repo-to-pdf

Make pdf from source code

Here is a sample

Install

npm install repo-to-pdf

Usage

Basic

npx repo-to-pdf [src/folder]

const { generateEbook } = require('repo-to-pdf')

/**
 * @typedef Options
 * @type {object}
 * @property {string} renderer - [native|node|calibre|wkhtmltopdf] node (puppeteer, default), native (built-in, dependency-free), calibre or wkhtmltopdf
 * @property {string} calibrePath - path of calibre's ebook-convert
 * @property {string} pdf_size - pdf size limit, in bytes
 * @property {string|string[]} white_list - list of file extensions to be included, separate by ','
 * @property {string|string[]} exclude_list - list of file paths to exclude, using Unix glob patterns
 * @property {string} format - [mobi|epub|pdf] can be either mobi, epub, pdf
 * @property {string} device - [desktop|tablet|mobile] style can be opt for desktop, tablet and mobile
 * @property {string} baseUrl - base url of CSS style files
 * @property {boolean} outline - include PDF outline/bookmarks (default true)
 * @property {number} concurrency - max number of parallel Puppeteer PDF jobs (default auto)
 * @property {boolean} footerPageNumber - show page numbers in the bottom-right footer with a rule
 * @property {boolean} footerChapterTitle - show the current section title in the bottom-right footer with a rule
 */

/**
 * Generate ebook from content folder with the given file format
 * @param {string} inputFolder - content folder
 * @param {string} outputFile - output file name
 * @param {string} title - title in ebook file
 * @param {Options} options
 */
generateEbook('./', 'test.pdf', 'repo-test', { renderer: 'node', pdf_size: 3 * 0.8 * 1000 * 1000, format: 'pdf', device: 'desktop', outline: true })

Renderers

| renderer | engine | deps | speed | output | | ------------- | ------------------------------------ | ------------------ | --------- | --------------- | | node | puppeteer (headless Chrome, default) | puppeteer + Chrome | slow | pdf | | native | built-in PDF writer | none | very fast | pdf | | wkhtmltopdf | wkhtmltopdf binary | wkhtmltopdf | medium | pdf | | calibre | calibre ebook-convert | calibre | medium | pdf, mobi, epub |

The native renderer reads the source code and writes a PDF directly, with no external dependencies or headless browser. It imitates the GitHub/HTML styling, applies highlight.js syntax highlighting, embeds PDF bookmarks (table of contents) and only references the built-in PDF fonts for Latin text so output files stay small. Characters outside WinAnsi (including CJK and most Unicode scripts) are rendered using embedded font subsets. A bundled CJK fallback font ships with the package; system fonts are preferred when available on Linux, macOS, and Windows.

Command Line Options

-d, --device [platform]
device [desktop(default)|mobile|tablet]
# npx repo-to-pdf ../repo -d mobile

-t, --title [name]
pdf filename
# npx repo-to-pdf ../repo -t MeinKampf.pdf

-w, --whitelist [wlist]
file format white list, split by ","
# npx repo-to-pdf ../repo -w js,md

-x, --exclude <patterns>
file paths to exclude, using Unix glob patterns, split by ","
# npx repo-to-pdf ../repo -x "**/*.test.js,src/generated/**"

-s, --size [size]
pdf file size limit, in MB, default 10 MB
# npx repo-to-pdf ../repo -s 10

-r, --renderer [native|node|calibre|wkhtmltopdf]
choose the render engine (default node). node uses puppeteer; native is the
built-in dependency-free PDF generator; calibre outputs pdf, mobi, epub
# npx repo-to-pdf ../repo -r native
# npx repo-to-pdf ../repo -r calibre

-f, --format [pdf|mobi|epub]
output format, either pdf, mobi, epub. mobi and epub are generated using calibre ebook-convert
# npx repo-to-pdf ../repo -f mobi

--no-outline
disable PDF outline/bookmarks generation (enabled by default)
# npx repo-to-pdf ../repo --no-outline

--footer-page-number
show page numbers at the bottom right of each page with a footer rule
# npx repo-to-pdf ../repo --footer-page-number

--footer-chapter-title
show the current section title at the bottom right of each page with a footer rule
# npx repo-to-pdf ../repo --footer-chapter-title

-p, --concurrency [num]
set max parallel Puppeteer PDF render jobs (auto uses up to 4 cores)
# npx repo-to-pdf ../repo -p 2

-c, --calibre [path]
path to ebook-convert, for MacOS, try /Applications/calibre.app/Contents/MacOS/ebook-convert; for linux, try /usr/bin/ebook-convert

For tablet, mobile

only supported by the renderer node

npx repo-to-pdf [src/folder] --device tablet
npx repo-to-pdf [src/folder] --device mobile

Testing

npm run test

Performance

cd test/data && wget https://github.com/redis/redis/archive/refs/tags/7.0.0.zip && unzip 7.0.0.zip

on M1 Macbook Air

# node (default): puppeteer / headless Chrome
time npx repo-to-pdf ./test/data/redis-7.0.0/src -s 3
2.09s user 0.36s system 2% cpu 1:42.14 total

# native: no headless browser, single process, much faster and lighter
time npx repo-to-pdf ./test/data/redis-7.0.0/src -s 3 -r native
# a few seconds, minimal CPU

time npx repo-to-pdf ./test/data/redis-7.0.0/src -s 3 -r wkhtmltopdf
43.78s user 0.84s system 93% cpu 47.787 total

Known issues

• The renderer wkhtmltopdf can split lines. It's a known unsolved issue of wkhtmltopdf: https://github.com/wkhtmltopdf/wkhtmltopdf/issues/2141, https://github.com/wkhtmltopdf/wkhtmltopdf/issues/1524
• To properly install puppeteer on Debian, you might need to install required libs: https://github.com/puppeteer/puppeteer/issues/290#issuecomment-322921352 and libgbm-dev