☰

General URL Cleaner

Cleans URL's from various popular sites. Also, makes sure the sites are using HTTPS.

Verze ze dne 16. 07. 2015. Zobrazit nejnovější verzi.

// ==UserScript==
// @run-at document-start
// @name           General URL Cleaner
// @namespace      
// @description    Cleans URL's from various popular sites. Also, makes sure the sites are using HTTPS.
// @include        /^https?://[a-z]*.google.(com|ca|co.uk|com.au|co.nz|ac|ad|ae|com.af|com.ag|com.ai|al|am|co.ao|com.ar)\/.*$/
// @include        /^https?://[a-z]*.google.(us|as|at|az|ba|com.bd|be|bf|bg|com.bh|bi|bj|com.bn|com.bo|com.br|bs|bt|co.bw)\/.*$/
// @include        /^https?://[a-z]*.google.(by|com.bz|com.kh|cc|cd|cf|cat|cg|ch|ci|co.ck|cl|cm|cn|com.co|co.cr|com.cu|cv)\/.*$/
// @include        /^https?://[a-z]*.google.(com.cy|cz|de|dj|dk|dm|com.do|dz|com.ec|ee|com.eg|es|com.et|fi|com.fj|fm|fr)\/.*$/
// @include        /^https?://[a-z]*.google.(ga|ge|gf|gg|com.gh|com.gi|gl|gm|gp|gr|com.gt|gy|com.hk|hn|hr|ht|hu|co.id|ir)\/.*$/
// @include        /^https?://[a-z]*.google.(iq|ie|co.il|im|co.in|io|is|it|je|com.jm|jo|co.jp|co.ke|ki|kg|co.kr|com.kw|kz)\/.*$/
// @include        /^https?://[a-z]*.google.(la|com.lb|com.lc|li|lk|co.ls|lt|lu|lv|com.ly|co.ma|md|me|mg|mk|ml|com.mm|mn)\/.*$/
// @include        /^https?://[a-z]*.google.(ms|com.mt|mu|mv|mw|com.mx|com.my|co.mz|com.na|ne|com.nf|com.ng|com.ni|nl|no)\/.*$/
// @include        /^https?://[a-z]*.google.(com.np|nr|nu|com.om|com.pa|com.pe|com.ph|com.pk|pl|com.pg|pn|com.pr|ps|com.py)\/.*$/
// @include        /^https?://[a-z]*.google.(vg|pt|com.qa|ro|rs|ru|rw|com.sa|com.sb|sc|se|com.sg|sh|si|sk|com.sl|sn|sm|so)\/.*$/
// @include        /^https?://[a-z]*.google.(st|com.sv|td|tg|co.th|com.tj|tk|tl|tm|to|tn|com.tn|com.tr|tt|com.tw|co.tz)\/.*$/
// @include        /^https?://[a-z]*.google.(com.ua|co.ug|com.uy|co.uz|com.vc|co.ve|co.vi|com.vn|vu|ws|co.za|co.zm|co.zw)\/.*$/
// @include        /^https?://[a-z]*.amazon.(cn|in|co.jp|fr|de|it|nl|es|co.uk|ca|com.mx|com|com.au|com.br)\/.*$/
// @include        /^https?://[a-z]*.newegg.(com|ca|cn)\/.*$/
// @include        /^https?://[a-z]*.ebay.(com.au|at|be|ca|fr|de|com.hk|in|ie|co.il|it|com.my|nl|co.za|ph|pl|com.sg|co.za|es|ch|co.th|co.uk|com|vn)\/.*$/
// @include        /^https?://[a-z]*.bing.com\/.*$/
// @include        /^https?://[a-z]*.youtube.com\/.*$/
// @include        /^https?://[a-z]*.dealtime.com\/.*$/
// @exclude        https://apis.google.com/*
// @exclude        https://www.google.com/recaptcha/api2/*
// @version        1.9.1.2
// @license        GPL version 3 or any later version; http://www.gnu.org/copyleft/gpl.html
// ==/UserScript==

// compile these regexes beforehand to improve efficiency
var bing = new RegExp(/^https?:\/\/www\.bing\.(.+?)\/search\?/);
var google = new RegExp(/^https?:\/\/[a-z]*\.google\.(.+?)\/[a-z]*\?/);
var googleImageRedirect = new RegExp(/^https?:\/\/www\.google\.(.+?)\/url\?/);
var youtube = new RegExp(/^https?:\/\/www\.youtube\.com\/watch/);
var ebay = new RegExp(/^https?:\/\/www\.ebay\.(.+?)\/itm/);
var ebaySearch = new RegExp(/^https?:\/\/www\.ebay\.(.+?)\/sch\//);
var amazondp = new RegExp(/^https?:\/\/www\.amazon\..*\/dp\//);
var amazongp = new RegExp(/^https?:\/\/www\.amazon\..*\/gp\/product\//);
var newegg = new RegExp(/^http:\/\/www\.newegg\.(com|ca)\/Product\/Product\.aspx/);
var dealtime = new RegExp(/http:\/\/stat\.dealtime\.com\/DealFrame\/DealFrame\.cmp\?/);

var googleSearchParameters = new RegExp(
    /\&(aqs|es_sm|channel|tab|num|hl|safe|tbo|sclient|sourceid|spell|site|sa|ei|client|complete|as_qdr|um|sa|tab|authuser|rlz|cad|rct|ved|usg|source|oe|oq|ie|dpr|gs_l|ved|tbas|sei|biw|bih|gpsrc|gfe_rd|gws_rd)\=[^&]*/g);

// Clean the current page URL
var newPageUrl = cleanUrl(document.URL);
if (newPageUrl != document.URL) location.replace(newPageUrl);

// Cleans links on the page
var links = document.links;
var excludeLinks = new RegExp(/(^$|^javascript\:|^mailto\:|^data\:)/); // don't do anything with links that are blank, javascript, email addresses, data

if (google.test(newPageUrl)) {
    document.addEventListener("DOMContentLoaded", cleanGooglePageLinks, false);
    window.onhashchange = googleInstant;
}
else {
    document.addEventListener("DOMContentLoaded", cleanPageLinks, false);
}

// Standard link cleaning function
function cleanPageLinks() {
    for (var i = links.length; i--;) {
        if (excludeLinks.test(links[i].href)) continue; // Links to skip
        links[i].href = cleanUrl(links[i].href); // Standard link cleaning
    }
    this.removeEventListener('DOMContentLoaded', cleanPageLinks, false); // We don't need to keep the event listener running
}

// Google search results link cleaning function
function cleanGooglePageLinks() {
    for (var i = links.length; i--;) {
        if (excludeLinks.test(links[i].href)) continue; // Links to skip
        links[i].removeAttribute('onmousedown'); // Remove search results redirection
        links[i].href = cleanUrl(links[i].href); // Standard link cleaning
    }
    this.removeEventListener('DOMContentLoaded', cleanGooglePageLinks, false); // We don't need to keep event listener running
}

// Google Instant document URL cleaning - if the search terms change, remove the extra stuff.
function googleInstant() {
    if (!document.URL.includes('#imgrc=')) {                                                        // Don't rewrite anything if an image is clicked in image searches
        var newSearchString = String(document.URL.match(/\#.*/)).replace(/^\#/,'');                 // The string after the hash, containing the new search terms
        var newSearchUrl = String(document.URL.replace(/search\?.*/, 'search?' + newSearchString)); // Remake the full URL with only the new search terms
        location.replace(newSearchUrl);
    }
}

// Main function for cleaning the url's
function cleanUrl(oldurl) {
    var newurl = oldurl;
    switch(true) {
        case googleImageRedirect.test(oldurl):
            newurl = decodeURIComponent(oldurl.replace(/^.*\&url\=/,'').replace(/\&(psig|ei|bvm)\=.*$/g,''));
            break;
        case google.test(oldurl):
            newurl = oldurl.replace('?','?&') // temporarily put an "&" after the "?" so that the regex below will always match
                           .replace(googleSearchParameters,'')
                           .replace('?&','?')
                           .replace(/^http\:/,'https:'); // always use https
            break;
        case bing.test(oldurl):
            newurl = oldurl.replace('?','?&')
                           .replace(/\&(go|qs|form|FORM|filt|pq|sc|sp|sk|qpvt)\=[^&]*/g,'')
                           .replace('?&','?')
                           .replace(/^http\:/,'https:');
            break;
        case youtube.test(oldurl):
            newurl = 'https://www.youtube.com/watch?' + oldurl.match(/v\=[^&]*/);
            break;
        case ebay.test(oldurl):
            // the split gets the domain name. Should be more efficient than a regex.
            newurl = 'http://' + oldurl.split('/')[2] + '/itm' + oldurl.match(/\/[0-9]{11,13}[^#?&\/]/);
            break;
        case ebaySearch.test(oldurl):
            newurl = oldurl.replace('?','?&')
                           .replace(/\&(\_osacat|\_odkw|\_from|rt|\_trksid|\_sacat)\=[^&]*/g,'')
                           .replace('?&','?');
            break;
        case amazondp.test(oldurl):
            newurl = 'http://' + oldurl.split('/')[2] + oldurl.match(/\/dp\/[A-Z0-9]{10}\/?/);
            break;
        case amazongp.test(oldurl):
            newurl = 'http://' + oldurl.split('/')[2] + oldurl.match(/\/gp\/product\/[A-Z0-9]{10}\/?/);
            break;
        case newegg.test(oldurl):
            newurl = 'http://' + oldurl.split('/')[2] + oldurl.match(/\/Product\/Product\.aspx\?Item\=[^&]*/);
            break;
        case dealtime.test(oldurl):
            newurl = decodeURIComponent(oldurl.replace(/.*\&url\=/,'').replace(/(\%26|)\&linkin_id\=.*$/,'')).replace(/\&(url|partner)\=[^&]*/g,'');
            break;
        default:
            break;
    }
    newurl = newurl.replace(/((\?|\&|)utm_(source|medium|campaign)\=[^&]*|\&amp\;)/g,'');
    return newurl;
}