pdf2json

Typesense operates on JSON documents. So you'd have to first extract the data from PDFs / office docs into JSON and then index it. There are libraries that let you do the conversion: https://github.com/modesty/pdf2json https://github.com/microsoft/Simplify-Docx

Typesense operates on JSON documents. So you'd have to first extract the data from PDFs / office docs into JSON and then index it. There are libraries that let you do the conversion: https://github.com/modesty/pdf2json https://github.com/microsoft/Simplify-Docx

Acroform support (so not the proprietary Adobe Livecycle XFA stuff) is relatively straightforward once you understand the underlying PDF specs. It's been around since PDF 1.2 (1996!!) At the rate I see development moving, pdf.js should be able to handle, say, every IRS form well before the end of the year, perhaps much sooner. If I had the skills or resources to help now, I would. I still don't understand why Mozilla has apparently not reached out to the larger business/governmental universe to see if they could partner in moving pdf.js forward. A single large insurance company might spend millions each year on licensing and implementing pdf "solutions". I can only imagine what the US government uses (or could use.) Has anyone from Mozilla spoken to 18F? Perhaps even a open (and not proprietary) JSON (instead of XML) based form standard f

Which is absurd and shortsighted-- 1. pdf.js has only seriously attempted to implement forms over just the last few weeks, and has made significant progress[1] not to mention someone did quite a bit of it on their own 4 years ago and no-one seemed interested in working with him. [2] 2. Mozilla has a strong interest in furthering web technologies, and pdf.js could be used in just about every enterprise LOB and government form based workflow (as they slowly move to web technologies), if it matured a bit more. Has Mozilla thought of how their unique position as a non-profit could enable them to partner in ways a corporation could not? The insurance industry on its own produces 10s if not 100s of millions of pages of PDFs every day...has anyone from Mozilla ever talked to ACORD (or even knows who they are?) [1] https://github.com/mozilla/pdf