Change hook call from macro to python

[alkabal]

Hi

I hope you a good year !

Another time i come back with the idea to move hook call from macro to python (and add hook_end).

I have noticed than try to use some outside/inside measurement do some move before calling the ocode so the hook is done after first move and imo this is not safe.

What is you opinion ?

Br

[kiall]

I'm not sure I understand the use case? You're trying to use a hook to perform measurements?

For me personally, I'd like to see as much as possible moved out of python and into the macros - as this will in my opinion make the code much more accessible to the average LinuxCNC user. Gcode is something we can expect the average LinuxCNC user to understand, while Python is not...

[alkabal]

Hi
For me a useful hook is done at start of each routine before any movement

I use hook as a pre safe check and probe/setter input management so I need to do this for each routine

also this is not wanted when a routine use several ocode call

The version in my fork is imo near to be finalized

[alkabal]

For me personally, I'd like to see as much as possible moved out of python and into the macros - as this will in my opinion make the code much more accessible to the average LinuxCNC user. Gcode is something we can expect the average LinuxCNC user to understand, while Python is not...

I will speak about moving the o<psng_hook> call from gcode to python if self.ocode("o<psng_hook> call.

[kiall]

For me personally, I'd like to see as much as possible moved out of python and into the macros - as this will in my opinion make the code much more accessible to the average LinuxCNC user. Gcode is something we can expect the average LinuxCNC user to understand, while Python is not...

I will speak about moving the o<psng_hook> call from gcode to python if self.ocode("o<psng_hook> call.

Let's take a concrete example - an X+ outside measurement:

probe-screen-ng/psng/python/workpiece_measurement.py

Lines 48 to 89 in 8c5f57b

	# X+
	def on_xp_released(self, gtkbutton, data=None):
	# move X - xy_clearance
	s = """G91
	G1 X-%f
	G90""" % (
	self.halcomp["ps_xy_clearance"]
	)
	if self.gcode(s) == -1:
	return
	if self.z_clearance_down() == -1:
	return
	# Start psng_xplus.ngc
	if self.ocode("o<psng_xplus> call") == -1:
	return
	a = self.probed_position_with_offsets()
	xres = float(a[0] + 0.5 * self.halcomp["ps_probe_diam"])
	self.display_result_xp(xres)
	self.lenght_x()
	self.add_history(
	gtkbutton.get_tooltip_text(),
	"XpLx",
	0,
	0,
	xres,
	self.lenght_x(),
	0,
	0,
	0,
	0,
	0,
	0,
	0,
	)
	# move Z to start point up
	if self.z_clearance_up() == -1:
	return
	# move to finded point
	s = "G1 X%f" % (xres)
	if self.gcode(s) == -1:
	return
	self.set_zerro("X")

You're wanting to move the hook call into that python so that the hook is ran prior the code on L50-59. I'm wanting to move the code on L50-59 from the python into the macro as there is no reason that code needs to be in the python.

Both solutions allow the hook to be executed prior to any movements, but one solution puts as much into the gcode as possible and should be far more approachable for the average LinuxCNC user who knows gcode, but does not know python.

[kiall]

I should add, that code hasn't already been moved because it's more complicated that just moving that code - as the macro is reused by several measurements - but IMO that is the approach that results in the most maintainable and approachable codebase.

[alkabal]

I understand your opinion imo one or other solution seem to be good with the choice i prefer to call from python even if you move L50-59 in ocode, mostly because imo hook need to be done only at each "def on_" rather than with all ocode call.

For common user i hope at the end to propose something that do not need to edit regular macro, and only think about use or not use hook.

If you check the code like L50-59 in other sequence you can see they are not allways the same so you need to expose several argument for passing the correct calc result + Axis letter.

[kiall]

If you check the code like L50-59 in other sequence you can see they are not allways the same so you need to expose several argument for passing the correct calc result + Axis letter.

This is what I was referring to with "that code hasn't already been moved because it's more complicated that just moving that code".

I understand your opinion imo one or other solution seem to be good with the choice i prefer to call from python even if you move L50-59 in ocode

The less python, the more LinuxCNC users can contribute to making it better - for me - making the probe screen maintainable is the #1 goal.

mostly because imo hook need to be done only at each "def on_" rather than with all ocode call.

Agreed - The hook should only be called once at the beginning of any action, but this is entirely possible without requiring that logic be split between Python and Gcode.

[alkabal]

Agreed - The hook should only be called once at the beginning of any action, but this is entirely possible without requiring that logic be split between Python and Gcode.

Have you change this with new version ?

[kiall]

Agreed - The hook should only be called once at the beginning of any action, but this is entirely possible without requiring that logic be split between Python and Gcode.

Have you change this with new version ?

Not yet... In the meantime, you can workaround this by checking #<_call_level> - if it's more than 2, return early from the hook... (I think 2 anyway - if it's zero based, you'll want to check for more than 1 instead.)

[alkabal]

thanks for the tips !