paper2data

paper2data is a small local Linux app for extracting calibrated data from vector plots in PDF papers. It renders PDF pages as SVG, lets you click the actual SVG paths and tick marks, and exports two-column .dat files or CSV.

It is intentionally SVG-only. If a paper embeds a plot as a raster image, there will be no selectable curve path to extract.

Install

On Ubuntu/Debian-like Linux systems:

sudo apt install nodejs npm mupdf-tools poppler-utils
git clone https://github.com/liambern/paper2data.git
cd paper2data
npm start

Open:

http://localhost:5177

The app has no npm package dependencies. It needs:

• node for the local server
• mutool from mupdf-tools to render PDF pages as SVG
• pdfinfo from poppler-utils to read page counts

If your system package manager provides an old Node version, install Node 20+ or newer and rerun npm start.

Workflow

1. Click Open PDF.
2. Pick the page containing the plot.
3. Use mouse wheel zoom and Pan mode to focus on the plot.
4. In Select mode, hover SVG paths to highlight them and click the curve you want. Shift-click adds multiple path segments.
5. In Calibrate mode, click X1, X2, Y1, and Y2 tick marks. When you click a selectable SVG tick/path, paper2data uses the center of that SVG object rather than the raw mouse point.
6. Enter the numeric axis values for X1, X2, Y1, and Y2.
7. Set optional Points, X min, and X max.
8. Export DAT or CSV.

Output

DAT output is the default and contains exactly two whitespace-separated columns with no header:

x y

CSV output contains:

x,y

Step controls the internal SVG-path sampling density. Points, when set, resamples the selected curve to exactly that many evenly spaced x-values over the requested range.

Notes

• Best results come from PDF figures that preserve plot curves as vector paths.
• Multi-line plots may represent one visible curve as several SVG path segments; use Shift-click and the selected-path list when needed.
• The app runs locally. Uploaded PDFs stay on your machine in .paper2data/.