vbauer
Added new captureOnCallback option
This pull request implements a solution to issue #11 .
I am not sure whether captureOnCallback is the best name, but callback was already taken.
If captureOnCallback is true, we use the onCallback callback to initiate screen capture (rather than checking if the document is ready).
I tested this feature using both phantom and slimer engines and it seems to work great.
I used the following url as the test page: https://presentation-ngwjsmdjlf.now.sh
<html>
<head></head>
<body>
<h1 id="heading">My Content!</h1>
<script>
var h1 = document.getElementById('heading');
// When the page load event fires, update text and make it red
window.addEventListener('load', function() {
h1.innerText = 'onLoad!';
h1.style.color = 'red';
});
function duration() {
var d = null;
var parts = location.search.split('?t=');
if (parts.length > 0) {
d = parseInt(parts[parts.length-1], 10);
}
return isNaN(d) ? 0 : d;
}
// When this timeout fires, update text, make it green,
// then tell PhantomJS we are ready!
setTimeout(function() {
h1.innerText = 'page is ready!';
h1.style.color = 'green';
if (typeof window.callPhantom === 'function') {
window.callPhantom({ hello: 'world' });
}
}, duration());
</script>
</body>
</html>
I am happy to make changes if needed.
This PR could definitely help folks with single page apps / dynamic UIs as well. (If they're willing to add window.callPhantom() in their code). I think issue #62 may have benefitted from a feature like this.
@jessegavin Sorry, I'm on a business trip now. I would be able to take a look a little bit later..
@vbauer , I really need this in my project, so if I can anyway throw in a vote for this, please consider this my vote:-)
