Fix extruder calculation with G2/G3

[ldursw]

When using this plugin with Arc Welder the extruded filament on arc moves was not calculated causing the final value to be lower than what the firmware reported.

This PR is a port from OllisGit/OctoPrint-FilamentManager#67 to SpoolManager.

[OllisGit]

Thanks a lot for the merge requests, I really appreciate that.

The newodometer.py was a copy of gcodeinterpreter.py Version OP 1.5.2.
I looked into the current implementation and I think the arc-movement is also not tracked yet.

https://github.com/OctoPrint/OctoPrint/blob/master/src/octoprint/util/gcodeInterpreter.py#L420

Did you also raise an issue in OP tracker?

[ldursw]

I'll open an issue there. Thanks for pointing out.

[bcat]

It looks like this is fixed in upstream OctoPrint now: OctoPrint/OctoPrint@b60c109. I can confirm that OctoPrint 1.7.2 shows the correct filament usage for a G-code file with arc moves. (In my case, I use the ArcWelder Cura plugin, so I don't have ArcWelder installed in OctoPrint, but the effect is the same.)

I just started using SpoolManager last night, and I haven't actually finished my first print with it yet, but it sounds like it will undercount the arc moves vs. what OctoPrint shows? (I can confirm that when this print finishes tomorrow, if that's useful.)

[OllisGit]

Hi l@dursw,
sorry I missed your PR to include in the last release 1.6.0.

I compared the OP-Implementation with your PR and is still valid, so your PR will be in the next release (not planed yet).

[OllisGit]

It took a while, but it is in the latest RC 1.7.0rc1.

@dursw, Thanks again!

[codingcatgirl]

Oh hey that's great news! I was about to ask for this, thanks for your work!