docling-core 1.1.1__tar.gz → 1.1.3__tar.gz

{docling_core-1.1.1 → docling_core-1.1.3}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: docling-core
-Version: 1.1.1
+Version: 1.1.3
 Summary: A python library to define and validate data types in Docling.
 Home-page: https://ds4sd.github.io/
 License: MIT

{docling_core-1.1.1 → docling_core-1.1.3}/docling_core/types/doc/base.py

@@ -6,7 +6,7 @@
 """Define common models across CCS objects."""
 from typing import Annotated, Literal, Optional, Union
 
-from pydantic import BaseModel, Field, StrictStr
+from pydantic import BaseModel, Field, PositiveInt, StrictStr
 
 from docling_core.search.mapping import es_field
 from docling_core.utils.alias import AliasModel
@@ -39,7 +39,7 @@ class S3Resource(BaseModel):
 
     mime: str
     path: str
-    page: Optional[int] = None
+    page: Optional[PositiveInt] = None
 
 
 class S3Data(AliasModel):
@@ -68,7 +68,7 @@ class Prov(AliasModel):
     """Provenance."""
 
     bbox: BoundingBox
-    page: int
+    page: PositiveInt
     span: Span
     ref_s3_data: Optional[StrictStr] = Field(
         default=None, alias="__ref_s3_data", json_schema_extra=es_field(suppress=True)
@@ -96,7 +96,7 @@ class PageDimensions(BaseModel):
     """Page dimensions."""
 
     height: float
-    page: int
+    page: PositiveInt
     width: float
 
 
@@ -128,37 +128,28 @@ class GlmTableCell(TableCell):
     )
 
 
-class Table(AliasModel):
-    """Table."""
+class BaseCell(AliasModel):
+    """Base cell."""
 
-    num_cols: int = Field(alias="#-cols")
-    num_rows: int = Field(alias="#-rows")
     bounding_box: Optional[BoundingBoxContainer] = Field(
         default=None, alias="bounding-box", json_schema_extra=es_field(suppress=True)
     )
-    data: Optional[list[list[Union[GlmTableCell, TableCell]]]] = None
-    model: Optional[str] = None
     prov: Optional[list[Prov]] = None
     text: Optional[str] = Field(
         default=None, json_schema_extra=es_field(term_vector="with_positions_offsets")
     )
     obj_type: str = Field(
-        alias="type",
-        json_schema_extra=es_field(type="keyword", ignore_above=8191),
+        alias="type", json_schema_extra=es_field(type="keyword", ignore_above=8191)
     )
 
 
-class BaseCell(AliasModel):
-    """Base cell."""
+class Table(BaseCell):
+    """Table."""
 
-    bounding_box: Optional[BoundingBoxContainer] = Field(
-        default=None, alias="bounding-box", json_schema_extra=es_field(suppress=True)
-    )
-    prov: Optional[list[Prov]] = None
-    text: Optional[str] = None
-    obj_type: str = Field(
-        alias="type", json_schema_extra=es_field(type="keyword", ignore_above=8191)
-    )
+    num_cols: int = Field(alias="#-cols")
+    num_rows: int = Field(alias="#-rows")
+    data: Optional[list[list[Union[GlmTableCell, TableCell]]]] = None
+    model: Optional[str] = None
 
 
 class BaseText(AliasModel):
@@ -196,4 +187,4 @@ class PageReference(BaseModel):
 
     hash: str = Field(json_schema_extra=es_field(type="keyword", ignore_above=8191))
     model: str = Field(json_schema_extra=es_field(suppress=True))
-    page: int = Field(json_schema_extra=es_field(type="short"))
+    page: PositiveInt = Field(json_schema_extra=es_field(type="short"))

{docling_core-1.1.1 → docling_core-1.1.3}/docling_core/types/doc/document.py

@@ -393,14 +393,28 @@ class ExportedCCSDocument(
 
         return data
 
-    def _resolve_ref(self, item: Ref) -> Optional[Table]:
-        """Return the resolved reference in case of table reference, otherwise None."""
-        result: Optional[Table] = None
+    def _resolve_ref(self, item: Ref) -> Optional[Union[BaseCell, BaseText]]:
+        """Return the resolved reference.
 
-        # NOTE: currently only resolves table refs & makes assumptions on ref parts
+        Resolved the Ref object within the document.
+        If the object is not found, None is returned.
+        """
+        result: Optional[Union[BaseCell, BaseText]] = None
+
+        # NOTE: currently only resolves refs explicitely, such that we can make
+        # assumptions on ref parts
         if item.obj_type == "table" and self.tables:
             parts = item.ref.split("/")
             result = self.tables[int(parts[2])]
+        elif item.obj_type == "figure" and self.figures:
+            parts = item.ref.split("/")
+            result = self.figures[int(parts[2])]
+        elif item.obj_type == "equation" and self.equations:
+            parts = item.ref.split("/")
+            result = self.equations[int(parts[2])]
+        elif item.obj_type == "footnote" and self.footnotes:
+            parts = item.ref.split("/")
+            result = self.footnotes[int(parts[2])]
 
         return result
 

{docling_core-1.1.1 → docling_core-1.1.3}/pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "docling-core"
-version = "1.1.1"
+version = "1.1.3"
 description = "A python library to define and validate data types in Docling."
 license = "MIT"
 authors = [