Using PhantomJS to embed all images of a webpage produces warnings but works
I'm trying to convert a webpage into a single file by embedding all the images (and other external resources once I passed this point). Here's how I run PhantomJs:
./phantomjs --web-security=false ./embed_images.js http://localhost/index.html > output.txt
And here's the embed_images.js:
var page = require('webpage').create(),
system = require('system'),
address;
if (system.args.length === 1) {
console.log('Usage: embed_images.js <some URL>');
phantom.exit(1);
}
else {
page.onConsoleMessage = function(msg) {
console.log(msg);
};
address = system.args[1];
page.open(address, function(status) {
page.evaluate(function() {
function embedImg(org) {
var img = new Image();
img.src = org.src;
img.onload = function() {
var canvas = document.createElement("canvas");
canvas.width = this.width;
canvas.height = this.height;
var ctx = canvas.getContext("2d");
ctx.drawImage(this, 0, 0);
var dataURL = canvas.toDataURL("image/png");
org.src = dataURL;
console.log(dataURL);
}
}
var imgs = document.getElementsByTagName("img");
for (var index=0; index < imgs.length; index++) {
embedImg(imgs[index]);
}
});
phantom.exit()
});
}
When I run the mentioned command, it results in a file like this:
Unsafe JavaScript attempt to access frame with URL from frame with URL file://./embed_images.js. Domains, protocols and ports must match.
Unsafe JavaScript attempt to access frame with URL about:blank from frame with URL file://./embed_images.js. Domains, protocols and ports must match.
There's multiple instances of the above error message. To test what's wrong, I ran the below code in my Chromium's console:
function embedImg(org) {
var img = new Image();
img.src = org.src;
img.onload = function() {
var canvas = document.createElement("canvas");
canvas.width = this.width;
canvas.height = this.height;
var ctx = canvas.getContext("2d");
ctx.drawImage(this, 0, 0);
var dataURL = canvas.toDataURL("image/png");
org.src = dataURL;
console.log(dataURL);
}
}
var imgs = document.getElementsByTagName("img");
for (var index=0; index < imgs.length; index++) {
embedImg(imgs[index]);
}
And it works just fine (my webpage doesn't reference any cross-domain images)! It will embed all the images into the HTML page. Does anyone know what might the problem be?
Here's my index.html file's content:
<!DOCTYPE html >
<html>
<head>
<meta charset="utf-8" />
</head>
<body>
<img src="1.png" >
</body>
</html>
And actual output (output.txt):
Unsafe JavaScript attempt to access frame with URL from frame with URL file://./embed_images.js. Domains, protocols and ports must match.
Unsafe JavaScript attempt to access frame with URL from frame with URL file://./embed_images.js. Domains, protocols and ports must match.
Unsafe JavaScript attempt to access frame with URL about:blank from frame with URL file://./embed_images.js. Domains, protocols and ports must match.
Unsafe JavaScript attempt to access frame with URL about:blank from frame with URL file://./embed_images.js. Domains, protocols and ports must match.
Unsafe JavaScript attempt to access frame with URL about:blank from frame with URL file://./embed_images.js. Domains, protocols and ports must match.
Unsafe JavaScript attempt to access frame with URL about:blank from frame with URL file://./embed_images.js. Domains, protocols and ports must match.
Unsafe JavaScript attempt to access frame with URL about:blank from frame with URL file://./embed_images.js. Domains, protocols and ports must match.
Unsafe JavaScript attempt to access frame with URL about:blank from frame with URL file://./embed_images.js. Domains, protocols and ports must match.
Unsafe JavaScript attempt to access frame with URL about:blank from frame with URL file://./embed_images.js. Domains, protocols and ports must match.
Unsafe JavaScript attempt to access frame with URL about:blank from frame with URL file://./embed_images.js. Domains, protocols and ports must match.
Unsafe JavaScript attempt to access frame with URL about:blank from frame with URL file://./embed_images.js. Domains, protocols and ports must match.
The strange thing is that while I've got just one image on my page, there are numerous error messages!
I'm using phantomjs-1.9.8-linux-x86_64.
Answer
Those notices are printed when phantom.exit is called. They don't cause any trouble, but are not nice when you need a clean PhantomJS output. In your case you can suppress the notices by "asynchronizing" phantom.exit like this:
setTimeout(function(){
phantom.exit();
}, 0);
I think the reason this is happening is because a large string is passed from the page context when phantom tries to exit.
I created a github issue for this.