Change url with ajax call - php

Question

I'm doing cross-domain GET to get cnn html like this:
$(function(){
var site = 'http://cnn.com';
$.get('proxy.php', { site:site }, function(data){
$(data).appendTo('#div');
}, 'html');
});
Im getting everything that I need except the url's sometimes are not complete urls but point to a certain path on there server like this:
/2013/01/24/business/samsung-record-fourth-quarter-2012-profits/index.html?hpt=hp_t3
So the problem is that if someone is clicking the link on my site the url will look like this:
http://MY-WEBSITE/2013/01/24/business/samsung-record-fourth-quarter-2012-profits/index.html?hpt=hp_t3
How can I get rid of my own url being inserted and replace it with 'cnn.com'?
I tried jquery split and replace but it doesn't work:
href = $(this).prop('href');
url = href.split('/');
href.replace(url[2], 'cnn.com');
I usually get an error in console 'split is not defined', when I fixed it the error moves on to 'url is not defined' and so on. Sometimes (with other code variations) I get no errors but it still doesn't work. I cant figure it out.

Answer 1

Looking at your code I am assuming you are using jQuery.
The problem is occurring because the source code on cnn.com seems to be using relative links. You can insert cnn.com at the beginning with the following jQuery
$(function() {
$('a').each(function() {
if ($(this).attr('href').indexOf('http') === 0) {
$(this).attr('href', 'http://www.cnn.com' + this.href);
}
});
});

Answer 2

You can simply check if the url is relative or not. Easiest way to do this is to check if it starts with http://.
var url = $(this).prop('href');
if (!(/^http/).test(url))
{
url = site + url; // prepend "http://cnn.com" to the address
}
alert(url); // url is now a full url
if you want a more generalised solution, you can use the regexp object on the site to determine if the prefix is there.
var site = "http://cnn.com";
var siteRegex = new RegExp("^" + site); // Regex - Starts with site
var url = $(this).prop('href');
if (!siteRegex.test(url))
{
url = site + url;
}
alert(url);

Answer 3

This little function can be used in a general purpose:
function RelativeToAbsoluteURL(root, url) {
var httpMatch = /https?:\/\//ig;
if (!httpMatch.test(url)) {
return root + (url.substr(0, 1) != "/" ? "/" + url : url);
} else {
return url;
}
}
Use:
RelativeToAbsoluteURL("http://www.cnn.com", "http://www.cnn.com/2013/01/24/business/samsung-record-fourth-quarter-2012-profits/index.html?hpt=hp_t3");
RelativeToAbsoluteURL("http://www.cnn.com", "/2013/01/24/business/samsung-record-fourth-quarter-2012-profits/index.html?hpt=hp_t3");
Both will output the same URL (http://www.cnn.com/2013/01/24/business/samsung-record-fourth-quarter-2012-profits/index.html?hpt=hp_t3)

Answer 4

DEMO
Either seems to work:
var href = $(this).attr('href');
var url = (href.indexOf("/")===0) ? "http://...."+href:href;
Possible alternative - this one will return the fully qualified URL from the href
var href = this.href; // actual href and not the property
var url =(href.indexOf(location.hostname)===7) ? href.replace(location.hostname,"www.cnn.com"):href;
Using prop
var href = $(this).prop('href');
var url = (href.indexOf("/")===0) ? "http://...."+href:href;