docling-core 2.12.0__py3-none-any.whl → 2.13.0__py3-none-any.whl

docling_core/types/doc/document.py

@@ -14,7 +14,7 @@ import warnings
 from io import BytesIO
 from pathlib import Path
 from typing import Any, Dict, Final, List, Literal, Optional, Tuple, Union
-from urllib.parse import unquote
+from urllib.parse import quote, unquote
 
 import pandas as pd
 import yaml
@@ -830,7 +830,7 @@ class PictureItem(FloatingItem):
             ):
                 return default_response
 
-            text = f"\n![Image]({str(self.image.uri)})\n"
+            text = f"\n![Image]({quote(str(self.image.uri))})\n"
             return text
 
         else:
@@ -884,7 +884,7 @@ class PictureItem(FloatingItem):
             ):
                 return default_response
 
-            img_text = f'<img src="{str(self.image.uri)}">'
+            img_text = f'<img src="{quote(str(self.image.uri))}">'
             return f"<figure>{caption_text}{img_text}</figure>"
 
         else:

docling_core/types/doc/labels.py

@@ -1,12 +1,12 @@
 """Models for the labels types."""
 
 from enum import Enum
+from typing import Tuple
 
 
 class DocItemLabel(str, Enum):
     """DocItemLabel."""
 
-    # DocLayNet v2
     CAPTION = "caption"
     FOOTNOTE = "footnote"
     FORMULA = "formula"
@@ -26,12 +26,34 @@ class DocItemLabel(str, Enum):
     KEY_VALUE_REGION = "key_value_region"
 
     # Additional labels for markup-based formats (e.g. HTML, Word)
-    PARAGRAPH = "paragraph"  # explicitly a paragraph and not arbitrary text
+    PARAGRAPH = "paragraph"
     REFERENCE = "reference"
 
-    def __str__(self):
-        """Get string value."""
-        return str(self.value)
+    @staticmethod
+    def get_color(label: "DocItemLabel") -> Tuple[int, int, int]:
+        """Return the RGB color associated with a given label."""
+        color_map = {
+            DocItemLabel.CAPTION: (255, 204, 153),
+            DocItemLabel.FOOTNOTE: (200, 200, 255),
+            DocItemLabel.FORMULA: (192, 192, 192),
+            DocItemLabel.LIST_ITEM: (153, 153, 255),
+            DocItemLabel.PAGE_FOOTER: (204, 255, 204),
+            DocItemLabel.PAGE_HEADER: (204, 255, 204),
+            DocItemLabel.PICTURE: (255, 204, 164),
+            DocItemLabel.SECTION_HEADER: (255, 153, 153),
+            DocItemLabel.TABLE: (255, 204, 204),
+            DocItemLabel.TEXT: (255, 255, 153),
+            DocItemLabel.TITLE: (255, 153, 153),
+            DocItemLabel.DOCUMENT_INDEX: (220, 220, 220),
+            DocItemLabel.CODE: (125, 125, 125),
+            DocItemLabel.CHECKBOX_SELECTED: (255, 182, 193),
+            DocItemLabel.CHECKBOX_UNSELECTED: (255, 182, 193),
+            DocItemLabel.FORM: (200, 255, 255),
+            DocItemLabel.KEY_VALUE_REGION: (183, 65, 14),
+            DocItemLabel.PARAGRAPH: (255, 255, 153),
+            DocItemLabel.REFERENCE: (176, 224, 230),
+        }
+        return color_map[label]
 
 
 class GroupLabel(str, Enum):

docling_core/types/legacy_doc/document.py

@@ -550,17 +550,18 @@ class ExportedCCSDocument(
 
                 elif (
                     isinstance(item, Table)
-                    and item.data
+                    and (item.data or item.text)
                     and item_type in main_text_labels
                 ):
 
                     md_table = ""
                     table = []
-                    for row in item.data:
-                        tmp = []
-                        for col in row:
-                            tmp.append(col.text)
-                        table.append(tmp)
+                    if item.data is not None:
+                        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:
@@ -579,7 +580,9 @@ class ExportedCCSDocument(
                     if item.text:
                         markdown_text = item.text
                     if not strict_text:
-                        markdown_text += "\n\n" + md_table
+                        markdown_text += (
+                            "\n\n" if len(markdown_text) > 0 else ""
+                        ) + md_table
 
                 elif isinstance(item, Figure) and item_type in main_text_labels:
 
@@ -587,7 +590,9 @@ class ExportedCCSDocument(
                     if item.text:
                         markdown_text = item.text
                     if not strict_text:
-                        markdown_text += f"\n{image_placeholder}"
+                        markdown_text += (
+                            "\n" if len(markdown_text) > 0 else ""
+                        ) + image_placeholder
 
                 if markdown_text:
                     md_texts.append(markdown_text)

{docling_core-2.12.0.dist-info → docling_core-2.13.0.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: docling-core
-Version: 2.12.0
+Version: 2.13.0
 Summary: A python library to define and validate data types in Docling.
 Home-page: https://ds4sd.github.io/
 License: MIT

{docling_core-2.12.0.dist-info → docling_core-2.13.0.dist-info}/RECORD

@@ -24,8 +24,8 @@ docling_core/types/__init__.py,sha256=MVRSgsk5focwGyAplh_TRR3dEecIXpd98g_u3zZ5HX
 docling_core/types/base.py,sha256=PusJskRVL19y-hq0BgXr5e8--QEqSqLnFNJ8UbOqW88,8318
 docling_core/types/doc/__init__.py,sha256=bEL4zKVOG7Wxm6xQrgF58mu-Teds9aSavuEAKVNhrTU,639
 docling_core/types/doc/base.py,sha256=_ttU8QI8wXDTQRUnN5n7L6D9wYFVLSAibxlFoMbgAsk,4557
-docling_core/types/doc/document.py,sha256=2W4wZunI0K_EOxNtY5jbKeyw7bYWKKNLiljxfN8anHc,91844
-docling_core/types/doc/labels.py,sha256=Pc5avKtGM2fv-w7mXinoxs9BkhktmFaJ6ACsgFiAAm4,1702
+docling_core/types/doc/document.py,sha256=ZHQBozH4-85p0YNBjzpTFURmTOenUrOwRyNr67shVcs,91865
+docling_core/types/doc/labels.py,sha256=yHVEuosoSNmywcTmnE5sSa7uXMtjgLt01OxaLSrPPxQ,2801
 docling_core/types/doc/tokens.py,sha256=uU_MYW_p7ypf7eYICFBvxdnVaPZ7CQnvZmbJ6oPrtEA,6134
 docling_core/types/doc/utils.py,sha256=YDOh_ZD1Y7OmCEDdCLJ_MO5K3HA67nc_acfhOK6WztU,1439
 docling_core/types/gen/__init__.py,sha256=C6TuCfvpSnSL5XDOFMcYHUY2-i08vvfOGRcdu6Af0pI,124
@@ -36,7 +36,7 @@ docling_core/types/legacy_doc/base.py,sha256=aBKBunw6M6nvEq4lqP1cfFWK3GpGa6PXwNQ
 docling_core/types/legacy_doc/doc_ann.py,sha256=CIQHW8yzu70bsMR9gtu7dqe4oz603Tq2eDDt9sh-tYo,1203
 docling_core/types/legacy_doc/doc_ocr.py,sha256=FfFqHAyMSbFt5cKeE7QLcxS0qUweBilBJoN9CH2TsQs,1394
 docling_core/types/legacy_doc/doc_raw.py,sha256=LrvQ9DhNjBRy98p_F9PUyHZeTGAxMKWqJzY4WJ7v-xs,3895
-docling_core/types/legacy_doc/document.py,sha256=A_cTYOjx6pNIICpOUm09YsfwPrIGDEZTKdetb2fx4PM,24273
+docling_core/types/legacy_doc/document.py,sha256=AW8AIBM19k-HtTmXPsFKagqd6gi9THJdB4RsPb1C5F0,24534
 docling_core/types/legacy_doc/tokens.py,sha256=uU_MYW_p7ypf7eYICFBvxdnVaPZ7CQnvZmbJ6oPrtEA,6134
 docling_core/types/nlp/__init__.py,sha256=hGcztAeVK7xkRBqRRvc4zbY4PGeJ0r0QrEsetnSx9nI,119
 docling_core/types/nlp/qa.py,sha256=TyZjubqkEoREv0YzmuLKlq4WW_TnJNj7BoBY1_r2a1E,2731
@@ -56,8 +56,8 @@ docling_core/utils/generate_jsonschema.py,sha256=uNX1O5XnjyB5nA66XqZXTt3YbGuR2ty
 docling_core/utils/legacy.py,sha256=xfp7U0JqjI60K3loWiNTk8w08_KfCUzTb2MNULBOIz4,24396
 docling_core/utils/validate.py,sha256=aQ11UbFyl8iD_N7yTTZmm_VVeXz8KcCyn3GLXgkfYRM,2049
 docling_core/utils/validators.py,sha256=azcrndLzhNkTWnbFSu9shJ5D3j_znnLrIFA5R8hzmGU,2798
-docling_core-2.12.0.dist-info/LICENSE,sha256=2M9-6EoQ1sxFztTOkXGAtwUDJvnWaAHdB9BYWVwGkIw,1087
-docling_core-2.12.0.dist-info/METADATA,sha256=Pvm-bnXtIVhPapkhV2z9ytH7CSskqmRtdoBroq-9gIU,5744
-docling_core-2.12.0.dist-info/WHEEL,sha256=sP946D7jFCHeNz5Iq4fL4Lu-PrWrFsgfLXbbkciIZwg,88
-docling_core-2.12.0.dist-info/entry_points.txt,sha256=oClcdb2L2RKx4jdqUykY16Kum_f0_whwWhGzIodyidc,216
-docling_core-2.12.0.dist-info/RECORD,,
+docling_core-2.13.0.dist-info/LICENSE,sha256=2M9-6EoQ1sxFztTOkXGAtwUDJvnWaAHdB9BYWVwGkIw,1087
+docling_core-2.13.0.dist-info/METADATA,sha256=p-9msa7jzKYkzczOr3T0kukBCghgsxs4zLQJJklH0Zc,5744
+docling_core-2.13.0.dist-info/WHEEL,sha256=sP946D7jFCHeNz5Iq4fL4Lu-PrWrFsgfLXbbkciIZwg,88
+docling_core-2.13.0.dist-info/entry_points.txt,sha256=oClcdb2L2RKx4jdqUykY16Kum_f0_whwWhGzIodyidc,216
+docling_core-2.13.0.dist-info/RECORD,,