Hi !
I try to look for this but in fact it seems more complicated than expected, and I didn't find a working up to date solution.
So we have layer bound property, which can return top left value of a layer.
But if the layer is smart layer (or a text) this may not work as I would have expect: indeed, it will return the value of the position of the first pixel, and not the actual smart layer position (determined by it's original boundaries).
Here is a screen capture to show the difference between the layer top left coordinate and the transform X Y position (which match the original file placement):
How do you get these X Y positions from a Photoshop script ?
Thanks ! 
1 Correct answer
Try this function
...function get_layer_bounds(layer)
{
try
{
if (layer.kind == LayerKind.SMARTOBJECT)
{
var layer0 = activeDocument.activeLayer;
activeDocument.activeLayer = layer;
var r = new ActionReference();
r.putEnumerated( charIDToTypeID( "Lyr " ), charIDToTypeID( "Ordn" ), charIDToTypeID( "Trgt" ) );
var d = executeActionGet(r).getObjectValue(stringIDToTypeID("smartObjectMore")).getList(stringIDToTypeID("tr
Explore related tutorials & articles
23
Replies
r-bin​
Hi, thanks for your very quick assistance (and your initial code snippet) !
This get_smart_object_corners() function doesn't take any parameter, how come ?
How to pass a Layer as parameter to it ?
My ultimate goal is to merge this code snippet to
As you can see L44, the function need to take a Layer as parameter.
Many thanks again!
(let me know if you prefer me to answer in the original thread).
I have never seen any code posted that can get a smart object layers associated transform setting. Never seen code to get any of the scaling values.
I see this function the gets the layers pixels bonds where activeDoctment.activelayer.bounds may return a larger area if the layer has a layer style that extends outside the layers pixel area, like a drop shadow .
I have seen the code that you posted to get the layer transform four corners to calculate a rotation angle in CC 2018 that may work if the layer transform simply sized and rotated the object by doing the right triangle math for the slope of a side after getting the layers four coners points. However that will not be correct if the layers associated transform has some perspective or other distortion. Also when I tested the function. The function did not return the transform's four corners.
I have never seen code that can get the values that Photoshop shows in the transform tools option bar. Fot the smart object layers assocated transform.
]
Perhaps you can but that is above my pay grade. I do not know JavaScript or java objects structures to be able to parse them or process them well. Do not know to retriever some Photoshop information using action manager code. I only modify Scriptlistener generated Action manager code you know the execute side of Action manger.
I'm older and retired I look for code from younger whippersnappers so I can hack on it. If I have to say it myself, I will, I may need to use a sledge hammer but I can hammer on javascript and often get it to do what I want it to do.
The first place I look for code is hear searching this forum. And I'm sadden when great scripters like Paul leave when other like Mike pass away while X is still maintaining Image Processor Pro but only drops in here once in a while. In other words I need you to stay here.
Thank you now that works well and has given me some other things to think about. It interesting to me how the layers bounds area and the transform points area ciffer, Don't know if I can make any use out of that.
// Select Smart Object Layers Associated transform area
app.activeDocument.selection.select(get_smart_object_corners());
////////////////////////////////////////////////////////////////////////////////////////// //
function get_smart_object_corners() {
try
{
var r = new ActionReference();
r.putEnumerated( charIDToTypeID( "Lyr " ), charIDToTypeID( "Ordn" ), charIDToTypeID( "Trgt" ) );
var d;
try { d = executeActionGet(r); } catch (e) { alert(e); return; }
try { d = d.getObjectValue(stringIDToTypeID("smartObjectMore"));} catch (e) { alert(e); return; }
try { d = d.getList(stringIDToTypeID("nonAffineTransform"));} catch (e) { alert(e); return; }
var ret = [[d.getDouble(0),d.getDouble(1)],
[d.getDouble(2),d.getDouble(3)],
[d.getDouble(4),d.getDouble(5)],
[d.getDouble(6),d.getDouble(7)]];
return ret;
}
catch (e) { alert(e); }
}
r-bin​
Many thanks for your expertise on this code snippet !! it seems to works nicely,
at least in my simple case (no perspective or complex transform on layer).
Without your generous help, it would have been impossible. 
Does it works for you for these complex scenarios ?
If yes, I'll mark r-bin answer last code snippet as the good one.
I agree with you that it is suprising that a simple property like that (it is shown in plain text in some text-input of the UI) needs to deal with such complex function on the script side. That was a very unexpected behavior, as nothing on the documentation explained that position would behave weirdly with smart objects.
At least we have a workaround now !
If you exclude Transform Warp and I think you should all the other Transform have the same user interface only their operation differs. How dragging with the mouse function differ. All record in the history palette as free transform even Transform Warp. Transform Warp though has a different UI interface and the same interface in the tools options bar there is a little icon that you can to toggle between the two transform UI. You can use both transforming a layer. So a transformed smart object transform can have setting for both interfaces. I would think all smart object layers will have standard free transform settings and some Smart object layers associated transform may have additional transform warp settings. The script seems the get the Free Transform four corners in a sequence that a selection using them will show when extreme distortion was used causing sides to crossover an other,
JJMack​
These are definitely complex cases, where the definition of Top Left XY may no be the same for everyone one
I consider this should be part of a wider feature request for getting XY of all Smart Layers corners.
I mark r-bin​ solution as a good one for now (as it does what I was looking for in my case), and I'll update if he or someone else come with a solution which could handle these cases.
Thanks for having taken the time to digg these possible issues !
JJMack​
It is this function that solves the problem I had and which I implemented in my script,
Your code snippet doesn't have Layer as function parameters as I requested in my second post, r-bin post was an answer to that.
But yes, your code snippet do contains a part of the solution as well... But I can't mark two posts the "the correct answer". :S
I marked your post as "useful" and "recommended" though
Thanks again for your contribution on this problem!
I just got a very particular cases,
How would you get X and Y position relative to the Artboard position ? Seems to be a quite complex API case.
I had a layer on a artboard which was offset by 961px, and all Y were offset by 961px. Idealy, I would prefer to have it not offset.
Thanks for your help
AdChoices