docling-core 2.34.2__py3-none-any.whl → 2.35.0__py3-none-any.whl

docling_core/transforms/visualizer/layout_visualizer.py

@@ -119,7 +119,10 @@ class LayoutVisualizer(BaseVisualizer):
                     )
 
     def _draw_doc_layout(
-        self, doc: DoclingDocument, images: Optional[dict[Optional[int], Image]] = None
+        self,
+        doc: DoclingDocument,
+        images: Optional[dict[Optional[int], Image]] = None,
+        included_content_layers: Optional[set[ContentLayer]] = None,
     ):
         """Draw the document clusters and optionaly the reading order."""
         clusters = []
@@ -128,6 +131,9 @@ class LayoutVisualizer(BaseVisualizer):
         if images is not None:
             my_images = images
 
+        if included_content_layers is None:
+            included_content_layers = {c for c in ContentLayer}
+
         # Initialise `my_images` beforehand: sometimes, you have the
         # page-images but no DocItems!
         for page_nr, page in doc.pages.items():
@@ -141,9 +147,7 @@ class LayoutVisualizer(BaseVisualizer):
         prev_image = None
         prev_page_nr = None
         for idx, (elem, _) in enumerate(
-            doc.iterate_items(
-                included_content_layers={ContentLayer.BODY, ContentLayer.FURNITURE}
-            )
+            doc.iterate_items(included_content_layers=included_content_layers)
         ):
             if not isinstance(elem, DocItem):
                 continue

docling_core/transforms/visualizer/table_visualizer.py

@@ -0,0 +1,135 @@
+"""Define classes for layout visualization."""
+
+import logging
+from copy import deepcopy
+from typing import Optional
+
+from PIL import ImageDraw
+from PIL.Image import Image
+from pydantic import BaseModel
+from typing_extensions import override
+
+from docling_core.transforms.visualizer.base import BaseVisualizer
+from docling_core.types.doc.document import ContentLayer, DoclingDocument, TableItem
+
+_log = logging.getLogger(__name__)
+
+
+class TableVisualizer(BaseVisualizer):
+    """Table visualizer."""
+
+    class Params(BaseModel):
+        """Table visualization parameters."""
+
+        # show_Label: bool = False
+        show_cells: bool = True
+        # show_rows: bool = False
+        # show_cols: bool = False
+
+    base_visualizer: Optional[BaseVisualizer] = None
+    params: Params = Params()
+
+    def _draw_table_cells(
+        self,
+        table: TableItem,
+        page_image: Image,
+        page_height: float,
+        scale_x: float,
+        scale_y: float,
+    ):
+        """Draw individual table cells."""
+        draw = ImageDraw.Draw(page_image, "RGBA")
+
+        for cell in table.data.table_cells:
+            if cell.bbox is not None:
+
+                tl_bbox = cell.bbox.to_top_left_origin(page_height=page_height)
+
+                cell_color = (256, 0, 0, 32)  # Transparent black for cells
+
+                cx0, cy0, cx1, cy1 = tl_bbox.as_tuple()
+                cx0 *= scale_x
+                cx1 *= scale_x
+                cy0 *= scale_y
+                cy1 *= scale_y
+
+                draw.rectangle(
+                    [(cx0, cy0), (cx1, cy1)],
+                    outline=(256, 0, 0, 128),
+                    fill=cell_color,
+                )
+
+    def _draw_doc_tables(
+        self,
+        doc: DoclingDocument,
+        images: Optional[dict[Optional[int], Image]] = None,
+        included_content_layers: Optional[set[ContentLayer]] = None,
+    ):
+        """Draw the document tables."""
+        my_images: dict[Optional[int], Image] = {}
+
+        if images is not None:
+            my_images = images
+
+        if included_content_layers is None:
+            included_content_layers = {c for c in ContentLayer}
+
+        # Initialise `my_images` beforehand: sometimes, you have the
+        # page-images but no DocItems!
+        for page_nr, page in doc.pages.items():
+            page_image = doc.pages[page_nr].image
+            if page_image is None or (pil_img := page_image.pil_image) is None:
+                raise RuntimeError("Cannot visualize document without images")
+            elif page_nr not in my_images:
+                image = deepcopy(pil_img)
+                my_images[page_nr] = image
+
+        for idx, (elem, _) in enumerate(
+            doc.iterate_items(included_content_layers=included_content_layers)
+        ):
+            if not isinstance(elem, TableItem):
+                continue
+            if len(elem.prov) == 0:
+                continue  # Skip elements without provenances
+
+            if len(elem.prov) == 1:
+
+                page_nr = elem.prov[0].page_no
+
+                if page_nr in my_images:
+                    image = my_images[page_nr]
+
+                    if self.params.show_cells:
+                        self._draw_table_cells(
+                            table=elem,
+                            page_height=doc.pages[page_nr].size.height,
+                            page_image=image,
+                            scale_x=image.width / doc.pages[page_nr].size.width,
+                            scale_y=image.height / doc.pages[page_nr].size.height,
+                        )
+
+                else:
+                    raise RuntimeError(f"Cannot visualize page-image for {page_nr}")
+
+            else:
+                _log.error("Can not yet visualise tables with multiple provenances")
+
+        return my_images
+
+    @override
+    def get_visualization(
+        self,
+        *,
+        doc: DoclingDocument,
+        **kwargs,
+    ) -> dict[Optional[int], Image]:
+        """Get visualization of the document as images by page."""
+        base_images = (
+            self.base_visualizer.get_visualization(doc=doc, **kwargs)
+            if self.base_visualizer
+            else None
+        )
+        return self._draw_doc_tables(
+            doc=doc,
+            images=base_images,
+        )

{docling_core-2.34.2.dist-info → docling_core-2.35.0.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: docling-core
-Version: 2.34.2
+Version: 2.35.0
 Summary: A python library to define and validate data types in Docling.
 Author-email: Cesar Berrospi Ramis <ceb@zurich.ibm.com>, Panos Vagenas <pva@zurich.ibm.com>, Michele Dolfi <dol@zurich.ibm.com>, Christoph Auer <cau@zurich.ibm.com>, Peter Staar <taa@zurich.ibm.com>
 Maintainer-email: Panos Vagenas <pva@zurich.ibm.com>, Michele Dolfi <dol@zurich.ibm.com>, Christoph Auer <cau@zurich.ibm.com>, Peter Staar <taa@zurich.ibm.com>, Cesar Berrospi Ramis <ceb@zurich.ibm.com>

{docling_core-2.34.2.dist-info → docling_core-2.35.0.dist-info}/RECORD

@@ -34,8 +34,9 @@ docling_core/transforms/serializer/html_styles.py,sha256=-jBwS4EU7yfKoz0GSoxhwx9
 docling_core/transforms/serializer/markdown.py,sha256=wfMNrjA4wMehWLCejAhEN1eQPRixUO1SyL6ojkKkzZY,20614
 docling_core/transforms/visualizer/__init__.py,sha256=gUfF25yiJ_KO46ZIUNqZQOZGy2PLx6gnnr6AZYxKHXI,35
 docling_core/transforms/visualizer/base.py,sha256=aEF7b3rHq6DVdX8zDYEPoq55BHDYe4Hh_97lBdcW4lY,555
-docling_core/transforms/visualizer/layout_visualizer.py,sha256=ulXxWGIl69-HMKDPFk_XKgNCgQeDNc969PVt_X0-drA,7823
+docling_core/transforms/visualizer/layout_visualizer.py,sha256=N3SA9sMkg2bEZ_2r52FpwRXcI3EJ2M5P9LYK4Az4jqQ,7968
 docling_core/transforms/visualizer/reading_order_visualizer.py,sha256=-ej5uLriNtr9C7YBHXMg8sZfB9Uc8cSRr1bJ8FVjpY8,5320
+docling_core/transforms/visualizer/table_visualizer.py,sha256=XlLMSROyRW2UtAjKTltcESSs_rdQNKjO3QvO7ET7uc0,4275
 docling_core/types/__init__.py,sha256=MVRSgsk5focwGyAplh_TRR3dEecIXpd98g_u3zZ5HXo,260
 docling_core/types/base.py,sha256=PusJskRVL19y-hq0BgXr5e8--QEqSqLnFNJ8UbOqW88,8318
 docling_core/types/doc/__init__.py,sha256=bysJn2iwjAHwThSWDPXEdVUUij7p_ax12_nx2_0CMdg,653
@@ -73,9 +74,9 @@ docling_core/utils/generate_jsonschema.py,sha256=uNX1O5XnjyB5nA66XqZXTt3YbGuR2ty
 docling_core/utils/legacy.py,sha256=DrI3QGoL755ZCIoKHF74-pTWm8R0zfFo2C2vB5dT2aY,24463
 docling_core/utils/validate.py,sha256=aQ11UbFyl8iD_N7yTTZmm_VVeXz8KcCyn3GLXgkfYRM,2049
 docling_core/utils/validators.py,sha256=azcrndLzhNkTWnbFSu9shJ5D3j_znnLrIFA5R8hzmGU,2798
-docling_core-2.34.2.dist-info/licenses/LICENSE,sha256=2M9-6EoQ1sxFztTOkXGAtwUDJvnWaAHdB9BYWVwGkIw,1087
-docling_core-2.34.2.dist-info/METADATA,sha256=SZQWH-WCvFQCz39nJqxUjv_t14TaI_WDThYzOki7EHc,6453
-docling_core-2.34.2.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-docling_core-2.34.2.dist-info/entry_points.txt,sha256=ER4zROQWkFMHIrY-oqY5E4HeCcCIg8dLkNztYGxdb7c,59
-docling_core-2.34.2.dist-info/top_level.txt,sha256=O-tcXpGiurlud-1ZxMq1b-OmrfAVA4sajcgWU32RtfA,13
-docling_core-2.34.2.dist-info/RECORD,,
+docling_core-2.35.0.dist-info/licenses/LICENSE,sha256=2M9-6EoQ1sxFztTOkXGAtwUDJvnWaAHdB9BYWVwGkIw,1087
+docling_core-2.35.0.dist-info/METADATA,sha256=Gube58hbnoDQGoeGmaK-yrMulAuKHiw7lUeGKxzSDsc,6453
+docling_core-2.35.0.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+docling_core-2.35.0.dist-info/entry_points.txt,sha256=ER4zROQWkFMHIrY-oqY5E4HeCcCIg8dLkNztYGxdb7c,59
+docling_core-2.35.0.dist-info/top_level.txt,sha256=O-tcXpGiurlud-1ZxMq1b-OmrfAVA4sajcgWU32RtfA,13
+docling_core-2.35.0.dist-info/RECORD,,