Draft 2d offset

[VVM]

OS: Debian GNU/Linux 11 (bullseye) (LXDE/LXDE)
Word size of FreeCAD: 64-bit
Version: 0.20.26720 (Git) AppImage
Build type: Release
Branch: (HEAD detached at 7bca26e)
Hash: 7bca26e7c14d4d43374ff7e0c8582735b249cbfd
Python version: 3.9.9
Qt version: 5.12.9
Coin version: 4.0.0
OCC version: 7.5.3
Locale: Russian/Russia (ru_RU)

Sorry, im not speak english.

2d_offset_1.png (15.51 KiB) Viewed 2070 times

2d_offset_2.png (13.81 KiB) Viewed 2070 times

[Roy_043]

File?

Both solutions seem valid to me, although I know that other CAD programs (DwgCAD) will do this differently.

[chrisb]

There is also a Форум на русском языке.

[paullee]

Wondering if the wire is not really closed, so it offset like the way it shows?

Need the model file for avoidance of doubt

[VVM]

File:

Draft_2D_offset.FCStd

(8.24 KiB) Downloaded 63 times

[Roy_043]

There is a bug in the isReallyClosed() code used by the Draft_Offset command. It only checks if the first vertex of the first edge and the last vertex of the last edge coincide. But since the direction of the edges can vary, there are in fact 4 possibilities.

https://github.com/FreeCAD/FreeCAD/blob ... #L272-L276

[Roy_043]

https://github.com/FreeCAD/FreeCAD/pull/5496

Edit:
Oops, it's not that simple. Will have to do some more checks.

[paullee]

There is a bug in the isReallyClosed() code used by the Draft_Offset command. It only checks if the first vertex of the first edge and the last vertex of the last edge coincide. But since the direction of the edges can vary, there are in fact 4 possibilities.

https://github.com/FreeCAD/FreeCAD/blob ... #L272-L276

Interesting

Should need to test Orientation then, would elaborate. In the meantime, this is history of revision for reference

https://github.com/FreeCAD/FreeCAD/comm ... c32b9856ce

https://github.com/FreeCAD/FreeCAD/comm ... 21581570b0#

[paullee]

The arc edge orientation is found 'Reversed' - a situation I remember usually happen 'within edit mode' (EDIT: of a Sketch) for an arc. The arc (EDIT: geometry 'within a Sketch in edit mode') is always counter-clockwire in edit mode to my memory.

So it seems needs to verify the Vertex for testing should be 'Forward', otherwise switch to another Vertex.

Code: Select all

s=Gui.Selection.getSelection()[0]
ss=s.Shape
ssw=ss.Wires[0]

import draftgeoutils.wires as wires
wires.isReallyClosed(ssw)
False

# wire.Vertexes[0] is not wire.Vertexes[-1] as previously noted
ssw.Vertexes[0].Point
Vector (22.016533000365484, -22.04014152497416, 0.0)

ssw.Vertexes[-1].Point
Vector (-22.203595540784868, -21.85167977200638, 0.0)

ss.Edges
[<Edge object at 0x7f8d1400bbf0>, <Edge object at 0x55572c84ba60>, <Edge object at 0x55572c5d3ed0>, <Edge object at 0x55572c1a2760>, <Edge object at 0x55572c1a27e0>, <Edge object at 0x55572c1a2860>, <Edge object at 0x55572c1a28e0>, <Edge object at 0x55572c807610>]

sse1=ss.Edges[0]
sse1.Vertexes[0].Point
Vector (22.016533000365484, -22.04014152497416, 0.0)

len(ss.Edges)
8

# last edge Vertexes[1] is not 1st edge Vertexes[0]
sse2=ss.Edges[7]
sse2.Vertexes[1].Point
Vector (-22.203595540784868, -21.85167977200638, 0.0)


# found it is last edge Vertexes[0] == 1st edge Vertexes[0]
sse2.Vertexes[0].Point
Vector (22.016533000365484, -22.04014152497416, 0.0)

sse2v0=sse2.Vertexes[0]
sse2v1=sse2.Vertexes[1]

# found the vertex orientation is 'Reversed' for the arc
sse2v0.Orientation
'Reversed'

sse2v1.Orientation
'Forward'

sse1v0=sse1.Vertexes[0]
sse1v1=sse1.Vertexes[1]
sse1v0.Orientation
'Forward'
sse1v1.Orientation
'Reversed'

[Roy_043]

I was unaware of the Orientation property of vertices. It is indeed interesting. But why do we have the Draft isReallyClosed() function? what is wrong with isClosed()?