leaflet map does not appear correctly until resize - scala

I am using Leaflet with scalajs-leaflet facade on Binding.scala, and the map initializes/appears incorrectly.
In order to reproduce the issue, i have prepared a lihaoyi/workbench page similar to that in scalajs-leaflet.
First, download the forked scalajs-leaflet from https://github.com/mcku/scalajs-leaflet
Run sbt in scalajs-leaflet directory.
Enter ~ example/fastOptJS in sbt. Now, a web server started at port 12345.
Open
http://localhost:12345/example/target/scala-2.12/classes/leaflet2binding-dev.html in a browser
The problem is the map container appears but the content (tiles, etc) is not correct. Map becomes fine after a small resize on window, which triggers _onResize handler of leaflet.
The container is in the Leaflet2Binding.scala file and has its size already specified prior to initialization:
val mapElement = <div id="mapid" style="width: 1000px; height: 600px;
position: relative; outline: currentcolor none medium;"
class="leaflet-container leaflet-touch leaflet-fade-anim
leaflet-grab leaflet-touch-drag leaflet-touch-zoom"
data:tabindex="0"></div>.asInstanceOf[HTMLElement]
It is possible to insert a line lmap.invalidateSize(true) in the following line before returning the element
https://github.com/mcku/scalajs-leaflet/blob/83b770bc76de450567ababf6c7d2af0700dd58c9/example/src/main/scala/example/Leaflet2Binding.scala#L39, but did not help in this case. Namely here:
#dom def renderMap = {
val mapElement = ... (same element as above)
.. some other initializations ..
lmap.invalidateSize(true) // true means, use animation
println("mapElement._leaflet_id " +mapElement.asInstanceOf[js.Dynamic]._leaflet_id) // prints non-null value, makes me think the container is initialized
mapElement
}
Any ideas? This is binding.scala specific, but it may be a leaflet issue as well.
EDIT Possible workaround
It looks like, the map element has its clientWidth property not available during the process. Which is understandable as the document is not "ready" yet. However, the css style.width is available and could be defined in px. In that case it is possible to patch leaflet to take css style width into account during computation.
it works if the style width is specified in px.
diff --git a/src/map/Map.js b/src/map/Map.js
index b94dd443..6544d7b7 100644
--- a/src/map/Map.js
+++ b/src/map/Map.js
## -903,8 +903,9 ## export var Map = Evented.extend({
getSize: function () {
if (!this._size || this._sizeChanged) {
this._size = new Point(
- this._container.clientWidth || 0,
- this._container.clientHeight || 0);
+
+ this._container.clientWidth || parseInt(this._container.style.width.replace("px",""),10) || 0,^M
+ this._container.clientHeight || parseInt(this._container.style.height.replace("px",""),10) || 0);;^M
this._sizeChanged = false;
}

Maybe lmap.invalidateSize(true) is called too early (DOM is not ready or repainted).
Make sure this does not happen. To prevent that I wrap this code like:
setTimeout(function () {
mapid.invalidateSize(true);
}, 100);
This must be done after every DOM repaint.

People that are still facing this issue for react-leaflet, making the component only render when zoom and center are initiated, and removed when the Map is not displaying:
useEffect(() => {
setCenter([50.5, 30.5]);
setZoom(14);
}, []);
And then in the JSX
{zoom && center.length > 0 && (
<div className={styles.map}>
<Map center={center} zoom={zoom} />
</div>
)}

Related

Modal for fullsize image with gatsby-image - limit height and width

What I want to achive
I am using gatsby and want to design an image gallery. Clicking on one of the images shall open a modal, which: (1) is showing the image in maximum possible size, so that it still fits into the screen and (2) is centered in the screen.
My Code
/* imagemodal.js */
import React from 'react'
import * as ImagemodalStyles from './imagemodal.module.css'
import { Modal } from 'react-bootstrap'
import Img from 'gatsby-image'
import { useStaticQuery, graphql } from 'gatsby'
export default function Imagemodal() {
const data = useStaticQuery(graphql`
query {
file(relativePath: { eq: "images/mytestimage.jpg" }) {
childImageSharp {
fluid(maxWidth: 1200) {
...GatsbyImageSharpFluid
}
}
}
}
`)
return (
<div>
<Modal
show={true}
centered
className={ImagemodalStyles.imageModal}
dialogClassName={ImagemodalStyles.imageModalDialog}
onHide={(e) => console.log(e)}
>
<Modal.Header closeButton />
<Modal.Body className={ImagemodalStyles.imageModalBody}>
<h1>TestInhalt</h1>
<Img fluid={data.file.childImageSharp.fluid} />
</Modal.Body>
</Modal>
</div>
)
}
/* imagemodal.module.scss */
.imageModalDialog {
display: inline-block;
width: auto;
}
.imageModal {
text-align: center;
}
.imageModalBody img {
max-height: calc(100vh - 225px);
}
The Problem
The image does not scale to the screen size. The image is either too big - so it flows over the vieport - or it is too small. Secondly, the modal size does not respond to the image size correctly and / or is not centered.
What I tried
I used this suggestion for the CSS: How to limit the height of the modal?
I tried as well dozens of other CSS parameter combinations. But I could not find a working solution.
I tried to format the gatsby-image directly with a style-tag.
I tried as well react-modal but had similar problems.
Does anyone have a good solution to show a gatsby-image in full screen size in a responsive modal? For me it is okay to use either the bootstrap-modal or react-modal - or any other suitable solution.
Edit
In the end I ended up with a workaround. I used react-image-lightbox and took the Image-Source from gatsby-image as the input for lightbox. My component gets the data from the graphQL query in the props via props.imageData.
This works quite well for me:
import Lightbox from 'react-image-lightbox';
...
export default function Imagegallery(props) {
...
const allImages = props.imageData.edges
const [indexImageToShow, setIndexImageToShow] = useState()
...
return(
<Lightbox
mainSrc={allImages[indexImageToShow].node.childrenImageSharp[0].fluid.src}
...
/>
Special thanks to #FerranBuireu to point me to the right direction
Assuming that the functionality works as expected, as it seems, it's a matter of CSS rules, not React/Gatsby issue. The following rule:
.imageModalBody img {
max-height: calc(100vh - 225px);
}
It Will never be applied properly, since gatsby-image creates an output of HTML structure of nested <div>, <picture> and <img> so your rule will be affected by the inherited and relativity of the HTML structure. In other words, you are not pointing to the image itself with that rule because of the result HTML structure.
You should point to the <Img>, which indeed, it's a wrapper, not an <img>.
return (
<div>
<Modal show={true} onHide={handleClose} centered className={ImagemodalStyles.imageModal} dialogClassName={ImagemodalStyles.imageModalDialog}>
<Modal.Header closeButton />
<Modal.Body>
<Img className={ImagemodalStyles.imageModalBody} fluid={props.data.file.childImageSharp.fluid} />
</Modal.Body>
</Modal>
</div>
)
The snippet above will add the (spot the difference, without img):
.imageModalBody {
max-height: calc(100vh - 225px);
}
To the wrapper, which may or may not fix the issue, but at least will apply the rule correctly. It's difficult to know what's wrong without a CodeSandbox but you will apply the styles correctly with this workaround.
Keep always in mind that when using gatsby-image, the <img> it's profound in the resultant HTML structure so your styles should apply to the outer wrapper of it.

How to set the zIndex layer order for geoJson layers?

I would like to have certain layers to be always on top of others, no matter in which order they are added to the map.
I am aware of bringToFront(), but it does not meet my requirements. I would like to set the zIndex dynamically based on properties.
Leaflet has the method setZIndex(), but this apparently does not work for geoJson layers:
https://jsfiddle.net/jw2srhwn/
Any ideas?
Cannot be done for vector geometries.
zIndex is a property of HTMLElements, and vector geometries (lines and polygons) are rendered as SVG elements, or programatically as <canvas> draw calls. Those two methods have no concept of zIndex, so the only thing that works is pushing elements to the top (or bottom) of the SVG group or <canvas> draw sequence.
Also, remind that L.GeoJSON is just a specific type of L.LayerGroup, in your case containing instances of L.Polygon. Furthermore, if you read Leaflet's documentation about the setZIndex() method on L.LayerGroup:
Calls setZIndex on every layer contained in this group, passing the z-index.
So, do L.Polygons have a setZIndex() method? No. So calling that in their containing group does nothing. It will have an effect on any L.GridLayers contained in that group, though.
Coming back to your problem:
I would like to have certain layers to be always on top of others, no matter in which order they are added to the map.
Looks like the thing you're looking for is map panes. Do read the map panes tutorial.
This is one of the reason for the implementation of user defined "panes" in Leaflet 1.0 (compared to versions 0.x).
Create panes: var myPane = map.createPane("myPaneName")
If necessary, set the class / z-index of the pane element: myPane.style.zIndex = 450 (refer to z-index values of built-in panes)
When creating your layers, specify their target pane option: L.rectangle(corners, { pane: "myPaneName" })
When building through the L.geoJSON factory, you can loop through your features with the onEachFeature option to clone your layers with specified target pane.
Demo: https://jsfiddle.net/3v7hd2vx/90/
For peoples who are searching about Z-Index
All path layers (so all except for markers) have no z-index because svg layers have a fix order. The first element is painted first. So the last element is painted on top.
#IvanSanchez described good why zIndex not working.
You can control the order with layer.bringToBack() or layer.bringToFront().
With that code you have more options to control the order of the layers.
L.Path.include({
getZIndex: function() {
var node = this._path;
var index = 0;
while ( (node = node.previousElementSibling) ) {
index++;
}
return index;
},
setZIndex: function(idx) {
var obj1 = this._path;
var parent = obj1.parentNode;
if(parent.childNodes.length < idx){
idx = parent.childNodes.length-1;
}
var obj2 = parent.childNodes[idx];
if(obj2 === undefined || obj2 === null){
return;
}
var next2 = obj2.nextSibling;
if (next2 === obj1) {
parent.insertBefore(obj1, obj2);
} else {
parent.insertBefore(obj2, obj1);
if (next2) {
parent.insertBefore(obj1, next2);
} else {
parent.appendChild(obj1);
}
}
},
oneUp: function(){
this.setZIndex(this.getZIndex()+1)
},
oneDown: function(){
this.setZIndex(this.getZIndex()-1)
}
});
Then you can call
polygon.oneUp()
polygon.oneDown()
polygon.setZIndex(2)
polygon.getZIndex()
And now layergroup.setZIndex(2) are working

Customizing fullpage.js to skip section(s) dynamically

The question is simple but i'm not able to make a script by myself for what i need...
I am actually using a script ( fullpage.js ) who toggle some classes into a container ( in my case switching from fp-viewing-1 to fp-viewing-x ) when you scroll down/up between sections.
I need to make a script that listen from this container and toggle a new class into a div ONLY when a class ( in my case fp-viewing-3 ) is added to this container ( from the fullpage.js script of course ).
Any way to make it?
I need to make a script that listen from this container
That's not the way to go for it.
If you want to use the status class, then just create a new class based on the previous ones as explained in this fullpage.js tutorial.
Create a conditional CSS class that will only get applied when its parent class matches your requirement.
Something like this, for example, would only apply the red color to element with myClass when you are in section 1 slide 0.
.fp-viewing-1-0 .myClass{
color: red;
}
Having:
<div id="fullpage">
<div class="section"></div>
<div class="section myClass"></div>
<div class="section"></div>
<div>
If for some other reason (use of plugins etc) you really need to add the class dynamically, then go for fullpage.js callbacks onLeave or afterLoad:
$('#fullpage').fullpage({
onLeave: function(index, nextIndex, direction){
var destination = $('.section').eq(nextIndex - 1);
destination.find('.my-element').addClass('myClass');
}
});
This is the solution to my problem.
Fullpage works as intended except for section 2.
Section 2 will be usable only scrolling down, the script ignore it when scrolling up.
$(document).ready(function() {
$('#application').fullpage({
onLeave: function(index, nextIndex, direction){
var destinationToIgnore = $('.fp-section').hasClass('ignore');
if(destinationToIgnore && direction =='up'){
var destination = nextIndex = 1
$.fn.fullpage.moveTo(destination);
}
},
afterLoad: function(anchorLink, index){
var loadedSection = $(this);
if(index !== 1){
$('.section-intro').removeClass('ignore');
}
if(index == 3){
$('.section-intro').addClass('ignore');
}
}
});
});

TinyMCE - applying a style over bullets and multiple paragraphs applies the style to each bullet & para - how do I avoid?

I'm trying to use the theme_advanced_styles command within TinyMCE to add classes to selections of text within the TinyMCE editor. The problem is that if the paragraph contains bullets, then the style is applied throughout them (as well as to each individual paragraph).
What I want is just for the entire selection I made to have the style class added to the start of it. Ie if my style class is 'expandCollapse' I want:
<p class="expandCollapse">some content... some content... some content... some content... som content... some content... some content...
<ul>
<li>asdsadsadsadsasda</li>
<li>asdsadsa</li>
<li>sada</li>
</ul>
asome content... some content... some content... some content... some content... some content... some content... some content... </p>
But what I get is:
<p class="expandCollapse">some content... some content... some content... some content... some content... some content... some content...
<ul>
<li class="expandCollapse">asdsadsadsadsasda</li>
<li class="expandCollapse">asdsadsa</li>
<li class="expandCollapse">sada</li>
</ul>
</p>
<p class="expandCollapse">asome content... some content... some content... some content... some content... some content... some content... some content... </p>
Any ideas anyone?!
So I had to answer my own question as I needed an answer very quickly. It appears the behaviour I was experiencing is intentional? and certainly not something that has been removed in the very latest versions of TinyMCE (both 3.x and 4.x after testing).
With this in mind I ended up having to make a plugin to do what I wanted.
I borrowed a huge amount of code by Peter Wilson, from a post he made here: http://www.tinymce.com/forum/viewtopic.php?id=20319 So thanks very much for this Peter!
I ended up slightly changing the rules from my original question in that my solution adds an outer wrapping div around all the content I want to select. This method also allowed me to reliably then grab the required areas of html with jQuery in my front-end site.
My version of Peter's code is just very slightly modified from the original in order to add a class to the DIV, rename it, use a different button etc.
The plugin works perfectly and allows for a div to be created wrapping any amount of content within TinyMCE. The divs inserted have the class name I need also applied to it.
Add 'customDiv' to your plugin AND button bar for it to appear.
(function() {
tinymce.create("tinymce.plugins.Div", {
init : function(editor, url) {
editor.addCommand("mceWrapDiv", function() {
var ed = this, s = ed.selection, dom = ed.dom, sb, eb, n, div, bm, r, i;
// Get start/end block
sb = dom.getParent(s.getStart(), dom.isBlock);
eb = dom.getParent(s.getEnd(), dom.isBlock);
// If the document is empty then there can't be anything to wrap.
if (!sb && !eb) {
return;
}
// If empty paragraph node then do not use bookmark
if (sb != eb || sb.childNodes.length > 1 || (sb.childNodes.length == 1 && sb.firstChild.nodeName != 'BR'))
bm = s.getBookmark();
// Move selected block elements into a new DIV - positioned before the first block
tinymce.each(s.getSelectedBlocks(s.getStart(), s.getEnd()), function(e) {
// If this is the first node then we need to create the DIV along with the following dummy paragraph.
if (!div) {
div = dom.create('div',{'class' : 'expandCollapse'});
e.parentNode.insertBefore(div, e);
// Insert an empty dummy paragraph to prevent people getting stuck in a nested block. The dummy has a '-'
// in it to prevent it being removed as an empty paragraph.
var dummy = dom.create('p');
e.parentNode.insertBefore(dummy, e);
//dummy.innerHTML = '-';
}
// Move this node to the new DIV
if (div!=null)
div.appendChild(dom.remove(e));
});
if (!bm) {
// Move caret inside empty block element
if (!tinymce.isIE) {
r = ed.getDoc().createRange();
r.setStart(sb, 0);
r.setEnd(sb, 0);
s.setRng(r);
} else {
s.select(sb);
s.collapse(1);
}
} else
s.moveToBookmark(bm);
});
editor.addButton("customDiv", {
//title: "<div>",
image: url + '/customdiv.gif',
cmd: "mceWrapDiv",
title : 'Wrap content in expand/collapse element'
});
}
});
tinymce.PluginManager.add("customDiv", tinymce.plugins.Div);
})();

Drag and Drop into Fabric.js canvas

How can I drop items (like image, or other object from other canvas) into canvas which is managed by fabricjs? I have found many examples how to move items inside canvas but I would like to drag and drop item from outer element into canvas.
Since you asked for an example and I haven't tried it out myself yet, here goes:
Example Fiddle
Markup
<div id="images">
<img draggable="true" src="http://i.imgur.com/8rmMZI3.jpg" width="250" height="250"></img>
<img draggable="true" src="http://i.imgur.com/q9aLMza.png" width="252" height="295"></img>
<img draggable="true" src="http://i.imgur.com/wMU4SFn.jpg" width="238" height="319"></img>
</div>
<div id="canvas-container">
<canvas id="canvas" width="800" height="600"></canvas>
</div>
JS Breakdown
1. Fabric.canvas instance
First we want our canvas, of course:
var canvas = new fabric.Canvas('c');
2. Feature Detection (optional)
Not sure this is necessary, since the fact that you have a canvas makes it very likely that the browser has Drag and Drop as well. Were you to use it, you can do so like this, using Modernizr:
if (Modernizr.draganddrop) {
// Browser supports HTML5 DnD.
// Bind the event listeners for the image elements
// Bind the event listeners for the canvas
} else {
// Replace with a fallback to a library solution.
alert("This browser doesn't support the HTML5 Drag and Drop API.");
}
3. Events
Again, unlike the source article I below, the source and target elements are different (in that articles's example, you just move divs around within the same parent container), so I failed to notice that some of the events are meant for the element being dragged, but most are bound to the element into which you are dropping.
NOTE: I know this is technically a question about Fabric.js, but it's really kind of a question about Drag and Drop in the context of adding objects to a <canvas> with Fabric.js, which is why I'm going a bit more in depth about the DnD stuff now.
For the <img>
dragstart (I added a class here to lower the opacity)
dragend (and removed that class here)
For #canvas-container:
dragenter (added a class to give the canvas container that nifty dotted line)
dragover: Here you can set the event.dataTransfer.dropEffect property to show one of the native cursor types. The default would be 'move' here, but I set it to 'copy' since I don't actually remove the <img> element (in fact in the fiddle you can, for example create several McClures).
dragleave (removed the dotted line here)
drop: The handler for this event creates and adds the fabric.Image object (see the fiddle).
if (Modernizr.draganddrop) {
// Browser supports HTML5 DnD.
// Bind the event listeners for the image elements
var images = document.querySelectorAll('#images img');
[].forEach.call(images, function (img) {
img.addEventListener('dragstart', handleDragStart, false);
img.addEventListener('dragend', handleDragEnd, false);
});
// Bind the event listeners for the canvas
var canvasContainer = document.getElementById('canvas-container');
canvasContainer.addEventListener('dragenter', handleDragEnter, false);
canvasContainer.addEventListener('dragover', handleDragOver, false);
canvasContainer.addEventListener('dragleave', handleDragLeave, false);
canvasContainer.addEventListener('drop', handleDrop, false);
} else {
// Replace with a fallback to a library solution.
alert("This browser doesn't support the HTML5 Drag and Drop API.");
}
Sources:
HTML5 Rocks - Native HTML5 Drag and Drop
Modernizr
Web Platform Docs > DOM > Properties - dropEffect
Web Platform Docs > DOM > Events
dragstart
dragend
dragenter
dragover
dragleave
drop
I had gone through fiddle of #natchiketa, And fixed the problem , just check this fiddle..
http://jsfiddle.net/Ahammadalipk/w8kkc/185/
window.onload = function () {
var canvas = new fabric.Canvas('canvas');
/*
NOTE: the start and end handlers are events for the <img> elements; the rest are bound to
the canvas container.
*/
function handleDragStart(e) {
[].forEach.call(images, function (img) {
img.classList.remove('img_dragging');
});
this.classList.add('img_dragging');
}
function handleDragOver(e) {
if (e.preventDefault) {
e.preventDefault();
}
e.dataTransfer.dropEffect = 'copy';
return false;
}
function handleDragEnter(e) {
this.classList.add('over');
}
function handleDragLeave(e) {
this.classList.remove('over');
}
function handleDrop(e) {
if (e.stopPropagation) {
e.stopPropagation(); // stops the browser from redirecting.
}
var img = document.querySelector('#images img.img_dragging');
var newImage = new fabric.Image(img, {
width: img.width,
height: img.height,
// Set the center of the new object based on the event coordinates relative
// to the canvas container.
left: e.layerX,
top: e.layerY
});
newImage.hasControls = newImage.hasBorders = false;
canvas.add(newImage);
return false;
}
function handleDragEnd(e) {
// this/e.target is the source node.
[].forEach.call(images, function (img) {
img.classList.remove('img_dragging');
});
}
if (Modernizr.draganddrop) {
var images = document.querySelectorAll('#images img');
[].forEach.call(images, function (img) {
img.addEventListener('dragstart', handleDragStart, false);
img.addEventListener('dragend', handleDragEnd, false);
});
var canvasContainer = document.getElementById("canvas-container");
canvasContainer.addEventListener('dragenter', handleDragEnter, false);
canvasContainer.addEventListener('dragover', handleDragOver, false);
canvasContainer.addEventListener('dragleave', handleDragLeave, false);
canvasContainer.addEventListener('drop', handleDrop, false);
} else {
alert("This browser doesn't support the HTML5 Drag and Drop API.");
}
}
Thanks
Well, the question is quite old^^
I have updateted the fiddle, no it will work in Firefox as well.
Fiddle
function handleDrop(e) {
// this / e.target is current target element.
e.preventDefault(); //I've altert this line for FireFox
As far as I understand, drag and drop is not already provided by fabricjs,
fiddling it will be quite interesting.
Well I m new to javascript and fabricJS but I think this fiddle might help
fiddle
html code
<div class="img_cont">
<img class="img" id="ele1" draggable="true" ondragstart="dragElement(event)" src="https://webkit.org/wp-content/uploads/STP-300x300.png">
<img class="img" id="ele2" draggable="true" ondragstart="dragElement(event)" src="https://webkit.org/wp-content/uploads/ephy-webkit-graphic.png">
<img class="img" id="ele3" draggable="true" ondragstart="dragElement(event)" src="https://res.cloudinary.com/css-tricks/image/upload/w_600,q_auto,f_auto/buysellads/uu/7/112766/1646327381-MC_CSSTricks_Logo_600x600-_1_.png">
<img class="img" id="ele4" draggable="true" ondragstart="dragElement(event)" src="https://miro.medium.com/max/1400/1*9hd_8qR0CMZ8L0pVbFLjDw.png">
</div>
<br>
<div id="canvas_cont" ondragover="allowDrop(event)" ondrop="dropElement(event)">
<canvas id="canvas" width="650" height="350" ></canvas>
</div>
javascript code
// allowDrop function called on ondragover event.
function allowDrop(e) {
e.preventDefault();
}
//dragElement function called on ondrag event.
function dragElement(e) {
e.dataTransfer.setData("id", e.target.id); //transfer the "data" i.e. id of the target dragged.
}
//Initializing fabric canvas on window load event.
var canvas;
window.onload = function(){
canvas = new fabric.Canvas(document.getElementById("canvas"));
}
//dropElement function called on ondrop event.
function dropElement(e) {
e.preventDefault();
var data = e.dataTransfer.getData("id"); //receiving the "data" i.e. id of the target dropped.
var imag = document.getElementById(data); //getting the target image info through its id.
var img = new fabric.Image(imag, { //initializing the fabric image.
left: e.layerX - 80, //positioning the target on exact position of mouse event drop through event.layerX,Y.
top: e.layerY - 40,
});
img.scaleToWidth(imag.width); //scaling the image height and width with target height and width, scaleToWidth, scaleToHeight fabric inbuilt function.
img.scaleToHeight(imag.height);
canvas.add(img);
}
The accepted answer no longer works.
This is for drag and drop from desktop using the dataTransfer interface.
canvas.on('drop', function(event) {
// prevent the file to open in new tab
event.e.stopPropagation();
event.e.stopImmediatePropagation();
event.e.preventDefault();
// Use DataTransfer interface to access the file(s)
if(event.e.dataTransfer.files.length > 0){
var files = event.e.dataTransfer.files;
for (var i = 0, f; f = files[i]; i++) {
// Only process image files.
if (f.type.match('image.*')) {
// Read the File objects in this FileList.
var reader = new FileReader();
// listener for the onload event
reader.onload = function(evt) {
// put image on canvas
fabric.Image.fromURL(evt.target.result, function(obj) {
obj.scaleToHeight(canvas.height);
obj.set('strokeWidth',0);
canvas.add(obj);
});
};
// Read in the image file as a data URL.
reader.readAsDataURL(f);
}
}
}
});
Resources
https://developer.mozilla.org/en-US/docs/Web/API/HTML_Drag_and_Drop_API/File_drag_and_drop