I'm now working on a painting app. It uses a html canvas. Users can draw shapes on the canvas.
Now a problem comes to me. I want to select the line that I drew on the canvas. When it is selected, I can drag it or delete it. How can I implement it ? Any good ideas?
It might be worth your while to have a look at https://github.com/canvimation/canvimation.github.com this contains the source code for a drawing application using canvas.
You should note that this application is being re-coded but there is not yet a working version using the new code on line. The new source code is in the stage1 branch. Hopefully this new code is easier to understand than the old code and the code referred to below comes from the stage1 branch.
Basically a shape object is created for each shape drawn which includes all data about eh shape, including path data and data giving the rectangular boundary around the shape. When the "boundarydrop" div is clicked then function checkBoundary is called and data about the shift key and cursor position are passed. Then for each shape the first check is to see if the cursor is within the boundary of the shape and if so then a more refined check is carried out. For a closed shape the check is if the cursor in inside the shape and for an open shape if the cursor is close to the path.
There are further complications depending whether the shift key is held down or not and which group the shape belongs to. However the basics are there.
Functions to check out
in index.html
shiftdown
getPosition
in scripts/drawboundary.js
checkBoundary
isIn
isOn
in scripts/shape.js
shape
A working online version of the application is at http://canvimation.github.com/ but this uses the older code in the master branch and there are some bugs but it will give you some idea of what the application does.
Hope this is of some help
There is a library called kinetic.js, with it you can keep track of the shapes you draw on the canvas with the select feature.
Here is the link: https://github.com/ericdrowell/KineticJS
Related
I am new to Lua scripting, and game development. So please I am just a noob in Lua.
I have searched the net for solutions to my problems, without any luck.
I use Photoshop, Corona, Dusk, json and Tiled on windows7.
I am creating a "board" like game, i.e. Setlers. I am using a world map, as the background. The background image of the game area is a world map (world.png file). I have no problem here.
I would like to create transparrent clickable objects matching the countrys borders on my gamemap with all parameters and values (I have added in Tiled) stored in the object. So When the player clicks on the country the transparrent object (on top of the map) is the one clicked and an eventlistener acts on the click.
In Tiled I can create all the objects I need, naming them + assigning parameters and other values.
If I add object.alpha value in Tiled, the alpha value is passed on to corona and working there.
How can I read these data from the json/tmx file in Corona and adding them to a lua table?
The way I am thinking to use the Tiled map and its objects, is to create one polyline trace of each country’s border (creating one object per country). Then place each “country traced object” on top of the world.png map, also naming the object with the countrys name like “object.name = TileBritannia” and also the other properties for use in game.
My problem is getting the objects info, like object.name, and an eventlistener reacting to a click on the object.
Is a polyline the right way to create a clickable area on a map, when I use a png file as a background image?
What is the best way to create a country border objects, in one layer or with all countries as individual object layers in Tiled.
Can I create one layer with sub objects and still access them in my code?
How do I get the object name and other properties, set in Tiled.
When I try to use the (local britannia = tiledMap:load("britannia.json")) the "load" is not working, getting a nil value.
I am looking for a code that will extract/get/read the object.name i.e. “objBritannia” or "TileBritannia". from the json/tmx file.
When I try to read the different parameters from the json file, I don't get the result I expect. I get the result = function: 046A73B0, was hoping for an object name of some sort.
Please provide links to or code example.
I have edited the question.
Thanks
For questions 1 and 2: I have not used Tiled, but based on Corona Tiled, you have the right strategy in mind. That page makes me think that you can just use tap event listener to detect tap. If you are having issues with the example on that web page, please update your question to be more specific. If tap event handling doesn't work (maybe you're talking about a different Tiled lib), look a Polygon fill and Point in Polygon detection, because that's basically what you need to do. Try some stuff from there. If it still doesn't work for you, then update your question with specifics otherwise it will be likely get closed (it is a little too broad as it is).
For #3, Lua is a dynamic language that supports adding properties to objects in one line. So upon the example on the Corona Tiled page, all you would have to do is
tiledMap = require("tiled")
local britannia = tiledMap:load("britannia.json")
britannia.name = "Britannnia"
local Zulu = tiledMap:load("zulu.json")
zulu.name = "zulu"
Naturally you will probably have a whole bunch so you will create a function that you call for each tile. It's not clear what map.layer["objBritannia "].nameIs("TileBritannia") is supposed to do so I can't comment.
We have a flash application that we are planning on converting to javascript. It's a pretty simple map application with an image as the background and a bunch of simple polygon movie clips that represent destinations on the map.
I would like to iterate through each movie clip and extract the shape into an array of x,y points to redraw the polygon using an external javascript function.
Is this possible with actionscript?
If you want to export the shape coordinates at author time, you can do try the JSFL script recommented by #strille or this one or export transparent images (if that's not too limiting for your application).
If you need to export the shapes at runtime, you can use the awesome as3swf library to decompile the swf and export the shapes. Have a look at the ShapeExport wiki as there are couple of handy exporters for js like JSCanvasShapeExporter and the more generic JSONShapeExporter
There are ways you can read the coordinates from an SWF. For instance, I've written a parser in PHP (link). Getting the data doesn't help though, as it turns out. The Flash painting model is different enough from the HTML5 one enough to make transfer exceeding difficult. The main obstacle I discovered is that in Flash, a path can be filled with two fill styles: one for area enclosed by the path, the other for enclosed area considered to be "outside" by the even-odd rule (e.g. the pentagon in the middle of a star). Since the HTML5 canvas let you specify only one fill style, you can't redraw shapes from Flash accurately. I was trying to create a tool that extract shapes as SVG and was getting a lot of gap and holes in the result.
Flash Player 11.6 introduced readGraphicsData() which does exactly what you ask for.
If you need to target an earlier version, then there's no simple way to read shape coordinates from a display object with ActionScript at runtime unfortunately.
If you just want to extract the shape coordinates once someone has written a jsfl script for Flash CS3 which looks like it might be able to help you out.
I need to be able to draw a thick patterend line between 2 points in AS3, I can't use the draw API because it doesn't all me to actually put detail (pattern etc) into the thickness of the line, I thought about perhaps using the line to create a bitmap version and then using that as a mask, but I remember many years ago seeing some examples that use a movieclip as a source for a line, but I can't find examples of that now at all, any ideas?
I've attached an example image of how I want the line to look.
I suspect Graphics.lineBitmapStyle() will do the trick. If you want to use a Sprite or MovieClip as the source, you'll have to draw() it to a BitmapData first. The example code on that later link should get you up and running with that.
This is hard to explain so I created a JS Fiddle to show what is going on: http://jsfiddle.net/dGdxS/ - (Webkit only) Just mouse over the canvas to draw - you may need to click around
I noticed the performance of the canvas was not what I expected, and when I set the line alpha I think I see why.
It looks like every new point I add to the drawing it is layering a completely new line over the previous line. The stroke alpha is set to 1%, and when you draw the older areas of the stroke begin to darken as the layers build up.
I assume I am doing something wrong. Do I need to clear the canvas before drawing to it at each new point?
HTML5 canvas does not come with an endPath method, but instead you can call context.beginPath() again to close the current path and begin a new one.
This is as opposed to calling context.closePath(), which will actually draw a line to the beginning of the current path, but not end it. If you were to draw again, the pen will simply continue drawing from that same point that closed the path. See this excellent related SO answer for more details and explanations.
Because you never call ctx.closePath();, so it will redraw the entire path from the beginning each time.
So you really want something similar to this:
http://jsfiddle.net/dGdxS/7/
Im looking for an easy to use method of assigning drag behavior to multiple objects (images, shapes etc) in canvas. Does anyone have a good way or know of any libraries for dragging objects around? Thanks
Creating your own mouse events takes a little work - ideally you should either create or use some kind of mini-library. I'm thinking of creating something like this in the near future. Anyway, I created a drag and drop demo on jsFiddle showing how to drag images - you can view it here.
You can create draggable images like this:
var myImage = new DragImage(sourcePath, x, y);
Let me know if you have any questions about this. Hope it helps.
EDIT
There was a bug when dragging multiple images. Here is a new version.
Another thing you might want to check out is easeljs it sort of in the style of AS3... mouseEvents dragging etc...
The HTML Canvas—unlike SVG or HTML—uses a non-retained (or immediate) graphics API. This means that when you draw something (like an image) to the canvas no knowledge of that thing remains. The only thing left is pixels on the canvas, blended with all the previous pixels. You can't really drag a subset of pixels; for one thing, the pixels that were 'under' them are gone. What you would have to do is:
Track the mousedown event and see if it's in the 'right' location for dragging. (You'll have to keep track of what images/objects are where and perform mouse hit detection.)
As the user drags the mouse, redraw the entire canvas from scratch, drawing the image in a new location each time based on the offset between the current mouse location and the initial mousedown location.
Some alternatives that I might suggest:
SVG
Pure HTML
Multiple layered canvases, and drag one transparent canvas over another.
The HTML Canvas is good for a lot of things. User interaction with "elements" that appear to be distinct (but are not) is not one of those things.
Update: Here are some examples showing dragging on the canvas:
http://developer.yahoo.com/yui/examples/dragdrop/dd-region.html
http://www.redsquirrel.com/dave/work/interactivecanvas/
http://langexplr.blogspot.com/2008/11/using-canvas-html-element.html
None of these have created a separate library for tracking your shapes for you, however.
KineticJS is one such Javascript Library that u can use exclusively for animations
Heres the Link html5canvastutorials
Canvas and jCanvas
You're definitely gonna want to check out jCanvas. It's a super clean wrapper for Canvas, which kicks open a lot of doors without adding code complexity. It makes things like this a breeze.
For example, here's a little sandbox of something close to what you're after, with dragging and redrawing built right in:
Drawing an Arrow Between Two Elements.
I ventured down the road of doing everything with DIVs and jQuery but it always fell short on interactivity and quality.
Hope that helps others, like me.
JP
As you create new objects whether they are windows, cards, shapes or images to be draggable, you can store them in an array of "objects currently not selected". When you click on them or select them or start dragging them you can remove them from the array of "objects not selected". This way you can control what can move in the event of a particular mousedown event or mousemove event by checking if it isn't selected. If it is selected it will not be in the "not selected" array and you can move the mouse pointer over other shapes while dragging shapes without them becoming dragged.
Creating arrays of objects you would like to drag also helps with hierarchy. Canvas draws the pixels belonging to the foremost object last. So if the objects are in an array you simply switch their instance as in element in the array say from objectArray[20] to objectArray[4] as you iterate through the array and draw the objects stored in the array elements you can change whether other objects are seen on top or behind other objects.