem):
 
             if img is not None:
                 imgb64 = self._image_to_base64(img)
-                text = f"\n![Image](data:image/png;base64,{imgb64})\n"
+                text = f"![Image](data:image/png;base64,{imgb64})"
 
                 return text
             else:
@@ -893,7 +899,7 @@ class PictureItem(FloatingItem):
             ):
                 return default_response
 
-            text = f"\n![Image]({quote(str(self.image.uri))})\n"
+            text = f"![Image]({quote(str(self.image.uri))})"
             return text
 
         else:
@@ -1101,7 +1107,9 @@ class TableItem(FloatingItem):
         return md_table
 
     def export_to_html(
-        self, doc: Optional["DoclingDocument"] = None, add_caption: bool = True
+        self,
+        doc: Optional["DoclingDocument"] = None,
+        add_caption: bool = True,
     ) -> str:
         """Export the table as html."""
         if doc is None:
@@ -1330,11 +1338,81 @@ class TableItem(FloatingItem):
         return body
 
 
-class KeyValueItem(DocItem):
+class GraphCell(BaseModel):
+    """GraphCell."""
+
+    label: GraphCellLabel
+
+    cell_id: int
+
+    text: str  # sanitized text
+    orig: str  # text as seen on document
+
+    prov: Optional[ProvenanceItem] = None
+
+    # in case you have a text, table or picture item
+    item_ref: Optional[RefItem] = None
+
+
+class GraphLink(BaseModel):
+    """GraphLink."""
+
+    label: GraphLinkLabel
+
+    source_cell_id: int
+    target_cell_id: int
+
+
+class GraphData(BaseModel):
+    """GraphData."""
+
+    cells: List[GraphCell] = Field(default_factory=list)
+    links: List[GraphLink] = Field(default_factory=list)
+
+    @field_validator("links")
+    @classmethod
+    def validate_links(cls, links, info):
+        """Ensure that each link is valid."""
+        cells = info.data.get("cells", [])
+
+        valid_cell_ids = {cell.cell_id for cell in cells}
+
+        for link in links:
+            if link.source_cell_id not in valid_cell_ids:
+                raise ValueError(
+                    f"Invalid source_cell_id {link.source_cell_id} in GraphLink"
+                )
+            if link.target_cell_id not in valid_cell_ids:
+                raise ValueError(
+                    f"Invalid target_cell_id {link.target_cell_id} in GraphLink"
+                )
+
+        return links
+
+
+class KeyValueItem(FloatingItem):
     """KeyValueItem."""
 
     label: typing.Literal[DocItemLabel.KEY_VALUE_REGION] = DocItemLabel.KEY_VALUE_REGION
 
+    graph: GraphData
+
+    def _export_to_markdown(self) -> str:
+        # TODO add actual implementation
+        return "<!-- missing-key-value-item -->"
+
+
+class FormItem(FloatingItem):
+    """FormItem."""
+
+    label: typing.Literal[DocItemLabel.FORM] = DocItemLabel.FORM
+
+    graph: GraphData
+
+    def _export_to_markdown(self) -> str:
+        # TODO add actual implementation
+        return "<!-- missing-form-item -->"
+
 
 ContentItem = Annotated[
     Union[
@@ -1446,7 +1524,9 @@ class DoclingDocument(BaseModel):
     )
 
     furniture: Annotated[GroupItem, Field(deprecated=True)] = GroupItem(
-        name="_root_", self_ref="#/furniture", content_layer=ContentLayer.FURNITURE
+        name="_root_",
+        self_ref="#/furniture",
+        content_layer=ContentLayer.FURNITURE,
     )  # List[RefItem] = []
     body: GroupItem = GroupItem(name="_root_", self_ref="#/body")  # List[RefItem] = []
 
@@ -1455,6 +1535,7 @@ class DoclingDocument(BaseModel):
     pictures: List[PictureItem] = []
     tables: List[TableItem] = []
     key_value_items: List[KeyValueItem] = []
+    form_items: List[FormItem] = []
 
     pages: Dict[int, PageItem] = {}  # empty as default
 
@@ -1851,6 +1932,68 @@ class DoclingDocument(BaseModel):
 
         return section_header_item
 
+    def add_key_values(
+        self,
+        graph: GraphData,
+        prov: Optional[ProvenanceItem] = None,
+        parent: Optional[NodeItem] = None,
+    ):
+        """add_key_values.
+
+        :param graph: GraphData:
+        :param prov: Optional[ProvenanceItem]:  (Default value = None)
+        :param parent: Optional[NodeItem]:  (Default value = None)
+        """
+        if not parent:
+            parent = self.body
+
+        key_value_index = len(self.key_value_items)
+        cref = f"#/key_value_items/{key_value_index}"
+
+        kv_item = KeyValueItem(
+            graph=graph,
+            self_ref=cref,
+            parent=parent.get_ref(),
+        )
+        if prov:
+            kv_item.prov.append(prov)
+
+        self.key_value_items.append(kv_item)
+        parent.children.append(RefItem(cref=cref))
+
+        return kv_item
+
+    def add_form(
+        self,
+        graph: GraphData,
+        prov: Optional[ProvenanceItem] = None,
+        parent: Optional[NodeItem] = None,
+    ):
+        """add_form.
+
+        :param graph: GraphData:
+        :param prov: Optional[ProvenanceItem]:  (Default value = None)
+        :param parent: Optional[NodeItem]:  (Default value = None)
+        """
+        if not parent:
+            parent = self.body
+
+        form_index = len(self.form_items)
+        cref = f"#/form_items/{form_index}"
+
+        form_item = FormItem(
+            graph=graph,
+            self_ref=cref,
+            parent=parent.get_ref(),
+        )
+        if prov:
+            form_item.prov.append(prov)
+
+        self.form_items.append(form_item)
+        parent.children.append(RefItem(cref=cref))
+
+        return form_item
+
     def num_pages(self):
         """num_pages."""
         return len(self.pages.values())
@@ -2009,7 +2152,8 @@ class DoclingDocument(BaseModel):
                             img.save(loc_path)
                             if reference_path is not None:
                                 obj_path = relative_path(
-                                    reference_path.resolve(), loc_path.resolve()
+                                    reference_path.resolve(),
+                                    loc_path.resolve(),
                                 )
                             else:
                                 obj_path = loc_path
@@ -2027,7 +2171,10 @@ class DoclingDocument(BaseModel):
         """Print_element_tree."""
         for ix, (item, level) in enumerate(self.iterate_items(with_groups=True)):
             if isinstance(item, GroupItem):
-                print(" " * level, f"{ix}: {item.label.value} with name={item.name}")
+                print(
+                    " " * level,
+                    f"{ix}: {item.label.value} with name={item.name}",
+                )
             elif isinstance(item, DocItem):
                 print(" " * level, f"{ix}: {item.label.value}")
 
@@ -2100,6 +2247,20 @@ class DoclingDocument(BaseModel):
         with open(filename, "w", encoding="utf-8") as fw:
             yaml.dump(out, fw, default_flow_style=default_flow_style)
 
+    @classmethod
+    def load_from_yaml(cls, filename: Path) -> "DoclingDocument":
+        """load_from_yaml.
+
+        Args:
+            filename: The filename to load a YAML-serialized DoclingDocument from.
+
+        Returns:
+            DoclingDocument: the loaded DoclingDocument
+        """
+        with open(filename, encoding="utf-8") as f:
+            data = yaml.load(f, Loader=yaml.FullLoader)
+        return DoclingDocument.model_validate(data)
+
     def export_to_dict(
         self,
         mode: str = "json",
@@ -2115,7 +2276,7 @@ class DoclingDocument(BaseModel):
         self,
         filename: Path,
         artifacts_dir: Optional[Path] = None,
-        delim: str = "\n",
+        delim: str = "\n\n",  # TODO: deprecate
         from_element: int = 0,
         to_element: int = sys.maxsize,
         labels: set[DocItemLabel] = DEFAULT_EXPORT_LABELS,
@@ -2158,7 +2319,7 @@ class DoclingDocument(BaseModel):
 
     def export_to_markdown(  # noqa: C901
         self,
-        delim: str = "\n",
+        delim: str = "\n\n",  # TODO deprecate
         from_element: int = 0,
         to_element: int = sys.maxsize,
         labels: set[DocItemLabel] = DEFAULT_EXPORT_LABELS,
@@ -2205,10 +2366,44 @@ class DoclingDocument(BaseModel):
         :returns: The exported Markdown representation.
         :rtype: str
         """
-        mdtexts: list[str] = []
-        list_nesting_level = 0  # Track the current list nesting level
-        previous_level = 0  # Track the previous item's level
-        in_list = False  # Track if we're currently processing list items
+        comps = self._get_markdown_components(
+            node=self.body,
+            from_element=from_element,
+            to_element=to_element,
+            labels=labels,
+            strict_text=strict_text,
+            escaping_underscores=escaping_underscores,
+            image_placeholder=image_placeholder,
+            image_mode=image_mode,
+            indent=indent,
+            text_width=text_width,
+            page_no=page_no,
+            included_content_layers=included_content_layers,
+            list_level=0,
+            is_inline_scope=False,
+            visited=set(),
+        )
+        return delim.join(comps)
+
+    def _get_markdown_components(  # noqa: C901
+        self,
+        node: NodeItem,
+        from_element: int,
+        to_element: int,
+        labels: set[DocItemLabel],
+        strict_text: bool,
+        escaping_underscores: bool,
+        image_placeholder: str,
+        image_mode: ImageRefMode,
+        indent: int,
+        text_width: int,
+        page_no: Optional[int],
+        included_content_layers: set[ContentLayer],
+        list_level: int,
+        is_inline_scope: bool,
+        visited: set[str],  # refs of visited items
+    ) -> list[str]:
+        components: list[str] = []  # components to concatenate
 
         # Our export markdown doesn't contain any emphasis styling:
         # Bold, Italic, or Bold-Italic
@@ -2243,137 +2438,138 @@ class DoclingDocument(BaseModel):
 
             return "".join(parts)
 
-        def _append_text(text: str, do_escape_html=True, do_escape_underscores=True):
+        def _ingest_text(text: str, do_escape_html=True, do_escape_underscores=True):
             if do_escape_underscores and escaping_underscores:
                 text = _escape_underscores(text)
             if do_escape_html:
                 text = html.escape(text, quote=False)
-            mdtexts.append(text)
+            if text:
+                components.append(text)
 
         for ix, (item, level) in enumerate(
             self.iterate_items(
-                self.body,
+                node,
                 with_groups=True,
                 page_no=page_no,
                 included_content_layers=included_content_layers,
             )
         ):
-            # If we've moved to a lower level, we're exiting one or more groups
-            if level < previous_level:
-                # Calculate how many levels we've exited
-                level_difference = previous_level - level
-                # Decrement list_nesting_level for each list group we've exited
-                list_nesting_level = max(0, list_nesting_level - level_difference)
-
-            previous_level = level  # Update previous_level for next iteration
+            if item.self_ref in visited:
+                continue
+            else:
+                visited.add(item.self_ref)
 
             if ix < from_element or to_element <= ix:
                 continue  # skip as many items as you want
 
-            if (isinstance(item, DocItem)) and (item.label not in labels):
+            elif (isinstance(item, DocItem)) and (item.label not in labels):
                 continue  # skip any label that is not whitelisted
 
-            # Handle newlines between different types of content
-            if (
-                len(mdtexts) > 0
-                and not isinstance(item, (ListItem, GroupItem))
-                and in_list
-            ):
-                mdtexts[-1] += "\n"
-                in_list = False
-
-            if isinstance(item, GroupItem) and item.label in [
-                GroupLabel.LIST,
-                GroupLabel.ORDERED_LIST,
-            ]:
-
-                if list_nesting_level == 0:  # Check if we're on the top level.
-                    # In that case a new list starts directly after another list.
-                    mdtexts.append("\n")  # Add a blank line
-
-                # Increment list nesting level when entering a new list
-                list_nesting_level += 1
-                in_list = True
-                continue
-
             elif isinstance(item, GroupItem):
-                continue
+                if item.label in [
+                    GroupLabel.LIST,
+                    GroupLabel.ORDERED_LIST,
+                ]:
+                    comps = self._get_markdown_components(
+                        node=item,
+                        from_element=from_element,
+                        to_element=to_element,
+                        labels=labels,
+                        strict_text=strict_text,
+                        escaping_underscores=escaping_underscores,
+                        image_placeholder=image_placeholder,
+                        image_mode=image_mode,
+                        indent=indent,
+                        text_width=text_width,
+                        page_no=page_no,
+                        included_content_layers=included_content_layers,
+                        list_level=list_level + 1,
+                        is_inline_scope=is_inline_scope,
+                        visited=visited,
+                    )
+                    # NOTE: assumes unordered (flag & marker currently in ListItem)
+                    indent_str = list_level * indent * " "
+                    text = "\n".join(
+                        [
+                            # avoid additional marker on already evaled sublists
+                            cpt if cpt and cpt[0] == " " else f"{indent_str}- {cpt}"
+                            for cpt in comps
+                        ]
+                    )
+                    _ingest_text(text=text)
+                elif item.label == GroupLabel.INLINE:
+                    comps = self._get_markdown_components(
+                        node=item,
+                        from_element=from_element,
+                        to_element=to_element,
+                        labels=labels,
+                        strict_text=strict_text,
+                        escaping_underscores=escaping_underscores,
+                        image_placeholder=image_placeholder,
+                        image_mode=image_mode,
+                        indent=indent,
+                        text_width=text_width,
+                        page_no=page_no,
+                        included_content_layers=included_content_layers,
+                        list_level=list_level,
+                        is_inline_scope=True,
+                        visited=visited,
+                    )
+                    _ingest_text(" ".join(comps))
+                else:
+                    continue
 
             elif isinstance(item, TextItem) and item.label in [DocItemLabel.TITLE]:
-                in_list = False
                 marker = "" if strict_text else "#"
                 text = f"{marker} {item.text}"
-                _append_text(text.strip() + "\n")
+                _ingest_text(text.strip())
 
             elif (
                 isinstance(item, TextItem)
                 and item.label in [DocItemLabel.SECTION_HEADER]
             ) or isinstance(item, SectionHeaderItem):
-                in_list = False
                 marker = ""
                 if not strict_text:
                     marker = "#" * level
                     if len(marker) < 2:
                         marker = "##"
-                text = f"{marker} {item.text}\n"
-                _append_text(text.strip() + "\n")
-
-            elif isinstance(item, CodeItem) and item.label in labels:
-                in_list = False
-                text = f"```\n{item.text}\n```\n"
-                _append_text(text, do_escape_underscores=False, do_escape_html=False)
-
-            elif isinstance(item, ListItem) and item.label in [DocItemLabel.LIST_ITEM]:
-                in_list = True
-                # Calculate indent based on list_nesting_level
-                # -1 because level 1 needs no indent
-                list_indent = " " * (indent * (list_nesting_level - 1))
-
-                marker = ""
-                if strict_text:
-                    marker = ""
-                elif item.enumerated:
-                    marker = item.marker
-                else:
-                    marker = "-"  # Markdown needs only dash as item marker.
+                text = f"{marker} {item.text}"
+                _ingest_text(text.strip())
 
-                text = f"{list_indent}{marker} {item.text}"
-                _append_text(text)
+            elif isinstance(item, CodeItem):
+                text = f"`{item.text}`" if is_inline_scope else f"```\n{item.text}\n```"
+                _ingest_text(text, do_escape_underscores=False, do_escape_html=False)
 
             elif isinstance(item, TextItem) and item.label in [DocItemLabel.FORMULA]:
-                in_list = False
                 if item.text != "":
-                    _append_text(
-                        f"$${item.text}$$\n",
+                    _ingest_text(
+                        f"${item.text}$" if is_inline_scope else f"$${item.text}$$",
                         do_escape_underscores=False,
                         do_escape_html=False,
                     )
                 elif item.orig != "":
-                    _append_text(
-                        "<!-- formula-not-decoded -->\n",
+                    _ingest_text(
+                        "<!-- formula-not-decoded -->",
                         do_escape_underscores=False,
                         do_escape_html=False,
                     )
 
-            elif isinstance(item, TextItem) and item.label in labels:
-                in_list = False
+            elif isinstance(item, TextItem):
                 if len(item.text) and text_width > 0:
                     text = item.text
                     wrapped_text = textwrap.fill(text, width=text_width)
-                    _append_text(wrapped_text + "\n")
+                    _ingest_text(wrapped_text)
                 elif len(item.text):
-                    text = f"{item.text}\n"
-                    _append_text(text)
+                    _ingest_text(item.text)
 
             elif isinstance(item, TableItem) and not strict_text:
-                in_list = False
-                _append_text(item.caption_text(self))
+                if caption_text := item.caption_text(self):
+                    _ingest_text(caption_text)
                 md_table = item.export_to_markdown()
-                _append_text("\n" + md_table + "\n")
+                _ingest_text(md_table)
 
             elif isinstance(item, PictureItem) and not strict_text:
-                in_list = False
-                _append_text(item.caption_text(self))
+                _ingest_text(item.caption_text(self))
 
                 line = item.export_to_markdown(
                     doc=self,
@@ -2381,19 +2577,17 @@ class DoclingDocument(BaseModel):
                     image_mode=image_mode,
                 )
 
-                _append_text(line, do_escape_html=False, do_escape_underscores=False)
+                _ingest_text(line, do_escape_html=False, do_escape_underscores=False)
 
-            elif isinstance(item, DocItem) and item.label in labels:
-                in_list = False
-                text = "<!-- missing-text -->"
-                _append_text(text, do_escape_html=False, do_escape_underscores=False)
+            elif isinstance(item, (KeyValueItem, FormItem)):
+                text = item._export_to_markdown()
+                _ingest_text(text, do_escape_html=False, do_escape_underscores=False)
 
-        mdtext = (delim.join(mdtexts)).strip()
-        mdtext = re.sub(
-            r"\n\n\n+", "\n\n", mdtext
-        )  # remove cases of double or more empty lines.
+            elif isinstance(item, DocItem):
+                text = "<!-- missing-text -->"
+                _ingest_text(text, do_escape_html=False, do_escape_underscores=False)
 
-        return mdtext
+        return components
 
     def export_to_text(  # noqa: C901
         self,
@@ -2519,7 +2713,11 @@ class DoclingDocument(BaseModel):
 
             return (in_ordered_list, html_texts)
 
-        head_lines = ["<!DOCTYPE html>", f'<html lang="{html_lang}">', html_head]
+        head_lines = [
+            "<!DOCTYPE html>",
+            f'<html lang="{html_lang}">',
+            html_head,
+        ]
         html_texts: list[str] = []
 
         prev_level = 0  # Track the previous item's level
@@ -2599,7 +2797,8 @@ class DoclingDocument(BaseModel):
                 section_level: int = min(item.level + 1, 6)
 
                 text = get_html_tag_with_text_direction(
-                    html_tag=f"h{section_level}", text=_prepare_tag_content(item.text)
+                    html_tag=f"h{section_level}",
+                    text=_prepare_tag_content(item.text),
                 )
                 html_texts.append(text)
 
@@ -2620,14 +2819,17 @@ class DoclingDocument(BaseModel):
                             "</figure>"
                         )
 
+                img_fallback = _image_fallback(item)
+
                 # If the formula is not processed correcty, use its image
                 if (
                     item.text == ""
                     and item.orig != ""
                     and image_mode == ImageRefMode.EMBEDDED
                     and len(item.prov) > 0
+                    and img_fallback is not None
                 ):
-                    text = _image_fallback(item)
+                    text = img_fallback
 
                 # Building a math equation in MathML format
                 # ref https://www.w3.org/TR/wai-aria-1.1/#math
@@ -2647,9 +2849,13 @@ class DoclingDocument(BaseModel):
                             "Malformed formula cannot be rendered. "
                             f"Error {err.__class__.__name__}, formula={math_formula}"
                         )
-                        if image_mode == ImageRefMode.EMBEDDED and len(item.prov) > 0:
-                            text = _image_fallback(item)
-                        else:
+                        if (
+                            image_mode == ImageRefMode.EMBEDDED
+                            and len(item.prov) > 0
+                            and img_fallback is not None
+                        ):
+                            text = img_fallback
+                        elif len(math_formula) > 0:
                             text = f"<pre>{math_formula}</pre>"
 
                 elif math_formula != "":
@@ -2856,13 +3062,19 @@ class DoclingDocument(BaseModel):
             self.iterate_items(
                 self.body,
                 with_groups=True,
-                included_content_layers={ContentLayer.BODY, ContentLayer.FURNITURE},
+                included_content_layers={
+                    ContentLayer.BODY,
+                    ContentLayer.FURNITURE,
+                },
             )
         ):
             # Close lists if we've moved to a lower nesting level
             if current_level < previous_level and ordered_list_stack:
                 ordered_list_stack = _close_lists(
-                    current_level, previous_level, ordered_list_stack, output_parts
+                    current_level,
+                    previous_level,
+                    ordered_list_stack,
+                    output_parts,
                 )
             previous_level = current_level
 
@@ -2970,7 +3182,10 @@ class DoclingDocument(BaseModel):
         return "".join(output_parts)
 
     def _export_to_indented_text(
-        self, indent="  ", max_text_len: int = -1, explicit_tables: bool = False
+        self,
+        indent="  ",
+        max_text_len: int = -1,
+        explicit_tables: bool = False,
     ):
         """Export the document to indented text to expose hierarchy."""
         result = []

docling_core/types/doc/labels.py

@@ -75,6 +75,7 @@ class GroupLabel(str, Enum):
     FORM_AREA = "form_area"
     KEY_VALUE_AREA = "key_value_area"
     COMMENT_SECTION = "comment_section"
+    INLINE = "inline"
 
     def __str__(self):
         """Get string value."""
@@ -140,6 +141,29 @@ class TableCellLabel(str, Enum):
         return str(self.value)
 
 
+class GraphCellLabel(str, Enum):
+    """GraphCellLabel."""
+
+    UNSPECIFIED = "unspecified"
+
+    KEY = "key"
+    VALUE = "value"
+
+    CHECKBOX = "checkbox"
+
+
+class GraphLinkLabel(str, Enum):
+    """GraphLinkLabel."""
+
+    UNSPECIFIED = "unspecified"
+
+    TO_VALUE = "to_value"
+    TO_KEY = "to_key"
+
+    TO_PARENT = "to_parent"
+    TO_CHILD = "to_child"
+
+
 class CodeLanguageLabel(str, Enum):
     """CodeLanguageLabel."""
 

{docling_core-2.19.1.dist-info → docling_core-2.21.0.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: docling-core
-Version: 2.19.1
+Version: 2.21.0
 Summary: A python library to define and validate data types in Docling.
 Home-page: https://ds4sd.github.io/
 License: MIT

{docling_core-2.19.1.dist-info → docling_core-2.21.0.dist-info}/RECORD

@@ -23,9 +23,9 @@ docling_core/transforms/chunker/hybrid_chunker.py,sha256=kokjDdxjc_gygOokQwYFVnH
 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=bysJn2iwjAHwThSWDPXEdVUUij7p_ax12_nx2_0CMdg,653
-docling_core/types/doc/base.py,sha256=lMRNq1DUK7K26L2VNZRqFaItCSZ6m9BdYTVaJA98PZQ,11495
-docling_core/types/doc/document.py,sha256=t1nk1GeR5_YvZhuWUVZkkBekp89vFB4RBtMuwD3Acw4,104373
-docling_core/types/doc/labels.py,sha256=cqH4DGN9lgZns6gOtL5urzZzUPGOjHJ75xQbIKSh_h8,5306
+docling_core/types/doc/base.py,sha256=22U1qDlD-2ICmgzbdZrjNayoPHnq4S1ks1GRoqB7y1Q,12542
+docling_core/types/doc/document.py,sha256=oaa8QetSQTiPOQIZEMpvzoxwPVBOn9DLUwE62tB707w,110223
+docling_core/types/doc/labels.py,sha256=0J9Gsqz-jQ4FP2yxs9wOxoTr3qg97BniFX7MJVziUmk,5684
 docling_core/types/doc/tokens.py,sha256=i73PXkmqXCLsQ5SddnJX8L9e_Ub2_K_DYSE-VE8NDq0,3925
 docling_core/types/doc/utils.py,sha256=SaiQD-WMMooFm1bMqwatU-IGhtG048iKJb-ppnJit_k,2250
 docling_core/types/gen/__init__.py,sha256=C6TuCfvpSnSL5XDOFMcYHUY2-i08vvfOGRcdu6Af0pI,124
@@ -56,8 +56,8 @@ docling_core/utils/generate_jsonschema.py,sha256=uNX1O5XnjyB5nA66XqZXTt3YbGuR2ty
 docling_core/utils/legacy.py,sha256=SqNQAxl97aHfoJEsC9vZcMJg5FNkmqKPFi-wdSrnfI0,24442
 docling_core/utils/validate.py,sha256=aQ11UbFyl8iD_N7yTTZmm_VVeXz8KcCyn3GLXgkfYRM,2049
 docling_core/utils/validators.py,sha256=azcrndLzhNkTWnbFSu9shJ5D3j_znnLrIFA5R8hzmGU,2798
-docling_core-2.19.1.dist-info/LICENSE,sha256=2M9-6EoQ1sxFztTOkXGAtwUDJvnWaAHdB9BYWVwGkIw,1087
-docling_core-2.19.1.dist-info/METADATA,sha256=Uz-AUOD2_itxSEVxatsPbCQ0pFBE3fMX-gXx0YLmsKw,5803
-docling_core-2.19.1.dist-info/WHEEL,sha256=sP946D7jFCHeNz5Iq4fL4Lu-PrWrFsgfLXbbkciIZwg,88
-docling_core-2.19.1.dist-info/entry_points.txt,sha256=oClcdb2L2RKx4jdqUykY16Kum_f0_whwWhGzIodyidc,216
-docling_core-2.19.1.dist-info/RECORD,,
+docling_core-2.21.0.dist-info/LICENSE,sha256=2M9-6EoQ1sxFztTOkXGAtwUDJvnWaAHdB9BYWVwGkIw,1087
+docling_core-2.21.0.dist-info/METADATA,sha256=gyy6KPX1dAJel3ysIg3Zt73Opw8WtKW2nSXxonkkRKc,5803
+docling_core-2.21.0.dist-info/WHEEL,sha256=sP946D7jFCHeNz5Iq4fL4Lu-PrWrFsgfLXbbkciIZwg,88
+docling_core-2.21.0.dist-info/entry_points.txt,sha256=oClcdb2L2RKx4jdqUykY16Kum_f0_whwWhGzIodyidc,216
+docling_core-2.21.0.dist-info/RECORD,,