docling-core 2.24.0__tar.gz → 2.25.0__tar.gz

{docling_core-2.24.0 → docling_core-2.25.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: docling-core
-Version: 2.24.0
+Version: 2.25.0
 Summary: A python library to define and validate data types in Docling.
 Home-page: https://github.com/docling-project
 License: MIT

{docling_core-2.24.0 → docling_core-2.25.0}/docling_core/types/doc/base.py

@@ -182,7 +182,10 @@ class BoundingBox(BaseModel):
     ) -> float:
         """intersection_over_self."""
         intersection_area = self.intersection_area_with(other=other)
-        return intersection_area / self.area()
+        if self.area() > 0:
+            return intersection_area / self.area()
+        else:
+            return 0.0
 
     def to_bottom_left_origin(self, page_height: float) -> "BoundingBox":
         """to_bottom_left_origin.

{docling_core-2.24.0 → docling_core-2.25.0}/docling_core/types/doc/document.py

@@ -542,25 +542,32 @@ class DocTagsDocument(BaseModel):
 
     @classmethod
     def from_doctags_and_image_pairs(
-        cls, doctags: List[Union[Path, str]], images: List[Union[Path, PILImage.Image]]
+        cls,
+        doctags: typing.Sequence[Union[Path, str]],
+        images: Optional[List[Union[Path, PILImage.Image]]],
     ):
         """from_doctags_and_image_pairs."""
-        if len(doctags) != len(images):
+        if images is not None and len(doctags) != len(images):
             raise ValueError("Number of page doctags must be equal to page images!")
         doctags_doc = cls()
 
         pages = []
-        for dt, img in zip(doctags, images):
+
+        for ix, dt in enumerate(doctags):
             if isinstance(dt, Path):
                 with dt.open("r") as fp:
                     dt = fp.read()
             elif isinstance(dt, str):
                 pass
 
-            if isinstance(img, Path):
-                img = PILImage.open(img)
-            elif isinstance(dt, PILImage.Image):
-                pass
+            img = None
+            if images is not None:
+                img = images[ix]
+
+                if isinstance(img, Path):
+                    img = PILImage.open(img)
+                elif isinstance(img, PILImage.Image):
+                    pass
 
             page = DocTagsPage(tokens=dt, image=img)
             pages.append(page)
@@ -568,6 +575,25 @@ class DocTagsDocument(BaseModel):
         doctags_doc.pages = pages
         return doctags_doc
 
+    @classmethod
+    def from_multipage_doctags_and_images(
+        cls,
+        doctags: Union[Path, str],
+        images: Optional[List[Union[Path, PILImage.Image]]],
+    ):
+        """From doctags with `<page_break>` and corresponding list of page images."""
+        if isinstance(doctags, Path):
+            with doctags.open("r") as fp:
+                doctags = fp.read()
+        dt_list = (
+            doctags.removeprefix(f"<{DocumentToken.DOCUMENT.value}>")
+            .removesuffix(f"</{DocumentToken.DOCUMENT.value}>")
+            .split(f"<{DocumentToken.PAGE_BREAK.value}>")
+        )
+        dt_list = [el.strip() for el in dt_list]
+
+        return cls.from_doctags_and_image_pairs(dt_list, images)
+
 
 class ProvenanceItem(BaseModel):
     """ProvenanceItem."""
@@ -722,7 +748,9 @@ class TextItem(DocItem):
     text: str  # sanitized representation
 
     formatting: Optional[Formatting] = None
-    hyperlink: Optional[Union[AnyUrl, Path]] = None
+    hyperlink: Optional[Union[AnyUrl, Path]] = Field(
+        union_mode="left_to_right", default=None
+    )
 
     @deprecated("Use export_to_doctags() instead.")
     def export_to_document_tokens(self, *args, **kwargs):

{docling_core-2.24.0 → docling_core-2.25.0}/docling_core/types/doc/page.py

@@ -1,5 +1,6 @@
 """Datastructures for PaginatedDocument."""
 
+import copy
 import json
 import logging
 import math
@@ -530,10 +531,16 @@ class SegmentedPdfPage(SegmentedPage):
         """
         cells = []
         for page_cell in self.iterate_cells(cell_unit):
-            cell_bbox = page_cell.to_bounding_box()
+            pc = copy.deepcopy(page_cell)
+            # Bring cell_bbox coord origin to the same as input bbox.coord_origin:
+            if page_cell.rect.coord_origin != bbox.coord_origin:
+                if bbox.coord_origin == CoordOrigin.TOPLEFT:
+                    pc.rect = pc.rect.to_top_left_origin(self.dimension.height)
+                elif bbox.coord_origin == CoordOrigin.BOTTOMLEFT:
+                    pc.rect = pc.rect.to_bottom_left_origin(self.dimension.height)
+            cell_bbox = pc.to_bounding_box()
             if cell_bbox.intersection_over_self(bbox) > ios:
-                cells.append(page_cell)
-
+                cells.append(pc)
         return cells
 
     def export_to_dict(self) -> Dict:

{docling_core-2.24.0 → docling_core-2.25.0}/pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "docling-core"
-version = "2.24.0"
+version = "2.25.0"
 description = "A python library to define and validate data types in Docling."
 license = "MIT"
 authors = [