Wkhtmltopdf - Shell utility to convert html to PDF using webkit.

I tried porting it to OpenBSD and found building it from source to be surprisingly complicated. IIRC, it wanted me to install their own fork of qt, just to render PDFs. Pass.

I went with htmldoc (http://www.htmldoc.org/) instead. Perhaps the rendered PDFs aren't as pretty, but for my purposes, it works quite well.

Use wkhtmltoimage to create images of book-quality syntax-highlighted source code:

http://superuser.com/questions/213217/convert-html-to-image

Use gvim to pick from 100 different ready-made colour schemes:

http://www.vi-improved.org/color_sampler_pack/

Shameless plug: I describe this technique in an appendix of my technical manual.

http://www.whitemagicsoftware.com/books/indispensable/

For example:

    #!/bin/bash

    echo Converting $1 to $1.html...

    # Schemes: pyte
    vim $1 -c "set nobackup" ${2:+"-c set number"} \
      -c ":colorscheme tango-morning" \
      -c ":TOhtml" \
      -c "wq"  -c ":q"

    echo Converting $1.html to $1.png...

    wkhtmltoimage --transparent --minimum-font-size 80 \
      --format png --quality 100 --width 4000 \
      $1.html $1.png

    mv $1.png $1.orig.png

    # For scripts that are under 80 characters, make the generated image always
    # the same size (4000 pixels = width + (border x 2) = 3950 + 25 x 2).
    LENGTH=$(awk 'BEGIN {x=0} {if( length($0)>x ) x=length()} END {print x;}' $1)

    if [ $LENGTH -le 80 ]; then
      EXTENT=true
    fi

    echo Trimming $1.png...
    convert -format png $1.orig.png -background none -antialias -trim +repage \
      -density 150x150 ${EXTENT:+-extent 3950x} \
      -bordercolor none -border 25 \
      $1.png

    echo Removing temporary files...
    rm -f $1.orig.png
    rm -f $1.html

We are using this. Seems to do fine. My only complaint is that trying to pronounce the name makes me choke on my own uuvula.

This is fantastic, thank you so much for posting this. I felt like pointing out another headless Webkit project, PhantomJS (http://www.phantomjs.org/), which can also perform HTML to PDF conversion.

edit: I've just run a comparison between wkhtmltopdf and PhantomJS and wkhtmltopdf is far superior. wkhtmltopdf produces correct output and bookmarks but has a slightly lower quality, where PhantomJS's output was not correct and without bookmarks but at a slightly higher quality. From my perspective correctness beats quality.

Very useful tool. Beware a tough time getting various fonts to work, though — and some characters (thin space anyone?) will just completely crash it.

I use the excellent Pandoc for my conversion needs: http://johnmacfarlane.net/pandoc/

On a related note: I'm looking for a command line utility (*nix) to go the other way. Any recommendations for a tool which extracts text from a PDF?

This is a great tool, I've been using it for a long time. Only downside is that on Ubuntu Server certain graphics libraries aren't in the PPA so you have to download a static binary instead of using apt-get. But it's a minor quibble for a great program.

I was briefly looking for exactly this, after our lawyer suggested manually printing out thousands of emails and scanning to PDF. Luckily my encouragement to use a PST file finally won out. Not a confidence inspiring moment.

We're using this on http://interstateapp.com to allow users to download a PDF version of their roadmap. The script works very well!

This tool works impressively well. I use it to convert my CV to PDF, as I don't trust my CSS to work everywhere.

Cool. Can you get an image out of it ?