OCR and Extraction

[mrutunjay-kinagi]

Does docling support Checkbox , form data and radio button data extraction from pdf ?

TIA!

[AdityaMannu1709]

Checkbox Tags are present when the output format is DocTags. You can find it there as data for each checkbox is enclosed in its paired tag.

[mrutunjay-kinagi]

Thanks for responding , I did go through the code , docling-parse is taking care of the parsing. Can you please point me to the documentation docling-parse about DocTags ?

[AdityaMannu1709]

You can observe the Doctag's output of different PDFs as specifically there is no discussion of Doctag structure in documentation. I had observed around 30 PDFs doctags using the below snippet -

pipeline_options = PdfPipelineOptions(do_table_structure=True, do_ocr=True)
pipeline_options.table_structure_options.mode = TableFormerMode.ACCURATE  

doc_converter = DocumentConverter(
    format_options={
        InputFormat.PDF: PdfFormatOption(pipeline_options=pipeline_options)
    }
)

result = doc_converter.convert(pdf_path)
print(result.document.export_to_document_tokens())  # output: "DocTags"