docling-core 0.2.0__tar.gz → 1.1.0__tar.gz

{docling_core-0.2.0 → docling_core-1.1.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: docling-core
-Version: 0.2.0
+Version: 1.1.0
 Summary: A python library to define and validate data types in Docling.
 Home-page: https://ds4sd.github.io/
 License: MIT
@@ -31,6 +31,7 @@ Requires-Dist: jsonschema (>=4.16.0,<5.0.0)
 Requires-Dist: poetry (>=1.8.3,<2.0.0)
 Requires-Dist: pydantic (>=2.6.0,<3.0.0)
 Requires-Dist: pyproject-toml (>=0.0.10,<0.0.11)
+Requires-Dist: tabulate (>=0.9.0,<0.10.0)
 Project-URL: Repository, https://github.com/DS4SD/docling-core
 Description-Content-Type: text/markdown
 

{docling_core-0.2.0 → docling_core-1.1.0}/docling_core/search/package.py

@@ -5,6 +5,7 @@
 
 """Models and methods to define a package model."""
 
+import importlib.metadata
 import re
 from typing import Final
 
@@ -27,7 +28,7 @@ class Package(BaseModel, extra="forbid"):
 
     name: StrictStr
     version: Annotated[str, StringConstraints(strict=True, pattern=VERSION_PATTERN)] = (
-        "0.1.0"
+        importlib.metadata.version("docling-core")
     )
 
     def __hash__(self):

{docling_core-0.2.0 → docling_core-1.1.0}/docling_core/types/base.py

@@ -39,6 +39,10 @@ PredicateKeyNameT = TypeVar("PredicateKeyNameT", bound=str)
 PredicateKeyTypeT = TypeVar("PredicateKeyTypeT", bound=str)
 ProvenanceTypeT = TypeVar("ProvenanceTypeT", bound=str)
 CollectionNameTypeT = TypeVar("CollectionNameTypeT", bound=str)
+Coordinates = Annotated[
+    list[float],
+    Field(min_length=2, max_length=2, json_schema_extra=es_field(type="geo_point")),
+]
 T = TypeVar("T", bound=Hashable)
 
 UniqueList = Annotated[
@@ -61,7 +65,7 @@ ACQUISITION_TYPE = Literal[
 
 
 class Identifier(AliasModel, Generic[IdentifierTypeT], extra="forbid"):
-    """Unique identifier of a Deep Search data object."""
+    """Unique identifier of a Docling data object."""
 
     type_: IdentifierTypeT = Field(
         alias="type",
@@ -81,7 +85,7 @@ class Identifier(AliasModel, Generic[IdentifierTypeT], extra="forbid"):
         alias="_name",
         title="_Name",
         description=(
-            "A unique identifier of the data object across Deep Search, consisting of "
+            "A unique identifier of the data object across Docling, consisting of "
             "the concatenation of type and value in lower case, separated by hash "
             "(#)."
         ),
@@ -118,7 +122,7 @@ class Log(AliasModel, extra="forbid"):
         json_schema_extra=es_field(type="keyword", ignore_above=8191),
     )
     agent: StrictStr = Field(
-        description="The Deep Search agent that performed the task, e.g., CCS or CXS.",
+        description="The Docling agent that performed the task, e.g., CCS or CXS.",
         json_schema_extra=es_field(type="keyword", ignore_above=8191),
     )
     type_: StrictStr = Field(
@@ -138,7 +142,7 @@ class Log(AliasModel, extra="forbid"):
 
 
 class FileInfoObject(AliasModel):
-    """Filing information for any data object to be stored in a Deep Search database."""
+    """Filing information for any data object to be stored in a Docling database."""
 
     filename: StrictStr = Field(
         description="The name of a persistent object that created this data object",
@@ -156,7 +160,7 @@ class FileInfoObject(AliasModel):
     document_hash: StrictStr = Field(
         description=(
             "A unique identifier of this data object within a collection of a "
-            "Deep Search database"
+            "Docling database"
         ),
         alias="document-hash",
         json_schema_extra=es_field(type="keyword", ignore_above=8191),
@@ -164,7 +168,7 @@ class FileInfoObject(AliasModel):
 
 
 class CollectionTypeEnum(str, Enum):
-    """Enumeration of valid Deep Search collection types."""
+    """Enumeration of valid Docling collection types."""
 
     generic = "Generic"
     document = "Document"

{docling_core-0.2.0 → docling_core-1.1.0}/docling_core/types/doc/document.py

@@ -3,7 +3,7 @@
 # SPDX-License-Identifier: MIT
 #
 
-"""Models for the Deep Search Document data type."""
+"""Models for the Docling Document data type."""
 
 from datetime import datetime
 from typing import Generic, Optional, Union
@@ -16,6 +16,7 @@ from pydantic import (
     StrictStr,
     model_validator,
 )
+from tabulate import tabulate
 
 from docling_core.search.mapping import es_field
 from docling_core.types.base import (
@@ -352,7 +353,7 @@ class ExportedCCSDocument(
         CollectionNameTypeT,
     ],
 ):
-    """Document model for Deep Search."""
+    """Document model for Docling."""
 
     obj_type: StrictStr = Field(
         "pdf-document",
@@ -391,3 +392,111 @@ class ExportedCCSDocument(
                     item["$ref"] = ref
 
         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
+
+        # NOTE: currently only resolves table refs & makes assumptions on ref parts
+        if item.obj_type == "table" and self.tables:
+            parts = item.ref.split("/")
+            result = self.tables[int(parts[2])]
+
+        return result
+
+    def export_to_markdown(
+        self,
+        delim: str = "\n\n",
+        main_text_start: int = 0,
+        main_text_stop: Optional[int] = None,
+    ) -> str:
+        r"""Serialize to Markdown.
+
+        Operates on a slice of the document's main_text as defined through arguments
+        main_text_start and main_text_stop; defaulting to the whole main_text.
+
+        Args:
+            delim (str, optional): Delimiter to use when concatenating the various
+                Markdown parts. Defaults to "\n\n".
+            main_text_start (int, optional): Main-text slicing start index (inclusive).
+                Defaults to 0.
+            main_text_end (Optional[int], optional): Main-text slicing stop index
+                (exclusive). Defaults to None.
+
+        Returns:
+            str: The exported Markdown representation.
+        """
+        has_title = False
+        prev_text = ""
+        md_texts: list[str] = []
+
+        if self.main_text is not None:
+            for orig_item in self.main_text[main_text_start:main_text_stop]:
+                markdown_text = ""
+
+                item = (
+                    self._resolve_ref(orig_item)
+                    if isinstance(orig_item, Ref)
+                    else orig_item
+                )
+                if item is None:
+                    continue
+
+                item_type = item.obj_type
+                if isinstance(item, BaseText) and item_type in {
+                    "title",
+                    "subtitle-level-1",
+                    "paragraph",
+                    "caption",
+                }:
+                    text = item.text
+
+                    # ignore repeated text
+                    if prev_text == text:
+                        continue
+                    else:
+                        prev_text = text
+
+                    # first title match
+                    if item_type == "title" and not has_title:
+                        markdown_text = f"# {text}"
+                        has_title = True
+
+                    # secondary titles
+                    elif item_type in {"title", "subtitle-level-1"} or (
+                        has_title and item_type == "title"
+                    ):
+                        markdown_text = f"## {text}"
+
+                    # normal text
+                    else:
+                        markdown_text = text
+
+                elif isinstance(item, Table) and item.data:
+                    table = []
+                    for row in item.data:
+                        tmp = []
+                        for col in row:
+                            tmp.append(col.text)
+                        table.append(tmp)
+
+                    if len(table) > 1 and len(table[0]) > 0:
+                        try:
+                            md_table = tabulate(
+                                table[1:], headers=table[0], tablefmt="github"
+                            )
+                        except ValueError:
+                            md_table = tabulate(
+                                table[1:],
+                                headers=table[0],
+                                tablefmt="github",
+                                disable_numparse=True,
+                            )
+
+                        markdown_text = md_table
+
+                if markdown_text:
+                    md_texts.append(markdown_text)
+
+        result = delim.join(md_texts)
+        return result

{docling_core-0.2.0 → docling_core-1.1.0}/docling_core/types/rec/attribute.py

@@ -6,7 +6,7 @@
 """Define the model Attribute."""
 from typing import Generic, Optional
 
-from pydantic import BaseModel, Field
+from pydantic import Field
 from typing_extensions import Annotated
 
 from docling_core.search.mapping import es_field
@@ -16,23 +16,20 @@ from docling_core.types.base import (
     PredicateKeyTypeT,
     PredicateValueTypeT,
     ProvenanceTypeT,
-    SubjectNameTypeT,
-    SubjectTypeT,
 )
 from docling_core.types.rec.base import ProvenanceItem
 from docling_core.types.rec.predicate import Predicate
+from docling_core.utils.alias import AliasModel
 
 
 class Attribute(
-    BaseModel,
+    AliasModel,
     Generic[
         IdentifierTypeT,
         PredicateValueTypeT,
         PredicateKeyNameT,
         PredicateKeyTypeT,
         ProvenanceTypeT,
-        SubjectTypeT,
-        SubjectNameTypeT,
     ],
     extra="forbid",
 ):

{docling_core-0.2.0 → docling_core-1.1.0}/docling_core/types/rec/predicate.py

@@ -5,7 +5,7 @@
 
 """Define the model Predicate."""
 from datetime import datetime
-from typing import Annotated, Generic, Optional, TypeVar
+from typing import Annotated, Generic, Optional
 
 from pydantic import (
     BaseModel,
@@ -17,16 +17,14 @@ from pydantic import (
 )
 
 from docling_core.search.mapping import es_field
+from docling_core.types.base import (
+    Coordinates,
+    PredicateKeyNameT,
+    PredicateKeyTypeT,
+    PredicateValueTypeT,
+)
 from docling_core.utils.alias import AliasModel
 
-PredicateValueTypeT = TypeVar("PredicateValueTypeT", bound=str)
-PredicateKeyNameT = TypeVar("PredicateKeyNameT", bound=str)
-PredicateKeyTypeT = TypeVar("PredicateKeyTypeT", bound=str)
-Coordinates = Annotated[
-    list[float],
-    Field(min_length=2, max_length=2, json_schema_extra=es_field(type="geo_point")),
-]
-
 
 class NumericalValue(BaseModel, extra="forbid"):
     """Model for numerical values."""
@@ -117,7 +115,7 @@ class PredicateValue(AliasModel, Generic[PredicateValueTypeT], extra="forbid"):
 
 
 class Predicate(
-    BaseModel,
+    AliasModel,
     Generic[PredicateValueTypeT, PredicateKeyNameT, PredicateKeyTypeT],
     extra="forbid",
 ):

{docling_core-0.2.0 → docling_core-1.1.0}/docling_core/types/rec/record.py

@@ -80,8 +80,6 @@ class Record(
                 PredicateKeyNameT,
                 PredicateKeyTypeT,
                 ProvenanceTypeT,
-                SubjectTypeT,
-                SubjectNameTypeT,
             ]
         ]
     ] = None

{docling_core-0.2.0 → docling_core-1.1.0}/docling_core/types/rec/subject.py

@@ -19,6 +19,10 @@ from docling_core.types.doc.base import S3Reference
 from docling_core.utils.alias import AliasModel
 
 
+class SubjectNameIdentifier(Identifier[SubjectNameTypeT], Generic[SubjectNameTypeT]):
+    """Identifier of subject names.""" ""
+
+
 class Subject(
     AliasModel,
     Generic[IdentifierTypeT, SubjectTypeT, SubjectNameTypeT],
@@ -53,7 +57,7 @@ class Subject(
         ),
         json_schema_extra=es_field(type="keyword", ignore_above=8191),
     )
-    names: list[Identifier[SubjectNameTypeT]] = Field(
+    names: list[SubjectNameIdentifier[SubjectNameTypeT]] = Field(
         description=(
             "List of given names for this subject. They may not be unique across "
             "different subjects."

{docling_core-0.2.0 → docling_core-1.1.0}/docling_core/utils/ds_generate_docs.py

@@ -44,7 +44,7 @@ def _prepare_directory(folder: str, clean: bool = False) -> None:
 
 
 def generate_collection_jsonschema(folder: str):
-    """Generate the JSON schema of Deep Search collections and export them to a folder.
+    """Generate the JSON schema of Docling collections and export them to a folder.
 
     Args:
         folder: The name of the directory.
@@ -58,7 +58,7 @@ def generate_collection_jsonschema(folder: str):
 
 
 def generate_collection_html(folder: str):
-    """Generate HTML pages documenting the data model of Deep Search collections.
+    """Generate HTML pages documenting the data model of Docling collections.
 
     The JSON schemas files need to be in a folder and the generated HTML pages will be
     written in the same folder.
@@ -79,7 +79,7 @@ def generate_collection_html(folder: str):
 
 
 def generate_collection_markdown(folder: str):
-    """Generate Markdown pages documenting the data model of Deep Search collections.
+    """Generate Markdown pages documenting the data model of Docling collections.
 
     The JSON schemas files need to be in a folder and the generated markdown pages will
     be written in the same folder.
@@ -101,7 +101,7 @@ def generate_collection_markdown(folder: str):
 
 
 def main() -> None:
-    """Generate the JSON Schema of Deep Search collections and export documentation."""
+    """Generate the JSON Schema of Docling collections and export documentation."""
     argparser = argparse.ArgumentParser()
     argparser.add_argument(
         "directory",

{docling_core-0.2.0 → docling_core-1.1.0}/pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "docling-core"
-version = "0.2.0"
+version = "1.1.0"
 description = "A python library to define and validate data types in Docling."
 license = "MIT"
 authors = [
@@ -53,6 +53,7 @@ jsonref = "^1.1.0"
 json-schema-for-humans = "^1.0.0"
 poetry = "^1.8.3"
 pyproject-toml = "^0.0.10"
+tabulate = "^0.9.0"
 
 [tool.poetry.group.dev.dependencies]
 black = "^24.4.2"
@@ -67,6 +68,7 @@ flake8-docstrings = "^1.6.0"
 pep8-naming = "^0.13.2"
 jsondiff = "^2.0.0"
 types-setuptools = "^70.3.0"
+python-semantic-release = "^7.32.2"
 
 [tool.setuptools.packages.find]
 where = ["docling_core/resources/schemas"]
@@ -110,5 +112,18 @@ python_version = "3.9"
 plugins = ["pydantic.mypy"]
 
 [[tool.mypy.overrides]]
-module = ["jsonref.*", "jsonschema.*", "json_schema_for_humans.*"]
+module = ["jsondiff.*", "jsonref.*", "jsonschema.*", "json_schema_for_humans.*", "tabulate.*"]
 ignore_missing_imports = true
+
+[tool.semantic_release]
+# for default values check:
+# https://github.com/python-semantic-release/python-semantic-release/blob/v7.32.2/semantic_release/defaults.cfg
+
+version_source = "tag_only"
+branch = "main"
+
+# configure types which should trigger minor and patch version bumps respectively
+# (note that they must be a subset of the configured allowed types):
+parser_angular_allowed_types = "build,chore,ci,docs,feat,fix,perf,style,refactor,test"
+parser_angular_minor_types = "feat"
+parser_angular_patch_types = "fix,perf"