@Anonymous wrote:
It will be great if we can do as fig. above, for acute angle & obtuse angle, we will draw the lines with different layer and use dimlinear for dimensioning. My purpose is calling out all chamfer corners by choose an area, the lisp will reconize the corners and automatically call out all dimensions.I alsmost use the lines (LWpolylines,I do not use)
For one thing, it would need to be DIMALIGNED or possibly DIMROTATED, not DIMLINEAR, which dimensions only in horizontal and vertical directions.
In any case, what you describe becomes hugely more complicated. It would be a lot easier if you used Polylines, because a routine would have some way of deciding which edges to consider in relation to each other, and because it would have some basis for putting the dimension line locations on the outside.
As a start, see >this< on that other thread for something that finds the virtual intersection between two selected edges, and determines the Chamfer distances from there to the near ends of those edges. But that requires picking on the two edges involved. If you select a bunch of independent Lines together, a routine would have a daunting challenge [if it's even possible] to decide which are the "main" sides and which are the chamfered corners. A Polyline at least has them in an order such that it could look at edges that are two segments apart.
Other questions arise:
Would all the chamfered corner pieces always be shorter than all the "main" edges? Since it would not be able to limit consideration to edges that are not orthogonal, the length comparison could be crucial to deciding which are the chamfered edges.
By the wording "choose an area," are you picturing picking within an area, or selecting the Lines that define it [or a Polyline that defines it, if you change to using those]? If by picking within an area, a routine would almost certainly use BOUNDARY or BPOLY to draw a Polyline to work with, anyway.
Kent Cooper, AIA