A command line utility that exports omnigraffle canvases files into various formats.

omnigraffle export
pip install omnigraffle-export==1.7.1


OmniGraffle Export tool

A command line tool that allows to export one or more canvases from OmniGraffle into various formats using OmniGraffle AppleScript interface.


In order to have it successfully installed and working, following is required:

You can either clone the repository and use the setup tool: install

Or using the PIP:

pip install omnigraffle_export


Usage: omnigraffle-export [options] <source> <target>

  -h, --help  show this help message and exit
  -c NAME     canvas name. If not given it will be guessed from the target
              filename unless it is a directory.
  -f FMT      format (one of: pdf, png, svg, eps). Guessed from the target
              filename suffix unless it is a directory. Defaults to pdf
  --force     force the export
  --debug     print out debug messages

If the target file exists it tries to determine whether the canvas has been changed. It does that by comparing the checkums. Since the PDF export always results into a different file, it uses the PDF subject attribute to store there a checksum that will be generated by exporting to a PNG format.


  • Export all canvases into directory figures using EPS

    $ omnigraffle-export -f eps schemas.graffle figures
  • Export canvas named FullModel into a FullModel.svg SVG file:

    $ omnigraffle-export schemas.graffle FullModel.svg
  • Export canvas name FullModel into a my_new_model.pdf PDF file:

    $ omnigraffle-export -c FullModel -f pdf schemas.graffle my_new_model


    $ omnigraffle-export -c FullModel schemas.graffle my_new_model.pdf

    Note: that the ’.pdf’ suffix will be automatically added in the first case.

  • Export all canvases into directory figures using EPS

    $ omnigraffle-export -f eps schemas.graffle figures

LaTeX Support Example

One of the main motivation for this package was to be able to quickly export OmniGraffle canvases and use them in LaTeX. One of the possible setup is following: every time a figure is included add some instruction so it can be later exported from OmniGraffle file. For example using a comment like:

% omnigraffle sources/schemas.graffle figures/CondorKernel.pdf

to export a canvas CondorKernel from sources/schemas.graffle into figures/CondorKernel.pdf as PDF.

  % omnigraflle: sources/schemas.graffle figures/CondorKernel.pdf

An example preprocesor in Python using frabric.

from fabric.api import *
import re

# latex files to process
fnames = ['UCGridRLDecisionModel.tex']
omnigraffle_re = re.compile(r'%\s*omnigraflle:\s*([^ ]+)\s+([^ ]+)')

def _convert(source, target):
  local('omnigraffle-export %s %s' % (source, target))

def schemas():
  Generate all schemas

  for fname in fnames:
    with open(fname) as f:
      for l in f:
          m = omnigraffle_re.match(l.strip())
          if m:

Export on Demand

The omnigraffle-export can be used either in a batch mode or in a more interactive way. For example it can be used to export the currently selected canvas into a file.

Following is an example Python script that will export currently active canvas into a PDF file that has the same name as the canvas and is placed in the same directory as the OmniGraffle document:

#!/usr/bin/env python

import os
import sys
import omnigraffle

og = omnigraffle.OmniGraffle()
schema = og.active_document()

schema_path = schema.path
schema_fname = os.path.basename(schema_path)
schema_dir = os.path.dirname(schema_path)
export_info_fname = os.path.join(schema_dir, '.' + schema_fname[0:schema_fname.rindex('.')] + '.omnigraffle_export')

canvas_name = schema.active_canvas_name()

format = 'pdf'
target_path =  os.path.join(schema_dir, canavs_name + '.' + format)

schema.export(canvas_name, target_path, format=format)

Fancier version can be download here.

Making a release

  • update the version number in
  • rerun tests $ python test
  • commit
  • push
  • register release $ python register
  • upload release $ python bdist upload

Developing omnigraffle-export

  • clone
  • stage $ python develop
  • make a new feature branch
  • code
  • do release
  • unstage $python develop --uninstall