From 5b03a1b99523e76a720a072dc133c5424f1aecf3 Mon Sep 17 00:00:00 2001
From: debaditya <debaditya2005hait@gmail.com>
Date: Sun, 26 Jul 2026 02:55:20 +0530
Subject: [PATCH] fix: skip image enrichment without pages

Signed-off-by: debaditya <debaditya2005hait@gmail.com>
---
 docling/models/base_model.py |  2 +-
 tests/test_backend_pptx.py   | 45 +++++++++++++++++++++++++++++++++++-
 2 files changed, 45 insertions(+), 2 deletions(-)

diff --git a/docling/models/base_model.py b/docling/models/base_model.py
index d48bc2d9c7..5001546092 100644
--- a/docling/models/base_model.py
+++ b/docling/models/base_model.py
@@ -191,7 +191,7 @@ def prepare_element(
             embedded_im = element.get_image(conv_res.document)
             if embedded_im is not None:
                 return ItemAndImageEnrichmentElement(item=element, image=embedded_im)
-            elif len(element.prov) == 0:
+            elif len(element.prov) == 0 or not conv_res.pages:
                 return None
 
         # Crop the image form the page
diff --git a/tests/test_backend_pptx.py b/tests/test_backend_pptx.py
index badd74603a..c3c79324a5 100644
--- a/tests/test_backend_pptx.py
+++ b/tests/test_backend_pptx.py
@@ -1,3 +1,4 @@
+from collections.abc import Iterable
 from pathlib import Path
 from types import SimpleNamespace
 
@@ -5,16 +6,21 @@
 from docling_core.types.doc import (
     ContentLayer,
     GroupItem,
+    NodeItem,
     PictureClassificationLabel,
+    PictureItem,
     TextItem,
 )
 
 from docling.backend.docx.drawingml.utils import get_libreoffice_cmd
 from docling.backend.mspowerpoint_backend import MsPowerpointDocumentBackend
 from docling.datamodel.backend_options import MsPowerpointBackendOptions
-from docling.datamodel.base_models import InputFormat
+from docling.datamodel.base_models import InputFormat, ItemAndImageEnrichmentElement
 from docling.datamodel.document import ConversionResult, DoclingDocument
+from docling.datamodel.pipeline_options import ConvertPipelineOptions
 from docling.document_converter import DocumentConverter, PowerpointFormatOption
+from docling.models.base_model import BaseItemAndImageEnrichmentModel
+from docling.pipeline.simple_pipeline import SimplePipeline
 
 from .test_data_gen_flag import GEN_TEST_DATA
 from .verify_utils import verify_document, verify_export
@@ -24,6 +30,27 @@
 CHART_PPTX = Path("./tests/data/pptx/sources/pptx_chart.pptx")
 
 
+class _PictureEnrichmentModel(BaseItemAndImageEnrichmentModel):
+    images_scale = 1.0
+
+    def is_processable(self, doc: DoclingDocument, element: NodeItem) -> bool:
+        return isinstance(element, PictureItem)
+
+    def __call__(
+        self,
+        doc: DoclingDocument,
+        element_batch: Iterable[ItemAndImageEnrichmentElement],
+    ) -> Iterable[NodeItem]:
+        for element in element_batch:
+            yield element.item
+
+
+class _ChartEnrichmentPipeline(SimplePipeline):
+    def __init__(self, pipeline_options: ConvertPipelineOptions) -> None:
+        super().__init__(pipeline_options)
+        self.enrichment_pipe = [_PictureEnrichmentModel()]
+
+
 @pytest.fixture(scope="module")
 def libreoffice_available() -> bool:
     """Return True when a working LibreOffice installation is detected."""
@@ -305,6 +332,22 @@ def test_chart_image_not_rendered_by_default():
     )
 
 
+def test_chart_enrichment_skips_image_when_pages_empty():
+    """Image enrichment skips native charts without an embedded or page image."""
+    format_options = {
+        InputFormat.PPTX: PowerpointFormatOption(pipeline_cls=_ChartEnrichmentPipeline)
+    }
+    converter = DocumentConverter(
+        allowed_formats=[InputFormat.PPTX], format_options=format_options
+    )
+
+    result = converter.convert(CHART_PPTX, raises_on_error=True)
+
+    pictures = list(result.document.pictures)
+    assert len(pictures) == 1
+    assert pictures[0].image is None
+
+
 def test_chart_image_rendering(libreoffice_available):
     """render_chart_images=True attaches a LibreOffice-rendered image.