Skip to main content

​
Overview

ConversionResult is returned by all DocumentConverter conversion methods. It extends ConversionAssets and contains the converted DoclingDocument along with metadata about the conversion process, including status, errors, timing information, and page-level data.

​
Class Definition

from docling.datamodel.document import ConversionResult

​
Inheritance

ConversionResult extends ConversionAssets and inherits all its attributes and methods.

​
Attributes

​
Core Attributes

​
document
DoclingDocument
The converted Docling document. This is the primary output of the conversion process.
​
input
InputDocument
The input document metadata, including file path, format, and validation status.
​
status
ConversionStatus
The status of the conversion. Possible values:
  • ConversionStatus.SUCCESS - Conversion completed successfully
  • ConversionStatus.PARTIAL_SUCCESS - Conversion completed with some issues
  • ConversionStatus.FAILURE - Conversion failed
  • ConversionStatus.PENDING - Conversion not yet started
  • ConversionStatus.SKIPPED - Document was skipped
​
errors
list[ErrorItem]
List of errors that occurred during conversion. Each ErrorItem contains:
  • component_type: The component where the error occurred
  • module_name: The module that raised the error
  • error_message: Description of the error

​
Metadata Attributes

​
pages
list[Page]
List of page-level metadata for paginated documents. Each Page contains information about dimensions, page number, and page-specific content.
​
timings
dict[str, ProfilingItem]
Timing information for different stages of the conversion process.
​
confidence
ConfidenceReport
Confidence scores for various aspects of the conversion.
​
assembled
AssembledUnit
Information about how the document was assembled from its components.
​
version
DoclingVersion
Version information about Docling and its dependencies used for the conversion.
​
timestamp
str
ISO timestamp when the conversion assets were saved.

​
Methods

​
save()

Serialize the full ConversionResult to a compressed ZIP archive. 
result.save(filename="output.zip", indent=2)
​
filename
Union[str, Path]
required
Path where the ZIP archive will be saved.
​
indent
int
default:"2"
Number of spaces for JSON indentation. Use None for compact output.
The saved archive contains:
  • document.json - The DoclingDocument
  • version.json - Version information
  • status.json - Conversion status
  • errors.json - Error list
  • pages.json - Page metadata
  • timings.json - Performance timings
  • confidence.json - Confidence scores
  • timestamp.json - Save timestamp

​
load()

Load a ConversionResult from a saved ZIP archive. 
result = ConversionResult.load("output.zip")
​
filename
Union[str, Path]
required
Path to the ZIP archive to load.
​
return
ConversionResult
The loaded ConversionResult instance.

​
Usage Examples

​
Basic Conversion

from docling.document_converter import DocumentConverter
from docling.datamodel.base_models import ConversionStatus

converter = DocumentConverter()
result = converter.convert("document.pdf")

# Check conversion status
if result.status == ConversionStatus.SUCCESS:
    print("Conversion successful!")
    print(f"Document name: {result.document.name}")
    print(f"Number of pages: {len(result.pages)}")
else:
    print(f"Conversion failed with status: {result.status}")
    for error in result.errors:
        print(f"Error: {error.error_message}")

​
Accessing the Document

result = converter.convert("document.pdf")

# Get the DoclingDocument
doc = result.document

# Export to different formats
markdown_text = doc.export_to_markdown()
html_text = doc.export_to_html()
json_dict = doc.export_to_dict()

# Access document structure
for item in doc.iterate_items():
    print(f"{item.label}: {item.get_text()}")

​
Error Handling

converter = DocumentConverter()

# Don't raise on errors
result = converter.convert("document.pdf", raises_on_error=False)

if result.status != ConversionStatus.SUCCESS:
    print(f"Conversion status: {result.status}")
    
    # Examine errors
    for error in result.errors:
        print(f"Component: {error.component_type}")
        print(f"Module: {error.module_name}")
        print(f"Message: {error.error_message}")

​
Saving and Loading Results

# Convert and save
result = converter.convert("document.pdf")
result.save(filename="conversion_output.zip")

# Later, load the result
from docling.datamodel.document import ConversionResult

loaded_result = ConversionResult.load("conversion_output.zip")
print(f"Loaded document: {loaded_result.document.name}")
print(f"Original conversion time: {loaded_result.timestamp}")

​
Batch Processing with Results

converter = DocumentConverter()
documents = ["doc1.pdf", "doc2.docx", "doc3.html"]

results = []
for result in converter.convert_all(documents, raises_on_error=False):
    results.append(result)
    
    if result.status == ConversionStatus.SUCCESS:
        print(f"✓ {result.input.file.name}")
    else:
        print(f"✗ {result.input.file.name}: {result.status}")

# Generate summary
success_count = sum(1 for r in results if r.status == ConversionStatus.SUCCESS)
print(f"\nSuccessfully converted {success_count}/{len(results)} documents")

​
Accessing Timing Information

result = converter.convert("document.pdf")

# View performance timings
for stage, timing in result.timings.items():
    print(f"{stage}: {timing.duration:.2f}s")

​
Page-Level Information

result = converter.convert("document.pdf")

# Access page metadata
for i, page in enumerate(result.pages, 1):
    print(f"Page {i}:")
    print(f"  Size: {page.size.width} x {page.size.height}")
    print(f"  Page number: {page.page_number}")

​
See Also

Build docs developers (and LLMs) love

Get started for freeTalk to us