docling-core 1.0.0__py3-none-any.whl → 1.1.0__py3-none-any.whl

docling_core/types/doc/document.py

@@ -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 (
@@ -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-1.0.0.dist-info → docling_core-1.1.0.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: docling-core
-Version: 1.0.0
+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-1.0.0.dist-info → docling_core-1.1.0.dist-info}/RECORD

@@ -20,7 +20,7 @@ docling_core/types/doc/base.py,sha256=-j4vVs3JZuaUjm0fHIkLU9TD_4IZXQuGouLrddEAwP
 docling_core/types/doc/doc_ann.py,sha256=8pV2efUglw19jxl4_oqB__mSxjWvtGIcllyCdqA-b2s,1196
 docling_core/types/doc/doc_ocr.py,sha256=6PC0C-OczF-MyfgRxEI1xs3PWgNOzi7i2yEQbTqZz0I,1387
 docling_core/types/doc/doc_raw.py,sha256=Y69G6IiauNDaoT-5el4xo1ypWpnBJQ75akGGkCMTZSc,3888
-docling_core/types/doc/document.py,sha256=cMduCiFkPVCmXQehvNkXqXtDiXJJtB72o7_LZXz_S6I,12549
+docling_core/types/doc/document.py,sha256=8L7F53Nr6U-z93zNJDit6nXtjiddysEmJ8KuWFO_iOA,16436
 docling_core/types/gen/__init__.py,sha256=C6TuCfvpSnSL5XDOFMcYHUY2-i08vvfOGRcdu6Af0pI,124
 docling_core/types/gen/generic.py,sha256=l4CZ4_Lb8ONG36WNJWbKX5hGKvTh_yU-hXp5hsm7uVU,844
 docling_core/types/nlp/__init__.py,sha256=hGcztAeVK7xkRBqRRvc4zbY4PGeJ0r0QrEsetnSx9nI,119
@@ -39,8 +39,8 @@ docling_core/utils/ds_generate_docs.py,sha256=0xGBagdC_PGjyeHXYZo90VnVrSTMZgHb0S
 docling_core/utils/ds_generate_jsonschema.py,sha256=EhNQutqWJFWuN-yl9UUPFZ7DJTvGqg54qBIvUMHTHdA,1647
 docling_core/utils/validate.py,sha256=3FmnxnKTDZC5J9OGxCL3U3DGRl0t0bBV1NcySXswdas,2031
 docling_core/utils/validators.py,sha256=fBdyWX4PvFh7o_d25ZTs4iwmeo75QTbrxsvXv2kXkTg,2777
-docling_core-1.0.0.dist-info/LICENSE,sha256=2M9-6EoQ1sxFztTOkXGAtwUDJvnWaAHdB9BYWVwGkIw,1087
-docling_core-1.0.0.dist-info/METADATA,sha256=RPsZbjVEs0aIfMYDmK25CxR1b77iRCqbu8WbodN4l9g,5174
-docling_core-1.0.0.dist-info/WHEEL,sha256=sP946D7jFCHeNz5Iq4fL4Lu-PrWrFsgfLXbbkciIZwg,88
-docling_core-1.0.0.dist-info/entry_points.txt,sha256=XHhtJEkdUuLxXSNxLdFIzx_siQ3z2UFQEKp-P8VYAE4,189
-docling_core-1.0.0.dist-info/RECORD,,
+docling_core-1.1.0.dist-info/LICENSE,sha256=2M9-6EoQ1sxFztTOkXGAtwUDJvnWaAHdB9BYWVwGkIw,1087
+docling_core-1.1.0.dist-info/METADATA,sha256=GMKoJSJxOTqMKGqLy6dLdZMkLkLia3tN-elzHJFbWRA,5216
+docling_core-1.1.0.dist-info/WHEEL,sha256=sP946D7jFCHeNz5Iq4fL4Lu-PrWrFsgfLXbbkciIZwg,88
+docling_core-1.1.0.dist-info/entry_points.txt,sha256=XHhtJEkdUuLxXSNxLdFIzx_siQ3z2UFQEKp-P8VYAE4,189
+docling_core-1.1.0.dist-info/RECORD,,