pptx2tikz

Convert PowerPoint (.pptx) slides to TikZ/LaTeX. Works as a command-line tool and as a JavaScript library (Node.js and browser).

Built on top of pptxtojson with extensions for SVG extraction, math color/font parsing, and GVML clipboard support.

Installation

npm install pptx2tikz

CLI usage

npx pptx2tikz slides.pptx                        # convert slide 1
npx pptx2tikz slides.pptx -s 3                   # convert slide 3
npx pptx2tikz slides.pptx -a                      # convert all slides
npx pptx2tikz slides.pptx --pdf output.pdf        # convert all and compile to PDF
npx pptx2tikz slides.pptx -s 2 -o slide2.tex     # specify output file

CLI options

| Option | Description | |---|---| | -s, --slide <n> | Slide number to convert (default: 1) | | -a, --all | Convert all slides, each into its own subfolder | | --pdf <file> | Convert all slides, compile, and stitch into one PDF | | -o, --output <file> | Output .tex file (default: output.tex) | | --out-dir <dir> | Output directory for --all mode (default: slides/) | | --no-images | Replace images with placeholder boxes | | --image-dir <dir> | Image output directory (default: images/) | | --compile | Run pdflatex after converting | | --include-layout | Include layout/master slide elements | | --save-svgs <dir> | Save extracted SVG files to <dir> for inspection | | --xcolor-rgb-convert | Convert colors to xcolor expressions |

The --pdf and --compile options require pdflatex to be installed and on your PATH.

Library usage

Node.js

import { parse, convertSlideToTikZ } from 'pptx2tikz'
import { installNodeSvgEnvironment } from 'svg2tikz/node-env'
import { readFileSync, writeFileSync } from 'fs'

installNodeSvgEnvironment() // required in Node for SVG-to-TikZ conversion

const buffer = readFileSync('slides.pptx')
const result = await parse(buffer.buffer)

const { tex, images } = convertSlideToTikZ(result.slides[0], result.size)
writeFileSync('slide.tex', tex)

Browser

import { parse, convertSlideToTikZ } from 'pptx2tikz'

const file = /* File from <input type="file"> */
const buffer = await file.arrayBuffer()
const result = await parse(buffer)

const { tex, images } = convertSlideToTikZ(result.slides[0], result.size)
console.log(tex)

convertSlideToTikZ(slide, size, options?)

| Option | Type | Description | |---|---|---| | noImages | boolean | Replace images with placeholder boxes | | imageDir | string | Directory prefix for \includegraphics paths (default: "images") | | includeLayoutElements | boolean | Include layout/master slide elements | | xcolorRgbConvert | boolean | Convert colors to xcolor expressions |

Returns { tex, images } where tex is the full LaTeX document string and images is an array of { filename, base64data, ext } objects for any raster images referenced by \includegraphics.

parse(arrayBuffer)

Parses a .pptx file. Returns:

{
  slides: [{ elements, layoutElements, fill, note, transition }],
  size: { width, height },   // in points
  themeColors: ['#...', ...]
}

parseClipboardGVML(arrayBuffer)

Parses a GVML clipboard ZIP (the format used when copying shapes from PowerPoint). Returns the same structure as parse, with a single slide.

Output format

Each slide is output as a self-contained LaTeX document using the standalone document class with a single tikzpicture environment. The slide dimensions are preserved in centimeters.

SVG images embedded in the .pptx are converted to native TikZ commands using svg2tikz. Raster images (PNG, JPEG) are saved as separate files and included via \includegraphics.

Updating the pptxtojson dependency

This repo vendors pptxtojson as a git subtree under pptxtojson/. To pull in upstream updates:

git subtree pull --prefix=pptxtojson https://github.com/pipipi-pikachu/pptxtojson.git master --squash

License

MIT