Joining parallel lines with single line
Accepted solution
I had an old one written in 1999.
You can try the code below:
;;;
;;; Joinends.lsp
;;;
;;; By 3wood, April 1999
(defun c:jdd(/ ss1 pt0 i lin0 pt1 pt2 b1 pt_s1 b2 pt_s2a dst0 pt_s2)
(while (= ss1 nil)
(prompt"\nSelect lines to be connected:")
(setq ss1(ssget '((0 . "line"))))
)
(setq pt0(getpoint"\nSelect a point to determine which ends will be connected: "))
(setq i 0)
(repeat (sslength ss1)
(setq lin0 (entget(ssname ss1 i)))
(setq pt1 (cdr(assoc 10 lin0)))
(setq pt2 (cdr(assoc 11 lin0)))
(if (<(distance pt1 pt0)(distance pt2 pt0))
(setq b1(append b1 (list pt1)))
(setq b1(append b1 (list pt2)))
)
(setq i (1+ i))
)
(while (>=(length b1) 2)
(setq pt_s1 (car b1))
(setq b2 (cdr b1))
(foreach pt_s2a b2
(if (= dst0 nil)
(setq pt_s2 pt_s2a
dst0(distance pt_s1 pt_s2a)
)
(if (<(distance pt_s1 pt_s2a) dst0)
(setq pt_s2 pt_s2a
dst0 (distance pt_s1 pt_s2a)
)
)
)
)
(setq dst0 nil)
(setq b1(append(cdr(member pt_s2 b2))(cdr(member pt_s2 (reverse b2)))))
(setq lin0(subst(cons 10 pt_s1)(assoc 10 lin0)lin0))
(setq lin0(subst(cons 11 pt_s2)(assoc 11 lin0)lin0))
(entmake lin0)
)
(princ)
)
Accepted solution
For a broader-scope solution, try CapEnd.lsp with its CE command, available >here<. The Line "capping" is the default, but it can also close such ends with an Arc, a circular-pipe cut-end symbol, or a cut line. It can be used on any object types with linearity that have ends [Lines, Arcs, partial Ellipses, open-ended Polylines and Splines, even Rays], and it verifies that the things selected line up and aim in the same direction at the selected ends. You do need to select each object, anywhere closer to the end you want capped. See more at the image in the link, and comments in the file.
Kent Cooper, AIA 
@Sea-Haven wrote:
Try this just drag 1st point - 2nd point over the 2 lines near the missing end.
[That's in answer to the original post from over 2 years ago, not to the latest reviving Message, since it does only one end-capping at a time.]
Apart from setting the current Layer to that of one of the Lines [what if they're on different Layers?] and not setting it back, I don't see any advantage to that over just having ENDpoint Osnap on [which I virtually always do anyway] and just using the plain old LINE command. You're going to need to pick twice near the desired ends, so just Osnap to them.
It can also be thrown off by geometric relationships. With the white being the existing Lines, and the dashed green being the drag-across line, this red result works as expected:
but in these kinds of situations it "sees" the wrong ends:
BUT it has the advantage over mine that it will at least do angled closures and connect ends of non-parallel Lines [if you need to do those] -- mine checks that the ends are parallel and line up in such a way that in the Line-closure option the result is perpendicular, because of its other closure options that don't make sense otherwise.
Kent Cooper, AIA
@MaccyDee wrote:
.... do you know of anyway to make this LSP or another LSP be able to close multiple pairs of lines at once? ....
That would be challenging. I assume you would want to select all of that stuff in a big window or something. So a routine would need to figure out which open ends are close to which other open ends -- probably by storing all ends of all non-closed Polylines [or Lines, if you might use those sometimes], eliminating any that have an end of another object at the same place, and doing distance comparisons. It may be possible, but I can't work on that right now.
Kent Cooper, AIA
And would they always be lined-up-and-parallel at their ends [the kind of relationships that CapEnd.lsp accepts], or might you sometimes have situations like these to close [red being the closure Lines]?
Kent Cooper, AIA
@Charlie.ford7Z4SW wrote:
....
Let me know if there is a way of putting a couple lines in this command to sort that problem
....
More than a "couple," but if Line cappings of open ends will do, this [add the blue stuff] seems to work in quickie testing:
....
(list of (- of))
)
(if (not (vlax-curve-isClosed obj))
(progn ; then
(setq OffsetResult2 (entlast))
(entdel OffsetResult2); temporarily, in order to:
(setq OffsetResult1 (entlast))
(entdel OffsetResult2); bring back in
(command
"_.line" "_non" (vlax-curve-getStartPoint OffsetResult1) "_non" (vlax-curve-getStartPoint OffsetResult2) ""
"_.line" "_non" (vlax-curve-getEndPoint OffsetResult1) "_non" (vlax-curve-getEndPoint OffsetResult2) ""
); command
); progn
); if
)
(vla-delete ss)
....
Kent Cooper, AIA
Or, if Arc [really Circle] closures are preferred, try this:
....
(list of (- of))
)
(if (not (vlax-curve-isClosed obj))
(progn ; then
(setq OffsetResult2 (entlast))
(command
"_.circle" "_non" (vlax-curve-getStartPoint obj) "_non" (vlax-curve-getStartPoint OffsetResult2) ""
"_.circle" "_non" (vlax-curve-getEndPoint obj) "_non" (vlax-curve-getEndPoint OffsetResult2) ""
); command
); progn
); if
)
(vla-delete ss)
....
Kent Cooper, AIA
@nathanERDLX wrote:
....
could be on my side but the 'download tip' button isn't functional anymore on Cadanalyst, ....
I just recently pointed that out to the people at Cadalyst, and they replied that they're working on fixing it. In the meantime, here's that one.
Kent Cooper, AIA
Hi there!
I was interested in using this LISP you edited, but its not working for me, I dont know why. The result of Charlie's code with your's edit looks like this for me. Dunno whats wrong? It puts only "ends" of offsets to new layer. Or maybe I am using it wrong. I though it create offset of all lines in layer and then create "cap" on ends?
EDIT: (I am using AC LT 2024)
