Community
Does anybody know of a routine that will allow to draw a polyline through existing blocks, in the order the blocks should be numbered? Therefore when the sequence has been specified by the polyline then user to define the number system required, and the blocks are automatically numbered in the sequence of the polyine start to finish. 
@kirwan59 wrote:..... draw a polyline through existing blocks, in the order the blocks should be numbered? Therefore when the sequence has been specified by the polyline then user to define the number system required, and the blocks are automatically numbered in the sequence of the polyine start to finish.
Will it keep the polyline?
What is your number system?
Mulitple attributes in one block?
Post a sample of your block.
Try this: <as a starting point>
(defun c:test (/ strt pt1 pt2 ptl blk)
(vl-load-com)
(setq strt (getint "\nStart Number:"))
(setq pt1 (getpoint "\nPick point:") ptl (list pt1))
(while (setq pt2 (getpoint pt1 "\nNext point"))(grdraw pt1 pt2 7)
(setq ptl (cons pt2 ptl) pt1 pt2)
)
(if ptl (progn
(entmakex (append (list (cons 0 "LWPOLYLINE")
(cons 100 "AcDbEntity")
(cons 100 "AcDbPolyline")
(cons 90 (length ptl))
)
(mapcar (function (lambda (p) (cons 10 p))) ptl)))
(foreach pts (reverse ptl)
(setq blk
(vlax-invoke
(vlax-get (vla-get-ActiveLayout
(vla-get-activedocument
(vlax-get-acad-object)))
'Block)
'InsertBlock
pts
"BlockName"
1 1 1 0))
(vla-put-textstring (car (vlax-invoke blk 'Getattributes)) (itoa strt))
(setq strt (1+ strt))
)
)
)
)
This code assume the block "blockname" is existing, only one attribute on the block, and the polyline will remain,
@kirwan59 wrote:...a routine that will allow to draw a polyline through existing blocks, in the order the blocks should be numbered? .... the blocks are automatically numbered in the sequence of the polyine start to finish.
If you're only using the Polyline to determine the order, can you do it without that, and just let the order of selection of the Blocks govern? This is a thread with a routine that numbers Block attributes in selection order, which you may be able to use as a starting point:
Search the Discussion Group further -- sequential numbering routines are out there in a variety of formats [using Attributes vs. placing Text, with Attributes as Blocks are inserted or with pre-existing Blocks, selecting objects vs. just designating locations, using different numbering systems, etc., etc.].
Kent Cooper, AIA 
@kirwan59 wrote:.... The number will either run top (T) to bottom (B), left (L) to right (R), or L to R, T to B, or B to T, L to R, etc. There are a number of ways the blocks will be numbered, however they will run in sequence, and will be contiguous. .... By using a pline draw through the blocks in the sequence that you want to order the attributes. ....
I think the ordering part of it could be accomplished without the need to draw a Polyline, if it's always regular in some direction such as you describe. This has a routine that does something like that for Text entities, based on insertion points:
It would need another argument and further code for the secondary direction of sorting [it sorts only in one direction, so if you call for Left to right, it groups columns but the order within each column may vary], but it could surely be adapted.
Just to clarify: by "top to bottom, left to right", do you mean that the left-most column of Blocks would be numbered first, from top to bottom, and then the next column to its right would be numbered from top to bottom, etc.?
1 4 7
2 5 8
3 6 9
Or do you mean the top-most row would be numbered first, from left to right, then the next row below it from left to right?
1 2 3
4 5 6
7 8 9
To make it unambiguous, you could build something with prompts such as "First row or column to be numbered [T/B/L/R] :" followed by something like (if they typed T or B) "End to start numbering rows [L/R] :" or (if they typed L or R) "End to start numbering columns [T/B] :".
EDIT:
Or, come to think of it, since you talk about them always being contiguous, would it be a switchback situation, such as this?
1 2 3
6 5 4
7 8 9
or this?
1 6 7
2 5 8
3 4 9
In a case like those, the second prompt would ask for the end at which to start numbering that designated initial row or column, rather than all rows or columns.
Kent Cooper, AIA
@Kent1Cooper wrote:....
I think the ordering part of it could be accomplished without the need to draw a Polyline, if it's always regular in some direction such as you describe. ....
The attached BlockSSSort.lsp [= Block Selection Set Sort] seems to do that, in limited testing. But it does need it to be a regular sequence, such as top-to-bottom/left-to-right, not a meandering order. And it assumes accurately aligned rows and columns as your description of a grid suggests [within the precision argument in the (equal) function, which you can adjust].
It leaves a list of Block entity names in sorted-in-both-directions-as-instructed order, in the blsort2 variable, and also returns that list. [If the answer to my previous post's sequencing question is that it should be a switchback sequence, adjustments would need to be made.]
This way, you don't need to draw any Polyline, nor find the Blocks by looking around at the Polyline's vertices, nor select the Blocks in any particular order. You can just designate the sort orders in the routine's arguments, and select the Blocks by any method(s) you prefer.
I'll leave to others, at least for the time being, the stepping through the sorted list and assigning the incremented text content to the Attribute [if that hasn't already been sufficiently addressed].
Kent Cooper, AIA
The way i see it, there are two ways you can do this.
1. Draw a polyline for your sequence as preparation (in grid like format) until your project manager tells you the sequence.which means for all the possible sequence you have a pline passing thru those location. then invoke a routine to insert the block that increments automatically according to your numbering system.
2. Like Kents suggestion use a position sort routine. Blocks are already in place, then invoke the routine to increment the numbers based on the conditions your PM's decision.
Read the syntax in Kents lisp file found at the header.
@kirwan59 wrote:....
I have tried to use your BlockSSSort.lsp but it loads successfully (APPLOAD), but when I try to invoke the command I get
'Command: BSSS
Unknown command "BSSS". Press F1 for help.'
Am I doing this correctly? ....
See the Usage note near the top of the file. (BSSS "R" "B") or (BSSS "T" "L") or ....
Kent Cooper, AIA
@kirwan59 wrote:... the problem is that the blocks could be inherant to the drawing when I receive it, this means I have no clue to what order the blocks were inserted. As to the number sequence and direction it can be either dependent upon how the Project Manager wants to start and complete numbering sequence. .... ...if the numbering sequence does jump from one area to another, then jumps back again that I can still control the sequencing as the polyline will be the determining factor! Plus I don't want to be picking blocks manually to sequence, especially if I have 600 blocks to do .... ...the starting point is how could you program for a function to return a list of blocks, that run in sequence, that are touching a polyline. I....
To sort them with (BSSS), it doesn't matter whether the Blocks were already there or you put them in, or in what order they were inserted; they just need to be there to be selected. The whole point of it is to sort them positionally regardless of insertion order or anything else.
It appears that this in Message 6:
"The number will either run top (T) to bottom (B), left (L) to right (R), or L to R, T to B, or B to T, L to R, etc."
is not always the case, but that's what I based my suggestion on.
I suspect that to find all Blocks that lie along a Polyline might be a challenge if they are not at the Polyline vertices, but maybe it's not insurmountable [see below]. If they're at the vertices, meaning that you would have to draw a Polyline with a vertex at each Block [presumably at its insertion point], I don't see that as being any less daunting than picking the Blocks manually in sequence, except that it would be easier to change the order later, with Pedit.
If:
a. the Blocks are small enough in relation to the spacing between rows and columns, and
b. there are not other kinds of things with insertion points [such as Text] around to complicate things, and
c. the Polyline doesn't go by any out-of-sequence Blocks when it jumps from one area to another,
then I guess it would be possible to run along the Polyline at smaller-than-Aperture-box increments, applying Osnap Insert, looking to find Blocks even if they're not all at vertices. It might well depend on there being some element of the Block at its insertion point [not for example, a square with the insertion point at the center but no drawing entity there], so that once Osnap finds an insertion point, it can look with something like (ssget "C") for the Block that goes with it.
Kent Cooper, AIA
@stevor wrote:....
Also, it may work better if another variable name is added to the local variable list, namely: blsort2.
Where is reads "blsort1 coord " make it "blsort1 blsort2 coord "
....
I deliberately did not localize the blsort2 variable, so that the list, stored in a variable for access, would be available after the routine is finished. But it would depend on how it might be incorporated into something larger. Since it returns that list as well as storing it in that variable, things could be set up with that variable localized within this subroutine, if appropropriate in the larger context. But I would be inclined to localize it at the level of the larger routine of which this might be a part, assuming that's the setup.
Kent Cooper, AIA
@kirwan59 wrote:This is Kent's header
(defun BSSS (first dir / bss bent blist blsort1 coord clist csort); = Block Selection Set Sort
I assume the routine is initialised by typing BSSS at the command prompt.
To elaborate a little on stevor's reply:
If it began with
(defun C:BSSS (/ ....)
with the C:, and with no arguments listed before the slash in the parentheses, then yes, you would invoke it by typing BSSS at the Command: prompt. That's the way a "command"-type function is defined.
But without the C:, and with arguments listed before that slash, it's not a command-type but a "sub-routine"-type function, and needs to be used with its name in parentheses, with the argument values supplied, as in the Usage comment in the file.
Kent Cooper, AIA
![Autodesk Design & Make Report](https://forums.autodesk.com/t5/image/serverpage/image-id/1350519iCB55845EBFC43016/image-size/large?v=v2&px=999" "Autodesk Design & Make Report")
