Merge remote-tracking branch 'origin/main' into md_table_improvements

Signed-off-by: Michael Honaker <Michael.Honaker@ibm.com>

.github/workflows/checks.yml

@@ -22,8 +22,8 @@ jobs:
         python-version: ['3.9', '3.10', '3.11', '3.12', '3.13']
     steps:
       - uses: actions/checkout@v4
-      - name: Install tesseract
-        run: sudo apt-get update && sudo apt-get install -y tesseract-ocr tesseract-ocr-eng tesseract-ocr-fra tesseract-ocr-deu tesseract-ocr-spa tesseract-ocr-script-latn libleptonica-dev libtesseract-dev pkg-config
+      - name: Install tesseract and ffmpeg
+        run: sudo apt-get update && sudo apt-get install -y ffmpeg tesseract-ocr tesseract-ocr-eng tesseract-ocr-fra tesseract-ocr-deu tesseract-ocr-spa tesseract-ocr-script-latn libleptonica-dev libtesseract-dev pkg-config
       - name: Set TESSDATA_PREFIX
         run: |
           echo "TESSDATA_PREFIX=$(dpkg -L tesseract-ocr-eng | grep tessdata$)" >> "$GITHUB_ENV"
@@ -60,7 +60,7 @@ jobs:
         run: |
           for file in docs/examples/*.py; do
             # Skip batch_convert.py
-            if [[ "$(basename "$file")" =~ ^(batch_convert|compare_vlm_models|minimal|minimal_vlm_pipeline|export_multimodal|custom_convert|develop_picture_enrichment|rapidocr_with_custom_models|offline_convert|pictures_description|pictures_description_api|vlm_pipeline_api_model).py ]]; then
+            if [[ "$(basename "$file")" =~ ^(batch_convert|compare_vlm_models|minimal|minimal_vlm_pipeline|minimal_asr_pipeline|export_multimodal|custom_convert|develop_picture_enrichment|rapidocr_with_custom_models|offline_convert|pictures_description|pictures_description_api|vlm_pipeline_api_model).py ]]; then
                 echo "Skipping $file"
                 continue
             fi

CHANGELOG.md

@@ -1,3 +1,23 @@
+## [v2.38.0](https://github.com/docling-project/docling/releases/tag/v2.38.0) - 2025-06-23
+
+### Feature
+
+* Support audio input ([#1763](https://github.com/docling-project/docling/issues/1763)) ([`1557e7c`](https://github.com/docling-project/docling/commit/1557e7ce3e036fb51eb118296f5cbff3b6dfbfa7))
+* **markdown:** Add formatting & improve inline support ([#1804](https://github.com/docling-project/docling/issues/1804)) ([`861abcd`](https://github.com/docling-project/docling/commit/861abcdcb0d406342b9566f81203b87cf32b7ad0))
+* Maximum image size for Vlm models ([#1802](https://github.com/docling-project/docling/issues/1802)) ([`215b540`](https://github.com/docling-project/docling/commit/215b540f6c078a72464310ef22975ebb6cde4f0a))
+
+### Fix
+
+* **docx:** Ensure list items have a list parent ([#1827](https://github.com/docling-project/docling/issues/1827)) ([`d26dac6`](https://github.com/docling-project/docling/commit/d26dac61a86b0af5b16686f78956ba047bcbddba))
+* **msword_backend:** Identify text in the same line after an image #1425 ([#1610](https://github.com/docling-project/docling/issues/1610)) ([`1350a8d`](https://github.com/docling-project/docling/commit/1350a8d3e5ea3c4b4d506757758880c8f78efd8c))
+* Ensure uninitialized pages are removed before assembling document ([#1812](https://github.com/docling-project/docling/issues/1812)) ([`dd7f64f`](https://github.com/docling-project/docling/commit/dd7f64ff28226cd9964fc4d8ba807b2c8a6358ef))
+* Formula conversion with page_range param set ([#1791](https://github.com/docling-project/docling/issues/1791)) ([`dbab30e`](https://github.com/docling-project/docling/commit/dbab30e92cc1d130ce7f9335ab9c46aa7a30930d))
+
+### Documentation
+
+* Update readme and add ASR example ([#1836](https://github.com/docling-project/docling/issues/1836)) ([`f3ae302`](https://github.com/docling-project/docling/commit/f3ae3029b8a6d6f0109383fbc82ebf9da3942afd))
+* Support running examples from root or subfolder ([#1816](https://github.com/docling-project/docling/issues/1816)) ([`64ac043`](https://github.com/docling-project/docling/commit/64ac043786efdece0c61827051a5b41dddf6c5d7))
+
 ## [v2.37.0](https://github.com/docling-project/docling/releases/tag/v2.37.0) - 2025-06-16
 
 ### Feature

README.md

@@ -28,14 +28,15 @@ Docling simplifies document processing, parsing diverse formats — including ad
 
 ## Features
 
-* 🗂️ Parsing of [multiple document formats][supported_formats] incl. PDF, DOCX, XLSX, HTML, images, and more
+* 🗂️  Parsing of [multiple document formats][supported_formats] incl. PDF, DOCX, PPTX, XLSX, HTML, WAV, MP3, images (PNG, TIFF, JPEG, ...), and more
 * 📑 Advanced PDF understanding incl. page layout, reading order, table structure, code, formulas, image classification, and more
 * 🧬 Unified, expressive [DoclingDocument][docling_document] representation format
-* ↪️ Various [export formats][supported_formats] and options, including Markdown, HTML, and lossless JSON
+* ↪️  Various [export formats][supported_formats] and options, including Markdown, HTML, [DocTags](https://arxiv.org/abs/2503.11576) and lossless JSON
 * 🔒 Local execution capabilities for sensitive data and air-gapped environments
 * 🤖 Plug-and-play [integrations][integrations] incl. LangChain, LlamaIndex, Crew AI & Haystack for agentic AI
 * 🔍 Extensive OCR support for scanned PDFs and images
-* 🥚 Support of several Visual Language Models ([SmolDocling](https://huggingface.co/ds4sd/SmolDocling-256M-preview))
+* 👓 Support of several Visual Language Models ([SmolDocling](https://huggingface.co/ds4sd/SmolDocling-256M-preview))
+* 🎙️  Support for Audio with Automatic Speech Recognition (ASR) models
 * 💻 Simple and convenient CLI
 
 ### Coming soon

docling/backend/md_backend.py

@@ -2,9 +2,10 @@ import logging
 import re
 import warnings
 from copy import deepcopy
+from enum import Enum
 from io import BytesIO
 from pathlib import Path
-from typing import List, Optional, Set, Union
+from typing import List, Literal, Optional, Set, Union
 
 import marko
 import marko.element
@@ -21,7 +22,8 @@ from docling_core.types.doc import (
 )
 from docling_core.types.doc.document import Formatting, OrderedList, UnorderedList
 from marko import Markdown
-from pydantic import AnyUrl, TypeAdapter
+from pydantic import AnyUrl, BaseModel, Field, TypeAdapter
+from typing_extensions import Annotated
 
 from docling.backend.abstract_backend import DeclarativeDocumentBackend
 from docling.backend.html_backend import HTMLDocumentBackend
@@ -35,6 +37,31 @@ _START_MARKER = f"#_#_{_MARKER_BODY}_START_#_#"
 _STOP_MARKER = f"#_#_{_MARKER_BODY}_STOP_#_#"
 
 
+class _PendingCreationType(str, Enum):
+    """CoordOrigin."""
+
+    HEADING = "heading"
+    LIST_ITEM = "list_item"
+
+
+class _HeadingCreationPayload(BaseModel):
+    kind: Literal["heading"] = "heading"
+    level: int
+
+
+class _ListItemCreationPayload(BaseModel):
+    kind: Literal["list_item"] = "list_item"
+
+
+_CreationPayload = Annotated[
+    Union[
+        _HeadingCreationPayload,
+        _ListItemCreationPayload,
+    ],
+    Field(discriminator="kind"),
+]
+
+
 class MarkdownDocumentBackend(DeclarativeDocumentBackend):
     def _shorten_underscore_sequences(self, markdown_text: str, max_length: int = 10):
         # This regex will match any sequence of underscores
@@ -155,6 +182,52 @@ class MarkdownDocumentBackend(DeclarativeDocumentBackend):
                 doc.add_table(data=table_data)
         return
 
+    def _create_list_item(
+        self,
+        doc: DoclingDocument,
+        parent_item: Optional[NodeItem],
+        text: str,
+        formatting: Optional[Formatting] = None,
+        hyperlink: Optional[Union[AnyUrl, Path]] = None,
+    ):
+        if not isinstance(parent_item, (OrderedList, UnorderedList)):
+            _log.warning("ListItem would have not had a list parent, adding one.")
+            parent_item = doc.add_unordered_list(parent=parent_item)
+        item = doc.add_list_item(
+            text=text,
+            enumerated=(isinstance(parent_item, OrderedList)),
+            parent=parent_item,
+            formatting=formatting,
+            hyperlink=hyperlink,
+        )
+        return item
+
+    def _create_heading_item(
+        self,
+        doc: DoclingDocument,
+        parent_item: Optional[NodeItem],
+        text: str,
+        level: int,
+        formatting: Optional[Formatting] = None,
+        hyperlink: Optional[Union[AnyUrl, Path]] = None,
+    ):
+        if level == 1:
+            item = doc.add_title(
+                text=text,
+                parent=parent_item,
+                formatting=formatting,
+                hyperlink=hyperlink,
+            )
+        else:
+            item = doc.add_heading(
+                text=text,
+                level=level - 1,
+                parent=parent_item,
+                formatting=formatting,
+                hyperlink=hyperlink,
+            )
+        return item
+
     def _iterate_elements(  # noqa: C901
         self,
         *,
@@ -162,6 +235,9 @@ class MarkdownDocumentBackend(DeclarativeDocumentBackend):
         depth: int,
         doc: DoclingDocument,
         visited: Set[marko.element.Element],
+        creation_stack: list[
+            _CreationPayload
+        ],  # stack for lazy item creation triggered deep in marko's AST (on RawText)
         parent_item: Optional[NodeItem] = None,
         formatting: Optional[Formatting] = None,
         hyperlink: Optional[Union[AnyUrl, Path]] = None,
@@ -177,28 +253,17 @@ class MarkdownDocumentBackend(DeclarativeDocumentBackend):
                 f" - Heading level {element.level}, content: {element.children[0].children}"  # type: ignore
             )
 
-            if len(element.children) == 1:
-                child = element.children[0]
-                snippet_text = str(child.children)  # type: ignore
-                visited.add(child)
-            else:
-                snippet_text = ""  # inline group will be created
-
-            if element.level == 1:
-                parent_item = doc.add_title(
-                    text=snippet_text,
-                    parent=parent_item,
+            if len(element.children) > 1:  # inline group will be created further down
+                parent_item = self._create_heading_item(
+                    doc=doc,
+                    parent_item=parent_item,
+                    text="",
+                    level=element.level,
                     formatting=formatting,
                     hyperlink=hyperlink,
                 )
             else:
-                parent_item = doc.add_heading(
-                    text=snippet_text,
-                    level=element.level - 1,
-                    parent=parent_item,
-                    formatting=formatting,
-                    hyperlink=hyperlink,
-                )
+                creation_stack.append(_HeadingCreationPayload(level=element.level))
 
         elif isinstance(element, marko.block.List):
             has_non_empty_list_items = False
@@ -224,22 +289,16 @@ class MarkdownDocumentBackend(DeclarativeDocumentBackend):
             self._close_table(doc)
             _log.debug(" - List item")
 
-            if len(child.children) == 1:
-                snippet_text = str(child.children[0].children)  # type: ignore
-                visited.add(child)
-            else:
-                snippet_text = ""  # inline group will be created
-            is_numbered = isinstance(parent_item, OrderedList)
-            if not isinstance(parent_item, (OrderedList, UnorderedList)):
-                _log.warning("ListItem would have not had a list parent, adding one.")
-                parent_item = doc.add_unordered_list(parent=parent_item)
-            parent_item = doc.add_list_item(
-                enumerated=is_numbered,
-                parent=parent_item,
-                text=snippet_text,
+            if len(child.children) > 1:  # inline group will be created further down
+                parent_item = self._create_list_item(
+                    doc=doc,
+                    parent_item=parent_item,
+                    text="",
                     formatting=formatting,
                     hyperlink=hyperlink,
                 )
+            else:
+                creation_stack.append(_ListItemCreationPayload())
 
         elif isinstance(element, marko.inline.Image):
             self._close_table(doc)
@@ -285,6 +344,31 @@ class MarkdownDocumentBackend(DeclarativeDocumentBackend):
                     self.md_table_buffer.append(snippet_text)
             elif snippet_text:
                 self._close_table(doc)
+
+                if creation_stack:
+                    while len(creation_stack) > 0:
+                        to_create = creation_stack.pop()
+                        if isinstance(to_create, _ListItemCreationPayload):
+                            parent_item = self._create_list_item(
+                                doc=doc,
+                                parent_item=parent_item,
+                                text=snippet_text,
+                                formatting=formatting,
+                                hyperlink=hyperlink,
+                            )
+                        elif isinstance(to_create, _HeadingCreationPayload):
+                            # not keeping as parent_item as logic for correctly tracking
+                            # that not implemented yet (section components not captured
+                            # as heading children in marko)
+                            self._create_heading_item(
+                                doc=doc,
+                                parent_item=parent_item,
+                                text=snippet_text,
+                                level=to_create.level,
+                                formatting=formatting,
+                                hyperlink=hyperlink,
+                            )
+                else:
                     doc.add_text(
                         label=DocItemLabel.TEXT,
                         parent=parent_item,
@@ -353,7 +437,6 @@ class MarkdownDocumentBackend(DeclarativeDocumentBackend):
             parent_item = doc.add_inline_group(parent=parent_item)
 
         processed_block_types = (
-            # marko.block.Heading,
             marko.block.CodeBlock,
             marko.block.FencedCode,
             marko.inline.RawText,
@@ -369,6 +452,7 @@ class MarkdownDocumentBackend(DeclarativeDocumentBackend):
                     depth=depth + 1,
                     doc=doc,
                     visited=visited,
+                    creation_stack=creation_stack,
                     parent_item=parent_item,
                     formatting=formatting,
                     hyperlink=hyperlink,
@@ -412,6 +496,7 @@ class MarkdownDocumentBackend(DeclarativeDocumentBackend):
                 doc=doc,
                 parent_item=None,
                 visited=set(),
+                creation_stack=[],
             )
             self._close_table(doc=doc)  # handle any last hanging table
 

docling/backend/msword_backend.py

@@ -14,7 +14,7 @@ from docling_core.types.doc import (
     TableCell,
     TableData,
 )
-from docling_core.types.doc.document import Formatting
+from docling_core.types.doc.document import Formatting, OrderedList, UnorderedList
 from docx import Document
 from docx.document import Document as DocxDocument
 from docx.oxml.table import CT_Tc
@@ -84,7 +84,7 @@ class MsWordDocumentBackend(DeclarativeDocumentBackend):
             self.valid = True
         except Exception as e:
             raise RuntimeError(
-                f"MsPowerpointDocumentBackend could not load document with hash {self.document_hash}"
+                f"MsWordDocumentBackend could not load document with hash {self.document_hash}"
             ) from e
 
     @override
@@ -251,9 +251,15 @@ class MsWordDocumentBackend(DeclarativeDocumentBackend):
                     self._handle_tables(element, docx_obj, doc)
                 except Exception:
                     _log.debug("could not parse a table, broken docx table")
-
+            # Check for Image
             elif drawing_blip:
                 self._handle_pictures(docx_obj, drawing_blip, doc)
+                # Check for Text after the Image
+                if (
+                    tag_name in ["p"]
+                    and element.find(".//w:t", namespaces=namespaces) is not None
+                ):
+                    self._handle_text_elements(element, docx_obj, doc)
             # Check for the sdt containers, like table of contents
             elif tag_name in ["sdt"]:
                 sdt_content = element.find(".//w:sdtContent", namespaces=namespaces)
@@ -268,6 +274,7 @@ class MsWordDocumentBackend(DeclarativeDocumentBackend):
                 self._handle_text_elements(element, docx_obj, doc)
             else:
                 _log.debug(f"Ignoring element in DOCX with tag: {tag_name}")
+
         return doc
 
     def _str_to_int(
@@ -390,7 +397,11 @@ class MsWordDocumentBackend(DeclarativeDocumentBackend):
             if isinstance(c, Hyperlink):
                 text = c.text
                 hyperlink = Path(c.address)
-                format = self._get_format_from_run(c.runs[0])
+                format = (
+                    self._get_format_from_run(c.runs[0])
+                    if c.runs and len(c.runs) > 0
+                    else None
+                )
             elif isinstance(c, Run):
                 text = c.text
                 hyperlink = None
@@ -578,7 +589,7 @@ class MsWordDocumentBackend(DeclarativeDocumentBackend):
         all_paragraphs = []
 
         # Sort paragraphs within each container, then process containers
-        for container_id, paragraphs in container_paragraphs.items():
+        for paragraphs in container_paragraphs.values():
             # Sort by vertical position within each container
             sorted_container_paragraphs = sorted(
                 paragraphs,
@@ -689,14 +700,13 @@ class MsWordDocumentBackend(DeclarativeDocumentBackend):
         doc: DoclingDocument,
     ) -> None:
         paragraph = Paragraph(element, docx_obj)
-
+        paragraph_elements = self._get_paragraph_elements(paragraph)
         text, equations = self._handle_equations_in_text(
             element=element, text=paragraph.text
         )
 
         if text is None:
             return
-        paragraph_elements = self._get_paragraph_elements(paragraph)
         text = text.strip()
 
         # Common styles for bullet and numbered lists.
@@ -912,6 +922,44 @@ class MsWordDocumentBackend(DeclarativeDocumentBackend):
         )
         return
 
+    def _add_formatted_list_item(
+        self,
+        doc: DoclingDocument,
+        elements: list,
+        marker: str,
+        enumerated: bool,
+        level: int,
+    ) -> None:
+        # This should not happen by construction
+        if not isinstance(self.parents[level], (OrderedList, UnorderedList)):
+            return
+        if len(elements) == 1:
+            text, format, hyperlink = elements[0]
+            doc.add_list_item(
+                marker=marker,
+                enumerated=enumerated,
+                parent=self.parents[level],
+                text=text,
+                formatting=format,
+                hyperlink=hyperlink,
+            )
+        else:
+            new_item = doc.add_list_item(
+                marker=marker,
+                enumerated=enumerated,
+                parent=self.parents[level],
+                text="",
+            )
+            new_parent = doc.add_group(label=GroupLabel.INLINE, parent=new_item)
+            for text, format, hyperlink in elements:
+                doc.add_text(
+                    label=DocItemLabel.TEXT,
+                    parent=new_parent,
+                    text=text,
+                    formatting=format,
+                    hyperlink=hyperlink,
+                )
+
     def _add_list_item(
         self,
         *,
@@ -921,6 +969,9 @@ class MsWordDocumentBackend(DeclarativeDocumentBackend):
         elements: list,
         is_numbered: bool = False,
     ) -> None:
+        # TODO: this method is always called with is_numbered. Numbered lists should be properly addressed.
+        if not elements:
+            return None
         enum_marker = ""
 
         level = self._get_level()
@@ -937,21 +988,9 @@ class MsWordDocumentBackend(DeclarativeDocumentBackend):
             if is_numbered:
                 enum_marker = str(self.listIter) + "."
                 is_numbered = True
-            new_parent = self._create_or_reuse_parent(
-                doc=doc,
-                prev_parent=self.parents[level],
-                paragraph_elements=elements,
+            self._add_formatted_list_item(
+                doc, elements, enum_marker, is_numbered, level
             )
-            for text, format, hyperlink in elements:
-                doc.add_list_item(
-                    marker=enum_marker,
-                    enumerated=is_numbered,
-                    parent=new_parent,
-                    text=text,
-                    formatting=format,
-                    hyperlink=hyperlink,
-                )
-
         elif (
             self._prev_numid() == numid
             and self.level_at_new_list is not None
@@ -981,20 +1020,12 @@ class MsWordDocumentBackend(DeclarativeDocumentBackend):
             if is_numbered:
                 enum_marker = str(self.listIter) + "."
                 is_numbered = True
-
-            new_parent = self._create_or_reuse_parent(
-                doc=doc,
-                prev_parent=self.parents[self.level_at_new_list + ilevel],
-                paragraph_elements=elements,
-            )
-            for text, format, hyperlink in elements:
-                doc.add_list_item(
-                    marker=enum_marker,
-                    enumerated=is_numbered,
-                    parent=new_parent,
-                    text=text,
-                    formatting=format,
-                    hyperlink=hyperlink,
+            self._add_formatted_list_item(
+                doc,
+                elements,
+                enum_marker,
+                is_numbered,
+                self.level_at_new_list + ilevel,
             )
         elif (
             self._prev_numid() == numid
@@ -1002,7 +1033,7 @@ class MsWordDocumentBackend(DeclarativeDocumentBackend):
             and prev_indent is not None
             and ilevel < prev_indent
         ):  # Close list
-            for k, v in self.parents.items():
+            for k in self.parents:
                 if k > self.level_at_new_list + ilevel:
                     self.parents[k] = None
 
@@ -1011,19 +1042,12 @@ class MsWordDocumentBackend(DeclarativeDocumentBackend):
             if is_numbered:
                 enum_marker = str(self.listIter) + "."
                 is_numbered = True
-            new_parent = self._create_or_reuse_parent(
-                doc=doc,
-                prev_parent=self.parents[self.level_at_new_list + ilevel],
-                paragraph_elements=elements,
-            )
-            for text, format, hyperlink in elements:
-                doc.add_list_item(
-                    marker=enum_marker,
-                    enumerated=is_numbered,
-                    parent=new_parent,
-                    text=text,
-                    formatting=format,
-                    hyperlink=hyperlink,
+            self._add_formatted_list_item(
+                doc,
+                elements,
+                enum_marker,
+                is_numbered,
+                self.level_at_new_list + ilevel,
             )
             self.listIter = 0
 
@@ -1033,21 +1057,10 @@ class MsWordDocumentBackend(DeclarativeDocumentBackend):
             if is_numbered:
                 enum_marker = str(self.listIter) + "."
                 is_numbered = True
-            new_parent = self._create_or_reuse_parent(
-                doc=doc,
-                prev_parent=self.parents[level - 1],
-                paragraph_elements=elements,
-            )
-            for text, format, hyperlink in elements:
-                # Add the list item to the parent group
-                doc.add_list_item(
-                    marker=enum_marker,
-                    enumerated=is_numbered,
-                    parent=new_parent,
-                    text=text,
-                    formatting=format,
-                    hyperlink=hyperlink,
+            self._add_formatted_list_item(
+                doc, elements, enum_marker, is_numbered, level - 1
             )
+
         return
 
     def _handle_tables(

docling/backend/noop_backend.py

@@ -0,0 +1,51 @@
+import logging
+from io import BytesIO
+from pathlib import Path
+from typing import Set, Union
+
+from docling.backend.abstract_backend import AbstractDocumentBackend
+from docling.datamodel.base_models import InputFormat
+from docling.datamodel.document import InputDocument
+
+_log = logging.getLogger(__name__)
+
+
+class NoOpBackend(AbstractDocumentBackend):
+    """
+    A no-op backend that only validates input existence.
+    Used e.g. for audio files where actual processing is handled by the ASR pipeline.
+    """
+
+    def __init__(self, in_doc: "InputDocument", path_or_stream: Union[BytesIO, Path]):
+        super().__init__(in_doc, path_or_stream)
+
+        _log.debug(f"NoOpBackend initialized for: {path_or_stream}")
+
+        # Validate input
+        try:
+            if isinstance(self.path_or_stream, BytesIO):
+                # Check if stream has content
+                self.valid = len(self.path_or_stream.getvalue()) > 0
+                _log.debug(
+                    f"BytesIO stream length: {len(self.path_or_stream.getvalue())}"
+                )
+            elif isinstance(self.path_or_stream, Path):
+                # Check if file exists
+                self.valid = self.path_or_stream.exists()
+                _log.debug(f"File exists: {self.valid}")
+            else:
+                self.valid = False
+        except Exception as e:
+            _log.error(f"NoOpBackend validation failed: {e}")
+            self.valid = False
+
+    def is_valid(self) -> bool:
+        return self.valid
+
+    @classmethod
+    def supports_pagination(cls) -> bool:
+        return False
+
+    @classmethod
+    def supported_formats(cls) -> Set[InputFormat]:
+        return set(InputFormat)

docling/cli/main.py

@@ -29,6 +29,15 @@ from docling.backend.docling_parse_v4_backend import DoclingParseV4DocumentBacke
 from docling.backend.pdf_backend import PdfDocumentBackend
 from docling.backend.pypdfium2_backend import PyPdfiumDocumentBackend
 from docling.datamodel.accelerator_options import AcceleratorDevice, AcceleratorOptions
+from docling.datamodel.asr_model_specs import (
+    WHISPER_BASE,
+    WHISPER_LARGE,
+    WHISPER_MEDIUM,
+    WHISPER_SMALL,
+    WHISPER_TINY,
+    WHISPER_TURBO,
+    AsrModelType,
+)
 from docling.datamodel.base_models import (
     ConversionStatus,
     FormatToExtensions,
@@ -37,12 +46,14 @@ from docling.datamodel.base_models import (
 )
 from docling.datamodel.document import ConversionResult
 from docling.datamodel.pipeline_options import (
+    AsrPipelineOptions,
     EasyOcrOptions,
     OcrOptions,
     PaginatedPipelineOptions,
     PdfBackend,
-    PdfPipeline,
     PdfPipelineOptions,
+    PipelineOptions,
+    ProcessingPipeline,
     TableFormerMode,
     VlmPipelineOptions,
 )
@@ -54,8 +65,14 @@ from docling.datamodel.vlm_model_specs import (
     SMOLDOCLING_TRANSFORMERS,
     VlmModelType,
 )
-from docling.document_converter import DocumentConverter, FormatOption, PdfFormatOption
+from docling.document_converter import (
+    AudioFormatOption,
+    DocumentConverter,
+    FormatOption,
+    PdfFormatOption,
+)
 from docling.models.factories import get_ocr_factory
+from docling.pipeline.asr_pipeline import AsrPipeline
 from docling.pipeline.vlm_pipeline import VlmPipeline
 
 warnings.filterwarnings(action="ignore", category=UserWarning, module="pydantic|torch")
@@ -296,13 +313,17 @@ def convert(  # noqa: C901
         ),
     ] = ImageRefMode.EMBEDDED,
     pipeline: Annotated[
-        PdfPipeline,
+        ProcessingPipeline,
         typer.Option(..., help="Choose the pipeline to process PDF or image files."),
-    ] = PdfPipeline.STANDARD,
+    ] = ProcessingPipeline.STANDARD,
     vlm_model: Annotated[
         VlmModelType,
         typer.Option(..., help="Choose the VLM model to use with PDF or image files."),
     ] = VlmModelType.SMOLDOCLING,
+    asr_model: Annotated[
+        AsrModelType,
+        typer.Option(..., help="Choose the ASR model to use with audio/video files."),
+    ] = AsrModelType.WHISPER_TINY,
     ocr: Annotated[
         bool,
         typer.Option(
@@ -450,12 +471,14 @@ def convert(  # noqa: C901
         ),
     ] = None,
 ):
+    log_format = "%(asctime)s\t%(levelname)s\t%(name)s: %(message)s"
+
     if verbose == 0:
-        logging.basicConfig(level=logging.WARNING)
+        logging.basicConfig(level=logging.WARNING, format=log_format)
     elif verbose == 1:
-        logging.basicConfig(level=logging.INFO)
+        logging.basicConfig(level=logging.INFO, format=log_format)
     else:
-        logging.basicConfig(level=logging.DEBUG)
+        logging.basicConfig(level=logging.DEBUG, format=log_format)
 
     settings.debug.visualize_cells = debug_visualize_cells
     settings.debug.visualize_layout = debug_visualize_layout
@@ -530,9 +553,12 @@ def convert(  # noqa: C901
             ocr_options.lang = ocr_lang_list
 
         accelerator_options = AcceleratorOptions(num_threads=num_threads, device=device)
-        pipeline_options: PaginatedPipelineOptions
+        # pipeline_options: PaginatedPipelineOptions
+        pipeline_options: PipelineOptions
 
-        if pipeline == PdfPipeline.STANDARD:
+        format_options: Dict[InputFormat, FormatOption] = {}
+
+        if pipeline == ProcessingPipeline.STANDARD:
             pipeline_options = PdfPipelineOptions(
                 allow_external_plugins=allow_external_plugins,
                 enable_remote_services=enable_remote_services,
@@ -574,7 +600,13 @@ def convert(  # noqa: C901
                 pipeline_options=pipeline_options,
                 backend=backend,  # pdf_backend
             )
-        elif pipeline == PdfPipeline.VLM:
+
+            format_options = {
+                InputFormat.PDF: pdf_format_option,
+                InputFormat.IMAGE: pdf_format_option,
+            }
+
+        elif pipeline == ProcessingPipeline.VLM:
             pipeline_options = VlmPipelineOptions(
                 enable_remote_services=enable_remote_services,
             )
@@ -600,13 +632,48 @@ def convert(  # noqa: C901
                 pipeline_cls=VlmPipeline, pipeline_options=pipeline_options
             )
 
-        if artifacts_path is not None:
-            pipeline_options.artifacts_path = artifacts_path
-
-        format_options: Dict[InputFormat, FormatOption] = {
+            format_options = {
                 InputFormat.PDF: pdf_format_option,
                 InputFormat.IMAGE: pdf_format_option,
             }
+
+        elif pipeline == ProcessingPipeline.ASR:
+            pipeline_options = AsrPipelineOptions(
+                # enable_remote_services=enable_remote_services,
+                # artifacts_path = artifacts_path
+            )
+
+            if asr_model == AsrModelType.WHISPER_TINY:
+                pipeline_options.asr_options = WHISPER_TINY
+            elif asr_model == AsrModelType.WHISPER_SMALL:
+                pipeline_options.asr_options = WHISPER_SMALL
+            elif asr_model == AsrModelType.WHISPER_MEDIUM:
+                pipeline_options.asr_options = WHISPER_MEDIUM
+            elif asr_model == AsrModelType.WHISPER_BASE:
+                pipeline_options.asr_options = WHISPER_BASE
+            elif asr_model == AsrModelType.WHISPER_LARGE:
+                pipeline_options.asr_options = WHISPER_LARGE
+            elif asr_model == AsrModelType.WHISPER_TURBO:
+                pipeline_options.asr_options = WHISPER_TURBO
+            else:
+                _log.error(f"{asr_model} is not known")
+                raise ValueError(f"{asr_model} is not known")
+
+            _log.info(f"pipeline_options: {pipeline_options}")
+
+            audio_format_option = AudioFormatOption(
+                pipeline_cls=AsrPipeline,
+                pipeline_options=pipeline_options,
+            )
+
+            format_options = {
+                InputFormat.AUDIO: audio_format_option,
+            }
+
+        if artifacts_path is not None:
+            pipeline_options.artifacts_path = artifacts_path
+            # audio_pipeline_options.artifacts_path = artifacts_path
+
         doc_converter = DocumentConverter(
             allowed_formats=from_formats,
             format_options=format_options,
@@ -614,6 +681,7 @@ def convert(  # noqa: C901
 
         start_time = time.time()
 
+        _log.info(f"paths: {input_doc_paths}")
         conv_results = doc_converter.convert_all(
             input_doc_paths, headers=parsed_headers, raises_on_error=abort_on_error
         )

docling/datamodel/asr_model_specs.py

@@ -0,0 +1,92 @@
+import logging
+from enum import Enum
+
+from pydantic import (
+    AnyUrl,
+)
+
+from docling.datamodel.accelerator_options import AcceleratorDevice
+from docling.datamodel.pipeline_options_asr_model import (
+    # AsrResponseFormat,
+    # ApiAsrOptions,
+    InferenceAsrFramework,
+    InlineAsrNativeWhisperOptions,
+    TransformersModelType,
+)
+
+_log = logging.getLogger(__name__)
+
+WHISPER_TINY = InlineAsrNativeWhisperOptions(
+    repo_id="tiny",
+    inference_framework=InferenceAsrFramework.WHISPER,
+    verbose=True,
+    timestamps=True,
+    word_timestamps=True,
+    temperatue=0.0,
+    max_new_tokens=256,
+    max_time_chunk=30.0,
+)
+
+WHISPER_SMALL = InlineAsrNativeWhisperOptions(
+    repo_id="small",
+    inference_framework=InferenceAsrFramework.WHISPER,
+    verbose=True,
+    timestamps=True,
+    word_timestamps=True,
+    temperatue=0.0,
+    max_new_tokens=256,
+    max_time_chunk=30.0,
+)
+
+WHISPER_MEDIUM = InlineAsrNativeWhisperOptions(
+    repo_id="medium",
+    inference_framework=InferenceAsrFramework.WHISPER,
+    verbose=True,
+    timestamps=True,
+    word_timestamps=True,
+    temperatue=0.0,
+    max_new_tokens=256,
+    max_time_chunk=30.0,
+)
+
+WHISPER_BASE = InlineAsrNativeWhisperOptions(
+    repo_id="base",
+    inference_framework=InferenceAsrFramework.WHISPER,
+    verbose=True,
+    timestamps=True,
+    word_timestamps=True,
+    temperatue=0.0,
+    max_new_tokens=256,
+    max_time_chunk=30.0,
+)
+
+WHISPER_LARGE = InlineAsrNativeWhisperOptions(
+    repo_id="large",
+    inference_framework=InferenceAsrFramework.WHISPER,
+    verbose=True,
+    timestamps=True,
+    word_timestamps=True,
+    temperatue=0.0,
+    max_new_tokens=256,
+    max_time_chunk=30.0,
+)
+
+WHISPER_TURBO = InlineAsrNativeWhisperOptions(
+    repo_id="turbo",
+    inference_framework=InferenceAsrFramework.WHISPER,
+    verbose=True,
+    timestamps=True,
+    word_timestamps=True,
+    temperatue=0.0,
+    max_new_tokens=256,
+    max_time_chunk=30.0,
+)
+
+
+class AsrModelType(str, Enum):
+    WHISPER_TINY = "whisper_tiny"
+    WHISPER_SMALL = "whisper_small"
+    WHISPER_MEDIUM = "whisper_medium"
+    WHISPER_BASE = "whisper_base"
+    WHISPER_LARGE = "whisper_large"
+    WHISPER_TURBO = "whisper_turbo"

docling/datamodel/base_models.py

@@ -49,6 +49,7 @@ class InputFormat(str, Enum):
     XML_USPTO = "xml_uspto"
     XML_JATS = "xml_jats"
     JSON_DOCLING = "json_docling"
+    AUDIO = "audio"
 
 
 class OutputFormat(str, Enum):
@@ -73,6 +74,7 @@ FormatToExtensions: Dict[InputFormat, List[str]] = {
     InputFormat.XLSX: ["xlsx", "xlsm"],
     InputFormat.XML_USPTO: ["xml", "txt"],
     InputFormat.JSON_DOCLING: ["json"],
+    InputFormat.AUDIO: ["wav", "mp3"],
 }
 
 FormatToMimeType: Dict[InputFormat, List[str]] = {
@@ -104,6 +106,7 @@ FormatToMimeType: Dict[InputFormat, List[str]] = {
     ],
     InputFormat.XML_USPTO: ["application/xml", "text/plain"],
     InputFormat.JSON_DOCLING: ["application/json"],
+    InputFormat.AUDIO: ["audio/x-wav", "audio/mpeg", "audio/wav", "audio/mp3"],
 }
 
 MimeTypeToFormat: dict[str, list[InputFormat]] = {
@@ -298,7 +301,7 @@ class OpenAiChatMessage(BaseModel):
 class OpenAiResponseChoice(BaseModel):
     index: int
     message: OpenAiChatMessage
-    finish_reason: str
+    finish_reason: Optional[str]
 
 
 class OpenAiResponseUsage(BaseModel):

docling/datamodel/document.py

@@ -249,7 +249,7 @@ class _DocumentConversionInput(BaseModel):
             backend: Type[AbstractDocumentBackend]
             if format not in format_options.keys():
                 _log.error(
-                    f"Input document {obj.name} does not match any allowed format."
+                    f"Input document {obj.name} with format {format} does not match any allowed format: ({format_options.keys()})"
                 )
                 backend = _DummyBackend
             else:
@@ -318,6 +318,8 @@ class _DocumentConversionInput(BaseModel):
         mime = mime or _DocumentConversionInput._detect_csv(content)
         mime = mime or "text/plain"
         formats = MimeTypeToFormat.get(mime, [])
+        _log.info(f"detected formats: {formats}")
+
         if formats:
             if len(formats) == 1 and mime not in ("text/plain"):
                 return formats[0]

docling/datamodel/pipeline_options.py

@@ -11,8 +11,13 @@ from pydantic import (
 )
 from typing_extensions import deprecated
 
+from docling.datamodel import asr_model_specs
+
 # Import the following for backwards compatibility
 from docling.datamodel.accelerator_options import AcceleratorDevice, AcceleratorOptions
+from docling.datamodel.pipeline_options_asr_model import (
+    InlineAsrOptions,
+)
 from docling.datamodel.pipeline_options_vlm_model import (
     ApiVlmOptions,
     InferenceFramework,
@@ -202,7 +207,7 @@ smolvlm_picture_description = PictureDescriptionVlmOptions(
 
 # GraniteVision
 granite_picture_description = PictureDescriptionVlmOptions(
-    repo_id="ibm-granite/granite-vision-3.1-2b-preview",
+    repo_id="ibm-granite/granite-vision-3.2-2b-preview",
     prompt="What is shown in this image?",
 )
 
@@ -260,6 +265,11 @@ class VlmPipelineOptions(PaginatedPipelineOptions):
     )
 
 
+class AsrPipelineOptions(PipelineOptions):
+    asr_options: Union[InlineAsrOptions] = asr_model_specs.WHISPER_TINY
+    artifacts_path: Optional[Union[Path, str]] = None
+
+
 class PdfPipelineOptions(PaginatedPipelineOptions):
     """Options for the PDF pipeline."""
 
@@ -297,6 +307,7 @@ class PdfPipelineOptions(PaginatedPipelineOptions):
     )
 
 
-class PdfPipeline(str, Enum):
+class ProcessingPipeline(str, Enum):
     STANDARD = "standard"
     VLM = "vlm"
+    ASR = "asr"

docling/datamodel/pipeline_options_asr_model.py

@@ -0,0 +1,57 @@
+from enum import Enum
+from typing import Any, Dict, List, Literal, Optional, Union
+
+from pydantic import AnyUrl, BaseModel
+from typing_extensions import deprecated
+
+from docling.datamodel.accelerator_options import AcceleratorDevice
+from docling.datamodel.pipeline_options_vlm_model import (
+    # InferenceFramework,
+    TransformersModelType,
+)
+
+
+class BaseAsrOptions(BaseModel):
+    kind: str
+    # prompt: str
+
+
+class InferenceAsrFramework(str, Enum):
+    # MLX = "mlx" # disabled for now
+    # TRANSFORMERS = "transformers" # disabled for now
+    WHISPER = "whisper"
+
+
+class InlineAsrOptions(BaseAsrOptions):
+    kind: Literal["inline_model_options"] = "inline_model_options"
+
+    repo_id: str
+
+    verbose: bool = False
+    timestamps: bool = True
+
+    temperature: float = 0.0
+    max_new_tokens: int = 256
+    max_time_chunk: float = 30.0
+
+    torch_dtype: Optional[str] = None
+    supported_devices: List[AcceleratorDevice] = [
+        AcceleratorDevice.CPU,
+        AcceleratorDevice.CUDA,
+        AcceleratorDevice.MPS,
+    ]
+
+    @property
+    def repo_cache_folder(self) -> str:
+        return self.repo_id.replace("/", "--")
+
+
+class InlineAsrNativeWhisperOptions(InlineAsrOptions):
+    inference_framework: InferenceAsrFramework = InferenceAsrFramework.WHISPER
+
+    language: str = "en"
+    supported_devices: List[AcceleratorDevice] = [
+        AcceleratorDevice.CPU,
+        AcceleratorDevice.CUDA,
+    ]
+    word_timestamps: bool = True

docling/document_converter.py

@@ -19,6 +19,7 @@ from docling.backend.md_backend import MarkdownDocumentBackend
 from docling.backend.msexcel_backend import MsExcelDocumentBackend
 from docling.backend.mspowerpoint_backend import MsPowerpointDocumentBackend
 from docling.backend.msword_backend import MsWordDocumentBackend
+from docling.backend.noop_backend import NoOpBackend
 from docling.backend.xml.jats_backend import JatsDocumentBackend
 from docling.backend.xml.uspto_backend import PatentUsptoDocumentBackend
 from docling.datamodel.base_models import (
@@ -41,6 +42,7 @@ from docling.datamodel.settings import (
     settings,
 )
 from docling.exceptions import ConversionError
+from docling.pipeline.asr_pipeline import AsrPipeline
 from docling.pipeline.base_pipeline import BasePipeline
 from docling.pipeline.simple_pipeline import SimplePipeline
 from docling.pipeline.standard_pdf_pipeline import StandardPdfPipeline
@@ -118,6 +120,11 @@ class PdfFormatOption(FormatOption):
     backend: Type[AbstractDocumentBackend] = DoclingParseV4DocumentBackend
 
 
+class AudioFormatOption(FormatOption):
+    pipeline_cls: Type = AsrPipeline
+    backend: Type[AbstractDocumentBackend] = NoOpBackend
+
+
 def _get_default_option(format: InputFormat) -> FormatOption:
     format_to_default_options = {
         InputFormat.CSV: FormatOption(
@@ -156,6 +163,7 @@ def _get_default_option(format: InputFormat) -> FormatOption:
         InputFormat.JSON_DOCLING: FormatOption(
             pipeline_cls=SimplePipeline, backend=DoclingJSONBackend
         ),
+        InputFormat.AUDIO: FormatOption(pipeline_cls=AsrPipeline, backend=NoOpBackend),
     }
     if (options := format_to_default_options.get(format)) is not None:
         return options

docling/models/readingorder_model.py

@@ -124,7 +124,7 @@ class ReadingOrderModel:
             page_no = page.page_no + 1
             size = page.size
 
-            assert size is not None
+            assert size is not None, "Page size is not initialized."
 
             out_doc.add_page(page_no=page_no, size=size)
 

docling/pipeline/asr_pipeline.py

@@ -0,0 +1,253 @@
+import logging
+import os
+import re
+from io import BytesIO
+from pathlib import Path
+from typing import List, Optional, Union, cast
+
+from docling_core.types.doc import DoclingDocument, DocumentOrigin
+
+# import whisper  # type: ignore
+# import librosa
+# import numpy as np
+# import soundfile as sf  # type: ignore
+from docling_core.types.doc.labels import DocItemLabel
+from pydantic import BaseModel, Field, validator
+
+from docling.backend.abstract_backend import AbstractDocumentBackend
+from docling.backend.noop_backend import NoOpBackend
+
+# from pydub import AudioSegment  # type: ignore
+# from transformers import WhisperForConditionalGeneration, WhisperProcessor, pipeline
+from docling.datamodel.accelerator_options import (
+    AcceleratorOptions,
+)
+from docling.datamodel.base_models import (
+    ConversionStatus,
+    FormatToMimeType,
+)
+from docling.datamodel.document import ConversionResult, InputDocument
+from docling.datamodel.pipeline_options import (
+    AsrPipelineOptions,
+)
+from docling.datamodel.pipeline_options_asr_model import (
+    InlineAsrNativeWhisperOptions,
+    # AsrResponseFormat,
+    InlineAsrOptions,
+)
+from docling.datamodel.pipeline_options_vlm_model import (
+    InferenceFramework,
+)
+from docling.datamodel.settings import settings
+from docling.pipeline.base_pipeline import BasePipeline
+from docling.utils.accelerator_utils import decide_device
+from docling.utils.profiling import ProfilingScope, TimeRecorder
+
+_log = logging.getLogger(__name__)
+
+
+class _ConversationWord(BaseModel):
+    text: str
+    start_time: Optional[float] = Field(
+        None, description="Start time in seconds from video start"
+    )
+    end_time: Optional[float] = Field(
+        None, ge=0, description="End time in seconds from video start"
+    )
+
+
+class _ConversationItem(BaseModel):
+    text: str
+    start_time: Optional[float] = Field(
+        None, description="Start time in seconds from video start"
+    )
+    end_time: Optional[float] = Field(
+        None, ge=0, description="End time in seconds from video start"
+    )
+    speaker_id: Optional[int] = Field(None, description="Numeric speaker identifier")
+    speaker: Optional[str] = Field(
+        None, description="Speaker name, defaults to speaker-{speaker_id}"
+    )
+    words: Optional[list[_ConversationWord]] = Field(
+        None, description="Individual words with time-stamps"
+    )
+
+    def __lt__(self, other):
+        if not isinstance(other, _ConversationItem):
+            return NotImplemented
+        return self.start_time < other.start_time
+
+    def __eq__(self, other):
+        if not isinstance(other, _ConversationItem):
+            return NotImplemented
+        return self.start_time == other.start_time
+
+    def to_string(self) -> str:
+        """Format the conversation entry as a string"""
+        result = ""
+        if (self.start_time is not None) and (self.end_time is not None):
+            result += f"[time: {self.start_time}-{self.end_time}] "
+
+        if self.speaker is not None:
+            result += f"[speaker:{self.speaker}] "
+
+        result += self.text
+        return result
+
+
+class _NativeWhisperModel:
+    def __init__(
+        self,
+        enabled: bool,
+        artifacts_path: Optional[Path],
+        accelerator_options: AcceleratorOptions,
+        asr_options: InlineAsrNativeWhisperOptions,
+    ):
+        """
+        Transcriber using native Whisper.
+        """
+        self.enabled = enabled
+
+        _log.info(f"artifacts-path: {artifacts_path}")
+        _log.info(f"accelerator_options: {accelerator_options}")
+
+        if self.enabled:
+            try:
+                import whisper  # type: ignore
+            except ImportError:
+                raise ImportError(
+                    "whisper is not installed. Please install it via `pip install openai-whisper` or do `uv sync --extra asr`."
+                )
+            self.asr_options = asr_options
+            self.max_tokens = asr_options.max_new_tokens
+            self.temperature = asr_options.temperature
+
+            self.device = decide_device(
+                accelerator_options.device,
+                supported_devices=asr_options.supported_devices,
+            )
+            _log.info(f"Available device for Whisper: {self.device}")
+
+            self.model_name = asr_options.repo_id
+            _log.info(f"loading _NativeWhisperModel({self.model_name})")
+            if artifacts_path is not None:
+                _log.info(f"loading {self.model_name} from {artifacts_path}")
+                self.model = whisper.load_model(
+                    name=self.model_name,
+                    device=self.device,
+                    download_root=str(artifacts_path),
+                )
+            else:
+                self.model = whisper.load_model(
+                    name=self.model_name, device=self.device
+                )
+
+            self.verbose = asr_options.verbose
+            self.timestamps = asr_options.timestamps
+            self.word_timestamps = asr_options.word_timestamps
+
+    def run(self, conv_res: ConversionResult) -> ConversionResult:
+        audio_path: Path = Path(conv_res.input.file).resolve()
+
+        try:
+            conversation = self.transcribe(audio_path)
+
+            # Ensure we have a proper DoclingDocument
+            origin = DocumentOrigin(
+                filename=conv_res.input.file.name or "audio.wav",
+                mimetype="audio/x-wav",
+                binary_hash=conv_res.input.document_hash,
+            )
+            conv_res.document = DoclingDocument(
+                name=conv_res.input.file.stem or "audio.wav", origin=origin
+            )
+
+            for citem in conversation:
+                conv_res.document.add_text(
+                    label=DocItemLabel.TEXT, text=citem.to_string()
+                )
+
+            conv_res.status = ConversionStatus.SUCCESS
+            return conv_res
+
+        except Exception as exc:
+            _log.error(f"Audio tranciption has an error: {exc}")
+
+        conv_res.status = ConversionStatus.FAILURE
+        return conv_res
+
+    def transcribe(self, fpath: Path) -> list[_ConversationItem]:
+        result = self.model.transcribe(
+            str(fpath), verbose=self.verbose, word_timestamps=self.word_timestamps
+        )
+
+        convo: list[_ConversationItem] = []
+        for _ in result["segments"]:
+            item = _ConversationItem(
+                start_time=_["start"], end_time=_["end"], text=_["text"], words=[]
+            )
+            if "words" in _ and self.word_timestamps:
+                item.words = []
+                for __ in _["words"]:
+                    item.words.append(
+                        _ConversationWord(
+                            start_time=__["start"],
+                            end_time=__["end"],
+                            text=__["word"],
+                        )
+                    )
+            convo.append(item)
+
+        return convo
+
+
+class AsrPipeline(BasePipeline):
+    def __init__(self, pipeline_options: AsrPipelineOptions):
+        super().__init__(pipeline_options)
+        self.keep_backend = True
+
+        self.pipeline_options: AsrPipelineOptions = pipeline_options
+
+        artifacts_path: Optional[Path] = None
+        if pipeline_options.artifacts_path is not None:
+            artifacts_path = Path(pipeline_options.artifacts_path).expanduser()
+        elif settings.artifacts_path is not None:
+            artifacts_path = Path(settings.artifacts_path).expanduser()
+
+        if artifacts_path is not None and not artifacts_path.is_dir():
+            raise RuntimeError(
+                f"The value of {artifacts_path=} is not valid. "
+                "When defined, it must point to a folder containing all models required by the pipeline."
+            )
+
+        if isinstance(self.pipeline_options.asr_options, InlineAsrNativeWhisperOptions):
+            asr_options: InlineAsrNativeWhisperOptions = (
+                self.pipeline_options.asr_options
+            )
+            self._model = _NativeWhisperModel(
+                enabled=True,  # must be always enabled for this pipeline to make sense.
+                artifacts_path=artifacts_path,
+                accelerator_options=pipeline_options.accelerator_options,
+                asr_options=asr_options,
+            )
+        else:
+            _log.error(f"No model support for {self.pipeline_options.asr_options}")
+
+    def _determine_status(self, conv_res: ConversionResult) -> ConversionStatus:
+        status = ConversionStatus.SUCCESS
+        return status
+
+    @classmethod
+    def get_default_options(cls) -> AsrPipelineOptions:
+        return AsrPipelineOptions()
+
+    def _build_document(self, conv_res: ConversionResult) -> ConversionResult:
+        _log.info(f"start _build_document in AsrPipeline: {conv_res.input.file}")
+        with TimeRecorder(conv_res, "doc_build", scope=ProfilingScope.DOCUMENT):
+            self._model.run(conv_res=conv_res)
+
+        return conv_res
+
+    @classmethod
+    def is_backend_supported(cls, backend: AbstractDocumentBackend):
+        return isinstance(backend, NoOpBackend)

docling/pipeline/base_pipeline.py

@@ -193,6 +193,17 @@ class PaginatedPipeline(BasePipeline):  # TODO this is a bad name.
                 )
                 raise e
 
+            # Filter out uninitialized pages (those with size=None) that may remain
+            # after timeout or processing failures to prevent assertion errors downstream
+            initial_page_count = len(conv_res.pages)
+            conv_res.pages = [page for page in conv_res.pages if page.size is not None]
+
+            if len(conv_res.pages) < initial_page_count:
+                _log.info(
+                    f"Filtered out {initial_page_count - len(conv_res.pages)} uninitialized pages "
+                    f"due to timeout or processing failures"
+                )
+
         return conv_res
 
     def _unload(self, conv_res: ConversionResult) -> ConversionResult:

docs/examples/batch_convert.py

@@ -121,14 +121,15 @@ def export_documents(
 def main():
     logging.basicConfig(level=logging.INFO)
 
+    data_folder = Path(__file__).parent / "../../tests/data"
     input_doc_paths = [
-        Path("./tests/data/pdf/2206.01062.pdf"),
-        Path("./tests/data/pdf/2203.01017v2.pdf"),
-        Path("./tests/data/pdf/2305.03393v1.pdf"),
-        Path("./tests/data/pdf/redp5110_sampled.pdf"),
+        data_folder / "pdf/2206.01062.pdf",
+        data_folder / "pdf/2203.01017v2.pdf",
+        data_folder / "pdf/2305.03393v1.pdf",
+        data_folder / "pdf/redp5110_sampled.pdf",
     ]
 
-    # buf = BytesIO(Path("./test/data/2206.01062.pdf").open("rb").read())
+    # buf = BytesIO((data_folder / "pdf/2206.01062.pdf").open("rb").read())
     # docs = [DocumentStream(name="my_doc.pdf", stream=buf)]
     # input = DocumentConversionInput.from_streams(docs)
 

docs/examples/custom_convert.py

@@ -16,7 +16,8 @@ _log = logging.getLogger(__name__)
 def main():
     logging.basicConfig(level=logging.INFO)
 
-    input_doc_path = Path("./tests/data/pdf/2206.01062.pdf")
+    data_folder = Path(__file__).parent / "../../tests/data"
+    input_doc_path = data_folder / "pdf/2206.01062.pdf"
 
     ###########################################################################
 

docs/examples/develop_formula_understanding.py

@@ -71,7 +71,8 @@ class ExampleFormulaUnderstandingPipeline(StandardPdfPipeline):
 def main():
     logging.basicConfig(level=logging.INFO)
 
-    input_doc_path = Path("./tests/data/pdf/2203.01017v2.pdf")
+    data_folder = Path(__file__).parent / "../../tests/data"
+    input_doc_path = data_folder / "pdf/2203.01017v2.pdf"
 
     pipeline_options = ExampleFormulaUnderstandingPipelineOptions()
     pipeline_options.do_formula_understanding = True

docs/examples/develop_picture_enrichment.py

@@ -76,7 +76,8 @@ class ExamplePictureClassifierPipeline(StandardPdfPipeline):
 def main():
     logging.basicConfig(level=logging.INFO)
 
-    input_doc_path = Path("./tests/data/pdf/2206.01062.pdf")
+    data_folder = Path(__file__).parent / "../../tests/data"
+    input_doc_path = data_folder / "pdf/2206.01062.pdf"
 
     pipeline_options = ExamplePictureClassifierPipelineOptions()
     pipeline_options.images_scale = 2.0

docs/examples/export_figures.py

@@ -16,7 +16,8 @@ IMAGE_RESOLUTION_SCALE = 2.0
 def main():
     logging.basicConfig(level=logging.INFO)
 
-    input_doc_path = Path("./tests/data/pdf/2206.01062.pdf")
+    data_folder = Path(__file__).parent / "../../tests/data"
+    input_doc_path = data_folder / "pdf/2206.01062.pdf"
     output_dir = Path("scratch")
 
     # Important: For operating with page images, we must keep them, otherwise the DocumentConverter

docs/examples/export_multimodal.py

@@ -19,7 +19,8 @@ IMAGE_RESOLUTION_SCALE = 2.0
 def main():
     logging.basicConfig(level=logging.INFO)
 
-    input_doc_path = Path("./tests/data/pdf/2206.01062.pdf")
+    data_folder = Path(__file__).parent / "../../tests/data"
+    input_doc_path = data_folder / "pdf/2206.01062.pdf"
     output_dir = Path("scratch")
 
     # Important: For operating with page images, we must keep them, otherwise the DocumentConverter

docs/examples/export_tables.py

@@ -12,7 +12,8 @@ _log = logging.getLogger(__name__)
 def main():
     logging.basicConfig(level=logging.INFO)
 
-    input_doc_path = Path("./tests/data/pdf/2206.01062.pdf")
+    data_folder = Path(__file__).parent / "../../tests/data"
+    input_doc_path = data_folder / "pdf/2206.01062.pdf"
     output_dir = Path("scratch")
 
     doc_converter = DocumentConverter()

docs/examples/full_page_ocr.py

@@ -9,7 +9,8 @@ from docling.document_converter import DocumentConverter, PdfFormatOption
 
 
 def main():
-    input_doc = Path("./tests/data/pdf/2206.01062.pdf")
+    data_folder = Path(__file__).parent / "../../tests/data"
+    input_doc_path = data_folder / "pdf/2206.01062.pdf"
 
     pipeline_options = PdfPipelineOptions()
     pipeline_options.do_ocr = True
@@ -32,7 +33,7 @@ def main():
         }
     )
 
-    doc = converter.convert(input_doc).document
+    doc = converter.convert(input_doc_path).document
     md = doc.export_to_markdown()
     print(md)
 

docs/examples/minimal_asr_pipeline.py

@@ -0,0 +1,56 @@
+from pathlib import Path
+
+from docling_core.types.doc import DoclingDocument
+
+from docling.datamodel import asr_model_specs
+from docling.datamodel.base_models import ConversionStatus, InputFormat
+from docling.datamodel.document import ConversionResult
+from docling.datamodel.pipeline_options import AsrPipelineOptions
+from docling.document_converter import AudioFormatOption, DocumentConverter
+from docling.pipeline.asr_pipeline import AsrPipeline
+
+
+def get_asr_converter():
+    """Create a DocumentConverter configured for ASR with whisper_turbo model."""
+    pipeline_options = AsrPipelineOptions()
+    pipeline_options.asr_options = asr_model_specs.WHISPER_TURBO
+
+    converter = DocumentConverter(
+        format_options={
+            InputFormat.AUDIO: AudioFormatOption(
+                pipeline_cls=AsrPipeline,
+                pipeline_options=pipeline_options,
+            )
+        }
+    )
+    return converter
+
+
+def asr_pipeline_conversion(audio_path: Path) -> DoclingDocument:
+    """ASR pipeline conversion using whisper_turbo"""
+    # Check if the test audio file exists
+    assert audio_path.exists(), f"Test audio file not found: {audio_path}"
+
+    converter = get_asr_converter()
+
+    # Convert the audio file
+    result: ConversionResult = converter.convert(audio_path)
+
+    # Verify conversion was successful
+    assert result.status == ConversionStatus.SUCCESS, (
+        f"Conversion failed with status: {result.status}"
+    )
+    return result.document
+
+
+if __name__ == "__main__":
+    audio_path = Path("tests/data/audio/sample_10s.mp3")
+
+    doc = asr_pipeline_conversion(audio_path=audio_path)
+    print(doc.export_to_markdown())
+
+    # Expected output:
+    #
+    # [time: 0.0-4.0]  Shakespeare on Scenery by Oscar Wilde
+    #
+    # [time: 5.28-9.96]  This is a LibriVox recording. All LibriVox recordings are in the public domain.

docs/examples/pictures_description_api.py

@@ -96,7 +96,8 @@ def watsonx_vlm_options():
 def main():
     logging.basicConfig(level=logging.INFO)
 
-    input_doc_path = Path("./tests/data/pdf/2206.01062.pdf")
+    data_folder = Path(__file__).parent / "../../tests/data"
+    input_doc_path = data_folder / "pdf/2206.01062.pdf"
 
     pipeline_options = PdfPipelineOptions(
         enable_remote_services=True  # <-- this is required!

docs/examples/run_with_accelerator.py

@@ -10,7 +10,8 @@ from docling.document_converter import DocumentConverter, PdfFormatOption
 
 
 def main():
-    input_doc = Path("./tests/data/pdf/2206.01062.pdf")
+    data_folder = Path(__file__).parent / "../../tests/data"
+    input_doc_path = data_folder / "pdf/2206.01062.pdf"
 
     # Explicitly set the accelerator
     # accelerator_options = AcceleratorOptions(
@@ -47,7 +48,7 @@ def main():
     settings.debug.profile_pipeline_timings = True
 
     # Convert the document
-    conversion_result = converter.convert(input_doc)
+    conversion_result = converter.convert(input_doc_path)
     doc = conversion_result.document
 
     # List with total time per document

docs/examples/tesseract_lang_detection.py

@@ -9,7 +9,8 @@ from docling.document_converter import DocumentConverter, PdfFormatOption
 
 
 def main():
-    input_doc = Path("./tests/data/pdf/2206.01062.pdf")
+    data_folder = Path(__file__).parent / "../../tests/data"
+    input_doc_path = data_folder / "pdf/2206.01062.pdf"
 
     # Set lang=["auto"] with a tesseract OCR engine: TesseractOcrOptions, TesseractCliOcrOptions
     # ocr_options = TesseractOcrOptions(lang=["auto"])
@@ -27,7 +28,7 @@ def main():
         }
     )
 
-    doc = converter.convert(input_doc).document
+    doc = converter.convert(input_doc_path).document
     md = doc.export_to_markdown()
     print(md)
 

docs/examples/translate.py

@@ -30,7 +30,8 @@ def translate(text: str, src: str = "en", dest: str = "de"):
 def main():
     logging.basicConfig(level=logging.INFO)
 
-    input_doc_path = Path("./tests/data/pdf/2206.01062.pdf")
+    data_folder = Path(__file__).parent / "../../tests/data"
+    input_doc_path = data_folder / "pdf/2206.01062.pdf"
     output_dir = Path("scratch")
 
     # Important: For operating with page images, we must keep them, otherwise the DocumentConverter

docs/examples/vlm_pipeline_api_model.py

@@ -95,8 +95,8 @@ def watsonx_vlm_options(model: str, prompt: str):
 def main():
     logging.basicConfig(level=logging.INFO)
 
-    # input_doc_path = Path("./tests/data/pdf/2206.01062.pdf")
-    input_doc_path = Path("./tests/data/pdf/2305.03393v1-pg9.pdf")
+    data_folder = Path(__file__).parent / "../../tests/data"
+    input_doc_path = data_folder / "pdf/2305.03393v1-pg9.pdf"
 
     pipeline_options = VlmPipelineOptions(
         enable_remote_services=True  # <-- this is required!

docs/index.md

@@ -20,14 +20,15 @@ Docling simplifies document processing, parsing diverse formats — including ad
 
 ## Features
 
-* 🗂️ Parsing of [multiple document formats][supported_formats] incl. PDF, DOCX, XLSX, HTML, images, and more
+* 🗂️  Parsing of [multiple document formats][supported_formats] incl. PDF, DOCX, PPTX, XLSX, HTML, WAV, MP3, images (PNG, TIFF, JPEG, ...), and more
 * 📑 Advanced PDF understanding incl. page layout, reading order, table structure, code, formulas, image classification, and more
 * 🧬 Unified, expressive [DoclingDocument][docling_document] representation format
-* ↪️ Various [export formats][supported_formats] and options, including Markdown, HTML, and lossless JSON
+* ↪️  Various [export formats][supported_formats] and options, including Markdown, HTML, [DocTags](https://arxiv.org/abs/2503.11576) and lossless JSON
 * 🔒 Local execution capabilities for sensitive data and air-gapped environments
 * 🤖 Plug-and-play [integrations][integrations] incl. LangChain, LlamaIndex, Crew AI & Haystack for agentic AI
 * 🔍 Extensive OCR support for scanned PDFs and images
-* 🥚 Support of several Visual Language Models ([SmolDocling](https://huggingface.co/ds4sd/SmolDocling-256M-preview)) 🔥
+* 👓 Support of several Visual Language Models ([SmolDocling](https://huggingface.co/ds4sd/SmolDocling-256M-preview))
+* 🎙️  Support for Audio with Automatic Speech Recognition (ASR) models
 * 💻 Simple and convenient CLI
 
 ### Coming soon

mkdocs.yml

@@ -80,6 +80,7 @@ nav:
       - "VLM pipeline with SmolDocling": examples/minimal_vlm_pipeline.py
       - "VLM pipeline with remote model": examples/vlm_pipeline_api_model.py
       - "VLM comparison": examples/compare_vlm_models.py
+      - "ASR pipeline with Whisper": examples/minimal_asr_pipeline.py      
       - "Figure export": examples/export_figures.py
       - "Table export": examples/export_tables.py
       - "Multimodal export": examples/export_multimodal.py

pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "docling"
-version = "2.37.0"  # DO NOT EDIT, updated automatically
+version = "2.38.0"  # DO NOT EDIT, updated automatically
 description = "SDK and CLI for parsing PDF, DOCX, HTML, and more, to a unified document representation for powering downstream workflows such as gen AI applications."
 license = "MIT"
 keywords = [
@@ -99,6 +99,9 @@ rapidocr = [
   # 'onnxruntime (>=1.7.0,<2.0.0) ; python_version >= "3.10"',
   # 'onnxruntime (>=1.7.0,<1.20.0) ; python_version < "3.10"',
 ]
+asr = [
+    "openai-whisper>=20240930",
+]
 
 [dependency-groups]
 dev = [
@@ -145,6 +148,9 @@ constraints = [
 package = true
 default-groups = "all"
 
+[tool.uv.sources]
+openai-whisper = { git = "https://github.com/openai/whisper.git", rev = "dd985ac4b90cafeef8712f2998d62c59c3e62d22" }
+
 [tool.setuptools.packages.find]
 include = ["docling*"]
 

tests/data/groundtruth/docling_v1/2305.03393v1-pg9.pages.json

@@ -2705,7 +2705,7 @@
               "b": 102.78223000000003,
               "coord_origin": "TOPLEFT"
             },
-            "confidence": 0.9373534917831421,
+            "confidence": 0.9373533725738525,
             "cells": [
               {
                 "index": 0,
@@ -2745,7 +2745,7 @@
               "b": 102.78223000000003,
               "coord_origin": "TOPLEFT"
             },
-            "confidence": 0.8858680725097656,
+            "confidence": 0.8858679533004761,
             "cells": [
               {
                 "index": 1,
@@ -13641,7 +13641,7 @@
               "b": 102.78223000000003,
               "coord_origin": "TOPLEFT"
             },
-            "confidence": 0.9373534917831421,
+            "confidence": 0.9373533725738525,
             "cells": [
               {
                 "index": 0,
@@ -13687,7 +13687,7 @@
               "b": 102.78223000000003,
               "coord_origin": "TOPLEFT"
             },
-            "confidence": 0.8858680725097656,
+            "confidence": 0.8858679533004761,
             "cells": [
               {
                 "index": 1,
@@ -26499,7 +26499,7 @@
               "b": 102.78223000000003,
               "coord_origin": "TOPLEFT"
             },
-            "confidence": 0.9373534917831421,
+            "confidence": 0.9373533725738525,
             "cells": [
               {
                 "index": 0,
@@ -26545,7 +26545,7 @@
               "b": 102.78223000000003,
               "coord_origin": "TOPLEFT"
             },
-            "confidence": 0.8858680725097656,
+            "confidence": 0.8858679533004761,
             "cells": [
               {
                 "index": 1,

tests/data/groundtruth/docling_v2/2203.01017v2.json

@@ -1,6 +1,6 @@
 {
   "schema_name": "DoclingDocument",
-  "version": "1.3.0",
+  "version": "1.4.0",
   "name": "2203.01017v2",
   "origin": {
     "mimetype": "application/pdf",
@@ -17863,7 +17863,8 @@
             }
           ]
         ]
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/1",
@@ -18753,7 +18754,8 @@
             }
           ]
         ]
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/2",
@@ -20117,7 +20119,8 @@
             }
           ]
         ]
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/3",
@@ -22266,7 +22269,8 @@
             }
           ]
         ]
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/4",
@@ -22927,7 +22931,8 @@
             }
           ]
         ]
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/5",
@@ -24050,7 +24055,8 @@
             }
           ]
         ]
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/6",
@@ -26307,7 +26313,8 @@
             }
           ]
         ]
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/7",
@@ -27600,7 +27607,8 @@
             }
           ]
         ]
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/8",
@@ -27635,7 +27643,8 @@
         "num_rows": 0,
         "num_cols": 0,
         "grid": []
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/9",
@@ -27670,7 +27679,8 @@
         "num_rows": 0,
         "num_cols": 0,
         "grid": []
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/10",
@@ -27705,7 +27715,8 @@
         "num_rows": 0,
         "num_cols": 0,
         "grid": []
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/11",
@@ -27740,7 +27751,8 @@
         "num_rows": 0,
         "num_cols": 0,
         "grid": []
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/12",
@@ -27783,7 +27795,8 @@
         "num_rows": 0,
         "num_cols": 0,
         "grid": []
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/13",
@@ -27818,7 +27831,8 @@
         "num_rows": 0,
         "num_cols": 0,
         "grid": []
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/14",
@@ -27853,7 +27867,8 @@
         "num_rows": 0,
         "num_cols": 0,
         "grid": []
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/15",
@@ -27888,7 +27903,8 @@
         "num_rows": 0,
         "num_cols": 0,
         "grid": []
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/16",
@@ -27931,7 +27947,8 @@
         "num_rows": 0,
         "num_cols": 0,
         "grid": []
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/17",
@@ -27966,7 +27983,8 @@
         "num_rows": 0,
         "num_cols": 0,
         "grid": []
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/18",
@@ -28001,7 +28019,8 @@
         "num_rows": 0,
         "num_cols": 0,
         "grid": []
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/19",
@@ -28036,7 +28055,8 @@
         "num_rows": 0,
         "num_cols": 0,
         "grid": []
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/20",
@@ -28071,7 +28091,8 @@
         "num_rows": 0,
         "num_cols": 0,
         "grid": []
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/21",
@@ -28106,7 +28127,8 @@
         "num_rows": 0,
         "num_cols": 0,
         "grid": []
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/22",
@@ -28141,7 +28163,8 @@
         "num_rows": 0,
         "num_cols": 0,
         "grid": []
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/23",
@@ -28176,7 +28199,8 @@
         "num_rows": 0,
         "num_cols": 0,
         "grid": []
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/24",
@@ -28211,7 +28235,8 @@
         "num_rows": 0,
         "num_cols": 0,
         "grid": []
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/25",
@@ -28246,7 +28271,8 @@
         "num_rows": 0,
         "num_cols": 0,
         "grid": []
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/26",
@@ -28281,7 +28307,8 @@
         "num_rows": 0,
         "num_cols": 0,
         "grid": []
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/27",
@@ -28324,7 +28351,8 @@
         "num_rows": 0,
         "num_cols": 0,
         "grid": []
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/28",
@@ -28359,7 +28387,8 @@
         "num_rows": 0,
         "num_cols": 0,
         "grid": []
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/29",
@@ -28394,7 +28423,8 @@
         "num_rows": 0,
         "num_cols": 0,
         "grid": []
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/30",
@@ -28429,7 +28459,8 @@
         "num_rows": 0,
         "num_cols": 0,
         "grid": []
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/31",
@@ -28464,7 +28495,8 @@
         "num_rows": 0,
         "num_cols": 0,
         "grid": []
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/32",
@@ -28499,7 +28531,8 @@
         "num_rows": 0,
         "num_cols": 0,
         "grid": []
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/33",
@@ -28542,7 +28575,8 @@
         "num_rows": 0,
         "num_cols": 0,
         "grid": []
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/34",
@@ -28577,7 +28611,8 @@
         "num_rows": 0,
         "num_cols": 0,
         "grid": []
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/35",
@@ -28612,7 +28647,8 @@
         "num_rows": 0,
         "num_cols": 0,
         "grid": []
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/36",
@@ -28647,7 +28683,8 @@
         "num_rows": 0,
         "num_cols": 0,
         "grid": []
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/37",
@@ -28682,7 +28719,8 @@
         "num_rows": 0,
         "num_cols": 0,
         "grid": []
-      }
+      },
+      "annotations": []
     }
   ],
   "key_value_items": [],

tests/data/groundtruth/docling_v2/2206.01062.json

@@ -1,6 +1,6 @@
 {
   "schema_name": "DoclingDocument",
-  "version": "1.3.0",
+  "version": "1.4.0",
   "name": "2206.01062",
   "origin": {
     "mimetype": "application/pdf",
@@ -23491,7 +23491,8 @@
             }
           ]
         ]
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/1",
@@ -26654,7 +26655,8 @@
             }
           ]
         ]
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/2",
@@ -29187,7 +29189,8 @@
             }
           ]
         ]
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/3",
@@ -31574,7 +31577,8 @@
             }
           ]
         ]
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/4",
@@ -34177,7 +34181,8 @@
             }
           ]
         ]
-      }
+      },
+      "annotations": []
     }
   ],
   "key_value_items": [],

tests/data/groundtruth/docling_v2/2305.03393v1-pg9.json

@@ -1,6 +1,6 @@
 {
   "schema_name": "DoclingDocument",
-  "version": "1.3.0",
+  "version": "1.4.0",
   "name": "2305.03393v1-pg9",
   "origin": {
     "mimetype": "application/pdf",
@@ -2104,7 +2104,8 @@
             }
           ]
         ]
-      }
+      },
+      "annotations": []
     }
   ],
   "key_value_items": [],

tests/data/groundtruth/docling_v2/2305.03393v1-pg9.pages.json

@@ -2705,7 +2705,7 @@
               "b": 102.78223000000003,
               "coord_origin": "TOPLEFT"
             },
-            "confidence": 0.9373534917831421,
+            "confidence": 0.9373533725738525,
             "cells": [
               {
                 "index": 0,
@@ -2745,7 +2745,7 @@
               "b": 102.78223000000003,
               "coord_origin": "TOPLEFT"
             },
-            "confidence": 0.8858680725097656,
+            "confidence": 0.8858679533004761,
             "cells": [
               {
                 "index": 1,
@@ -13641,7 +13641,7 @@
               "b": 102.78223000000003,
               "coord_origin": "TOPLEFT"
             },
-            "confidence": 0.9373534917831421,
+            "confidence": 0.9373533725738525,
             "cells": [
               {
                 "index": 0,
@@ -13687,7 +13687,7 @@
               "b": 102.78223000000003,
               "coord_origin": "TOPLEFT"
             },
-            "confidence": 0.8858680725097656,
+            "confidence": 0.8858679533004761,
             "cells": [
               {
                 "index": 1,
@@ -26499,7 +26499,7 @@
               "b": 102.78223000000003,
               "coord_origin": "TOPLEFT"
             },
-            "confidence": 0.9373534917831421,
+            "confidence": 0.9373533725738525,
             "cells": [
               {
                 "index": 0,
@@ -26545,7 +26545,7 @@
               "b": 102.78223000000003,
               "coord_origin": "TOPLEFT"
             },
-            "confidence": 0.8858680725097656,
+            "confidence": 0.8858679533004761,
             "cells": [
               {
                 "index": 1,

tests/data/groundtruth/docling_v2/2305.03393v1.doctags.txt

@@ -60,6 +60,8 @@
 <page_header><loc_159><loc_59><loc_366><loc_64>Optimized Table Tokenization for Table Structure Recognition</page_header>
 <page_header><loc_389><loc_59><loc_393><loc_64>7</page_header>
 <picture><loc_135><loc_103><loc_367><loc_177><caption><loc_110><loc_79><loc_393><loc_98>Fig. 3. OTSL description of table structure: A - table example; B - graphical representation of table structure; C - mapping structure on a grid; D - OTSL structure encoding; E - explanation on cell encoding</caption></picture>
+<unordered_list><list_item><loc_273><loc_172><loc_349><loc_176>4 - 2d merges: "C", "L", "U", "X"</list_item>
+</unordered_list>
 <section_header_level_1><loc_110><loc_193><loc_202><loc_198>4.2 Language Syntax</section_header_level_1>
 <text><loc_110><loc_205><loc_297><loc_211>The OTSL representation follows these syntax rules:</text>
 <unordered_list><list_item><loc_114><loc_219><loc_393><loc_232>1. Left-looking cell rule : The left neighbour of an "L" cell must be either another "L" cell or a "C" cell.</list_item>

tests/data/groundtruth/docling_v2/2305.03393v1.md

@@ -84,6 +84,8 @@ Fig. 3. OTSL description of table structure: A - table example; B - graphical re
 
 <!-- image -->
 
+- 4 - 2d merges: "C", "L", "U", "X"
+
 ## 4.2 Language Syntax
 
 The OTSL representation follows these syntax rules:

tests/data/groundtruth/docling_v2/amt_handbook_sample.json

@@ -1,6 +1,6 @@
 {
   "schema_name": "DoclingDocument",
-  "version": "1.3.0",
+  "version": "1.4.0",
   "name": "amt_handbook_sample",
   "origin": {
     "mimetype": "application/pdf",

tests/data/groundtruth/docling_v2/code_and_formula.json

@@ -1,6 +1,6 @@
 {
   "schema_name": "DoclingDocument",
-  "version": "1.3.0",
+  "version": "1.4.0",
   "name": "code_and_formula",
   "origin": {
     "mimetype": "application/pdf",

tests/data/groundtruth/docling_v2/csv-comma-in-cell.csv.json

@@ -1,6 +1,6 @@
 {
   "schema_name": "DoclingDocument",
-  "version": "1.3.0",
+  "version": "1.4.0",
   "name": "csv-comma-in-cell",
   "origin": {
     "mimetype": "text/csv",
@@ -538,7 +538,8 @@
             }
           ]
         ]
-      }
+      },
+      "annotations": []
     }
   ],
   "key_value_items": [],

tests/data/groundtruth/docling_v2/csv-comma.csv.json

@@ -1,6 +1,6 @@
 {
   "schema_name": "DoclingDocument",
-  "version": "1.3.0",
+  "version": "1.4.0",
   "name": "csv-comma",
   "origin": {
     "mimetype": "text/csv",
@@ -1788,7 +1788,8 @@
             }
           ]
         ]
-      }
+      },
+      "annotations": []
     }
   ],
   "key_value_items": [],

tests/data/groundtruth/docling_v2/csv-inconsistent-header.csv.json

@@ -1,6 +1,6 @@
 {
   "schema_name": "DoclingDocument",
-  "version": "1.3.0",
+  "version": "1.4.0",
   "name": "csv-inconsistent-header",
   "origin": {
     "mimetype": "text/csv",
@@ -526,7 +526,8 @@
             }
           ]
         ]
-      }
+      },
+      "annotations": []
     }
   ],
   "key_value_items": [],

tests/data/groundtruth/docling_v2/csv-pipe.csv.json

@@ -1,6 +1,6 @@
 {
   "schema_name": "DoclingDocument",
-  "version": "1.3.0",
+  "version": "1.4.0",
   "name": "csv-pipe",
   "origin": {
     "mimetype": "text/csv",
@@ -1788,7 +1788,8 @@
             }
           ]
         ]
-      }
+      },
+      "annotations": []
     }
   ],
   "key_value_items": [],

tests/data/groundtruth/docling_v2/csv-semicolon.csv.json

@@ -1,6 +1,6 @@
 {
   "schema_name": "DoclingDocument",
-  "version": "1.3.0",
+  "version": "1.4.0",
   "name": "csv-semicolon",
   "origin": {
     "mimetype": "text/csv",
@@ -1788,7 +1788,8 @@
             }
           ]
         ]
-      }
+      },
+      "annotations": []
     }
   ],
   "key_value_items": [],

tests/data/groundtruth/docling_v2/csv-tab.csv.json

@@ -1,6 +1,6 @@
 {
   "schema_name": "DoclingDocument",
-  "version": "1.3.0",
+  "version": "1.4.0",
   "name": "csv-tab",
   "origin": {
     "mimetype": "text/csv",
@@ -1788,7 +1788,8 @@
             }
           ]
         ]
-      }
+      },
+      "annotations": []
     }
   ],
   "key_value_items": [],

tests/data/groundtruth/docling_v2/csv-too-few-columns.csv.json

@@ -1,6 +1,6 @@
 {
   "schema_name": "DoclingDocument",
-  "version": "1.3.0",
+  "version": "1.4.0",
   "name": "csv-too-few-columns",
   "origin": {
     "mimetype": "text/csv",
@@ -526,7 +526,8 @@
             }
           ]
         ]
-      }
+      },
+      "annotations": []
     }
   ],
   "key_value_items": [],

tests/data/groundtruth/docling_v2/csv-too-many-columns.csv.json

@@ -1,6 +1,6 @@
 {
   "schema_name": "DoclingDocument",
-  "version": "1.3.0",
+  "version": "1.4.0",
   "name": "csv-too-many-columns",
   "origin": {
     "mimetype": "text/csv",
@@ -610,7 +610,8 @@
             }
           ]
         ]
-      }
+      },
+      "annotations": []
     }
   ],
   "key_value_items": [],

tests/data/groundtruth/docling_v2/equations.docx.json

@@ -1,6 +1,6 @@
 {
   "schema_name": "DoclingDocument",
-  "version": "1.3.0",
+  "version": "1.4.0",
   "name": "equations",
   "origin": {
     "mimetype": "application/vnd.openxmlformats-officedocument.wordprocessingml.document",
@@ -250,7 +250,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -280,7 +281,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -322,7 +324,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -436,7 +439,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -466,7 +470,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -520,7 +525,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -634,7 +640,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {

tests/data/groundtruth/docling_v2/example_01.html.json

@@ -1,6 +1,6 @@
 {
   "schema_name": "DoclingDocument",
-  "version": "1.3.0",
+  "version": "1.4.0",
   "name": "example_01",
   "origin": {
     "mimetype": "text/html",

tests/data/groundtruth/docling_v2/example_02.html.json

@@ -1,6 +1,6 @@
 {
   "schema_name": "DoclingDocument",
-  "version": "1.3.0",
+  "version": "1.4.0",
   "name": "example_02",
   "origin": {
     "mimetype": "text/html",

tests/data/groundtruth/docling_v2/example_03.html.json

@@ -1,6 +1,6 @@
 {
   "schema_name": "DoclingDocument",
-  "version": "1.3.0",
+  "version": "1.4.0",
   "name": "example_03",
   "origin": {
     "mimetype": "text/html",
@@ -637,7 +637,8 @@
             }
           ]
         ]
-      }
+      },
+      "annotations": []
     }
   ],
   "key_value_items": [],

tests/data/groundtruth/docling_v2/example_04.html.json

@@ -1,6 +1,6 @@
 {
   "schema_name": "DoclingDocument",
-  "version": "1.3.0",
+  "version": "1.4.0",
   "name": "example_04",
   "origin": {
     "mimetype": "text/html",
@@ -325,7 +325,8 @@
             }
           ]
         ]
-      }
+      },
+      "annotations": []
     }
   ],
   "key_value_items": [],

tests/data/groundtruth/docling_v2/example_05.html.json

@@ -1,6 +1,6 @@
 {
   "schema_name": "DoclingDocument",
-  "version": "1.3.0",
+  "version": "1.4.0",
   "name": "example_05",
   "origin": {
     "mimetype": "text/html",
@@ -325,7 +325,8 @@
             }
           ]
         ]
-      }
+      },
+      "annotations": []
     }
   ],
   "key_value_items": [],

tests/data/groundtruth/docling_v2/example_06.html.json

@@ -1,6 +1,6 @@
 {
   "schema_name": "DoclingDocument",
-  "version": "1.3.0",
+  "version": "1.4.0",
   "name": "example_06",
   "origin": {
     "mimetype": "text/html",

tests/data/groundtruth/docling_v2/example_07.html.json

@@ -1,6 +1,6 @@
 {
   "schema_name": "DoclingDocument",
-  "version": "1.3.0",
+  "version": "1.4.0",
   "name": "example_07",
   "origin": {
     "mimetype": "text/html",

tests/data/groundtruth/docling_v2/example_08.html.json

@@ -1,6 +1,6 @@
 {
   "schema_name": "DoclingDocument",
-  "version": "1.3.0",
+  "version": "1.4.0",
   "name": "example_08",
   "origin": {
     "mimetype": "text/html",
@@ -661,7 +661,8 @@
             }
           ]
         ]
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/1",
@@ -1330,7 +1331,8 @@
             }
           ]
         ]
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/2",
@@ -1999,7 +2001,8 @@
             }
           ]
         ]
-      }
+      },
+      "annotations": []
     }
   ],
   "key_value_items": [],

tests/data/groundtruth/docling_v2/inline_and_formatting.md.md

@@ -11,10 +11,12 @@ Create your feature branch: `git checkout -b feature/AmazingFeature` .
 3. Commit your changes ( `git commit -m 'Add some AmazingFeature'` )
 4. Push to the branch ( `git push origin feature/AmazingFeature` )
 5. Open a Pull Request
+6. **Whole list item has same formatting**
+7. List item has *mixed or partial* formatting
 
-## 
+*# Whole heading is italic*
 
-*Second* section
+<<<<<<< HEAD
 
 - **First** : Lorem ipsum.
 - **Second** : Dolor `sit` amet.
@@ -22,3 +24,13 @@ Create your feature branch: `git checkout -b feature/AmazingFeature` .
 | Bold Heading   | Italic Heading   |
 |----------------|------------------|
 | data a         | data b           |
+
+Some *`formatted_code`*
+
+## 
+
+*Partially formatted* heading to\_escape `not_to_escape`
+
+[$$E=mc^2$$](https://en.wikipedia.org/wiki/Albert_Einstein)
+
+origin/main

tests/data/groundtruth/docling_v2/inline_and_formatting.md.yaml

@@ -5,10 +5,14 @@ body:
   - $ref: '#/groups/0'
   - $ref: '#/groups/1'
   - $ref: '#/groups/2'
-  - $ref: '#/texts/27'
+  - $ref: '#/texts/32'
+  - $ref: '#/texts/33'
   - $ref: '#/groups/8'
-  - $ref: '#/groups/11'
   - $ref: '#/tables/0'
+  - $ref: '#/groups/11'
+  - $ref: '#/texts/44'
+  - $ref: '#/texts/48'
+  - $ref: '#/texts/49'
   content_layer: body
   label: unspecified
   name: _root_
@@ -49,6 +53,8 @@ groups:
   - $ref: '#/texts/18'
   - $ref: '#/texts/22'
   - $ref: '#/texts/26'
+  - $ref: '#/texts/27'
+  - $ref: '#/texts/28'
   content_layer: body
   label: ordered_list
   name: list
@@ -96,17 +102,18 @@ groups:
     $ref: '#/texts/22'
   self_ref: '#/groups/6'
 - children:
-  - $ref: '#/texts/28'
   - $ref: '#/texts/29'
+  - $ref: '#/texts/30'
+  - $ref: '#/texts/31'
   content_layer: body
   label: inline
   name: group
   parent:
-    $ref: '#/texts/27'
+    $ref: '#/texts/28'
   self_ref: '#/groups/7'
 - children:
-  - $ref: '#/texts/30'
-  - $ref: '#/texts/33'
+  - $ref: '#/texts/34'
+  - $ref: '#/texts/37'
   content_layer: body
   label: list
   name: list
@@ -114,36 +121,48 @@ groups:
     $ref: '#/body'
   self_ref: '#/groups/8'
 - children:
-  - $ref: '#/texts/31'
-  - $ref: '#/texts/32'
-  content_layer: body
-  label: inline
-  name: group
-  parent:
-    $ref: '#/texts/30'
-  self_ref: '#/groups/9'
-- children:
-  - $ref: '#/texts/34'
   - $ref: '#/texts/35'
   - $ref: '#/texts/36'
-  - $ref: '#/texts/37'
   content_layer: body
   label: inline
   name: group
   parent:
-    $ref: '#/texts/33'
+    $ref: '#/texts/34'
+  self_ref: '#/groups/9'
+- children:
+  - $ref: '#/texts/38'
+  - $ref: '#/texts/39'
+  - $ref: '#/texts/40'
+  - $ref: '#/texts/41'
+  content_layer: body
+  label: inline
+  name: group
+  parent:
+    $ref: '#/texts/37'
   self_ref: '#/groups/10'
-- children: []
+- children:
+  - $ref: '#/texts/42'
+  - $ref: '#/texts/43'
   content_layer: body
   label: inline
   name: group
   parent:
     $ref: '#/body'
   self_ref: '#/groups/11'
+- children:
+  - $ref: '#/texts/45'
+  - $ref: '#/texts/46'
+  - $ref: '#/texts/47'
+  content_layer: body
+  label: inline
+  name: group
+  parent:
+    $ref: '#/texts/44'
+  self_ref: '#/groups/12'
 key_value_items: []
 name: inline_and_formatting
 origin:
-  binary_hash: 15980020574215496313
+  binary_hash: 1036526097556828366
   filename: inline_and_formatting.md
   mimetype: text/markdown
 pages: {}
@@ -613,18 +632,47 @@ texts:
   self_ref: '#/texts/26'
   text: Open a Pull Request
   word_items_ids: []
+- children: []
+  content_layer: body
+  enumerated: true
+  formatting:
+    bold: true
+    italic: false
+    script: baseline
+    strikethrough: false
+    underline: false
+  label: list_item
+  marker: '-'
+  orig: Whole list item has same formatting
+  parent:
+    $ref: '#/groups/2'
+  prov: []
+  self_ref: '#/texts/27'
+  text: Whole list item has same formatting
+  word_items_ids: []
 - children:
   - $ref: '#/groups/7'
   content_layer: body
-  label: section_header
-  level: 1
+  enumerated: true
+  label: list_item
+  marker: '-'
   orig: ''
   parent:
-    $ref: '#/body'
+    $ref: '#/groups/2'
   prov: []
-  self_ref: '#/texts/27'
+  self_ref: '#/texts/28'
   text: ''
   word_items_ids: []
+- children: []
+  content_layer: body
+  label: text
+  orig: List item has
+  parent:
+    $ref: '#/groups/7'
+  prov: []
+  self_ref: '#/texts/29'
+  text: List item has
+  word_items_ids: []
 - children: []
   content_layer: body
   formatting:
@@ -634,22 +682,48 @@ texts:
     strikethrough: false
     underline: false
   label: text
-  orig: Second
+  orig: mixed or partial
   parent:
     $ref: '#/groups/7'
   prov: []
-  self_ref: '#/texts/28'
-  text: Second
+  self_ref: '#/texts/30'
+  text: mixed or partial
   word_items_ids: []
 - children: []
   content_layer: body
   label: text
-  orig: section
+  orig: formatting
   parent:
     $ref: '#/groups/7'
   prov: []
-  self_ref: '#/texts/29'
-  text: section
+  self_ref: '#/texts/31'
+  text: formatting
+  word_items_ids: []
+- children: []
+  content_layer: body
+  formatting:
+    bold: false
+    italic: true
+    script: baseline
+    strikethrough: false
+    underline: false
+  label: title
+  orig: Whole heading is italic
+  parent:
+    $ref: '#/body'
+  prov: []
+  self_ref: '#/texts/32'
+  text: Whole heading is italic
+  word_items_ids: []
+- children: []
+  content_layer: body
+  label: text
+  orig: <<<<<<< HEAD
+  parent:
+    $ref: '#/body'
+  prov: []
+  self_ref: '#/texts/33'
+  text: <<<<<<< HEAD
   word_items_ids: []
 - children:
   - $ref: '#/groups/9'
@@ -661,7 +735,7 @@ texts:
   parent:
     $ref: '#/groups/8'
   prov: []
-  self_ref: '#/texts/30'
+  self_ref: '#/texts/34'
   text: ''
   word_items_ids: []
 - children: []
@@ -677,7 +751,7 @@ texts:
   parent:
     $ref: '#/groups/9'
   prov: []
-  self_ref: '#/texts/31'
+  self_ref: '#/texts/35'
   text: First
   word_items_ids: []
 - children: []
@@ -687,7 +761,7 @@ texts:
   parent:
     $ref: '#/groups/9'
   prov: []
-  self_ref: '#/texts/32'
+  self_ref: '#/texts/36'
   text: ': Lorem ipsum.'
   word_items_ids: []
 - children:
@@ -700,7 +774,7 @@ texts:
   parent:
     $ref: '#/groups/8'
   prov: []
-  self_ref: '#/texts/33'
+  self_ref: '#/texts/37'
   text: ''
   word_items_ids: []
 - children: []
@@ -716,7 +790,7 @@ texts:
   parent:
     $ref: '#/groups/10'
   prov: []
-  self_ref: '#/texts/34'
+  self_ref: '#/texts/38'
   text: Second
   word_items_ids: []
 - children: []
@@ -726,7 +800,7 @@ texts:
   parent:
     $ref: '#/groups/10'
   prov: []
-  self_ref: '#/texts/35'
+  self_ref: '#/texts/39'
   text: ': Dolor'
   word_items_ids: []
 - captions: []
@@ -740,7 +814,7 @@ texts:
     $ref: '#/groups/10'
   prov: []
   references: []
-  self_ref: '#/texts/36'
+  self_ref: '#/texts/40'
   text: sit
   word_items_ids: []
 - children: []
@@ -750,7 +824,110 @@ texts:
   parent:
     $ref: '#/groups/10'
   prov: []
-  self_ref: '#/texts/37'
+  self_ref: '#/texts/41'
   text: amet.
   word_items_ids: []
+- children: []
+  content_layer: body
+  label: text
+  orig: Some
+  parent:
+    $ref: '#/groups/11'
+  prov: []
+  self_ref: '#/texts/42'
+  text: Some
+  word_items_ids: []
+- captions: []
+  children: []
+  code_language: unknown
+  content_layer: body
+  footnotes: []
+  formatting:
+    bold: false
+    italic: true
+    script: baseline
+    strikethrough: false
+    underline: false
+  label: code
+  orig: formatted_code
+  parent:
+    $ref: '#/groups/11'
+  prov: []
+  references: []
+  self_ref: '#/texts/43'
+  text: formatted_code
+  word_items_ids: []
+- children:
+  - $ref: '#/groups/12'
+  content_layer: body
+  label: section_header
+  level: 1
+  orig: ''
+  parent:
+    $ref: '#/body'
+  prov: []
+  self_ref: '#/texts/44'
+  text: ''
+  word_items_ids: []
+- children: []
+  content_layer: body
+  formatting:
+    bold: false
+    italic: true
+    script: baseline
+    strikethrough: false
+    underline: false
+  label: text
+  orig: Partially formatted
+  parent:
+    $ref: '#/groups/12'
+  prov: []
+  self_ref: '#/texts/45'
+  text: Partially formatted
+  word_items_ids: []
+- children: []
+  content_layer: body
+  label: text
+  orig: heading to_escape
+  parent:
+    $ref: '#/groups/12'
+  prov: []
+  self_ref: '#/texts/46'
+  text: heading to_escape
+  word_items_ids: []
+- captions: []
+  children: []
+  code_language: unknown
+  content_layer: body
+  footnotes: []
+  label: code
+  orig: not_to_escape
+  parent:
+    $ref: '#/groups/12'
+  prov: []
+  references: []
+  self_ref: '#/texts/47'
+  text: not_to_escape
+  word_items_ids: []
+- children: []
+  content_layer: body
+  hyperlink: https://en.wikipedia.org/wiki/Albert_Einstein
+  label: text
+  orig: $$E=mc^2$$
+  parent:
+    $ref: '#/body'
+  prov: []
+  self_ref: '#/texts/48'
+  text: $$E=mc^2$$
+  word_items_ids: []
+- children: []
+  content_layer: body
+  label: text
+  orig: origin/main
+  parent:
+    $ref: '#/body'
+  prov: []
+  self_ref: '#/texts/49'
+  text: origin/main
+  word_items_ids: []
 version: 1.4.0

tests/data/groundtruth/docling_v2/ipa20180000016.json

@@ -1,6 +1,6 @@
 {
   "schema_name": "DoclingDocument",
-  "version": "1.3.0",
+  "version": "1.4.0",
   "name": "ipa20180000016.xml",
   "origin": {
     "mimetype": "application/xml",
@@ -6005,7 +6005,8 @@
             }
           ]
         ]
-      }
+      },
+      "annotations": []
     }
   ],
   "key_value_items": [],

tests/data/groundtruth/docling_v2/ipa20200022300.json

@@ -1,6 +1,6 @@
 {
   "schema_name": "DoclingDocument",
-  "version": "1.3.0",
+  "version": "1.4.0",
   "name": "ipa20200022300.xml",
   "origin": {
     "mimetype": "application/xml",

tests/data/groundtruth/docling_v2/lorem_ipsum.docx.json

@@ -1,6 +1,6 @@
 {
   "schema_name": "DoclingDocument",
-  "version": "1.3.0",
+  "version": "1.4.0",
   "name": "lorem_ipsum",
   "origin": {
     "mimetype": "application/vnd.openxmlformats-officedocument.wordprocessingml.document",
@@ -66,7 +66,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -96,7 +97,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -126,7 +128,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -156,7 +159,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -186,7 +190,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     }
   ],

tests/data/groundtruth/docling_v2/multi_page.json

@@ -1,6 +1,6 @@
 {
   "schema_name": "DoclingDocument",
-  "version": "1.3.0",
+  "version": "1.4.0",
   "name": "multi_page",
   "origin": {
     "mimetype": "application/pdf",

tests/data/groundtruth/docling_v2/pa20010031492.json

@@ -1,6 +1,6 @@
 {
   "schema_name": "DoclingDocument",
-  "version": "1.3.0",
+  "version": "1.4.0",
   "name": "pa20010031492.xml",
   "origin": {
     "mimetype": "application/xml",
@@ -2127,7 +2127,8 @@
             }
           ]
         ]
-      }
+      },
+      "annotations": []
     }
   ],
   "key_value_items": [],

tests/data/groundtruth/docling_v2/pftaps057006474.json

@@ -1,6 +1,6 @@
 {
   "schema_name": "DoclingDocument",
-  "version": "1.3.0",
+  "version": "1.4.0",
   "name": "pftaps057006474.txt",
   "origin": {
     "mimetype": "text/plain",

tests/data/groundtruth/docling_v2/pg06442728.json

@@ -1,6 +1,6 @@
 {
   "schema_name": "DoclingDocument",
-  "version": "1.3.0",
+  "version": "1.4.0",
   "name": "pg06442728.xml",
   "origin": {
     "mimetype": "application/xml",

tests/data/groundtruth/docling_v2/picture_classification.json

@@ -1,6 +1,6 @@
 {
   "schema_name": "DoclingDocument",
-  "version": "1.3.0",
+  "version": "1.4.0",
   "name": "picture_classification",
   "origin": {
     "mimetype": "application/pdf",

tests/data/groundtruth/docling_v2/powerpoint_bad_text.pptx.json

@@ -1,6 +1,6 @@
 {
   "schema_name": "DoclingDocument",
-  "version": "1.3.0",
+  "version": "1.4.0",
   "name": "powerpoint_bad_text",
   "origin": {
     "mimetype": "application/vnd.ms-powerpoint",

tests/data/groundtruth/docling_v2/powerpoint_sample.pptx.json

@@ -1,6 +1,6 @@
 {
   "schema_name": "DoclingDocument",
-  "version": "1.3.0",
+  "version": "1.4.0",
   "name": "powerpoint_sample",
   "origin": {
     "mimetype": "application/vnd.ms-powerpoint",
@@ -2199,7 +2199,8 @@
             }
           ]
         ]
-      }
+      },
+      "annotations": []
     }
   ],
   "key_value_items": [],

tests/data/groundtruth/docling_v2/powerpoint_with_image.pptx.json

@@ -1,6 +1,6 @@
 {
   "schema_name": "DoclingDocument",
-  "version": "1.3.0",
+  "version": "1.4.0",
   "name": "powerpoint_with_image",
   "origin": {
     "mimetype": "application/vnd.ms-powerpoint",

tests/data/groundtruth/docling_v2/redp5110_sampled.json

@@ -1,6 +1,6 @@
 {
   "schema_name": "DoclingDocument",
-  "version": "1.3.0",
+  "version": "1.4.0",
   "name": "redp5110_sampled",
   "origin": {
     "mimetype": "application/pdf",
@@ -12471,7 +12471,8 @@
             }
           ]
         ]
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/1",
@@ -13096,7 +13097,8 @@
             }
           ]
         ]
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/2",
@@ -15356,7 +15358,8 @@
             }
           ]
         ]
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/3",
@@ -15713,7 +15716,8 @@
             }
           ]
         ]
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/4",
@@ -16918,7 +16922,8 @@
             }
           ]
         ]
-      }
+      },
+      "annotations": []
     }
   ],
   "key_value_items": [],

tests/data/groundtruth/docling_v2/right_to_left_01.json

@@ -1,6 +1,6 @@
 {
   "schema_name": "DoclingDocument",
-  "version": "1.3.0",
+  "version": "1.4.0",
   "name": "right_to_left_01",
   "origin": {
     "mimetype": "application/pdf",

tests/data/groundtruth/docling_v2/right_to_left_02.json

@@ -1,6 +1,6 @@
 {
   "schema_name": "DoclingDocument",
-  "version": "1.3.0",
+  "version": "1.4.0",
   "name": "right_to_left_02",
   "origin": {
     "mimetype": "application/pdf",

tests/data/groundtruth/docling_v2/right_to_left_03.json

@@ -1,6 +1,6 @@
 {
   "schema_name": "DoclingDocument",
-  "version": "1.3.0",
+  "version": "1.4.0",
   "name": "right_to_left_03",
   "origin": {
     "mimetype": "application/pdf",

tests/data/groundtruth/docling_v2/sample_sales_data.xlsm.json

@@ -1,6 +1,6 @@
 {
   "schema_name": "DoclingDocument",
-  "version": "1.3.0",
+  "version": "1.4.0",
   "name": "sample_sales_data",
   "origin": {
     "mimetype": "application/vnd.openxmlformats-officedocument.spreadsheetml.sheet",
@@ -2136,7 +2136,8 @@
             }
           ]
         ]
-      }
+      },
+      "annotations": []
     }
   ],
   "key_value_items": [],

tests/data/groundtruth/docling_v2/tablecell.docx.json

@@ -1,6 +1,6 @@
 {
   "schema_name": "DoclingDocument",
-  "version": "1.3.0",
+  "version": "1.4.0",
   "name": "tablecell",
   "origin": {
     "mimetype": "application/vnd.openxmlformats-officedocument.wordprocessingml.document",
@@ -78,7 +78,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       },
       "enumerated": false,
       "marker": "-"
@@ -98,7 +99,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       },
       "enumerated": false,
       "marker": "-"
@@ -130,7 +132,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -172,7 +175,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     }
   ],
@@ -419,7 +423,8 @@
             }
           ]
         ]
-      }
+      },
+      "annotations": []
     }
   ],
   "key_value_items": [],

tests/data/groundtruth/docling_v2/test-01.xlsx.json

@@ -1,6 +1,6 @@
 {
   "schema_name": "DoclingDocument",
-  "version": "1.3.0",
+  "version": "1.4.0",
   "name": "test-01",
   "origin": {
     "mimetype": "application/vnd.openxmlformats-officedocument.spreadsheetml.sheet",
@@ -681,7 +681,8 @@
             }
           ]
         ]
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/1",
@@ -1599,7 +1600,8 @@
             }
           ]
         ]
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/2",
@@ -2005,7 +2007,8 @@
             }
           ]
         ]
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/3",
@@ -2411,7 +2414,8 @@
             }
           ]
         ]
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/4",
@@ -2893,7 +2897,8 @@
             }
           ]
         ]
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/5",
@@ -3375,7 +3380,8 @@
             }
           ]
         ]
-      }
+      },
+      "annotations": []
     }
   ],
   "key_value_items": [],

tests/data/groundtruth/docling_v2/test_emf_docx.docx.json

@@ -1,6 +1,6 @@
 {
   "schema_name": "DoclingDocument",
-  "version": "1.3.0",
+  "version": "1.4.0",
   "name": "test_emf_docx",
   "origin": {
     "mimetype": "application/vnd.openxmlformats-officedocument.wordprocessingml.document",
@@ -60,7 +60,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -78,7 +79,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -96,7 +98,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -114,7 +117,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     }
   ],

tests/data/groundtruth/docling_v2/textbox.docx.itxt

@@ -11,83 +11,82 @@ item-0 at level 0: unspecified: group _root_
 ＊ Blisters
 ＊ Headache
 ＊ Sore throat
-  item-9 at level 1: list_item: 
+  item-9 at level 1: paragraph: 
   item-10 at level 1: paragraph: 
-  item-11 at level 1: paragraph: 
-  item-12 at level 1: section: group textbox
-    item-13 at level 2: paragraph: If a caregiver suspects that wit ... the same suggested reportable symptoms
+  item-11 at level 1: section: group textbox
+    item-12 at level 2: paragraph: If a caregiver suspects that wit ... the same suggested reportable symptoms
+  item-13 at level 1: paragraph: 
   item-14 at level 1: paragraph: 
   item-15 at level 1: paragraph: 
   item-16 at level 1: paragraph: 
-  item-17 at level 1: paragraph: 
-  item-18 at level 1: section: group textbox
-    item-19 at level 2: paragraph: Yes
+  item-17 at level 1: section: group textbox
+    item-18 at level 2: paragraph: Yes
+  item-19 at level 1: paragraph: 
   item-20 at level 1: paragraph: 
-  item-21 at level 1: paragraph: 
-  item-22 at level 1: section: group textbox
-    item-23 at level 2: list: group list
-      item-24 at level 3: list_item: A report must be submitted withi ... saster Prevention Information Network.
-      item-25 at level 3: list_item: A report must also be submitted  ... d Infectious Disease Reporting System.
-    item-26 at level 2: paragraph: 
-  item-27 at level 1: list: group list
-    item-28 at level 2: list_item: 
+  item-21 at level 1: section: group textbox
+    item-22 at level 2: list: group list
+      item-23 at level 3: list_item: A report must be submitted withi ... saster Prevention Information Network.
+      item-24 at level 3: list_item: A report must also be submitted  ... d Infectious Disease Reporting System.
+    item-25 at level 2: paragraph: 
+  item-26 at level 1: list: group list
+    item-27 at level 2: list_item: 
+  item-28 at level 1: paragraph: 
   item-29 at level 1: paragraph: 
   item-30 at level 1: paragraph: 
   item-31 at level 1: paragraph: 
   item-32 at level 1: paragraph: 
-  item-33 at level 1: paragraph: 
-  item-34 at level 1: section: group textbox
-    item-35 at level 2: paragraph: Health Bureau:
-    item-36 at level 2: paragraph: Upon receiving a report from the ... rt to the Centers for Disease Control.
-    item-37 at level 2: list: group list
-      item-38 at level 3: list_item: If necessary, provide health edu ... vidual to undergo specimen collection.
-      item-39 at level 3: list_item: Implement appropriate epidemic p ...  the Communicable Disease Control Act.
-    item-40 at level 2: paragraph: 
-  item-41 at level 1: list: group list
-    item-42 at level 2: list_item: 
-  item-43 at level 1: paragraph: 
-  item-44 at level 1: section: group textbox
-    item-45 at level 2: paragraph: Department of Education:
+  item-33 at level 1: section: group textbox
+    item-34 at level 2: paragraph: Health Bureau:
+    item-35 at level 2: paragraph: Upon receiving a report from the ... rt to the Centers for Disease Control.
+    item-36 at level 2: list: group list
+      item-37 at level 3: list_item: If necessary, provide health edu ... vidual to undergo specimen collection.
+      item-38 at level 3: list_item: Implement appropriate epidemic p ...  the Communicable Disease Control Act.
+    item-39 at level 2: paragraph: 
+  item-40 at level 1: list: group list
+    item-41 at level 2: list_item: 
+  item-42 at level 1: paragraph: 
+  item-43 at level 1: section: group textbox
+    item-44 at level 2: paragraph: Department of Education:
 Collabo ... vention measures at all school levels.
+  item-45 at level 1: paragraph: 
   item-46 at level 1: paragraph: 
   item-47 at level 1: paragraph: 
   item-48 at level 1: paragraph: 
   item-49 at level 1: paragraph: 
   item-50 at level 1: paragraph: 
   item-51 at level 1: paragraph: 
-  item-52 at level 1: paragraph: 
-  item-53 at level 1: section: group textbox
-    item-54 at level 2: inline: group group
-      item-55 at level 3: paragraph: The Health Bureau will handle
-      item-56 at level 3: paragraph: reporting and specimen collection
-      item-57 at level 3: paragraph: .
-    item-58 at level 2: paragraph: 
+  item-52 at level 1: section: group textbox
+    item-53 at level 2: inline: group group
+      item-54 at level 3: paragraph: The Health Bureau will handle
+      item-55 at level 3: paragraph: reporting and specimen collection
+      item-56 at level 3: paragraph: .
+    item-57 at level 2: paragraph: 
+  item-58 at level 1: paragraph: 
   item-59 at level 1: paragraph: 
   item-60 at level 1: paragraph: 
-  item-61 at level 1: paragraph: 
-  item-62 at level 1: section: group textbox
-    item-63 at level 2: paragraph: Whether the epidemic has eased.
-    item-64 at level 2: paragraph: 
-  item-65 at level 1: paragraph: 
-  item-66 at level 1: section: group textbox
-    item-67 at level 2: paragraph: Whether the test results are pos ... legally designated infectious disease.
-    item-68 at level 2: paragraph: No
+  item-61 at level 1: section: group textbox
+    item-62 at level 2: paragraph: Whether the epidemic has eased.
+    item-63 at level 2: paragraph: 
+  item-64 at level 1: paragraph: 
+  item-65 at level 1: section: group textbox
+    item-66 at level 2: paragraph: Whether the test results are pos ... legally designated infectious disease.
+    item-67 at level 2: paragraph: No
+  item-68 at level 1: paragraph: 
   item-69 at level 1: paragraph: 
-  item-70 at level 1: paragraph: 
-  item-71 at level 1: section: group textbox
-    item-72 at level 2: paragraph: Yes
-  item-73 at level 1: paragraph: 
-  item-74 at level 1: section: group textbox
-    item-75 at level 2: paragraph: Yes
+  item-70 at level 1: section: group textbox
+    item-71 at level 2: paragraph: Yes
+  item-72 at level 1: paragraph: 
+  item-73 at level 1: section: group textbox
+    item-74 at level 2: paragraph: Yes
+  item-75 at level 1: paragraph: 
   item-76 at level 1: paragraph: 
-  item-77 at level 1: paragraph: 
-  item-78 at level 1: section: group textbox
-    item-79 at level 2: paragraph: Case closed.
-    item-80 at level 2: paragraph: 
-    item-81 at level 2: paragraph: The Health Bureau will carry out ... ters for Disease Control if necessary.
-  item-82 at level 1: paragraph: 
-  item-83 at level 1: section: group textbox
-    item-84 at level 2: paragraph: No
+  item-77 at level 1: section: group textbox
+    item-78 at level 2: paragraph: Case closed.
+    item-79 at level 2: paragraph: 
+    item-80 at level 2: paragraph: The Health Bureau will carry out ... ters for Disease Control if necessary.
+  item-81 at level 1: paragraph: 
+  item-82 at level 1: section: group textbox
+    item-83 at level 2: paragraph: No
+  item-84 at level 1: paragraph: 
   item-85 at level 1: paragraph: 
   item-86 at level 1: paragraph: 
-  item-87 at level 1: paragraph: 

tests/data/groundtruth/docling_v2/textbox.docx.json

@@ -1,6 +1,6 @@
 {
   "schema_name": "DoclingDocument",
-  "version": "1.3.0",
+  "version": "1.4.0",
   "name": "textbox",
   "origin": {
     "mimetype": "application/vnd.openxmlformats-officedocument.wordprocessingml.document",
@@ -36,10 +36,10 @@
         "$ref": "#/texts/7"
       },
       {
-        "$ref": "#/texts/8"
+        "$ref": "#/groups/2"
       },
       {
-        "$ref": "#/groups/2"
+        "$ref": "#/texts/9"
       },
       {
         "$ref": "#/texts/10"
@@ -50,17 +50,14 @@
       {
         "$ref": "#/texts/12"
       },
-      {
-        "$ref": "#/texts/13"
-      },
       {
         "$ref": "#/groups/3"
       },
       {
-        "$ref": "#/texts/15"
+        "$ref": "#/texts/14"
       },
       {
-        "$ref": "#/texts/16"
+        "$ref": "#/texts/15"
       },
       {
         "$ref": "#/groups/4"
@@ -68,6 +65,9 @@
       {
         "$ref": "#/groups/6"
       },
+      {
+        "$ref": "#/texts/20"
+      },
       {
         "$ref": "#/texts/21"
       },
@@ -80,9 +80,6 @@
       {
         "$ref": "#/texts/24"
       },
-      {
-        "$ref": "#/texts/25"
-      },
       {
         "$ref": "#/groups/7"
       },
@@ -90,11 +87,14 @@
         "$ref": "#/groups/9"
       },
       {
-        "$ref": "#/texts/32"
+        "$ref": "#/texts/31"
       },
       {
         "$ref": "#/groups/10"
       },
+      {
+        "$ref": "#/texts/33"
+      },
       {
         "$ref": "#/texts/34"
       },
@@ -114,10 +114,10 @@
         "$ref": "#/texts/39"
       },
       {
-        "$ref": "#/texts/40"
+        "$ref": "#/groups/11"
       },
       {
-        "$ref": "#/groups/11"
+        "$ref": "#/texts/44"
       },
       {
         "$ref": "#/texts/45"
@@ -125,56 +125,53 @@
       {
         "$ref": "#/texts/46"
       },
-      {
-        "$ref": "#/texts/47"
-      },
       {
         "$ref": "#/groups/13"
       },
       {
-        "$ref": "#/texts/50"
+        "$ref": "#/texts/49"
       },
       {
         "$ref": "#/groups/14"
       },
       {
-        "$ref": "#/texts/53"
+        "$ref": "#/texts/52"
       },
       {
-        "$ref": "#/texts/54"
+        "$ref": "#/texts/53"
       },
       {
         "$ref": "#/groups/15"
       },
       {
-        "$ref": "#/texts/56"
+        "$ref": "#/texts/55"
       },
       {
         "$ref": "#/groups/16"
       },
       {
-        "$ref": "#/texts/58"
+        "$ref": "#/texts/57"
       },
       {
-        "$ref": "#/texts/59"
+        "$ref": "#/texts/58"
       },
       {
         "$ref": "#/groups/17"
       },
       {
-        "$ref": "#/texts/63"
+        "$ref": "#/texts/62"
       },
       {
         "$ref": "#/groups/18"
       },
+      {
+        "$ref": "#/texts/64"
+      },
       {
         "$ref": "#/texts/65"
       },
       {
         "$ref": "#/texts/66"
-      },
-      {
-        "$ref": "#/texts/67"
       }
     ],
     "content_layer": "body",
@@ -223,7 +220,7 @@
       },
       "children": [
         {
-          "$ref": "#/texts/9"
+          "$ref": "#/texts/8"
         }
       ],
       "content_layer": "body",
@@ -237,7 +234,7 @@
       },
       "children": [
         {
-          "$ref": "#/texts/14"
+          "$ref": "#/texts/13"
         }
       ],
       "content_layer": "body",
@@ -254,7 +251,7 @@
           "$ref": "#/groups/5"
         },
         {
-          "$ref": "#/texts/19"
+          "$ref": "#/texts/18"
         }
       ],
       "content_layer": "body",
@@ -268,10 +265,10 @@
       },
       "children": [
         {
-          "$ref": "#/texts/17"
+          "$ref": "#/texts/16"
         },
         {
-          "$ref": "#/texts/18"
+          "$ref": "#/texts/17"
         }
       ],
       "content_layer": "body",
@@ -285,7 +282,7 @@
       },
       "children": [
         {
-          "$ref": "#/texts/20"
+          "$ref": "#/texts/19"
         }
       ],
       "content_layer": "body",
@@ -299,16 +296,16 @@
       },
       "children": [
         {
-          "$ref": "#/texts/26"
+          "$ref": "#/texts/25"
         },
         {
-          "$ref": "#/texts/27"
+          "$ref": "#/texts/26"
         },
         {
           "$ref": "#/groups/8"
         },
         {
-          "$ref": "#/texts/30"
+          "$ref": "#/texts/29"
         }
       ],
       "content_layer": "body",
@@ -322,10 +319,10 @@
       },
       "children": [
         {
-          "$ref": "#/texts/28"
+          "$ref": "#/texts/27"
         },
         {
-          "$ref": "#/texts/29"
+          "$ref": "#/texts/28"
         }
       ],
       "content_layer": "body",
@@ -339,7 +336,7 @@
       },
       "children": [
         {
-          "$ref": "#/texts/31"
+          "$ref": "#/texts/30"
         }
       ],
       "content_layer": "body",
@@ -353,7 +350,7 @@
       },
       "children": [
         {
-          "$ref": "#/texts/33"
+          "$ref": "#/texts/32"
         }
       ],
       "content_layer": "body",
@@ -370,7 +367,7 @@
           "$ref": "#/groups/12"
         },
         {
-          "$ref": "#/texts/44"
+          "$ref": "#/texts/43"
         }
       ],
       "content_layer": "body",
@@ -383,14 +380,14 @@
         "$ref": "#/groups/11"
       },
       "children": [
+        {
+          "$ref": "#/texts/40"
+        },
         {
           "$ref": "#/texts/41"
         },
         {
           "$ref": "#/texts/42"
-        },
-        {
-          "$ref": "#/texts/43"
         }
       ],
       "content_layer": "body",
@@ -404,10 +401,10 @@
       },
       "children": [
         {
-          "$ref": "#/texts/48"
+          "$ref": "#/texts/47"
         },
         {
-          "$ref": "#/texts/49"
+          "$ref": "#/texts/48"
         }
       ],
       "content_layer": "body",
@@ -421,10 +418,10 @@
       },
       "children": [
         {
-          "$ref": "#/texts/51"
+          "$ref": "#/texts/50"
         },
         {
-          "$ref": "#/texts/52"
+          "$ref": "#/texts/51"
         }
       ],
       "content_layer": "body",
@@ -438,7 +435,7 @@
       },
       "children": [
         {
-          "$ref": "#/texts/55"
+          "$ref": "#/texts/54"
         }
       ],
       "content_layer": "body",
@@ -452,7 +449,7 @@
       },
       "children": [
         {
-          "$ref": "#/texts/57"
+          "$ref": "#/texts/56"
         }
       ],
       "content_layer": "body",
@@ -465,14 +462,14 @@
         "$ref": "#/body"
       },
       "children": [
+        {
+          "$ref": "#/texts/59"
+        },
         {
           "$ref": "#/texts/60"
         },
         {
           "$ref": "#/texts/61"
-        },
-        {
-          "$ref": "#/texts/62"
         }
       ],
       "content_layer": "body",
@@ -486,7 +483,7 @@
       },
       "children": [
         {
-          "$ref": "#/texts/64"
+          "$ref": "#/texts/63"
         }
       ],
       "content_layer": "body",
@@ -510,7 +507,8 @@
         "bold": true,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -528,7 +526,8 @@
         "bold": true,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -558,7 +557,8 @@
         "bold": true,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -588,7 +588,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       },
       "enumerated": false,
       "marker": "-"
@@ -600,12 +601,10 @@
       },
       "children": [],
       "content_layer": "body",
-      "label": "list_item",
+      "label": "paragraph",
       "prov": [],
       "orig": "",
-      "text": "",
-      "enumerated": false,
-      "marker": "-"
+      "text": ""
     },
     {
       "self_ref": "#/texts/7",
@@ -621,18 +620,6 @@
     },
     {
       "self_ref": "#/texts/8",
-      "parent": {
-        "$ref": "#/body"
-      },
-      "children": [],
-      "content_layer": "body",
-      "label": "paragraph",
-      "prov": [],
-      "orig": "",
-      "text": ""
-    },
-    {
-      "self_ref": "#/texts/9",
       "parent": {
         "$ref": "#/groups/2"
       },
@@ -646,9 +633,22 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
+    {
+      "self_ref": "#/texts/9",
+      "parent": {
+        "$ref": "#/body"
+      },
+      "children": [],
+      "content_layer": "body",
+      "label": "paragraph",
+      "prov": [],
+      "orig": "",
+      "text": ""
+    },
     {
       "self_ref": "#/texts/10",
       "parent": {
@@ -687,18 +687,6 @@
     },
     {
       "self_ref": "#/texts/13",
-      "parent": {
-        "$ref": "#/body"
-      },
-      "children": [],
-      "content_layer": "body",
-      "label": "paragraph",
-      "prov": [],
-      "orig": "",
-      "text": ""
-    },
-    {
-      "self_ref": "#/texts/14",
       "parent": {
         "$ref": "#/groups/3"
       },
@@ -712,9 +700,22 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
+    {
+      "self_ref": "#/texts/14",
+      "parent": {
+        "$ref": "#/body"
+      },
+      "children": [],
+      "content_layer": "body",
+      "label": "paragraph",
+      "prov": [],
+      "orig": "",
+      "text": ""
+    },
     {
       "self_ref": "#/texts/15",
       "parent": {
@@ -729,18 +730,6 @@
     },
     {
       "self_ref": "#/texts/16",
-      "parent": {
-        "$ref": "#/body"
-      },
-      "children": [],
-      "content_layer": "body",
-      "label": "paragraph",
-      "prov": [],
-      "orig": "",
-      "text": ""
-    },
-    {
-      "self_ref": "#/texts/17",
       "parent": {
         "$ref": "#/groups/5"
       },
@@ -754,13 +743,14 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       },
       "enumerated": false,
       "marker": "-"
     },
     {
-      "self_ref": "#/texts/18",
+      "self_ref": "#/texts/17",
       "parent": {
         "$ref": "#/groups/5"
       },
@@ -774,13 +764,14 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       },
       "enumerated": false,
       "marker": "-"
     },
     {
-      "self_ref": "#/texts/19",
+      "self_ref": "#/texts/18",
       "parent": {
         "$ref": "#/groups/4"
       },
@@ -792,7 +783,7 @@
       "text": ""
     },
     {
-      "self_ref": "#/texts/20",
+      "self_ref": "#/texts/19",
       "parent": {
         "$ref": "#/groups/6"
       },
@@ -805,6 +796,18 @@
       "enumerated": false,
       "marker": "-"
     },
+    {
+      "self_ref": "#/texts/20",
+      "parent": {
+        "$ref": "#/body"
+      },
+      "children": [],
+      "content_layer": "body",
+      "label": "paragraph",
+      "prov": [],
+      "orig": "",
+      "text": ""
+    },
     {
       "self_ref": "#/texts/21",
       "parent": {
@@ -855,18 +858,6 @@
     },
     {
       "self_ref": "#/texts/25",
-      "parent": {
-        "$ref": "#/body"
-      },
-      "children": [],
-      "content_layer": "body",
-      "label": "paragraph",
-      "prov": [],
-      "orig": "",
-      "text": ""
-    },
-    {
-      "self_ref": "#/texts/26",
       "parent": {
         "$ref": "#/groups/7"
       },
@@ -880,11 +871,12 @@
         "bold": true,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
-      "self_ref": "#/texts/27",
+      "self_ref": "#/texts/26",
       "parent": {
         "$ref": "#/groups/7"
       },
@@ -898,11 +890,12 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
-      "self_ref": "#/texts/28",
+      "self_ref": "#/texts/27",
       "parent": {
         "$ref": "#/groups/8"
       },
@@ -916,13 +909,14 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       },
       "enumerated": false,
       "marker": "-"
     },
     {
-      "self_ref": "#/texts/29",
+      "self_ref": "#/texts/28",
       "parent": {
         "$ref": "#/groups/8"
       },
@@ -936,13 +930,14 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       },
       "enumerated": false,
       "marker": "-"
     },
     {
-      "self_ref": "#/texts/30",
+      "self_ref": "#/texts/29",
       "parent": {
         "$ref": "#/groups/7"
       },
@@ -954,7 +949,7 @@
       "text": ""
     },
     {
-      "self_ref": "#/texts/31",
+      "self_ref": "#/texts/30",
       "parent": {
         "$ref": "#/groups/9"
       },
@@ -968,7 +963,7 @@
       "marker": "-"
     },
     {
-      "self_ref": "#/texts/32",
+      "self_ref": "#/texts/31",
       "parent": {
         "$ref": "#/body"
       },
@@ -980,7 +975,7 @@
       "text": ""
     },
     {
-      "self_ref": "#/texts/33",
+      "self_ref": "#/texts/32",
       "parent": {
         "$ref": "#/groups/10"
       },
@@ -994,9 +989,22 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
+    {
+      "self_ref": "#/texts/33",
+      "parent": {
+        "$ref": "#/body"
+      },
+      "children": [],
+      "content_layer": "body",
+      "label": "paragraph",
+      "prov": [],
+      "orig": "",
+      "text": ""
+    },
     {
       "self_ref": "#/texts/34",
       "parent": {
@@ -1071,18 +1079,6 @@
     },
     {
       "self_ref": "#/texts/40",
-      "parent": {
-        "$ref": "#/body"
-      },
-      "children": [],
-      "content_layer": "body",
-      "label": "paragraph",
-      "prov": [],
-      "orig": "",
-      "text": ""
-    },
-    {
-      "self_ref": "#/texts/41",
       "parent": {
         "$ref": "#/groups/12"
       },
@@ -1096,11 +1092,12 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
-      "self_ref": "#/texts/42",
+      "self_ref": "#/texts/41",
       "parent": {
         "$ref": "#/groups/12"
       },
@@ -1114,11 +1111,12 @@
         "bold": true,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
-      "self_ref": "#/texts/43",
+      "self_ref": "#/texts/42",
       "parent": {
         "$ref": "#/groups/12"
       },
@@ -1132,13 +1130,26 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
+    {
+      "self_ref": "#/texts/43",
+      "parent": {
+        "$ref": "#/groups/11"
+      },
+      "children": [],
+      "content_layer": "body",
+      "label": "paragraph",
+      "prov": [],
+      "orig": "",
+      "text": ""
+    },
     {
       "self_ref": "#/texts/44",
       "parent": {
-        "$ref": "#/groups/11"
+        "$ref": "#/body"
       },
       "children": [],
       "content_layer": "body",
@@ -1173,18 +1184,6 @@
     },
     {
       "self_ref": "#/texts/47",
-      "parent": {
-        "$ref": "#/body"
-      },
-      "children": [],
-      "content_layer": "body",
-      "label": "paragraph",
-      "prov": [],
-      "orig": "",
-      "text": ""
-    },
-    {
-      "self_ref": "#/texts/48",
       "parent": {
         "$ref": "#/groups/13"
       },
@@ -1198,11 +1197,12 @@
         "bold": true,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
-      "self_ref": "#/texts/49",
+      "self_ref": "#/texts/48",
       "parent": {
         "$ref": "#/groups/13"
       },
@@ -1214,7 +1214,7 @@
       "text": ""
     },
     {
-      "self_ref": "#/texts/50",
+      "self_ref": "#/texts/49",
       "parent": {
         "$ref": "#/body"
       },
@@ -1226,7 +1226,7 @@
       "text": ""
     },
     {
-      "self_ref": "#/texts/51",
+      "self_ref": "#/texts/50",
       "parent": {
         "$ref": "#/groups/14"
       },
@@ -1240,11 +1240,12 @@
         "bold": true,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
-      "self_ref": "#/texts/52",
+      "self_ref": "#/texts/51",
       "parent": {
         "$ref": "#/groups/14"
       },
@@ -1258,9 +1259,22 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
+    {
+      "self_ref": "#/texts/52",
+      "parent": {
+        "$ref": "#/body"
+      },
+      "children": [],
+      "content_layer": "body",
+      "label": "paragraph",
+      "prov": [],
+      "orig": "",
+      "text": ""
+    },
     {
       "self_ref": "#/texts/53",
       "parent": {
@@ -1275,18 +1289,6 @@
     },
     {
       "self_ref": "#/texts/54",
-      "parent": {
-        "$ref": "#/body"
-      },
-      "children": [],
-      "content_layer": "body",
-      "label": "paragraph",
-      "prov": [],
-      "orig": "",
-      "text": ""
-    },
-    {
-      "self_ref": "#/texts/55",
       "parent": {
         "$ref": "#/groups/15"
       },
@@ -1300,11 +1302,12 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
-      "self_ref": "#/texts/56",
+      "self_ref": "#/texts/55",
       "parent": {
         "$ref": "#/body"
       },
@@ -1316,7 +1319,7 @@
       "text": ""
     },
     {
-      "self_ref": "#/texts/57",
+      "self_ref": "#/texts/56",
       "parent": {
         "$ref": "#/groups/16"
       },
@@ -1330,9 +1333,22 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
+    {
+      "self_ref": "#/texts/57",
+      "parent": {
+        "$ref": "#/body"
+      },
+      "children": [],
+      "content_layer": "body",
+      "label": "paragraph",
+      "prov": [],
+      "orig": "",
+      "text": ""
+    },
     {
       "self_ref": "#/texts/58",
       "parent": {
@@ -1347,18 +1363,6 @@
     },
     {
       "self_ref": "#/texts/59",
-      "parent": {
-        "$ref": "#/body"
-      },
-      "children": [],
-      "content_layer": "body",
-      "label": "paragraph",
-      "prov": [],
-      "orig": "",
-      "text": ""
-    },
-    {
-      "self_ref": "#/texts/60",
       "parent": {
         "$ref": "#/groups/17"
       },
@@ -1372,11 +1376,12 @@
         "bold": true,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
-      "self_ref": "#/texts/61",
+      "self_ref": "#/texts/60",
       "parent": {
         "$ref": "#/groups/17"
       },
@@ -1388,7 +1393,7 @@
       "text": ""
     },
     {
-      "self_ref": "#/texts/62",
+      "self_ref": "#/texts/61",
       "parent": {
         "$ref": "#/groups/17"
       },
@@ -1402,11 +1407,12 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
-      "self_ref": "#/texts/63",
+      "self_ref": "#/texts/62",
       "parent": {
         "$ref": "#/body"
       },
@@ -1418,7 +1424,7 @@
       "text": ""
     },
     {
-      "self_ref": "#/texts/64",
+      "self_ref": "#/texts/63",
       "parent": {
         "$ref": "#/groups/18"
       },
@@ -1432,9 +1438,22 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
+    {
+      "self_ref": "#/texts/64",
+      "parent": {
+        "$ref": "#/body"
+      },
+      "children": [],
+      "content_layer": "body",
+      "label": "paragraph",
+      "prov": [],
+      "orig": "",
+      "text": ""
+    },
     {
       "self_ref": "#/texts/65",
       "parent": {
@@ -1458,18 +1477,6 @@
       "prov": [],
       "orig": "",
       "text": ""
-    },
-    {
-      "self_ref": "#/texts/67",
-      "parent": {
-        "$ref": "#/body"
-      },
-      "children": [],
-      "content_layer": "body",
-      "label": "paragraph",
-      "prov": [],
-      "orig": "",
-      "text": ""
     }
   ],
   "pictures": [],

tests/data/groundtruth/docling_v2/unit_test_01.html.json

@@ -1,6 +1,6 @@
 {
   "schema_name": "DoclingDocument",
-  "version": "1.3.0",
+  "version": "1.4.0",
   "name": "unit_test_01",
   "origin": {
     "mimetype": "text/html",

tests/data/groundtruth/docling_v2/unit_test_formatting.docx.itxt

@@ -17,14 +17,16 @@ item-0 at level 0: unspecified: group _root_
     item-16 at level 2: list_item: Italic bullet 1
     item-17 at level 2: list_item: Bold bullet 2
     item-18 at level 2: list_item: Underline bullet 3
-    item-19 at level 2: inline: group group
-      item-20 at level 3: list_item: Some
-      item-21 at level 3: list_item: italic
-      item-22 at level 3: list_item: bold
-      item-23 at level 3: list_item: underline
-    item-24 at level 2: list: group list
-      item-25 at level 3: inline: group group
-        item-26 at level 4: list_item: Nested
-        item-27 at level 4: list_item: italic
-        item-28 at level 4: list_item: bold
-  item-29 at level 1: paragraph: 
+    item-19 at level 2: list_item: 
+      item-20 at level 3: inline: group group
+        item-21 at level 4: text: Some
+        item-22 at level 4: text: italic
+        item-23 at level 4: text: bold
+        item-24 at level 4: text: underline
+    item-25 at level 2: list: group list
+      item-26 at level 3: list_item: 
+        item-27 at level 4: inline: group group
+          item-28 at level 5: text: Nested
+          item-29 at level 5: text: italic
+          item-30 at level 5: text: bold
+  item-31 at level 1: paragraph: 

tests/data/groundtruth/docling_v2/unit_test_formatting.docx.json

@@ -1,6 +1,6 @@
 {
   "schema_name": "DoclingDocument",
-  "version": "1.3.0",
+  "version": "1.4.0",
   "name": "unit_test_formatting",
   "origin": {
     "mimetype": "application/vnd.openxmlformats-officedocument.wordprocessingml.document",
@@ -42,7 +42,7 @@
         "$ref": "#/groups/1"
       },
       {
-        "$ref": "#/texts/23"
+        "$ref": "#/texts/25"
       }
     ],
     "content_layer": "body",
@@ -98,7 +98,7 @@
           "$ref": "#/texts/15"
         },
         {
-          "$ref": "#/groups/2"
+          "$ref": "#/texts/16"
         },
         {
           "$ref": "#/groups/3"
@@ -111,12 +111,9 @@
     {
       "self_ref": "#/groups/2",
       "parent": {
-        "$ref": "#/groups/1"
-      },
-      "children": [
-        {
         "$ref": "#/texts/16"
       },
+      "children": [
         {
           "$ref": "#/texts/17"
         },
@@ -125,6 +122,9 @@
         },
         {
           "$ref": "#/texts/19"
+        },
+        {
+          "$ref": "#/texts/20"
         }
       ],
       "content_layer": "body",
@@ -138,7 +138,7 @@
       },
       "children": [
         {
-          "$ref": "#/groups/4"
+          "$ref": "#/texts/21"
         }
       ],
       "content_layer": "body",
@@ -148,17 +148,17 @@
     {
       "self_ref": "#/groups/4",
       "parent": {
-        "$ref": "#/groups/3"
+        "$ref": "#/texts/21"
       },
       "children": [
-        {
-          "$ref": "#/texts/20"
-        },
-        {
-          "$ref": "#/texts/21"
-        },
         {
           "$ref": "#/texts/22"
+        },
+        {
+          "$ref": "#/texts/23"
+        },
+        {
+          "$ref": "#/texts/24"
         }
       ],
       "content_layer": "body",
@@ -182,7 +182,8 @@
         "bold": false,
         "italic": true,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -200,7 +201,8 @@
         "bold": true,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -218,7 +220,8 @@
         "bold": false,
         "italic": false,
         "underline": true,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -236,7 +239,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       },
       "hyperlink": "https:/github.com/DS4SD/docling"
     },
@@ -255,7 +259,8 @@
         "bold": true,
         "italic": true,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       },
       "hyperlink": "https:/github.com/DS4SD/docling"
     },
@@ -274,7 +279,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -292,7 +298,8 @@
         "bold": false,
         "italic": true,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -310,7 +317,8 @@
         "bold": true,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -328,7 +336,8 @@
         "bold": false,
         "italic": false,
         "underline": true,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -346,7 +355,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -364,7 +374,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       },
       "hyperlink": "https:/github.com/DS4SD/docling"
     },
@@ -383,7 +394,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -413,7 +425,8 @@
         "bold": false,
         "italic": true,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       },
       "enumerated": false,
       "marker": "-"
@@ -433,7 +446,8 @@
         "bold": true,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       },
       "enumerated": false,
       "marker": "-"
@@ -453,7 +467,8 @@
         "bold": false,
         "italic": false,
         "underline": true,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       },
       "enumerated": false,
       "marker": "-"
@@ -461,20 +476,18 @@
     {
       "self_ref": "#/texts/16",
       "parent": {
-        "$ref": "#/groups/2"
+        "$ref": "#/groups/1"
       },
-      "children": [],
+      "children": [
+        {
+          "$ref": "#/groups/2"
+        }
+      ],
       "content_layer": "body",
       "label": "list_item",
       "prov": [],
-      "orig": "Some",
-      "text": "Some",
-      "formatting": {
-        "bold": false,
-        "italic": false,
-        "underline": false,
-        "strikethrough": false
-      },
+      "orig": "",
+      "text": "",
       "enumerated": false,
       "marker": "-"
     },
@@ -485,18 +498,17 @@
       },
       "children": [],
       "content_layer": "body",
-      "label": "list_item",
+      "label": "text",
       "prov": [],
-      "orig": "italic",
-      "text": "italic",
+      "orig": "Some",
+      "text": "Some",
       "formatting": {
         "bold": false,
-        "italic": true,
+        "italic": false,
         "underline": false,
-        "strikethrough": false
-      },
-      "enumerated": false,
-      "marker": "-"
+        "strikethrough": false,
+        "script": "baseline"
+      }
     },
     {
       "self_ref": "#/texts/18",
@@ -505,18 +517,17 @@
       },
       "children": [],
       "content_layer": "body",
-      "label": "list_item",
+      "label": "text",
       "prov": [],
-      "orig": "bold",
-      "text": "bold",
+      "orig": "italic",
+      "text": "italic",
       "formatting": {
-        "bold": true,
-        "italic": false,
+        "bold": false,
+        "italic": true,
         "underline": false,
-        "strikethrough": false
-      },
-      "enumerated": false,
-      "marker": "-"
+        "strikethrough": false,
+        "script": "baseline"
+      }
     },
     {
       "self_ref": "#/texts/19",
@@ -525,7 +536,26 @@
       },
       "children": [],
       "content_layer": "body",
-      "label": "list_item",
+      "label": "text",
+      "prov": [],
+      "orig": "bold",
+      "text": "bold",
+      "formatting": {
+        "bold": true,
+        "italic": false,
+        "underline": false,
+        "strikethrough": false,
+        "script": "baseline"
+      }
+    },
+    {
+      "self_ref": "#/texts/20",
+      "parent": {
+        "$ref": "#/groups/2"
+      },
+      "children": [],
+      "content_layer": "body",
+      "label": "text",
       "prov": [],
       "orig": "underline",
       "text": "underline",
@@ -533,48 +563,25 @@
         "bold": false,
         "italic": false,
         "underline": true,
-        "strikethrough": false
-      },
-      "enumerated": false,
-      "marker": "-"
-    },
-    {
-      "self_ref": "#/texts/20",
-      "parent": {
-        "$ref": "#/groups/4"
-      },
-      "children": [],
-      "content_layer": "body",
-      "label": "list_item",
-      "prov": [],
-      "orig": "Nested",
-      "text": "Nested",
-      "formatting": {
-        "bold": false,
-        "italic": false,
-        "underline": false,
-        "strikethrough": false
-      },
-      "enumerated": false,
-      "marker": "-"
+        "strikethrough": false,
+        "script": "baseline"
+      }
     },
     {
       "self_ref": "#/texts/21",
       "parent": {
-        "$ref": "#/groups/4"
+        "$ref": "#/groups/3"
       },
-      "children": [],
+      "children": [
+        {
+          "$ref": "#/groups/4"
+        }
+      ],
       "content_layer": "body",
       "label": "list_item",
       "prov": [],
-      "orig": "italic",
-      "text": "italic",
-      "formatting": {
-        "bold": false,
-        "italic": true,
-        "underline": false,
-        "strikethrough": false
-      },
+      "orig": "",
+      "text": "",
       "enumerated": false,
       "marker": "-"
     },
@@ -585,7 +592,45 @@
       },
       "children": [],
       "content_layer": "body",
-      "label": "list_item",
+      "label": "text",
+      "prov": [],
+      "orig": "Nested",
+      "text": "Nested",
+      "formatting": {
+        "bold": false,
+        "italic": false,
+        "underline": false,
+        "strikethrough": false,
+        "script": "baseline"
+      }
+    },
+    {
+      "self_ref": "#/texts/23",
+      "parent": {
+        "$ref": "#/groups/4"
+      },
+      "children": [],
+      "content_layer": "body",
+      "label": "text",
+      "prov": [],
+      "orig": "italic",
+      "text": "italic",
+      "formatting": {
+        "bold": false,
+        "italic": true,
+        "underline": false,
+        "strikethrough": false,
+        "script": "baseline"
+      }
+    },
+    {
+      "self_ref": "#/texts/24",
+      "parent": {
+        "$ref": "#/groups/4"
+      },
+      "children": [],
+      "content_layer": "body",
+      "label": "text",
       "prov": [],
       "orig": "bold",
       "text": "bold",
@@ -593,13 +638,12 @@
         "bold": true,
         "italic": false,
         "underline": false,
-        "strikethrough": false
-      },
-      "enumerated": false,
-      "marker": "-"
+        "strikethrough": false,
+        "script": "baseline"
+      }
     },
     {
-      "self_ref": "#/texts/23",
+      "self_ref": "#/texts/25",
       "parent": {
         "$ref": "#/body"
       },

tests/data/groundtruth/docling_v2/unit_test_headers.docx.json

@@ -1,6 +1,6 @@
 {
   "schema_name": "DoclingDocument",
-  "version": "1.3.0",
+  "version": "1.4.0",
   "name": "unit_test_headers",
   "origin": {
     "mimetype": "application/vnd.openxmlformats-officedocument.wordprocessingml.document",
@@ -138,7 +138,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -168,7 +169,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -239,7 +241,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -269,7 +272,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -343,7 +347,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -373,7 +378,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -447,7 +453,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -477,7 +484,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -566,7 +574,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -596,7 +605,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -667,7 +677,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -697,7 +708,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -771,7 +783,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -801,7 +814,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {

tests/data/groundtruth/docling_v2/unit_test_headers_numbered.docx.json

@@ -1,6 +1,6 @@
 {
   "schema_name": "DoclingDocument",
-  "version": "1.3.0",
+  "version": "1.4.0",
   "name": "unit_test_headers_numbered",
   "origin": {
     "mimetype": "application/vnd.openxmlformats-officedocument.wordprocessingml.document",
@@ -214,7 +214,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -244,7 +245,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -315,7 +317,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -345,7 +348,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -419,7 +423,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -449,7 +454,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -523,7 +529,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -553,7 +560,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -620,7 +628,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -650,7 +659,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -721,7 +731,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -751,7 +762,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -825,7 +837,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -855,7 +868,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {

tests/data/groundtruth/docling_v2/unit_test_lists.docx.json

@@ -1,6 +1,6 @@
 {
   "schema_name": "DoclingDocument",
-  "version": "1.3.0",
+  "version": "1.4.0",
   "name": "unit_test_lists",
   "origin": {
     "mimetype": "application/vnd.openxmlformats-officedocument.wordprocessingml.document",
@@ -370,7 +370,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -400,7 +401,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -450,7 +452,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       },
       "enumerated": false,
       "marker": "-"
@@ -470,7 +473,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       },
       "enumerated": false,
       "marker": "-"
@@ -490,7 +494,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       },
       "enumerated": false,
       "marker": "-"
@@ -542,7 +547,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       },
       "enumerated": false,
       "marker": "-"
@@ -562,7 +568,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       },
       "enumerated": false,
       "marker": "-"
@@ -582,7 +589,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       },
       "enumerated": false,
       "marker": "-"
@@ -634,7 +642,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       },
       "enumerated": false,
       "marker": "-"
@@ -654,7 +663,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       },
       "enumerated": false,
       "marker": "-"
@@ -674,7 +684,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       },
       "enumerated": false,
       "marker": "-"
@@ -694,7 +705,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       },
       "enumerated": false,
       "marker": "-"
@@ -714,7 +726,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       },
       "enumerated": false,
       "marker": "-"
@@ -734,7 +747,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       },
       "enumerated": false,
       "marker": "-"
@@ -786,7 +800,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       },
       "enumerated": false,
       "marker": "-"
@@ -806,7 +821,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       },
       "enumerated": false,
       "marker": "-"
@@ -826,7 +842,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       },
       "enumerated": false,
       "marker": "-"
@@ -878,7 +895,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       },
       "enumerated": false,
       "marker": "-"
@@ -898,7 +916,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       },
       "enumerated": false,
       "marker": "-"
@@ -918,7 +937,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       },
       "enumerated": false,
       "marker": "-"
@@ -938,7 +958,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       },
       "enumerated": false,
       "marker": "-"
@@ -996,7 +1017,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       },
       "enumerated": false,
       "marker": "-"
@@ -1016,7 +1038,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       },
       "enumerated": false,
       "marker": "-"
@@ -1036,7 +1059,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       },
       "enumerated": false,
       "marker": "-"
@@ -1056,7 +1080,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       },
       "enumerated": false,
       "marker": "-"
@@ -1076,7 +1101,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       },
       "enumerated": false,
       "marker": "-"
@@ -1096,7 +1122,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       },
       "enumerated": false,
       "marker": "-"

tests/data/groundtruth/docling_v2/wiki_duck.html.json

@@ -1,6 +1,6 @@
 {
   "schema_name": "DoclingDocument",
-  "version": "1.3.0",
+  "version": "1.4.0",
   "name": "wiki_duck",
   "origin": {
     "mimetype": "text/html",
@@ -8489,7 +8489,8 @@
             }
           ]
         ]
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/1",
@@ -8648,7 +8649,8 @@
             }
           ]
         ]
-      }
+      },
+      "annotations": []
     }
   ],
   "key_value_items": [],

tests/data/groundtruth/docling_v2/word_image_anchors.docx.itxt

@@ -0,0 +1,16 @@
+item-0 at level 0: unspecified: group _root_
+  item-1 at level 1: paragraph: Transcript
+  item-2 at level 1: paragraph: February 20, 2025, 8:32PM
+  item-3 at level 1: picture
+  item-4 at level 1: inline: group group
+    item-5 at level 2: paragraph: This is test 1
+    item-6 at level 2: paragraph: 0:08
+Correct, he is not.
+  item-7 at level 1: paragraph: 
+  item-8 at level 1: picture
+  item-9 at level 1: inline: group group
+    item-10 at level 2: paragraph: This is test 2
+    item-11 at level 2: paragraph: 0:16
+Yeah, exactly.
+  item-12 at level 1: paragraph: 
+  item-13 at level 1: paragraph: 

tests/data/groundtruth/docling_v2/word_image_anchors.docx.json

@@ -0,0 +1,292 @@
+{
+  "schema_name": "DoclingDocument",
+  "version": "1.4.0",
+  "name": "word_image_anchors",
+  "origin": {
+    "mimetype": "application/vnd.openxmlformats-officedocument.wordprocessingml.document",
+    "binary_hash": 2428692234257307633,
+    "filename": "word_image_anchors.docx"
+  },
+  "furniture": {
+    "self_ref": "#/furniture",
+    "children": [],
+    "content_layer": "furniture",
+    "name": "_root_",
+    "label": "unspecified"
+  },
+  "body": {
+    "self_ref": "#/body",
+    "children": [
+      {
+        "$ref": "#/texts/0"
+      },
+      {
+        "$ref": "#/texts/1"
+      },
+      {
+        "$ref": "#/pictures/0"
+      },
+      {
+        "$ref": "#/groups/0"
+      },
+      {
+        "$ref": "#/texts/4"
+      },
+      {
+        "$ref": "#/pictures/1"
+      },
+      {
+        "$ref": "#/groups/1"
+      },
+      {
+        "$ref": "#/texts/7"
+      },
+      {
+        "$ref": "#/texts/8"
+      }
+    ],
+    "content_layer": "body",
+    "name": "_root_",
+    "label": "unspecified"
+  },
+  "groups": [
+    {
+      "self_ref": "#/groups/0",
+      "parent": {
+        "$ref": "#/body"
+      },
+      "children": [
+        {
+          "$ref": "#/texts/2"
+        },
+        {
+          "$ref": "#/texts/3"
+        }
+      ],
+      "content_layer": "body",
+      "name": "group",
+      "label": "inline"
+    },
+    {
+      "self_ref": "#/groups/1",
+      "parent": {
+        "$ref": "#/body"
+      },
+      "children": [
+        {
+          "$ref": "#/texts/5"
+        },
+        {
+          "$ref": "#/texts/6"
+        }
+      ],
+      "content_layer": "body",
+      "name": "group",
+      "label": "inline"
+    }
+  ],
+  "texts": [
+    {
+      "self_ref": "#/texts/0",
+      "parent": {
+        "$ref": "#/body"
+      },
+      "children": [],
+      "content_layer": "body",
+      "label": "paragraph",
+      "prov": [],
+      "orig": "Transcript",
+      "text": "Transcript",
+      "formatting": {
+        "bold": true,
+        "italic": false,
+        "underline": false,
+        "strikethrough": false,
+        "script": "baseline"
+      }
+    },
+    {
+      "self_ref": "#/texts/1",
+      "parent": {
+        "$ref": "#/body"
+      },
+      "children": [],
+      "content_layer": "body",
+      "label": "paragraph",
+      "prov": [],
+      "orig": "February 20, 2025, 8:32PM",
+      "text": "February 20, 2025, 8:32PM",
+      "formatting": {
+        "bold": false,
+        "italic": false,
+        "underline": false,
+        "strikethrough": false,
+        "script": "baseline"
+      }
+    },
+    {
+      "self_ref": "#/texts/2",
+      "parent": {
+        "$ref": "#/groups/0"
+      },
+      "children": [],
+      "content_layer": "body",
+      "label": "paragraph",
+      "prov": [],
+      "orig": "This is test 1",
+      "text": "This is test 1",
+      "formatting": {
+        "bold": true,
+        "italic": false,
+        "underline": false,
+        "strikethrough": false,
+        "script": "baseline"
+      }
+    },
+    {
+      "self_ref": "#/texts/3",
+      "parent": {
+        "$ref": "#/groups/0"
+      },
+      "children": [],
+      "content_layer": "body",
+      "label": "paragraph",
+      "prov": [],
+      "orig": "0:08\nCorrect, he is not.",
+      "text": "0:08\nCorrect, he is not.",
+      "formatting": {
+        "bold": false,
+        "italic": false,
+        "underline": false,
+        "strikethrough": false,
+        "script": "baseline"
+      }
+    },
+    {
+      "self_ref": "#/texts/4",
+      "parent": {
+        "$ref": "#/body"
+      },
+      "children": [],
+      "content_layer": "body",
+      "label": "paragraph",
+      "prov": [],
+      "orig": "",
+      "text": ""
+    },
+    {
+      "self_ref": "#/texts/5",
+      "parent": {
+        "$ref": "#/groups/1"
+      },
+      "children": [],
+      "content_layer": "body",
+      "label": "paragraph",
+      "prov": [],
+      "orig": "This is test 2",
+      "text": "This is test 2",
+      "formatting": {
+        "bold": true,
+        "italic": false,
+        "underline": false,
+        "strikethrough": false,
+        "script": "baseline"
+      }
+    },
+    {
+      "self_ref": "#/texts/6",
+      "parent": {
+        "$ref": "#/groups/1"
+      },
+      "children": [],
+      "content_layer": "body",
+      "label": "paragraph",
+      "prov": [],
+      "orig": "0:16\nYeah, exactly.",
+      "text": "0:16\nYeah, exactly.",
+      "formatting": {
+        "bold": false,
+        "italic": false,
+        "underline": false,
+        "strikethrough": false,
+        "script": "baseline"
+      }
+    },
+    {
+      "self_ref": "#/texts/7",
+      "parent": {
+        "$ref": "#/body"
+      },
+      "children": [],
+      "content_layer": "body",
+      "label": "paragraph",
+      "prov": [],
+      "orig": "",
+      "text": ""
+    },
+    {
+      "self_ref": "#/texts/8",
+      "parent": {
+        "$ref": "#/body"
+      },
+      "children": [],
+      "content_layer": "body",
+      "label": "paragraph",
+      "prov": [],
+      "orig": "",
+      "text": ""
+    }
+  ],
+  "pictures": [
+    {
+      "self_ref": "#/pictures/0",
+      "parent": {
+        "$ref": "#/body"
+      },
+      "children": [],
+      "content_layer": "body",
+      "label": "picture",
+      "prov": [],
+      "captions": [],
+      "references": [],
+      "footnotes": [],
+      "image": {
+        "mimetype": "image/png",
+        "dpi": 72,
+        "size": {
+          "width": 100.0,
+          "height": 100.0
+        },
+        "uri": "data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAGQAAABkCAYAAABw4pVUAAAAz0lEQVR4nO3bUW0CURRF0TukQvDSauBr0mACE1VBAzYQg5Lpdw0wO2EtA+cl+/6+GQAAAAAAAAAAAADe1DIR53X9mcNcdhnf5nm93Y8T8DElyzyuv/evlx/CMqeJOOz9AP4TJEaQGEFiBIkRJEaQGEFiBIkRJEaQGEFiBIkRJEaQGEFiBIkRJEaQGEFiBIkRJEaQGEFiBIkRJEaQGEFiBIkRJEaQGEFiBIkRJEaQGEFiWp8+t/k8f6/bDrvPl28CAAAAAAAAAAAAAAAAzLv5A5bTEG2TIIlOAAAAAElFTkSuQmCC"
+      },
+      "annotations": []
+    },
+    {
+      "self_ref": "#/pictures/1",
+      "parent": {
+        "$ref": "#/body"
+      },
+      "children": [],
+      "content_layer": "body",
+      "label": "picture",
+      "prov": [],
+      "captions": [],
+      "references": [],
+      "footnotes": [],
+      "image": {
+        "mimetype": "image/png",
+        "dpi": 72,
+        "size": {
+          "width": 100.0,
+          "height": 100.0
+        },
+        "uri": "data:image/png;base64,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"
+      },
+      "annotations": []
+    }
+  ],
+  "tables": [],
+  "key_value_items": [],
+  "form_items": [],
+  "pages": {}
+}

tests/data/groundtruth/docling_v2/word_image_anchors.docx.md

@@ -0,0 +1,13 @@
+**Transcript**
+
+February 20, 2025, 8:32PM
+
+<!-- image -->
+
+**This is test 1** 0:08
+Correct, he is not.
+
+<!-- image -->
+
+**This is test 2** 0:16
+Yeah, exactly.

tests/data/groundtruth/docling_v2/word_sample.docx.json

@@ -1,6 +1,6 @@
 {
   "schema_name": "DoclingDocument",
-  "version": "1.3.0",
+  "version": "1.4.0",
   "name": "word_sample",
   "origin": {
     "mimetype": "application/vnd.openxmlformats-officedocument.wordprocessingml.document",
@@ -106,7 +106,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -149,7 +150,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -167,7 +169,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -217,7 +220,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -235,7 +239,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       },
       "enumerated": false,
       "marker": "-"
@@ -255,7 +260,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       },
       "enumerated": false,
       "marker": "-"
@@ -275,7 +281,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       },
       "enumerated": false,
       "marker": "-"
@@ -295,7 +302,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -313,7 +321,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       },
       "enumerated": false,
       "marker": "-"
@@ -333,7 +342,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       },
       "enumerated": false,
       "marker": "-"
@@ -353,7 +363,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       },
       "enumerated": false,
       "marker": "-"
@@ -373,7 +384,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -426,7 +438,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -444,7 +457,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -462,7 +476,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -492,7 +507,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -510,7 +526,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       },
       "enumerated": false,
       "marker": "-"
@@ -530,7 +547,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       },
       "enumerated": false,
       "marker": "-"
@@ -550,7 +568,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       },
       "enumerated": false,
       "marker": "-"
@@ -897,7 +916,8 @@
             }
           ]
         ]
-      }
+      },
+      "annotations": []
     }
   ],
   "key_value_items": [],

tests/data/groundtruth/docling_v2/word_tables.docx.json

@@ -1,6 +1,6 @@
 {
   "schema_name": "DoclingDocument",
-  "version": "1.3.0",
+  "version": "1.4.0",
   "name": "word_tables",
   "origin": {
     "mimetype": "application/vnd.openxmlformats-officedocument.wordprocessingml.document",
@@ -119,7 +119,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -149,7 +150,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -179,7 +181,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -209,7 +212,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -239,7 +243,8 @@
         "bold": false,
         "italic": false,
         "underline": false,
-        "strikethrough": false
+        "strikethrough": false,
+        "script": "baseline"
       }
     },
     {
@@ -510,7 +515,8 @@
             }
           ]
         ]
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/1",
@@ -729,7 +735,8 @@
             }
           ]
         ]
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/2",
@@ -1020,7 +1027,8 @@
             }
           ]
         ]
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/3",
@@ -1387,7 +1395,8 @@
             }
           ]
         ]
-      }
+      },
+      "annotations": []
     },
     {
       "self_ref": "#/tables/4",
@@ -2398,7 +2407,8 @@
             }
           ]
         ]
-      }
+      },
+      "annotations": []
     }
   ],
   "key_value_items": [],

tests/data/md/inline_and_formatting.md

@@ -11,12 +11,22 @@ Create your feature branch: `git checkout -b feature/AmazingFeature`.
 3. Commit your changes (`git commit -m 'Add some AmazingFeature'`)
 4. Push to the branch (`git push origin feature/AmazingFeature`)
 5. Open a Pull Request
+6. **Whole list item has same formatting**
+7. List item has *mixed or partial* formatting
 
-## *Second* section  <!-- inline groups in headings not yet supported by serializers -->
+# *Whole heading is italic*
 
+<<<<<<< HEAD
 - **First**: Lorem ipsum.
 - **Second**: Dolor `sit` amet.
 
 | **Bold Heading** | *Italic Heading* |
 |------------------|------------------|
 | data a           | data b           |
+=======
+Some *`formatted_code`*
+
+## *Partially formatted* heading to_escape `not_to_escape`
+
+[$$E=mc^2$$](https://en.wikipedia.org/wiki/Albert_Einstein)
+>>>>>>> origin/main

tests/data/webp/groundtruth/docling_v2/webp-test.json

@@ -1,6 +1,6 @@
 {
   "schema_name": "DoclingDocument",
-  "version": "1.3.0",
+  "version": "1.4.0",
   "name": "webp-test",
   "origin": {
     "mimetype": "application/pdf",

tests/data_scanned/groundtruth/docling_v2/ocr_test.json

@@ -1,6 +1,6 @@
 {
   "schema_name": "DoclingDocument",
-  "version": "1.3.0",
+  "version": "1.4.0",
   "name": "ocr_test",
   "origin": {
     "mimetype": "application/pdf",