Fix F5.3 check

[cpresser]

This fixes a regression from #334.

---

My testing when making #334 was not good enough. For some cases the weird +0.00005 math breaks the rounding really bad. I remove that block completely, the rounding is simpler and better. At least my current testing shows that it now works again as intended.

@evanshultz can you please take a look. This breaks travis (false positives). So I would like to address that right away.

See https://github.com/KiCad/kicad-footprints/pull/2438#issuecomment-694928293

[chschlue]

Two questions:

• What is this function supposed to do? If I'm not totally mistaken, a millimeter is 1e3 microns, not 1e6.
• Isn't the int cast redundant? According to https://docs.python.org/3/library/functions.html#round, round() with a single parameter returns an int anyway.

[cpresser]

ad1. To check if lines are on a grid, their coordinates are converted to micrometer nanometer. That should return an integer so a modulo division can be used to determine if they are on the grid. I agree that 1E3 and 1E6 seems wrong. But it is wrong at two places, so the code works :smile: GRID = int(KLC_CRTYD_GRID * 1E6) Proposed fix, rename the function to mmToNanometer()

ad2. I was not aware, I will change that.

[cpresser]

I did a force-push to keep all changes in one commit. This is so small, adding a 2nd commit seemed wrong.

[evanshultz]

I'm fine to do it. Anybody object to merging now as is?