[css] Click through div to underlying elements

I have a div that has background:transparent, along with border. Underneath this div, I have more elements.

Currently, I'm able to click the underlying elements when I click outside of the overlay div. However, I'm unable to click the underlying elements when clicking directly on the overlay div.

I want to be able to click through this div so that I can click on the underlying elements.

My Problem

This question is related to css

The answer is


  1. Hide overlaying the element
  2. Determine cursor coordinates
  3. Get element on those coordinates
  4. Trigger click on element
  5. Show overlaying element again
$('#elementontop').click(e => {
    $('#elementontop').hide();
    $(document.elementFromPoint(e.clientX, e.clientY)).trigger("click");
    $('#elementontop').show();
});

I'm adding this answer because I didn’t see it here in full. I was able to do this using elementFromPoint. So basically:

  • attach a click to the div you want to be clicked through
  • hide it
  • determine what element the pointer is on
  • fire the click on the element there.
var range-selector= $("")
    .css("position", "absolute").addClass("range-selector")
    .appendTo("")
    .click(function(e) {
        _range-selector.hide();

        $(document.elementFromPoint(e.clientX,e.clientY)).trigger("click");
    });

In my case the overlaying div is absolutely positioned—I am not sure if this makes a difference. This works on IE8/9, Safari Chrome and Firefox at least.


I think the event.stopPropagation(); should be mentioned here as well. Add this to the Click function of your button.

Prevents the event from bubbling up the DOM tree, preventing any parent handlers from being notified of the event.


An easier way would be to inline the transparent background image using Data URIs as follows:

.click-through {
    pointer-events: none;
    background: url(data:image/gif;base64,R0lGODlhAQABAIAAAAAAAP///yH5BAEAAAAALAAAAAABAAEAAAIBRAA7);
}

Allowing the user to click through a div to the underlying element depends on the browser. All modern browsers, including Firefox, Chrome, Safari, and Opera, understand pointer-events:none.

For IE, it depends on the background. If the background is transparent, clickthrough works without you needing to do anything. On the other hand, for something like background:white; opacity:0; filter:Alpha(opacity=0);, IE needs manual event forwarding.

See a JSFiddle test and CanIUse pointer events.


Yes, you CAN force overlapping layers to pass through (ignore) click events.
PLUS you CAN have specific children excluded from this behavior...

You can do this, using pointer-events

pointer-events influences the reaction to click-, tap-, scroll- und hover events.


In a layer that should ignore / pass-through mentioned events you set

pointer-events: none; 

Children of that unresponsive layer that need to react mouse / tap events again need:

pointer-events: auto; 

That second part is very helpful if you work with multiple overlapping div layers (probably some parents being transparent), where you need to be able to click on child elements and only that child elements.

Example usage:

    <style>
        .parent {
            pointer-events:none;        
        }
        .child {
            pointer-events:auto;
        }
    </style>
    
    <div class="parent">
      <a href="#">I'm unresponsive</a>
      <a href="#" class="child">I'm clickable again, wohoo !</a>
    </div>

it doesn't work that way. the work around is to manually check the coordinates of the mouse click against the area occupied by each element.

area occupied by an element can found found by 1. getting the location of the element with respect to the top left of the page, and 2. the width and the height. a library like jQuery makes this pretty simple, although it can be done in plain js. adding an event handler for mousemove on the document object will provide continuous updates of the mouse position from the top and left of the page. deciding if the mouse is over any given object consists of checking if the mouse position is between the left, right, top and bottom edges of an element.


Nope, you can't click ‘through’ an element. You can get the co-ordinates of the click and try to work out what element was underneath the clicked element, but this is really tedious for browsers that don't have document.elementFromPoint. Then you still have to emulate the default action of clicking, which isn't necessarily trivial depending on what elements you have under there.

Since you've got a fully-transparent window area, you'll probably be better off implementing it as separate border elements around the outside, leaving the centre area free of obstruction so you can really just click straight through.


Another idea to try (situationally) would be to:

  1. Put the content you want in a div;
  2. Put the non-clicking overlay over the entire page with a z-index higher,
  3. make another cropped copy of the original div
  4. overlay and abs position the copy div in the same place as the original content you want to be clickable with an even higher z-index?

Any thoughts?


This is not a precise answer for the question but may help in finding a workaround for it.

I had an image I was hiding on page load and displaying when waiting on an AJAX call then hiding again however...

I found the only way to display my image when loading the page then make it disappear and be able to click things where the image was located before hiding it was to put the image into a DIV, make the size of the DIV 10x10 pixels or small enough to prevent it causing an issue then hiding the containing div. This allowed the image to overflow the div while visible and when the div was hidden, only the divs area was affected by inability to click objects beneath and not the whole size of the image the DIV contained and was displaying.

I tried all the methods to hide the image including CSS display=none/block, opacity=0, hiding the image with hidden=true. All of them resulted in my image being hidden but the area where it was displayed to act like there was a cover over the stuff underneath so clicks and so on wouldn't act on the underlying objects. Once the image was inside a tiny DIV and I hid the tiny DIV, the entire area occupied by the image was clear and only the tiny area under the DIV I hid was affected but as I made it small enough (10x10 pixels), the issue was fixed (sort of).

I found this to be a dirty workaround for what should be a simple issue but I was not able to find any way to hide the object in its native format without a container. My object was in the form of etc. If anyone has a better way, please let me know.


Just wrap a tag around all the HTML extract, for example

<a href="/categories/1">
  <img alt="test1" class="img-responsive" src="/assets/photo.jpg" />
  <div class="caption bg-orange">
    <h2>
      test1
    </h2>
  </div>
</a>

in my example my caption class has hover effects, that with pointer-events:none; you just will lose

wrapping the content will keep your hover effects and you can click in all the picture, div included, regards!


I think that you can consider changing your markup. If I am not wrong, you'd like to put an invisible layer above the document and your invisible markup may be preceding your document image (is this correct?).

Instead, I propose that you put the invisible right after the document image but changing the position to absolute.

Notice that you need a parent element to have position: relative and then you will be able to use this idea. Otherwise your absolute layer will be placed just in the top left corner.

An absolute position element is positioned relative to the first parent element that has a position other than static. If no such element is found, the containing block is html

Hope this helps. See here for more information about CSS positioning.


I couldn't always use pointer-events: none in my scenario, because I wanted both the overlay and the underlying element(s) to be clickable / selectable.

The DOM structure looked like this:

<div id="outerElement">
   <div id="canvas-wrapper">
      <canvas id="overlay"></canvas>
   </div>
   <!-- Omitted: element(s) behind canvas that should still be selectable -->
</div>

(The outerElement, canvas-wrapper and canvas elements have the same size.)

To make the elements behind the canvas act normally (e.g. selectable, editable), I used the following code:

canvasWrapper.style.pointerEvents = 'none';

outerElement.addEventListener('mousedown', event => {
    const clickedOnElementInCanvas = yourCheck // TODO: check if the event *would* click a canvas element.
    if (!clickedOnElementInCanvas) {

        // if necessary, add logic to deselect your canvas elements ...

        wrapper.style.pointerEvents = 'none';
        return true;
    }

    // Check if we emitted the event ourselves (avoid endless loop)
    if (event.isTrusted) {
        // Manually forward element to the canvas
        const mouseEvent = new MouseEvent(event.type, event);
        canvas.dispatchEvent(mouseEvent);
        mouseEvent.stopPropagation();
    }
    return true;
});

Some canvas objects also came with input fields, so I had to allow keyboard events, too. To do this, I had to update the pointerEvents property based on whether a canvas input field was currently focused or not:

onCanvasModified(canvas, () => {
    const inputFieldInCanvasActive = // TODO: Check if an input field of the canvas is active.
    wrapper.style.pointerEvents = inputFieldInCanvasActive  ? 'auto' : 'none';
});

I needed to do this and decided to take this route:

$('.overlay').click(function(e){
    var left = $(window).scrollLeft();
    var top = $(window).scrollTop();

    //hide the overlay for now so the document can find the underlying elements
    $(this).css('display','none');
    //use the current scroll position to deduct from the click position
    $(document.elementFromPoint(e.pageX-left, e.pageY-top)).click();
    //show the overlay again
    $(this).css('display','block');
});

You can place an AP overlay like...

#overlay {
  position: absolute;
  top: -79px;
  left: -60px;
  height: 80px;
  width: 380px;
  z-index: 2;
  background: url(fake.gif);
}
<div id="overlay"></div>

just put it over where you dont want ie cliked. Works in all.


I currently work with canvas speech balloons. But because the balloon with the pointer is wrapped in a div, some links under it aren't click able anymore. I cant use extjs in this case. See basic example for my speech balloon tutorial requires HTML5

So I decided to collect all link coordinates from inside the balloons in an array.

var clickarray=[];
function getcoo(thatdiv){
         thatdiv.find(".link").each(function(){
                 var offset=$(this).offset();           
                 clickarray.unshift([(offset.left),
                                     (offset.top),
                                     (offset.left+$(this).width()),
                                     (offset.top+$(this).height()),
                                     ($(this).attr('name')),
                                     1]);
                                     });
         }

I call this function on each (new) balloon. It grabs the coordinates of the left/top and right/down corners of a link.class - additionally the name attribute for what to do if someone clicks in that coordinates and I loved to set a 1 which means that it wasn't clicked jet. And unshift this array to the clickarray. You could use push too.

To work with that array:

$("body").click(function(event){
          event.preventDefault();//if it is a a-tag
          var x=event.pageX;
          var y=event.pageY;
          var job="";
          for(var i in clickarray){
              if(x>=clickarray[i][0] && x<=clickarray[i][2] && y>=clickarray[i][1] && y<=clickarray[i][3] && clickarray[i][5]==1){
                 job=clickarray[i][4];
                 clickarray[i][5]=0;//set to allready clicked
                 break;
                }
             }
          if(job.length>0){   
             // --do some thing with the job --
            }
          });

This function proofs the coordinates of a body click event or whether it was already clicked and returns the name attribute. I think it is not necessary to go deeper, but you see it is not that complicate. Hope in was enlish...