The text inside the block moves to the boundary？
I don't like to sound discouraging, but that seems like a near-impossible task to automate.
The only way I can even start to think about how to do it would be to have a routine run an 'IntersectWith function on each piece of Text in comparison with each of the Block's perimeter Polylines, to find out whether it overlaps any of them. Where it does, presumably it would have to compare the intersection locations with the Block's overall extents to figure out which way to Move the Text. It would Move it a little, try again, and repeat until it no longer intersects with that boundary object.
But several complications arise:
In the case of the "1" at the top-most "peak" of the ST2 Block, the X direction that would get the Text closer in its X coordinate to the X coordinate of the Block's middle would be leftward, so it would presumably Move it in the wrong direction, and end up with it outside the boundary when it finally doesn't intersect that edge any more.
If it can somehow [and I don't have any idea how yet] be made to figure out a way to go in the right direction, in similar situations [not quite true here] it could easily Move a piece of Text until it clears the one edge, but then it could intersect another edge. So it would have to compare the Text to every edge again, and presumably Move it [in this case] downward a little [closer to the Block midpoint's Y coordinate] and go through the whole exercise again, hoping that eventually it will get it clear of one edge without crossing another one.
Then there's the fact that the IntersectWith method works with the rectangular bounding box of the Text. Again, in the case of that "1", because of the angle of the boundary where it intersects it, if a routine Moves it to the right, it would have to Move it farther to get no intersection than is necessary for the character itself to clear the boundary. So it could end up Moving it more than is really required, and causing an intersection with another boundary object unnecessarily.
And what if a Text object is originally fully outside the boundary, and doesn't intersect any edge? How would a routine know it had to be Moved at all, let alone in what direction? That might be possible if the boundary of each Block was one Polyline [there are routines around here to determine whether or not a location is inside a closed Polyline], but when they're separate pieces like this, some of them concave and some convex in relation to the overall shape, it's really hard to imagine how a determination could be made.
I'm stumped.... I think the time it would take you go do it manually, even on a pretty large number of such Blocks, would be far less than the time it would take someone to come up with code to do it automatically.
Kent Cooper, AIA 
@john.uhden wrote:
I don't get it. You seem to be saying that the text moves (itself?) to the boundary, but the graphic shows it in the middle of the boundary.
Whatever the case, maybe you want to make the text an attribute of the block, middle justified, so that it stays relatively where you want it with respect to the block.
....
It's not the obvious larger ones in the middles, but the tiny ones in the corners.
In some cases, just changing their justifications from Left to Middle might do it, but A) that wouldn't work for all of them [e.g. the lower right corner one of ST1, which is currently inside but would then hit a boundary], and B) it would depend on the number of characters in each.
Kent Cooper, AIA
@1127204185 wrote:
Can that friend help write a lisp?
If by "that friend" you mean me, I still feel the same as what I said in the last paragraph of my first Reply [with correction -- "...the time it would take you to do it manually..."].
Kent Cooper, AIA
@Scottu2 wrote:
Perhaps the method would be...
First to find the corners (endpoint of the polyline) and the center of the object (average X's and Y's).
Then change the text to middle and move them by insert point to nearest end polyline end point within some +/- tolerance.
Then move the text 2 or 3 times its text height toward the center point.
I suspect the viability of that would depend a lot on the actual shape of the boundary. It doesn't look to me as though 3x the height would Move them far enough at the top or bottom ends of the ST2 shape in the original image. And at the left corner of that one, basing the Move displacement on the Text height when the direction would be essentially horizontal might succeed for short strings, but if the string lengths vary, could well not Move longer ones far enough to clear.
So many variant possibilities -- it seems unlikely that a universal solution is possible.
But I'm curious to know whether they are in fact using Acad LT, if @devitg's command-line excerpt really means the drawing was done in that. If so, there's no point in thinking about an AutoLisp approach. But @1127204185 did ask for that in Post 4, which I hope means they would be able to use such a routine.
Kent Cooper, AIA
@1127204185 wrote:
why
Because LT is for "LighT," and a light-weight version of any program is going to have fewer capabilities than the full version -- that's why it's less expensive. One of those differences in this case is that AutoLisp capability is simply not included in LT -- you need to have full AutoCAD, or one of the overlay programs built on full AutoCAD, to get it.
Kent Cooper, AIA
@roland.r71 wrote:
....
Which doesn't mean he actually uses the LT.
....
That was my uncertainty in the last paragraph of Post 10. We haven't heard specifically from the OP that they do use LT, but I assume so from Posts 12 & 14.
Kent Cooper, AIA
@Kent1Cooper wrote:
-- "...the time it would take you to do it manually..."].
Same here.
It makes me wonder why the need for a program and the "TEXT" is not even an attribute that would change from time to time. Maybe there's more to this than what the OP is telling us.
