Question

I have a function that uses regex to return root domain of the given url.

http://jsfiddle.net/hSpsT/

function cleanUp(url) {
  url = url.replace(new RegExp(/^\s+/),""); // START
  url = url.replace(new RegExp(/\s+$/),""); // END

  // IF FOUND, CONVERT BACK SLASHES TO FORWARD SLASHES
  url = url.replace(new RegExp(/\\/g),"/");

  // IF THERE, REMOVES 'http://', 'https://' or 'ftp://' FROM THE START
  url = url.replace(new RegExp(/^http\:\/\/|^https\:\/\/|^ftp\:\/\//i),"");

  // IF THERE, REMOVES 'www.' FROM THE START OF THE STRING
  url = url.replace(new RegExp(/^www\./i),"");
  //remove slash from end
  url = url.replace(new RegExp(/\/$/i),"");    
  return url;
}

But it uses multi regex and we are worried about the performance. Is there a better way to do the same in a one line regex?

Note:

document.location.host does not seem to work in my case.

Answer 1

Extract hostname name from string

Try:

function cleanUp(url) {
    var url = $.trim(url);
    if(url.search(/^https?\:\/\//) != -1)
        url = url.match(/^https?\:\/\/([^\/?#]+)(?:[\/?#]|$)/i, "");
    else
        url = url.match(/^([^\/?#]+)(?:[\/?#]|$)/i, "");
    return url[1];
}

alert(cleanUp('  http://www.google.com/about.html'));
alert(cleanUp('  www.google.com/about.html'));