docling 2.55.0__py3-none-any.whl → 2.55.1__py3-none-any.whl

docling/backend/md_backend.py

@@ -249,7 +249,10 @@ class MarkdownDocumentBackend(DeclarativeDocumentBackend):
 
         # Iterates over all elements in the AST
         # Check for different element types and process relevant details
-        if isinstance(element, marko.block.Heading) and len(element.children) > 0:
+        if (
+            isinstance(element, marko.block.Heading)
+            or isinstance(element, marko.block.SetextHeading)
+        ) and len(element.children) > 0:
             self._close_table(doc)
             _log.debug(
                 f" - Heading level {element.level}, content: {element.children[0].children}"  # type: ignore

docling/cli/main.py

@@ -355,6 +355,13 @@ def convert(  # noqa: C901
             help="Replace any existing text with OCR generated text over the full content.",
         ),
     ] = False,
+    tables: Annotated[
+        bool,
+        typer.Option(
+            ...,
+            help="If enabled, the table structure model will be used to extract table information.",
+        ),
+    ] = True,
     ocr_engine: Annotated[
         str,
         typer.Option(
@@ -591,7 +598,7 @@ def convert(  # noqa: C901
                 accelerator_options=accelerator_options,
                 do_ocr=ocr,
                 ocr_options=ocr_options,
-                do_table_structure=True,
+                do_table_structure=tables,
                 do_code_enrichment=enrich_code,
                 do_formula_enrichment=enrich_formula,
                 do_picture_description=enrich_picture_description,

docling/models/readingorder_model.py

@@ -9,6 +9,7 @@ from docling_core.types.doc import (
     NodeItem,
     ProvenanceItem,
     RefItem,
+    RichTableCell,
     TableData,
 )
 from docling_core.types.doc.document import ContentLayer
@@ -103,6 +104,22 @@ class ReadingOrderModel:
             else:
                 doc.add_text(parent=doc_item, label=c_label, text=c_text, prov=c_prov)
 
+    def _create_rich_cell_group(
+        self, element: BasePageElement, doc: DoclingDocument, table_item: NodeItem
+    ) -> RefItem:
+        """Create a group containing all child elements for a rich table cell."""
+        group_name = f"rich_cell_group_{len(doc.tables)}_0_0"
+        group_element = doc.add_group(
+            label=GroupLabel.UNSPECIFIED,
+            name=group_name,
+            parent=table_item,
+        )
+
+        # Add all child elements to the group
+        self._add_child_elements(element, group_element, doc)
+
+        return group_element.get_ref()
+
     def _readingorder_elements_to_docling_doc(
         self,
         conv_res: ConversionResult,
@@ -197,11 +214,21 @@ class ReadingOrderModel:
                             )
 
             elif isinstance(element, Table):
-                tbl_data = TableData(
-                    num_rows=element.num_rows,
-                    num_cols=element.num_cols,
-                    table_cells=element.table_cells,
-                )
+                # Check if table has no structure prediction
+                if element.num_rows == 0 and element.num_cols == 0:
+                    # Only create 1x1 table if there are children to put in it
+                    if element.cluster.children:
+                        # Create minimal 1x1 table with rich cell containing all children
+                        tbl_data = TableData(num_rows=1, num_cols=1, table_cells=[])
+                    else:
+                        # Create empty table with no structure
+                        tbl_data = TableData(num_rows=0, num_cols=0, table_cells=[])
+                else:
+                    tbl_data = TableData(
+                        num_rows=element.num_rows,
+                        num_cols=element.num_cols,
+                        table_cells=element.table_cells,
+                    )
 
                 prov = ProvenanceItem(
                     page_no=element.page_no + 1,
@@ -231,6 +258,30 @@ class ReadingOrderModel:
 
                         tbl.footnotes.append(new_footnote_item.get_ref())
 
+                # Handle case where table has no structure prediction but has children
+                if (
+                    element.num_rows == 0
+                    and element.num_cols == 0
+                    and element.cluster.children
+                ):
+                    # Create rich cell containing all child elements
+                    rich_cell_ref = self._create_rich_cell_group(element, out_doc, tbl)
+
+                    # Create rich table cell spanning the entire 1x1 table
+                    rich_cell = RichTableCell(
+                        text="",  # Empty text since content is in the group
+                        row_span=1,
+                        col_span=1,
+                        start_row_offset_idx=0,
+                        end_row_offset_idx=1,
+                        start_col_offset_idx=0,
+                        end_col_offset_idx=1,
+                        column_header=False,
+                        row_header=False,
+                        ref=rich_cell_ref,
+                    )
+                    out_doc.add_table_cell(table_item=tbl, cell=rich_cell)
+
                 # TODO: Consider adding children of Table.
 
             elif isinstance(element, FigureElement):

{docling-2.55.0.dist-info → docling-2.55.1.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: docling
-Version: 2.55.0
+Version: 2.55.1
 Summary: SDK and CLI for parsing PDF, DOCX, HTML, and more, to a unified document representation for powering downstream workflows such as gen AI applications.
 Author-email: Christoph Auer <cau@zurich.ibm.com>, Michele Dolfi <dol@zurich.ibm.com>, Maxim Lysak <mly@zurich.ibm.com>, Nikos Livathinos <nli@zurich.ibm.com>, Ahmed Nassar <ahn@zurich.ibm.com>, Panos Vagenas <pva@zurich.ibm.com>, Peter Staar <taa@zurich.ibm.com>
 License-Expression: MIT

{docling-2.55.0.dist-info → docling-2.55.1.dist-info}/RECORD

@@ -11,7 +11,7 @@ docling/backend/docling_parse_backend.py,sha256=9rUo1vPxX6QLzGqF-2B2iEYglZg6YQ3U
 docling/backend/docling_parse_v2_backend.py,sha256=3ckTfke8IICjaImlIzc3TRhG7KDuxDDba0AuCEcjA-M,9500
 docling/backend/docling_parse_v4_backend.py,sha256=xCBbaaXjNNrOaod9tmBuCbe5mL_ipmTNG2XOxVbGG3w,7891
 docling/backend/html_backend.py,sha256=r2m3aIKwwr8Vv2Fxri1FaZFvd4EWvTQlmSPwXeD79zg,47796
-docling/backend/md_backend.py,sha256=zrOUYoIYudUfigwnXRQocb_M4G_ptYfblNgr6BNTYQw,22678
+docling/backend/md_backend.py,sha256=TWboEPHl93pqI_Go1a3XpP-KpzI3d17xo5ZW42Ul0kY,22764
 docling/backend/mets_gbs_backend.py,sha256=EA8sY6tbmGiysKGYPPZiNlK-i7Adn8bLTo-7Ym15hTU,12774
 docling/backend/msexcel_backend.py,sha256=5JRbPwOjR1r45AMeIts1rj6InbOgLBf_CtAhvNPVmsQ,19157
 docling/backend/mspowerpoint_backend.py,sha256=wJgB2JStEPfD7MPpWQlpPN7bffPxaHFUnKD4wj8SLxU,15114
@@ -31,7 +31,7 @@ docling/backend/xml/jats_backend.py,sha256=_BWpQQg3SlsHAOOj0v2qRJoVqaQzL91GqN1tK
 docling/backend/xml/uspto_backend.py,sha256=Tv4CE7V5_QwxTNJPl90CAd_mAbwaLGy8S6s6evh1Xow,70910
 docling/chunking/__init__.py,sha256=h83TDs0AuOV6oEPLAPrn9dpGKiU-2Vg6IRNo4cv6GDA,346
 docling/cli/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-docling/cli/main.py,sha256=UX-5fRGVP_yGxTQez0x1PNnaNKRgWdcXGoPCHy-0uFM,32887
+docling/cli/main.py,sha256=glci4i1KYphr_8WueEnrSQF08xvtGxXBi6EvWDGewHU,33091
 docling/cli/models.py,sha256=rw_2JfeJ-k_iOLpz3JfgL1QbJY__W9nE23nHdov6VfU,6252
 docling/cli/tools.py,sha256=QhtRxQG0TVrfsMqdv5i7J0_qQy1ZZyWYnHPwJl7b5oY,322
 docling/datamodel/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
@@ -61,7 +61,7 @@ docling/models/picture_description_api_model.py,sha256=o3EkV5aHW_6WzE_fdj_VRnNCr
 docling/models/picture_description_base_model.py,sha256=kLthLhdlgwhootQ4_xhhcAk6A-vso5-qcsFJ3TcYfO0,2991
 docling/models/picture_description_vlm_model.py,sha256=Uja_BQSk7F-U1J2hm4yeLguirUzKYv1K8zRyw1IYomY,4150
 docling/models/rapid_ocr_model.py,sha256=anUVUwaj9Wubgu4FnHdYMuOVkQP_hJiLY1qRToelBoc,7700
-docling/models/readingorder_model.py,sha256=_usJdpM4GMWeGGneEwLLxa9grIGQb0XnNMugV72jGbY,14911
+docling/models/readingorder_model.py,sha256=-j-UuvnsYWqZvY0gByKz0bjcBwOhWQTHerCopig_jVs,17266
 docling/models/table_structure_model.py,sha256=7g_mFf1YzfF8PXQfefNu6XYZu7TzJAn86zKb6IEUdCg,12518
 docling/models/tesseract_ocr_cli_model.py,sha256=I3Gn28Y-LD8OfvyCElN9fLiNgpo2sT0uMkVt258253s,12881
 docling/models/tesseract_ocr_model.py,sha256=GdI5Cjfi87qcehVbM3wdKRvKkl_F9A4bwTUbjXZCJYA,10745
@@ -101,9 +101,9 @@ docling/utils/orientation.py,sha256=jTyLxyT31FlOodZoBMlADHNQK2lAWKYVs5z7pXd_6Cg,
 docling/utils/profiling.py,sha256=YaMGoB9MMZpagF9mb5ndoHj8Lpb9aIdb7El-Pl7IcFs,1753
 docling/utils/utils.py,sha256=kJtIYuzXeOyJHYlxmLAo7dGM5rEsDa1i84qEsUj1nio,1908
 docling/utils/visualization.py,sha256=tY2ylE2aiQKkmzlSLnFW-HTfFyqUUMguW18ldd1PLfo,2868
-docling-2.55.0.dist-info/licenses/LICENSE,sha256=mBb7ErEcM8VS9OhiGHnQ2kk75HwPhr54W1Oiz3965MY,1088
-docling-2.55.0.dist-info/METADATA,sha256=e1RK_bATZ2Q_Ie9kC6uHFCj99D7pkW678jxk_l0CHxk,11252
-docling-2.55.0.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-docling-2.55.0.dist-info/entry_points.txt,sha256=hzVlbeE0aMSTQ9S0-NTYN0Hmgsn6qL_EA2qX4UbkAuY,149
-docling-2.55.0.dist-info/top_level.txt,sha256=vkIywP-USjFyYo1AIRQbWQQaL3xB5jf8vkCYdTIfNic,8
-docling-2.55.0.dist-info/RECORD,,
+docling-2.55.1.dist-info/licenses/LICENSE,sha256=mBb7ErEcM8VS9OhiGHnQ2kk75HwPhr54W1Oiz3965MY,1088
+docling-2.55.1.dist-info/METADATA,sha256=Q0vntiJGbmAM9ONc6x8a-CeCJWpLDJPCN5k7_hhn6sA,11252
+docling-2.55.1.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+docling-2.55.1.dist-info/entry_points.txt,sha256=hzVlbeE0aMSTQ9S0-NTYN0Hmgsn6qL_EA2qX4UbkAuY,149
+docling-2.55.1.dist-info/top_level.txt,sha256=vkIywP-USjFyYo1AIRQbWQQaL3xB5jf8vkCYdTIfNic,8
+docling-2.55.1.dist-info/RECORD,,