docling-core 2.17.0__tar.gz → 2.17.1__tar.gz

{docling_core-2.17.0 → docling_core-2.17.1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: docling-core
-Version: 2.17.0
+Version: 2.17.1
 Summary: A python library to define and validate data types in Docling.
 Home-page: https://ds4sd.github.io/
 License: MIT

{docling_core-2.17.0 → docling_core-2.17.1}/docling_core/types/doc/document.py

@@ -5,6 +5,7 @@ import copy
 import hashlib
 import html
 import json
+import logging
 import mimetypes
 import os
 import re
@@ -20,6 +21,7 @@ from xml.etree.cElementTree import SubElement, tostring
 from xml.sax.saxutils import unescape
 
 import latex2mathml.converter
+import latex2mathml.exceptions
 import pandas as pd
 import yaml
 from PIL import Image as PILImage
@@ -44,6 +46,8 @@ from docling_core.types.doc.labels import CodeLanguageLabel, DocItemLabel, Group
 from docling_core.types.doc.tokens import DocumentToken, TableToken
 from docling_core.types.doc.utils import relative_path
 
+_logger = logging.getLogger(__name__)
+
 Uint64 = typing.Annotated[int, Field(ge=0, le=(2**64 - 1))]
 LevelNumber = typing.Annotated[int, Field(ge=1, le=100)]
 CURRENT_VERSION: Final = "1.0.0"
@@ -2487,34 +2491,47 @@ class DoclingDocument(BaseModel):
                 )
                 text = ""
 
-                # If the formula is not processed correcty, use its image
-                if (
-                    item.text == ""
-                    and item.orig != ""
-                    and image_mode == ImageRefMode.EMBEDDED
-                    and len(item.prov) > 0
-                ):
+                def _image_fallback(item: TextItem):
                     item_image = item.get_image(doc=self)
                     if item_image is not None:
                         img_ref = ImageRef.from_pil(item_image, dpi=72)
-                        text = (
+                        return (
                             "<figure>"
                             f'<img src="{img_ref.uri}" alt="{item.orig}" />'
                             "</figure>"
                         )
 
+                # If the formula is not processed correcty, use its image
+                if (
+                    item.text == ""
+                    and item.orig != ""
+                    and image_mode == ImageRefMode.EMBEDDED
+                    and len(item.prov) > 0
+                ):
+                    text = _image_fallback(item)
+
                 # Building a math equation in MathML format
                 # ref https://www.w3.org/TR/wai-aria-1.1/#math
                 elif formula_to_mathml:
-                    mathml_element = latex2mathml.converter.convert_to_element(
-                        math_formula, display="block"
-                    )
-                    annotation = SubElement(
-                        mathml_element, "annotation", dict(encoding="TeX")
-                    )
-                    annotation.text = math_formula
-                    mathml = unescape(tostring(mathml_element, encoding="unicode"))
-                    text = f"<div>{mathml}</div>"
+                    try:
+                        mathml_element = latex2mathml.converter.convert_to_element(
+                            math_formula, display="block"
+                        )
+                        annotation = SubElement(
+                            mathml_element, "annotation", dict(encoding="TeX")
+                        )
+                        annotation.text = math_formula
+                        mathml = unescape(tostring(mathml_element, encoding="unicode"))
+                        text = f"<div>{mathml}</div>"
+                    except Exception as err:
+                        _logger.warning(
+                            "Malformed formula cannot be rendered. "
+                            f"Error {err.__class__.__name__}, formula={math_formula}"
+                        )
+                        if image_mode == ImageRefMode.EMBEDDED and len(item.prov) > 0:
+                            text = _image_fallback(item)
+                        else:
+                            text = f"<pre>{math_formula}</pre>"
 
                 elif math_formula != "":
                     text = f"<pre>{math_formula}</pre>"

{docling_core-2.17.0 → docling_core-2.17.1}/pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "docling-core"
-version = "2.17.0"
+version = "2.17.1"
 description = "A python library to define and validate data types in Docling."
 license = "MIT"
 authors = [