Can I use MD original file use PageIndex parser & gen index?

[lwyBZss8924d]

May I ask, if I want to directly use the original file in Markdown format without the original page numbers, can it still be processed based on PageIndex to obtain a structured node hierarchy and summary? Looking forward to your guidance, thank you 🙏

[zmtomorrow]

Hi, this feature will be updated by the end of this week. We will keep you updated.

Thanks

[lwyBZss8924d]

Great, THANKS

[lwyBZss8924d]

I have a small idea or a question that I would like to seriously ask you:
If during content parsing of the original document, I can extract and segment the embedded "tables" & "images" in the original document and convert them into:

• tables BBOX Annotations = {}
• images BBOX Annotations = {}
• images BBOX summary = {}

Then, when generating the "PageIndex" later on, can we still utilize these converted "tables" & "images" or their summaries? Or use metadata for PageIndex? (I'm not sure if currently when we directly parse the original PDF document to generate the "PageIndex," do we also separately handle "tables" & "images"?)

[zmtomorrow]

Hi, I think that is really a good point. Thanks for raising this.

Ideally, we could treat any table or image as a sub-node in the tree with some metadata like

{
label: (img/table/graph)
page_number: xx
description: xxx
}

Feel free to give any feedback on this design. If you think it is useful, we will plan to add that to our package and also to the cloud.

[lwyBZss8924d]

Hi, I think that is really a good point. Thanks for raising this.

Ideally, we could treat any table or image as a sub-node in the tree with some metadata like

{
label: (img/table/graph)
page_number: xx
description: xxx
}

Feel free to give any feedback on this design. If you think it is useful, we will plan to add that to our package and also to the cloud.

Yes, Awesome! looking forward to PageIndex adding tag or summary options for assets like tables, images, or code blocks parsed from the layout structure. Thank you for your hard work, 🫡