Don't feel like coding? Use the Inkscape extension.
This library's intended purpose is to laser-cut svg images. However, it is structured such that it can be easily expanded to parse other image formats or compile to different numerical control languages.
Svg to Gcode is available on pip. To install it, execute:
pip install svg-to-gcode
Of course, you could also just download the sourcecode.
The module is divided in three sub-modules:
- svg_to_gcode.geometry offers a general representation of geometric curves.
- svg_to_gcode.parser parses svg files, converting them to geometric curves.
- svg_to_gcode.compiler transforms geometric curves into gcode.
If all you need is to compile an svg image to gcode, for a standard cnc machine, this is all the code you need. Just remember to select your own cutting and movement speeds.
from svg_to_gcode.svg_parser import parse_file
from svg_to_gcode.compiler import Compiler, interfaces
# Instantiate a compiler, specifying the interface type and the speed at which the tool should move. pass_depth controls
# how far down the tool moves after every pass. Set it to 0 if your machine does not support Z axis movement.
gcode_compiler = Compiler(interfaces.Gcode, movement_speed=1000, cutting_speed=300, pass_depth=5)
curves = parse_file("drawing.svg") # Parse an svg file into geometric curves
gcode_compiler.append_curves(curves)
gcode_compiler.compile_to_file("drawing.gcode", passes=2)
Interfaces exist to abstract commands used by the compiler. In this way, you can compile for a non-standard printer or to a completely new numerical control language without modifying the compiler. You can easily write custom interfaces to perform additional operations (like powering a fan) or to modify the gcode commands used to perform existing operations (some DIY laser cutters, for example, control the laser diode from the fan output).
The code bellow implements a custom interface which powers on a fan every time the laser is powered on.
from svg_to_gcode.svg_parser import parse_file
from svg_to_gcode.compiler import Compiler, interfaces
from svg_to_gcode.formulas import linear_map
class CustomInterface(interfaces.Gcode):
def __init__(self):
super().__init__()
self.fan_speed = 1
# Override the laser_off method such that it also powers off the fan.
def laser_off(self):
return "M107;\n" + "M5;" # Turn off the fan + turn off the laser
# Override the set_laser_power method
def set_laser_power(self, power):
if power < 0 or power > 1:
raise ValueError(f"{power} is out of bounds. Laser power must be given between 0 and 1. "
f"The interface will scale it correctly.")
return f"M106 S255\n" + f"M3 S{linear_map(0, 255, power)};" # Turn on the fan + change laser power
# Instantiate a compiler, specifying the custom interface and the speed at which the tool should move.
gcode_compiler = Compiler(CustomInterface, movement_speed=1000, cutting_speed=300, pass_depth=5)
curves = parse_file("drawing.svg") # Parse an svg file into geometric curves
gcode_compiler.append_curves(curves)
gcode_compiler.compile_to_file("drawing.gcode")
Before compiling, you could append or modify geometric curves. I'm not sure why you would want to, but you can. The code below draws a fractal and compiles it to gcode.
Ups. Looks like this example was never filled in...
Gcode only supports liner and circular arcs. Currently I've only implemented a line segment approximation. As such, geometric curves are compiled to a chain of line-segments. The exact length of the segments is adjusted dynamically such that it never diverges from the original curve by more then the value specified by TOLERANCES['approximation'].
The default value is 0.1. Smaller values improve accuracy, larger ones result in shorter gcode files.
from svg_to_gcode import TOLERANCES
TOLERANCES['approximation'] = 0.01
For now, this library only converts svgs to gcode files. However, its modular design makes it simple to support other formats. If you're looking to support a specific format, pull requests are always welcome. Just make sure to read CONTRIBUTING.md to get a feeling for the internal structure and best practices.