GeoPatterns

Generate beautiful SVG patterns from a string. This is a Python-port of Jason Long's Ruby library.

Installation

GeoPatterns requires Python 3.11 or newer and is installable via pip:

$ pip install geopatterns

Usage

Create a new pattern by initializing GeoPattern() with a string. The string is hashed, and the hash decides the pattern, the background color and the size of the shapes.

>>> from geopatterns import GeoPattern
>>> pattern = GeoPattern('A string for your consideration.')

To use a specific pattern, or a subset of the available patterns:

>>> GeoPattern('A string for your consideration.', patterns='sine_waves')
>>> GeoPattern('A string for your consideration.', patterns=['sine_waves', 'xes'])

To pick a base color (its hue and saturation still shift with the string), or to use an exact color:

>>> GeoPattern('A string for your consideration.', base_color='#fc0')
>>> GeoPattern('A string for your consideration.', color='#fc0')

Available patterns are:

• chevrons
• concentric_circles
• diamonds
• hexagons
• mosaic_squares
• nested_squares
• octagons
• overlapping_circles
• overlapping_rings
• plaid
• plus_signs
• sine_waves
• squares
• tessellation
• triangles
• xes

generator= still works but is deprecated, use patterns=. The old bricks, rings and sinewaves generators are gone. Upstream replaced them with octagons, concentric_circles and sine_waves, and passing the old names to generator= draws those instead, with a warning.

Get the SVG string:

>>> print(pattern.svg_string)
<svg xmlns="http://www.w3.org/2000/svg" ...

Get the Base64-encoded string (as bytes):

>>> print(pattern.base64_string)
b'PHN2ZyB4bWxucz0iaHR0cDov...

In the case of the Base64-encoded string, you can use it in CSS as follows:

body {
  background-image: url('data:image/svg+xml;base64,PHN2ZyB4bWxucz...zdmc+');
}

You can use cairosvg to save the SVG string as a PNG image. First, install cairosvg:

pip install cairosvg

And then run:

>>> import cairosvg
>>> from geopatterns import GeoPattern
>>> pattern = GeoPattern('A string for your consideration.', patterns='xes')
>>> cairosvg.svg2png(bytestring=pattern.svg_string, write_to="output.png")

If you just want to visualize the pattern, you can use cairosvg with PIL:

>>> from io import BytesIO
>>> import matplotlib.pyplot as plt
>>> from PIL import Image
>>> import cairosvg
>>> from geopatterns import GeoPattern
>>> pattern = GeoPattern('A string for your consideration.', patterns='xes')
>>> png = cairosvg.svg2png(bytestring=pattern.svg_string)
>>> image = Image.open(BytesIO(png))
>>> plt.imshow(image)
>>> plt.show()

Compatibility

For the same string and options, the SVG matches what geo_pattern 1.5.0 produces. The tests check this against upstream's fixtures and against hashes generated from its Ruby code.

Development

$ uv run pytest

tests/fixtures/hashes.json comes from upstream. To regenerate it, check out geo_pattern, install its color gem, and run:

$ ruby -I<upstream>/lib tests/ruby/generate_hashes.rb <upstream> > tests/fixtures/hashes.json