mikesimons/figma-fig-tools

fig parser

Hacky scripts to investigate the .fig file format.

Format

There are two formats of .fig file floating around. The older one has: 4 sections:

• Header
• kiwib schema (zlib deflated)
• data (zlib deflated)
• PNG thumbnail

Each section (except the header) consist of 2 fields:

• 4 byte length (little endian)
• data (length bytes)

The newer format is just a zip file containing:

• canvas.fig - A fig file of the older format except without the trailing PNG thumbnail
• thumbnail.png - The thumbnail that used to be embedded in the .fig
• meta.json - A bunch of boring metadata
• images/ - A folder that I've never seen populated but presumably contains... images

Installing

You'll need a relatively recent ruby (e.g. 2.7) and node + npm.

• Firstly install npm packages with npm install.
• Secondly, go grab a .fig file from figma
• Thirdly, run ./figunpack <file>

The unpacked contents will be output to a directory called <file>.out/.

In addition to unpacking and inflating the file contents we convert the kiwi encoded data in to JSON and write this to data.json. If you're only interested in the figma data itself, this is the interesting bit.