extract table bbox from DOCX

[aixiuxiuxiu]

I need to read a DOCX file that contains many tables and retrieve the bounding boxes (bbox) of the cells, but the bbox is always returning as None.

doc_converter =  DocumentConverter(
    format_options={
        InputFormat.DOCX: WordFormatOption(
        pipeline_cls=SimplePipeline, do_ocr = True # , backend=MsWordDocumentBackend 
        )
            }
)

[cau-git]

@aixiuxiuxiu Our native MS Word backend is not able to provide bounding boxes, because the DOCX format simply does not include such information. If you have a need for visual layout information, the current recommendation is to print your Word files to PDF (or automate it with some asset like unoserver) and feed it through Docling as PDF.