koobas.hobune.stream
Finding the starting and ending coordinates of a flick / swipe on iOS
I'm having a hard time with Unity's iOS touch controls.
I'd like to Instantiate an object using the flick / swipe command with the object instantiating at the flick / swipe starting point and pointing at the flick / swipe end point.
The best I've managed is to use Input.touchCount > 0 to get startPoint = Input.GetTouch(0).position. The whole thing is nestled in Update so it constantly gets recalculated which is probably going to cause problems later.
For endPoint, I've written if(Input.GetTouch(0).phase == TouchPhase.Ended || Input.GetTouch(0).phase == TouchPhase.Canceled) then touchEnd = Input.GetTouch(0).position.
It doesn't work of course, so I was wondering - how can I get the starting points and ending points of a swipe?
I haven't programmed any swipe gestures yet, but here's how I plan to do it:
- define a speed threshold: the minimum speed (or delta distance) that determines a swipe
- each frame that has touchCount == 1, check if the speed is greater than the threshold
- if it is (and it wasn't the last frame) store the start position
- if it is not (and it was the last frame) store the end position
- then you would do whatever you do after a swipe
Ok... I'm bored.
Here's some tested exactly twice code.
I'm sure it needs tweaking, but it's a good start:
var swipeThresh : float = 1.2; var swipeStart : Vector2 = Vector2.zero; var swipeEnd : Vector2 = Vector2.zero; var swipeWasActive : boolean = false;
function Update () { if ( Input.touchCount == 1 ) { processSwipe(); } }
function OnGUI () { processSwipe(); drawStartBox(); drawEndBox(); }
function processSwipe () { if ( Input.touchCount != 1 ) { return; }
var theTouch : Touch = Input.touches[0];
/* skip the frame if deltaPosition is zero */
if ( theTouch.deltaPosition == Vector2.zero ) {
return;
}
var speedVec : Vector2 = theTouch.deltaPosition * theTouch.deltaTime;
var theSpeed : float = speedVec.magnitude;
var swipeIsActive : boolean = ( theSpeed > swipeThresh );
if ( swipeIsActive ) {
if ( ! swipeWasActive ) {
swipeStart = theTouch.position;
}
}
else {
if ( swipeWasActive ) {
swipeEnd = theTouch.position;
Debug.Log("Swipe Complete");
}
}
swipeWasActive = swipeIsActive;
}
function drawStartBox () { if ( swipeStart == Vector2.zero ) { return; }
/* don't forget to invert the y-coordinate */
var theY = Screen.height - swipeStart.y;
var theX = swipeStart.x;
var theW = 140;
var theH = 40;
var theRect : Rect = Rect(theX, theY, theW, theH);
GUI.Label(theRect, "Start");
}
function drawEndBox () { if ( swipeEnd == Vector2.zero ) { return; }
/* don't forget to invert the y-coordinate */
var theY = Screen.height - swipeEnd.y;
var theX = swipeEnd.x;
var theW = 140;
var theH = 40;
var theRect : Rect = Rect(theX, theY, theW, theH);
GUI.Label(theRect, "End");
}
I just gave this a try and it works surprisingly well...
I used a speed threshhold of 1.2, which can be adjusted in the Inspector.
Hope this helps!
I'm going to need to tinker with this code a bit to get it to work in my project - but it does what I've asked. Thanks for taking the time to reply and the helpful input!
As you don't know whether something is a flick until after the user has both touched the screen and released, you need to save the touch data for a period of time. Using just Touch.deltaPosition only allows you to look one frame back, which is (hopefully) an extremely short amount of time.
The basic approach is to create a data structure that has all of the inputs of a specific finger id. You will want to save the Touch data and a timestamp. When any Touch is TouchPhase.Ended, then look through all that saved data to determine whether you have a flick. To determine whether you have a flick look backwards through your list of points and decide if the speed of input is a flick.
I don't understand why all this would be necessary (and my approach wouldn't work), but I guess I'll find out when I go to add it to my project soon. I don't have an iPad build set up right now...
At the very least for your approach you should iterate through all of the touches ins$$anonymous$$d of just looking at the list. When dealing with touches Input.mousePosition is the average position of all the touches on the screen, so if you detect a swipe with finger 1 and there are two other fingers on the screen, your start & end positions will not be at the location of the swipe. Additionally the start and end positions are a Vector2 not a float, so you want to save both the x & y positions, not just the x.
Finally, we found for our uses that the Touch.deltaPosition was not enough information to make a reliable decision regarding about whether there was a flick or not and what direction the user intended the flick to go in.
That's just sample/pseudo code thrown together to suggest a possible approach...
Input.mousePosition is a typo and was meant to be theTouch.position....
Since it's just example code I only deal with one touch and one direction for simplicity.
Obviously I'm not trying to provide a complete solution here...
Wow, looks like it's a lot more complex than I thought. Didn't seem like it would be so hard since there are a ton of developers out there using flick-based games, but when you're trying to distinguish between taps, holds, and other input methods then I guess it's necessary.
Will try a few more things out before marking an answer. Thanks for your input!
I wish I could do an iOS build right now... I had to restore my iPad recently and am having troubles with the provisioning stuff.
I would love to play with this... I just tested my code with Unity remote (after a few changes) and was pleasantly surprised with how well it seemed like it might be working.
It's hard to really test with Unity Remote....
