Users can paste an URL in a textbox on my site. When they do, I want to get that URL via jQuery AJAX and read the opengraph metadata from it. How can I do this?
I read this post How to read Open Graph and meta tags from a webpage with a url but the link in it is broken and it's more advanced than I need and not in jQuery :)
I don't need anything else but the opengraph metadata, so no parsing of structures etc.
Here's an example of a page: http://www.ebay.com/itm/Microsoft-Surface-Pro-3-12-Tablet-256GB-SSD-Intel-Core-i7-Haswell-8GB-RAM-/281656969697
So one of the fields I'd like to extract is <meta property="og:image" content="http://i.ebayimg.com/images/i/281656969697-0-1/s-l1000.jpg" ></meta>, to be precise the value http://i.ebayimg.com/images/i/281656969697-0-1/s-l1000.jpg
What I have now is copied from here: http://icant.co.uk/articles/crossdomain-ajax-with-jquery/error-handling.html
See my comment marked @Flo where I want to extract the open graph data, but I don't know how to parse the JSON response.
<a href="www.ebay.com/itm/Microsoft-Surface-Pro-3-12-Tablet-256GB-SSD-Intel-Core-i7-Haswell-8GB-RAM-/281656969697" class="ajaxtrigger">Load Ajax Content</a>
<div id="target"></div>
<script language="javascript" type="text/javascript">
$(function () {
$('.ajaxtrigger').click(function () {
var container = $('#target');
container.attr('tabIndex', '-1');
var trigger = $(this);
var url = trigger.attr('href');
if (!trigger.hasClass('loaded')) {
trigger.append('<span></span>');
trigger.addClass('loaded');
var msg = trigger.find('span').last();
} else {
var msg = trigger.find('span').last();
}
doAjax(url, msg, container);
return false;
});
});
function doAjax(url, msg, container) {
// if the URL starts with http
if (url.match('^http')) {
// assemble the YQL call
msg.removeClass('error');
msg.html(' (loading...)');
$.getJSON("//query.yahooapis.com/v1/public/yql?" +
"q=SELECT%20*%20FROM%20html%20WHERE%20url=%27" +
encodeURIComponent(url) +
"%27%20AND%20xpath=%27descendant-or-self::meta%27&format=json&callback=?",
function (data) {
if (data.results[0]) {
var data = filterData(data.results[0]);
//@Flo: get metadata from result, but now???
msg.html(' (ready.)');
container.
html(data).
focus().
effect("highlight", {}, 1000);
} else {
msg.html(' (error!)');
msg.addClass('error');
var errormsg = '<p>Error: could not load the page.</p>';
container.
html(errormsg).
focus().
effect('highlight', { color: '#c00' }, 1000);
}
}
);
} else {
$.ajax({
url: url,
timeout: 5000,
success: function (data) {
msg.html(' (ready.)');
container.
html(data).
focus().
effect("highlight", {}, 1000);
},
error: function (req, error) {
msg.html(' (error!)');
msg.addClass('error');
if (error === 'error') { error = req.statusText; }
var errormsg = 'There was a communication error: ' + error;
container.
html(errormsg).
focus().
effect('highlight', { color: '#c00' }, 1000);
},
beforeSend: function (data) {
msg.removeClass('error');
msg.html(' (loading...)');
}
});
}
}
function filterData(data) {
// filter all the nasties out
// no body tags
data = data.replace(/<?\/body[^>]*>/g, '');
// no linebreaks
data = data.replace(/[\r|\n]+/g, '');
// no comments
data = data.replace(/<--[\S\s]*?-->/g, '');
// no noscript blocks
data = data.replace(/<noscript[^>]*>[\S\s]*?<\/noscript>/g, '');
// no script blocks
data = data.replace(/<script[^>]*>[\S\s]*?<\/script>/g, '');
// no self closing scripts
data = data.replace(/<script.*\/>/, '');
// [... add as needed ...]
return data;
}
</script>
The object returned by this query is:
Object {query: Object}
query: Object
count: 33
created: "2015-05-02T04:36:46Z"
lang: "en-US"
results: Object
meta: Array[33]
0: Object
name: "viewport"
__proto__: Object
1: Object
content: "main"
name: "layout"
__proto__: Object
How can I filter this response to return the og:image value?
