docling-core 2.2.2__py3-none-any.whl → 2.3.0__py3-none-any.whl

docling_core/types/doc/document.py

@@ -4,6 +4,7 @@ import base64
 import mimetypes
 import re
 import sys
+import textwrap
 import typing
 from io import BytesIO
 from typing import Any, Dict, Final, List, Literal, Optional, Tuple, Union
@@ -99,12 +100,166 @@ class PictureMiscData(BaseModel):
     content: Dict[str, Any]
 
 
+class ChartLine(BaseModel):
+    """Represents a line in a line chart.
+
+    Attributes:
+        label (str): The label for the line.
+        values (List[Tuple[float, float]]): A list of (x, y) coordinate pairs
+            representing the line's data points.
+    """
+
+    label: str
+    values: List[Tuple[float, float]]
+
+
+class ChartBar(BaseModel):
+    """Represents a bar in a bar chart.
+
+    Attributes:
+        label (str): The label for the bar.
+        values (float): The value associated with the bar.
+    """
+
+    label: str
+    values: float
+
+
+class ChartStackedBar(BaseModel):
+    """Represents a stacked bar in a stacked bar chart.
+
+    Attributes:
+        label (List[str]): The labels for the stacked bars. Multiple values are stored
+            in cases where the chart is "double stacked," meaning bars are stacked both
+            horizontally and vertically.
+        values (List[Tuple[str, int]]): A list of values representing different segments
+            of the stacked bar along with their label.
+    """
+
+    label: List[str]
+    values: List[Tuple[str, int]]
+
+
+class ChartSlice(BaseModel):
+    """Represents a slice in a pie chart.
+
+    Attributes:
+        label (str): The label for the slice.
+        value (float): The value represented by the slice.
+    """
+
+    label: str
+    value: float
+
+
+class ChartPoint(BaseModel):
+    """Represents a point in a scatter chart.
+
+    Attributes:
+        value (Tuple[float, float]): A (x, y) coordinate pair representing a point in a
+            chart.
+    """
+
+    value: Tuple[float, float]
+
+
+class PictureChartData(BaseModel):
+    """Base class for picture chart data.
+
+    Attributes:
+        title (str): The title of the chart.
+    """
+
+    title: str
+
+
+class PictureLineChartData(PictureChartData):
+    """Represents data of a line chart.
+
+    Attributes:
+        kind (Literal["line_chart_data"]): The type of the chart.
+        x_axis_label (str): The label for the x-axis.
+        y_axis_label (str): The label for the y-axis.
+        lines (List[ChartLine]): A list of lines in the chart.
+    """
+
+    kind: Literal["line_chart_data"] = "line_chart_data"
+    x_axis_label: str
+    y_axis_label: str
+    lines: List[ChartLine]
+
+
+class PictureBarChartData(PictureChartData):
+    """Represents data of a bar chart.
+
+    Attributes:
+        kind (Literal["bar_chart_data"]): The type of the chart.
+        x_axis_label (str): The label for the x-axis.
+        y_axis_label (str): The label for the y-axis.
+        bars (List[ChartBar]): A list of bars in the chart.
+    """
+
+    kind: Literal["bar_chart_data"] = "bar_chart_data"
+    x_axis_label: str
+    y_axis_label: str
+    bars: List[ChartBar]
+
+
+class PictureStackedBarChartData(PictureChartData):
+    """Represents data of a stacked bar chart.
+
+    Attributes:
+        kind (Literal["stacked_bar_chart_data"]): The type of the chart.
+        x_axis_label (str): The label for the x-axis.
+        y_axis_label (str): The label for the y-axis.
+        stacked_bars (List[ChartStackedBar]): A list of stacked bars in the chart.
+    """
+
+    kind: Literal["stacked_bar_chart_data"] = "stacked_bar_chart_data"
+    x_axis_label: str
+    y_axis_label: str
+    stacked_bars: List[ChartStackedBar]
+
+
+class PicturePieChartData(PictureChartData):
+    """Represents data of a pie chart.
+
+    Attributes:
+        kind (Literal["pie_chart_data"]): The type of the chart.
+        slices (List[ChartSlice]): A list of slices in the pie chart.
+    """
+
+    kind: Literal["pie_chart_data"] = "pie_chart_data"
+    slices: List[ChartSlice]
+
+
+class PictureScatterChartData(PictureChartData):
+    """Represents data of a scatter chart.
+
+    Attributes:
+        kind (Literal["scatter_chart_data"]): The type of the chart.
+        x_axis_label (str): The label for the x-axis.
+        y_axis_label (str): The label for the y-axis.
+        points (List[ChartPoint]): A list of points in the scatter chart.
+    """
+
+    kind: Literal["scatter_chart_data"] = "scatter_chart_data"
+    x_axis_label: str
+    y_axis_label: str
+    points: List[ChartPoint]
+
+
 PictureDataType = Annotated[
     Union[
         PictureClassificationData,
         PictureDescriptionData,
         PictureMoleculeData,
         PictureMiscData,
+        PictureLineChartData,
+        PictureBarChartData,
+        PictureStackedBarChartData,
+        PicturePieChartData,
+        PictureScatterChartData,
     ],
     Field(discriminator="kind"),
 ]
@@ -1125,6 +1280,7 @@ class DoclingDocument(BaseModel):
         image_placeholder: str = "<!-- image -->",
         image_mode: ImageRefMode = ImageRefMode.PLACEHOLDER,
         indent: int = 4,
+        text_width: int = -1,
     ) -> str:
         r"""Serialize to Markdown.
 
@@ -1207,8 +1363,8 @@ class DoclingDocument(BaseModel):
             elif isinstance(item, TextItem) and item.label in [DocItemLabel.TITLE]:
                 in_list = False
                 marker = "" if strict_text else "#"
-                text = f"{marker} {item.text}\n"
-                mdtexts.append(text.strip())
+                text = f"{marker} {item.text}"
+                mdtexts.append(text.strip() + "\n")
 
             elif (
                 isinstance(item, TextItem)
@@ -1251,7 +1407,10 @@ class DoclingDocument(BaseModel):
 
             elif isinstance(item, TextItem) and item.label in labels:
                 in_list = False
-                if len(item.text):
+                if len(item.text) and text_width > 0:
+                    wrapped_text = textwrap.fill(text, width=text_width)
+                    mdtexts.append(wrapped_text + "\n")
+                elif len(item.text):
                     text = f"{item.text}\n"
                     mdtexts.append(text)
 

docling_core/types/doc/labels.py

@@ -29,6 +29,10 @@ class DocItemLabel(str, Enum):
     PARAGRAPH = "paragraph"  # explicitly a paragraph and not arbitrary text
     REFERENCE = "reference"
 
+    def __str__(self):
+        """Get string value."""
+        return str(self.value)
+
 
 class GroupLabel(str, Enum):
     """GroupLabel."""
@@ -43,6 +47,10 @@ class GroupLabel(str, Enum):
     SHEET = "sheet"
     SLIDE = "slide"
 
+    def __str__(self):
+        """Get string value."""
+        return str(self.value)
+
 
 class TableCellLabel(str, Enum):
     """TableCellLabel."""
@@ -51,3 +59,7 @@ class TableCellLabel(str, Enum):
     ROW_HEADER = "row_header"
     ROW_SECTION = "row_section"
     BODY = "body"
+
+    def __str__(self):
+        """Get string value."""
+        return str(self.value)

{docling_core-2.2.2.dist-info → docling_core-2.3.0.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: docling-core
-Version: 2.2.2
+Version: 2.3.0
 Summary: A python library to define and validate data types in Docling.
 Home-page: https://ds4sd.github.io/
 License: MIT

{docling_core-2.2.2.dist-info → docling_core-2.3.0.dist-info}/RECORD

@@ -21,8 +21,8 @@ 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=zvx631U_yQCcJam83hNdDanXEYnO3eN-CCw9vDr6S-I,4442
-docling_core/types/doc/document.py,sha256=SrOXpO6iCIYFkhWW-pksd4C4PeJ2jubKI5m34K_lTac,51902
-docling_core/types/doc/labels.py,sha256=mzmSd072A-qW3IThswHxwIHV8IoyTCbHHlNOrisinRA,1335
+docling_core/types/doc/document.py,sha256=XF43-v9oflV-E5r2k2quoKvq8qBp5mAB_VunshY9b10,56356
+docling_core/types/doc/labels.py,sha256=A8vWP82VAeXO1rlCO0oDKo_Hb8uDeQe0myOTY3P03hk,1596
 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/legacy_doc/__init__.py,sha256=Pzj_8rft6SJTVTCHgXRwHtuZjL6LK_6dcBWjikL9biY,125
@@ -49,8 +49,8 @@ docling_core/utils/generate_docs.py,sha256=BdKAoduWXOc7YMvcmlhjoJOFlUxij1ybxglj6
 docling_core/utils/generate_jsonschema.py,sha256=uNX1O5XnjyB5nA66XqZXTt3YbGuR2tyi_OhHepHYtZg,1654
 docling_core/utils/validate.py,sha256=3FmnxnKTDZC5J9OGxCL3U3DGRl0t0bBV1NcySXswdas,2031
 docling_core/utils/validators.py,sha256=azcrndLzhNkTWnbFSu9shJ5D3j_znnLrIFA5R8hzmGU,2798
-docling_core-2.2.2.dist-info/LICENSE,sha256=2M9-6EoQ1sxFztTOkXGAtwUDJvnWaAHdB9BYWVwGkIw,1087
-docling_core-2.2.2.dist-info/METADATA,sha256=680OW3ffU0R_QL5UdvoUASJBOyLj7YKlfeDcftd3Kkw,5432
-docling_core-2.2.2.dist-info/WHEEL,sha256=sP946D7jFCHeNz5Iq4fL4Lu-PrWrFsgfLXbbkciIZwg,88
-docling_core-2.2.2.dist-info/entry_points.txt,sha256=jIxlWv3tnO04irlZc0zfhqJIgz1bg9Hha4AkaLWSdUA,177
-docling_core-2.2.2.dist-info/RECORD,,
+docling_core-2.3.0.dist-info/LICENSE,sha256=2M9-6EoQ1sxFztTOkXGAtwUDJvnWaAHdB9BYWVwGkIw,1087
+docling_core-2.3.0.dist-info/METADATA,sha256=1cthKA2Ke9tujdMQzh4sNnBB1K90yFNQ9TCvD2MhKDI,5432
+docling_core-2.3.0.dist-info/WHEEL,sha256=sP946D7jFCHeNz5Iq4fL4Lu-PrWrFsgfLXbbkciIZwg,88
+docling_core-2.3.0.dist-info/entry_points.txt,sha256=jIxlWv3tnO04irlZc0zfhqJIgz1bg9Hha4AkaLWSdUA,177
+docling_core-2.3.0.dist-info/RECORD,,