docling 2.20.0__tar.gz → 2.21.0__tar.gz

{docling-2.20.0 → docling-2.21.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: docling
-Version: 2.20.0
+Version: 2.21.0
 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.
 Home-page: https://github.com/DS4SD/docling
 License: MIT
@@ -28,7 +28,7 @@ Provides-Extra: vlm
 Requires-Dist: beautifulsoup4 (>=4.12.3,<4.13.0)
 Requires-Dist: certifi (>=2024.7.4)
 Requires-Dist: deepsearch-glm (>=1.0.0,<2.0.0)
-Requires-Dist: docling-core[chunking] (>=2.17.2,<3.0.0)
+Requires-Dist: docling-core[chunking] (>=2.18.0,<3.0.0)
 Requires-Dist: docling-ibm-models (>=3.3.0,<4.0.0)
 Requires-Dist: docling-parse (>=3.3.0,<4.0.0)
 Requires-Dist: easyocr (>=1.7,<2.0)

{docling-2.20.0 → docling-2.21.0}/docling/models/ds_glm_model.py

@@ -4,7 +4,12 @@ from pathlib import Path
 from typing import List, Union
 
 from deepsearch_glm.andromeda_nlp import nlp_model
-from docling_core.types.doc import BoundingBox, CoordOrigin, DoclingDocument
+from docling_core.types.doc import (
+    BoundingBox,
+    CoordOrigin,
+    DocItemLabel,
+    DoclingDocument,
+)
 from docling_core.types.legacy_doc.base import BoundingBox as DsBoundingBox
 from docling_core.types.legacy_doc.base import (
     Figure,
@@ -71,12 +76,15 @@ class GlmModel:
         )
 
         main_text: List[Union[Ref, BaseText]] = []
+        page_headers: List[Union[Ref, BaseText]] = []
+        page_footers: List[Union[Ref, BaseText]] = []
+
         tables: List[DsSchemaTable] = []
         figures: List[Figure] = []
 
         page_no_to_page = {p.page_no: p for p in conv_res.pages}
 
-        for element in conv_res.assembled.elements:
+        for element in conv_res.assembled.body:
             # Convert bboxes to lower-left origin.
             target_bbox = DsBoundingBox(
                 element.cluster.bbox.to_bottom_left_origin(
@@ -238,6 +246,53 @@ class GlmModel:
                     )
                 )
 
+        # We can throw in headers and footers at the end of the legacy doc
+        # since the reading-order will re-sort it later.
+        for element in conv_res.assembled.headers:
+            # Convert bboxes to lower-left origin.
+            target_bbox = DsBoundingBox(
+                element.cluster.bbox.to_bottom_left_origin(
+                    page_no_to_page[element.page_no].size.height
+                ).as_tuple()
+            )
+
+            if isinstance(element, TextElement):
+
+                tel = BaseText(
+                    text=element.text,
+                    obj_type=layout_label_to_ds_type.get(element.label),
+                    name=element.label,
+                    prov=[
+                        Prov(
+                            bbox=target_bbox,
+                            page=element.page_no + 1,
+                            span=[0, len(element.text)],
+                        )
+                    ],
+                )
+                if element.label == DocItemLabel.PAGE_HEADER:
+                    index = len(page_headers)
+                    ref_str = f"#/page-headers/{index}"
+                    main_text.append(
+                        Ref(
+                            name=element.label,
+                            obj_type=layout_label_to_ds_type.get(element.label),
+                            ref=ref_str,
+                        ),
+                    )
+                    page_headers.append(tel)
+                elif element.label == DocItemLabel.PAGE_FOOTER:
+                    index = len(page_footers)
+                    ref_str = f"#/page-footers/{index}"
+                    main_text.append(
+                        Ref(
+                            name=element.label,
+                            obj_type=layout_label_to_ds_type.get(element.label),
+                            ref=ref_str,
+                        ),
+                    )
+                    page_footers.append(tel)
+
         page_dimensions = [
             PageDimensions(page=p.page_no + 1, height=p.size.height, width=p.size.width)
             for p in conv_res.pages
@@ -252,6 +307,8 @@ class GlmModel:
             tables=tables,
             figures=figures,
             page_dimensions=page_dimensions,
+            page_headers=page_headers,
+            page_footers=page_footers,
         )
 
         return ds_doc
@@ -264,6 +321,7 @@ class GlmModel:
             glm_doc = self.model.apply_on_doc(ds_doc_dict)
 
             docling_doc: DoclingDocument = to_docling_document(glm_doc)  # Experimental
+            1 == 1
 
         # DEBUG code:
         def draw_clusters_and_cells(ds_document, page_no, show: bool = False):

{docling-2.20.0 → docling-2.21.0}/docling/utils/glm_utils.py

@@ -15,6 +15,7 @@ from docling_core.types.doc import (
     TableCell,
     TableData,
 )
+from docling_core.types.doc.document import ContentLayer
 
 
 def resolve_item(paths, obj):
@@ -311,6 +312,15 @@ def to_docling_document(doc_glm, update_name_label=False) -> DoclingDocument:
                 current_list = None
 
                 doc.add_text(label=DocItemLabel.FORMULA, text="", orig=text, prov=prov)
+            elif label in [DocItemLabel.PAGE_HEADER, DocItemLabel.PAGE_FOOTER]:
+                current_list = None
+
+                doc.add_text(
+                    label=DocItemLabel(name_label),
+                    text=text,
+                    prov=prov,
+                    content_layer=ContentLayer.FURNITURE,
+                )
             else:
                 current_list = None
 

{docling-2.20.0 → docling-2.21.0}/pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "docling"
-version = "2.20.0"  # DO NOT EDIT, updated automatically
+version = "2.21.0"  # 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."
 authors = ["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 = "MIT"
@@ -26,7 +26,7 @@ packages = [{include = "docling"}]
 ######################
 python = "^3.9"
 pydantic = "^2.0.0"
-docling-core = {extras = ["chunking"], version = "^2.17.2"}
+docling-core = {extras = ["chunking"], version = "^2.18.0"}
 docling-ibm-models = "^3.3.0"
 deepsearch-glm = "^1.0.0"
 docling-parse = "^3.3.0"