mirror of
https://git.mirrors.martin98.com/https://github.com/petaflot/pygcode
synced 2025-04-23 06:10:08 +08:00

fixed canned drilling cycle machine processing bugs improved tests for linear movement
======= pygcode ======= GCODE Parser for Python Currently in development, ``pygcode`` is a low-level GCode interpreter for python. Installation ============ Install using ``pip`` ``pip install pygcode`` or `download directly from PyPi <https://pypi.python.org/pypi/pygcode>`__ Usage ===== Writing GCode ------------- Writing gcode from python object instances to text :: >>> from pygcode import * >>> gcodes = [ ... GCodeRapidMove(Z=5), ... GCodeStartSpindleCW(), ... GCodeRapidMove(X=10, Y=20), ... GCodeFeedRate(200), ... GCodeLinearMove(Z=-1.5), ... GCodeRapidMove(Z=5), ... GCodeStopSpindle(), ... ] >>> print('\n'.join(str(g) for g in gcodes)) G00 Z5 M03 G00 X10 Y20 F200 G01 Z-1.5 G00 Z5 M05 To plot along a lines of vectors, you could write... :: >>> from pygcode import * >>> from euclid import Vector3 >>> vectors = [ ... Vector3(0, 0, 0), ... Vector3(10, 0, 0), ... Vector3(10, 20, 0), ... Vector3(10, 20, 3), ... Vector3(0, 20, 3), ... Vector3(0, 0, 3), ... Vector3(0, 0, 0) ... ] >>> to_coords = lambda v: {'X': v.x, 'Y': v.y, 'Z': v.z} >>> for v in vectors: ... print("%s" % GCodeLinearMove(**to_coords(v))) G01 X0 Y0 Z0 G01 X10 Y0 Z0 G01 X10 Y20 Z0 G01 X10 Y20 Z3 G01 X0 Y20 Z3 G01 X0 Y0 Z3 G01 X0 Y0 Z0 Reading / Interpreting GCode ---------------------------- To read gcode from a file, utilise the ``Line`` class. Each ``Line`` instance contains a ``Block`` and an optional ``Comment``. The ``Block`` contains a list of gcodes you're after. :: from pygcode import Line with open('part.gcode', 'r') as fh: for line_text in fh.readlines(): line = Line(line_text) print(line) # will print the line (with cosmetic changes) line.block.gcodes # is your list of gcodes line.block.modal_params # are all parameters not assigned to a gcode, assumed to be motion modal parameters if line.comment: line.comment.text # your comment text To elaborate, here are some line examples :: >>> from pygcode import Line >>> line = Line('G01 x1 y2 f100 s1000 ; blah') >>> print(line) G01 X1 Y2 F100 S1000 ; blah >>> print(line.block) G01 X1 Y2 F100 S1000 >>> sorted(line.block.gcodes) [<GCodeFeedRate: F100>, <GCodeSpindleSpeed: S1000>, <GCodeLinearMove: G01{X1, Y2}>] >>> print(line.comment) ; blah Interpreting what a line of gcode does depends on the machine it's running on, and also that machine's state (or 'mode') The simple line of a rapid move to ``x=10, y=10`` may be ``G00 X10 Y10``. However, if the machine in question is in "Incremental Motion" mode ``G91`` then the machine will only end up at ``x=10, y=10`` if it started at ``x=0, y=0`` So, GCode interpretation is done via a virtual machine: :: >>> from pygcode import Machine, GCodeRapidMove >>> m = Machine() >>> m.pos <Position: X0 Y0 Z0> >>> g = GCodeRapidMove(X=10, Y=20) >>> m.process_gcodes(g) >>> m.pos <Position: X10 Y20 Z0> >>> m.process_gcodes(g) >>> m.pos <Position: X10 Y20 Z0> # same position; machine in absolute mode >>> m.mode.distance <GCodeAbsoluteDistanceMode: G90> # see >>> m.process_gcodes(GCodeIncrementalDistanceMode()) >>> m.process_gcodes(g) # same gcode as above >>> m.pos <Position: X20 Y40 Z0> all valid ``m.mode`` attributes can be found with ``from pygcode.gcodes import MODAL_GROUP_MAP; MODAL_GROUP_MAP.keys()`` Also note that the order codes are interpreted is important. For example, the following code is WRONG :: from pygcode import Machine, Line m = Machine() line = Line('G0 x10 y10 G91') m.process_gcodes(*line.block.gcodes) # WRONG! This will process the movement to ``x=10, y=10``, and **then** it will change the distance mode to *Incremental*... there are 2 ways to do this correctly. - ``m.process_gcodes(*sorted(line.block.gcodes))``, or simply - ``m.process_block(line.block)`` sorting a list of gcodes will sort them in execution order (as specified by `LinuxCNC's order of execution <http://linuxcnc.org/docs/html/gcode/overview.html#_g_code_order_of_execution>`__). ``process_block`` does this automatically. If you need to process & change one type of gcode (usually a movement), you must split a list of gcodes into those executed before, and after the one in question. :: from pygcode import GCodeRapidMove, GCodeLinearMove from pygcode import Machine, Line, split_gcodes m = Machine() line = Line('M0 G0 x10 y10 G91') (befores, (g,), afters) = split_gcodes(line.block.gcodes, (GCodeRapidMove, GCodeLinearMove)) m.process_gcodes(*sorted(befores)) if g.X is not None: g.X += 100 # shift linear movements (rapid or otherwise) m.process_gcodes(g) m.process_gcodes(*sorted(afters)) For a more practical use of machines & interpreting gcode, have a look at `pygcode-normalize.py <https://github.com/fragmuffin/pygcode/blob/master/scripts/pygcode-normalize.py>`__ At the time of writing this, that script converts arcs to linear codes, and expands drilling cycles to basic movements (so my `GRBL <https://github.com/gnea/grbl>`__ machine can understand them) Development =========== This library came from my own needs to interpret and convert erroneous arcs to linear segments, and to expand canned drilling cycles, but also as a means to *learn* GCode. As such there is no direct plan for further development, however I'm interested in what you'd like to use it for, and cater for that. Generally, in terms of what to support, I'm following the lead of: - `GRBL <https://github.com/gnea/grbl>`__ and - `LinuxCNC <http://linuxcnc.org/>`__ More support will come with increased interest. So that is... if you don't like what it does, or how it's documented, make some noise in the `issue section <https://github.com/fragmuffin/pygcode/issues>`__. if you get in early, you may get some free labour out of me ;) Supported G-Codes ----------------- All GCodes supported by `LinuxCNC <http://linuxcnc.org>`__ can be written, and parsed by ``pygcode``. Few GCodes are accurately interpreted by a virtual CNC ``Machine`` instance. Supported movements are currently; - linear movements - arc movements - canned drilling cycles
Languages
Python
91.8%
nesC
6.1%
Shell
2%