Is there anything that can detect drag gesture sta...

# compose

[Vivek Sharma]

Is there anything that can detect drag gesture started and stopped
I am currently using it like this:

Modifier.pointerInput(Unit) {
    detectTransformGestures { centroid, pan, zoom, rotation ->
        offset += pan
        scale *= zoom
    }
}

I am zooming the image and I need to know when the zoom drag is ended like we removed the fingers?
Further info 👇

[Vivek Sharma]

I want to use zooming + panning/dragging
But when I use detectTransformGestures , I can't control onDragEnd: () -> Unit , there is no such param

[Vivek Sharma]

But when I use detectDragGestures , I have onDragEnd: () -> Unit , But I can't control zooming , only I can control panning/dragging

[Albert Chang]

You can use both. See my implementation.

[Michael Paus]

I stumbled over the same issue some time ago and I think it is an omission in the API which makes things unnecessarily complicated.

[Vivek Sharma]

@Michael Paus yeah I think so 😄, but there are some things to look in the above repo

[Albert Chang]

IMO current API with coroutines integrated is much better compared to touch event handling in the view system as it allows you to separate logics of handling different gestures while in the view system you have to mix all the logics. If you want to detect multiple gestures, just use multiple detect*Gestures functions. Nothing counterintuitive.

[Vivek Sharma]

I tried using it like

Modifier.pointerInput(Unit) {
    detectTransformGestures { centroid, pan, zoom, rotation ->
        offset += pan
        scale *= zoom
    }
    detectDragGestures {
       // used onDragEnd() something here
    }
}

But this thing didn't work and only the lambda fun used first was working only

[Vivek Sharma]

@Albert Chang I looked into your Repo, this thing looks complex but I have no other option I guess

[Albert Chang]

As those are suspend functions, you need to use coroutineScope and launch as shown here. If you are familiar with coroutines, there’s nothing special here.

[Michael Paus]

@Vivek Sharma Somewhere else here in kotlinlang I was told that putting multiple detect*s into a single pointerInput is wrong. One should always do it like this:

Modifier
.pointerInput(Unit) {
    detectTransformGestures { centroid, pan, zoom, rotation ->
        offset += pan
        scale *= zoom
    }
}
.pointerInput(Unit) {
    detectDragGestures {
       // used onDragEnd() something here
    }
}

But this won’t change anything here because detectTransformGestures and detectDragGestures are mutually exclusive.

[Vivek Sharma]

I also tried using this before asking here!
And I found that they don't behave as we want and here only one thing was working , other was not

[Albert Chang]

That’s why I’m using Modifier.transformable() instead of detectTransformGestures.

[Vivek Sharma]

But can we control onDragEnd something like that in Modifier.transformable() ?

[Albert Chang]

You can't so I said you can use both.

[Albert Chang]

@Michael Paus
To be clear, detecting multiple gestures in the same PointerInputScope is valid as documented here.

More than one awaitPointerEventScope can run concurrently in the same PointerInputScope by using kotlinx.coroutines.launch. `block`s are dispatched to in the order in which they were installed.

Also see design note here.

[Michael Paus]

@Albert Chang This contradicts the following answer here https://kotlinlang.slack.com/archives/CJLTWPH7S/p1639070958379400?thread_ts=1638996490.330900&cid=CJLTWPH7S and also this issue in the issue tracker: https://issuetracker.google.com/issues/209937353?pli=1