tabulapdf/tabula ? reverse-engineered prompt

Build me a local web app that lets me upload a PDF and extract tables from it into CSV, with a simple browser interface that makes it easy to preview the PDF, select the table area, and download the cleaned data.

It should work best on text based PDFs, not scanned ones, and it should clearly tell me when a file probably won’t work. Keep everything running on my own machine so the PDF never gets sent anywhere. I want it to open in the browser on localhost, and I should be able to run it easily from source for testing.

Make the experience straightforward and friendly, with sensible defaults, progress while it processes the file, and an easy way to export the extracted data. If you need to check current docs or best practices online, go ahead and do that.