Package ca.weblite.codename1.js

The Codename One JS Bridge package includes classes that facilitate the interaction between Java and Javascript in a Codename One application.

See: Description

Interface Summary

Interface	Description
JSFunction	The JSFunction interface allows you to implement functions in Java that can be called from Javascript.

Class Summary

Class	Description
JavascriptContext	Represents a Javascript context of a single BrowserComponent.
JSObject	A Java Wrapper around a Javascript object.

Package ca.weblite.codename1.js Description

The Codename One JS Bridge package includes classes that facilitate the interaction between Java and Javascript in a Codename One application. It allows both Java to Javascript communication and the reverse via a mechanism similar to the one employed by Phone Gap/Apache Cordova.

Requirements

Currently the CodenameOne bridge will only run on platforms that include a native browser component. This includes only iOS and Android currently. Blackberry will be supported soon, but its native browser component is still quite buggy and is disabled by default.

Installation

To use the Codename One JS Bridge, you simply need to include the ca.weblite.codename1.js package in your project. The easiest way to do that currently is to check it out from SVN and copy the directory into the src directory of your project.

E.g. Step 1: Check out from SVN:

$ svn co svn checkout http://codenameone-incubator.googlecode.com/svn/trunk/shannah/CodenameOneTests/src/ca/weblite/codename1/js js

This will check out the js package into your current working directory. Next, create the necessary directory structure in your project's src directory:

$ cd /path/to/myproject/src
$ mkdir ca ca/weblite ca/weblite/codename1
            

Finally, copy the js directory that you checked out of SVN into your source directory:

$ cp -r /path/to/js /path/to/myproject/src/ca/weblite/codename1/
            

Usage

The class lays the foundation by enabling you to call Javascript code directly from Java. It provides automatic type conversion between Java and Javascript types as follows:

Java to Javascript

Java Type	Javascript Type
String	String
Double/Integer/Float/Long	Number
Boolean	Boolean
JSObject	Object
null	null
Other	Not Allowed

Javascript to Java

Javascript Type	Java Type
Number	Double
String	String
Boolean	Boolean
Object	JSObject
Function	JSObject
Array	JSObject
null	null
undefined	null

Note that this conversion table is more verbose than necessary, since Javascript functions and arrays are, in fact Objects themselves, so those rows are redundant. All Javascript objects are converted to JSObjects.

Getting Started

In order to start interacting with a Javascript environment, you need to create a WebBrowser and load a page. Then inside the WebBrowser's onLoad() handler, you can create a JavascriptContext on the internal BrowserComponent object:

Getting Values

The JavascriptContext.get(String) method is used to get values from Javascript. It takes an arbitrary Javascript expression, executes it, and returns the result, after converting it to the corresponding Java type. E.g. if the result is a String, it will return a String and if it is a Number it will return a java Double object. If the result is an Object, it will return a JSObject.

The following is a simple example that retrieves the document content, which is a string:

If you run this example in the simulator, you'll see something like the following:

Note: You don't see the WebBrowser in the background here due to a bug in the simulator. If you run this example on a device, you will see the WebBrowser component in the background.

Returning Numeric Values

As mentioned in the conversion table above, numeric values are automatically converted to java.lang.Double objects. The following example, returns the width and height properties of the window for use in Java.

The result, when run in the simulator would be something like:

Returning Objects

The previous examples involved only primitive return values. The JavascriptContext abstraction, in these cases, doesn't offer a whole lot of added-value over just using the BrowserComponent.executeJavascriptAndReturnString() method. The real value is when we are dealing with objects.

The following example obtains a reference to the window object and wraps it in a proxy JSObject class so that we can work directly with the window object:

This code produces the exact same result as the previous example. The difference is the intermediary step of wrapping the window object in a JSObject, and obtaining the outerHeight and outerWidth properties directly via that proxy object.

You can obtain a JSObject proxy for any Javascript object, even ones that you create on the fly. The following example creates an anonymous object with some keys and values and uses a JSObject proxy to interact with this object from Java.

The result is as follows:

See Working With Objects for more information about working with the JSObject class.

Returning Functions and Arrays

In Javascript, functions and arrays are just objects, so these are also encapsulated as JSObject proxies. See Working with Arrays and Workin with Functions for more details on how to work with these values via the JSObject interface.

Setting Values

Just as you can get values from Javascript using JavascriptContext's get() method, you can also set values via JavascriptContext.set(String,Object).

The following is a simple example that sets the location, and causes it to redirect to a different page:

If you run this example, you should see your browser display the Codename One website after a redirect.

Setting Object Values

The previous example showed us setting a primitive String value. You can do the same with other primitives like numbers and booleans, but you can also set Object values using the set() method.

The following example creates an anonymous Javascript object, wraps it in a JSObject proxy, sets some values on it, then sets the object as a property of the top-level window object.

As a result, you should see the following content set as the body of the HTML page in the WebBrowser. Note that we can refer to the "steve" object that we just set directly/globally because the "window" object's properties are always available directly through the global namespace in Javascript.

Working with Objects

Previously examples showed how to obtain a JSObject proxy to a Javascript object. There are 4 ways to get a JSObject:

1. Create an anonymous object:
   JSObject obj = (JSObject)ctx.get("{}");
2. Reference an existing object directly:
   JSObject obj = (JSObject)ctx.get("window.location");
3. As the result of a Javascript expression or function call:
   JSObject obj = (JSObject)ctx.get("document.getElementById('mydiv')")
4. Retrieve an Object member variable from another JSObject:
   JSObject obj = (JSObject)otherObj.get("aPropertyThatIsAnObject")

JSObjects are essentially just proxies around a Javascript object. Any calls to retrieve properties from a JSObject are just sent directly to the Javascript context, and the result returned. The JSObject object doesn't store copies the javascript object's properties. It just retrieves them as they are required via the get(String) method.

Getting Simple Values

You can always retrieve the properties of an object using the get(String) method. It takes the name of the property as a parameter, and returns its value, converted to the appropriate Java type. (e.g. if it is a String, it returns a String, if it is a number it returns a Double, and if it is an Object, it returns an object.

E.g.

String name = (String)obj.get("name");
Double age = (Double)obj.get("age");

Is equivalent to the following javascript:
var name = obj.name; var age = obj.age;
Assuming that the obj variable in Java is a proxy for the same obj variable in the javascript example.

Getting Nested Object Values

Often, in Javascript, an object contains a heirarchy of nested child objects. E.g.

var obj = {
    name : 'Steve',
    position : {
        x : 100,
        y : 105,
        z : -25
    }
}

In this case you may want to obtain the x coordinate of the nested position object. JSObject allows you to use the dot '.' notation for referencing sub-properties like this. E.g.

Double x = (Double)obj.get("position.x")

This feature raises the issue of how, then, to access properties that contain a '.' in its name. E.g.

var obj = {
    name : 'Steve',
    position : {
        x : 100,
        y : 105,
        z : -25
    },
    'position.x' : 200
}

In this example there is a top-level property named 'position.x' as well as a property at the component address position.x. This is a contrived example that is meant to be somewhat confusing in order to demonstrate how to differentiate between requests for properties in the child object heirarchy and top-level properties that happen to include a '.' in the property name.

We can force the retrieval of a top-level property by wrapping the key in single quotes:

Double x1 = (Double)obj.get("'position.x'")

This would return 200 for the above example, whereas:

Double x2 = (Double)obj.get("position.x")

Would return 100

.

Setting Object Values

The set(String,Object) method works the same as the JavascriptContext.set(String,Object) method except that it treats the local object as the root node. It allows you to easily set properties on the object. Values set here should be provided using Java values as they will automatically be converted to the appropriate associated Javascript type. If you are setting an Object as a value, then you'll need to set it as a JSObject and not a string representation of the object. This is because Strings will just be converted to Javascript strings.

Properties set via the set(String,Object) method modify the underlying Javascript object directly so that the change is immediately effective inside the javascript environment.

Just as with the get(String) method, you can set the values of direct properties or nested properties using the dot '.' notation. And just like get(String), you can force setting a direct property in cases where the property name includes a '.', by wrapping the key inside single quotes.

E.g.:

// Create a team object , and leave city null for now.
JSObject blueJays = (JSObject)ctx.get("{name : 'Blue Jays', city : null}");

// Create a city object and leave country null for now.
JSObject toronto = (JSObject)ctx.get("{name : 'Toronto', country : null}");

// Create a country object
JSObject canada = (JSObject)ctx.get("{name : 'Canada'}");

// Set the team's city to toronto
blueJays.set("city", toronto);

// Set toronto's country to canada
toronto.set("country", canada);

// Retrieve the name of the country where the blue jays play
String countryName = (String)blueJays.get("city.country.name");
       // Should contain "Canada"

// Change the name of Canada to "Canuck land" using nested 
// dot notation on the blueJays object.

blueJays.set("city.country.name", "Canuck land");

String blueJaysCountry = (String)blueJays.get("city.country.name");
String torontoCountry = (String)tornoto.get("country.name");
String canadaName = (String)canada.get("name");

//Note that all of these should be equal and contain "Canuck land"

Calling Object Methods

The ca.weblite.codename1.js.JSObject#call(String,Object) method allows you to call javascript methods that are members of the underlying object. It arguments are passed as an Object[] array. These will be automatically converted from the Java type to the corresponding Javascript type. Java type conversion are the same as using the JavascriptContext.set(String,Object) method.

The following example shows an object with a simple add() method that just adds two numbers together:

JSObject obj = (JSObject)ctx.get("{ add : function(a,b){ return a+b;}}");
Double result = (Double)obj.call("add", 
    new Object[]{new Integer(1), new Integer(3)}
);

// Result should be 4.0

Working with Arrays

In javascript, arrays are just objects that include a special ability to be iterated. You can use the alternate version of get(int) which takes an int as a parameter to retrieve the elements of an array.

For example, consider the following javascript object:

var obj = {
    name : 'Blue Jays',
    players : [
        { name : 'George Bell', age : 31},
        { name : 'Tony Fernandez', age : 34},
        { name : 'Lloyd Moseby', age : 29}
    ]
}

        

Then assuming we have a JSObject proxy for this object, we could loop through the players array and output the name and age of each player as follows:

JObject players = (JObject)obj.get("players");
int len = ((Double)players.get("length")).intValue();
for ( int i=0; i<len; i++){
    JSObject player = (JSObject)players.get(i);
    Log.p("Name : "+player.get("name")+" age : "+player.get("age"));
}
                

Calling Java Methods from Javascript

So far, our examples have been limited to Java calling into Javascript. However, it may be useful to be able to also go the other way: call java methods from Javascript. Some applications of this might include:

• Capturing video, or using the camera on the phone
• Getting the phone location
• Accessing the file system or storage API

The Codename One JS bridge supports javascript to java method calling by way of the JSFunction interface and the set(String,Object) methods on the JSObject class. You can implement a JSFunction object and register it as a callback with a JSObject, then you will be able to execute this object's apply method via a Javascript proxy.

As an example, let's implement a simple logging function:

JSObject logger = (JSObject)ctx.get("{}");
logger.set("log", new JSFunction(){

    public void apply(JSObject self, Object[] args) {
        String msg = (String)args[0];
        Log.p("[Javascript Logger] "+msg);
    }

});

ctx.set("window.logger", logger);


c.executeAndReturnString("logger.log('This is a test message');");

If you execute this code in the simulator, you'll see the following output in the console:

[EDT] 0:0:0,0 - [Javascript Logger] This is a test message

Running it on a device will yield similar output in the device log file.

Let's step through this code to see what is happening. First we create a new, empty javascript object and wrap it in a JSObject proxy. Then we use the JSObject's set(String,Object) method to add an anonymous JSFunction object to it with the propery of "log". This step registers a method proxy on the Javascript side that acts just like a normal javascript method, but which actually triggers the JSFunction's JSFunction.apply(JSObject,Object[]) method.

We then set this logger object to the global javascript scope by making it a direct child of the window object. Finally we issue a Javascript method call to logger.log(). This is what effectively calls the apply() method on our JSFunction object.

Caveats

JSFunction callbacks are executed asynchronously so as to prevent deadlocks. This means that you cannot return a value from this method using a return statement (hence the reason why the interface definition for JSFunction.apply(JSObject,Object[]) is void.

If you want to return a value back to Javascript, then you'll need to do it by providing a callback function as one of the parameters, and call this callback method from inside the JSFunction.apply(JSObject,Object[]) method upon completion.

Example: Passing a Javascript Callback to Your Callback

Since JSFunction callbacks are executed asynchronously, if you want to be able to return a result back to Javascript, you will need to do this via a Javascript callback. This is quite a common pattern in Javascript since it is single threaded and relies upon non-blocking patterns.

As an example, let's create a JSFunction callback that adds two numbers together and returns the result to Javascript via a callback:

First we will create the JSFunction object to perform the addition, as follows:

WebBrowser b = new WebBrowser(){
    protected void onLoad(String url){
        JSObject window = (JSObject)ctx.get("window");
        window.set("addAsync", new JSFunction(){

            public void apply(JSObject self, final Object[] args) {
                Double a = (Double)args[0];
                Double b = (Double)args[1];
                JSObject callback = (JSObject)args[2];

                double result = a.doubleValue() + b.doubleValue();
                callback.call(new Object[]{new Double(result)});

            }

        });
    }
};

b.setURL("jar:///ca/weblite/codename1/tests/AddAsync.html");
                

In this snippet, we start by obtaining a reference to the "window" object. We then add a method to this object named "addAsync". This method is a JSFunction object that we implement inline. The apply() method of the JSFunction object is the Java method that will be executed when the addAsync method is called from Javascript. In this case the addAsync method expects three parameters:

1. The two numbers that are being added together.
2. A Javascript callback method that will be executed when completed and passed the result of the addition as a parameter.

Notice that all numerical arguments are converted to Java Double objects, and the callback function is converted to a JSObject object. Also notice the use of callback.call(), which just calls the callback as a function itself. With this variant of the call() method, the window object is used as this. Notice also that we pass the result inside an Object[] array. This array will be expanded to the direct javascript function parameters. (i.e. it will not pass an array as the parameter to the javascript method, the array elements are extracted and passed individually.

Now, let's look at the HTML page contents for our example:

<!DOCTYPE html>
<html>
    <head>
        <title>Addition Example</title>
    </head>
    <body>
        <div>Addition Example</div>
        <p><input type="text" size="4" id="input1"/> +
            <input type="text" size="4" id="input2"/> =
            <span id="result"></span>
        </p>
        <p><button id="calculate">Calculate</button></p>
        <script src="AddAsync.js"></script>
    </body>
</html>
                        

Our HTML simply includes two text fields (to input the values to be added together), a button to initiate the calculation, and a <span> tag where the result will be placed when the calculation is complete.

Finally it includes the AddAsync.js Javascript file (which is placed in the same directory as the AddAsync.html file. Its contents are as follows:

document
    .getElementById('calculate')
    .addEventListener('click', function(){
        var aField = document.getElementById('input1');
        var bField = document.getElementById('input2');
        var a = parseFloat(aField.value);
        var b = parseFloat(bField.value);
        window.addAsync(a, b, function(result){
           document.getElementById('result').innerHTML = result;
        });
    }, true);

                        

This script attaches an event handler to the calculate button that gets the values from the two input fields and passes it to the window.addAsync() method for calculation. The addAsync() method is actually our java JSFunction that we implemented earlier.

One small word about the placement of these files: This example is taken from a class ca.weblite.codename1.tests.CodenameOneTests. The AddAsync.html and AddAsync.js files are included in the same directory as the CodenameOneTests.java file ( i.e. /ca/weblite/codename1/tests). We used the WebBrowser's setURL() method to load the AddAsync.html file from an absolute path using jar: protocol. Currently this is the best way of loading local HTML files into a WebBrowser object (i.e. use the jar: protocol and provide an absolute path).

The result of running this app is as follows:

Example: Exposing the Camera to Javascript

The following creates a Javascript function for taking photos on a mobile device. It involves a simple webpage with a "Capture" button. When the user clicks this button, it will dispatch a function call to CodenameOne to access the device's camera. After the user takes a picture, CodenameOne will execute a Javascript callback to add the picture to the webpage.

The HTML page source is as follows:

This loads the CameraExample.js script:

document.getElementById('capture')
    .addEventListener('click', function(){
        camera.capture(function(url){
            if ( url == null ){
                // No image was provided
                return;
            }
            var results = document.getElementById('results');
            results.appendChild(document.createTextNode("testing"));
            var img = document.createElement('img');
            img.setAttribute('src',url);
            img.setAttribute('width', '100');
            img.setAttribute('height', '100');
            results.appendChild(img);
        })
    }, true);

The CameraExample.js script attaches a listener to the 'click' event of the "Capture" button which simply calls the camera.capture() method, which is actually a JSFunction that has been registered with the Javascript runtime. This actually calls into Java.

We pass a callback function into camera.capture() which will be executed upon successfully completion of the camera. This is a common programming pattern in Javascript. If a non-null URL is passed to this callback function, it is expected to be the URL of the image (it will be a local file URL.

The Java code that powers this example is as follows:

This example puts together most of the features of the CodenameOne-JS library.

1. It creates a new JSObject from within Java code to serve as the camera object.
2. It registers a JSFunction callback as a Javascript function which is added as a method to the camera object.
3. It shows the use of the call() method of JSObject to call the callback function that was provided by Javascript from inside the JSFunction's apply() method.