I'm the author of this project, someone else posted it on my behalf, let me explain my story:
My bank requested 17 PDF documents for a mortgage application.
I got tired of opening and closing files and keeping track of things. So I decided to merge all files into a single, long PDF and send it to my bank, but that caused even more confusion.
To solve this issue, I came up with an idea: what if I could extend the standard PDF structure by embedding custom JSON metadata to indicate where files start and where they end?
I called it `.pdfx`, but it's simply a backwards-compatible PDF file, that when opened with traditional PDF viewers (Chrome, Acrobat), it renders as a regular, continuous document. But when opened in the PDFx runtime, a react-based viewer I built, the app parses the metadata coordinates and reconstructs a 2D free floating canvas, where the pages of a file extend horizontally, while files themselves grow vertically.
This view allows for a greater spatial understanding, batch processing, and a pleasant drag-n-drop user experience. Plus, additional support for SVGs, PNGs and technical drawings that are converted into PDFs when dropped onto the app.