Related
I want to add GET parameter to all URLs in special string (like html content of a website) .
For example :
Before:
$content = '... register ... login ...';
After:
$content = '... register ... login ...';
I think that this is only done with a regular expression , for this reason I wrote this function :
function makeLinks($str)
{
$str = preg_replace('#((https?://)?([-\w]+\.[-\w\.]+)+\w(:\d+)?(/([-\w/_\.]*(\?\S+)?)?)*)#', '$1?wid=${wid}', $str);
return $str;
}
But this pattern having problems! for example :
http://google.com?foo=bar => http://google.com?wid=${wid}?foo=bar
Please help me.
Try This:
function makeLinks($str)
{
$str = preg_replace_callback('/\b((?:https?|ftp):\/\/(?:[-A-Z0-9.]+)(?:\/[-A-Z0-9+&##\/%=~_|!:,.;]*)?)(?:\?([A-Z0-9+&##\/%=~_|!:,.;]*))?/i', 'modify_url', $str);
return $str;
}
function modify_url($matches) {
$query = isset($matches[2]) ? $matches[2]:'';
$result = $matches[1].'?'.$query;
if(!empty($query)) $result .= '&';
return $result.'wid=${wid}';
}
Optionally you can just add # without affecting the ending result. I hate to use them, but here it is in case you want to use them:
function modify_url($matches) {
$result = $matches[1].'?'.#$matches[2];
if(!#empty($matches[2])) $result .= '&';
return $result.'wid=${wid}';
}
Idealy, you should extract the urls and parse them, but this solution should work.
I think there may be a short way. My solution :
function makeLinks($str) {
preg_match_all('|(https?:\/\/(www\.)?[-a-zA-Z0-9#:%._\+~#=]{2,256}\.[a-z]{2,6}\b([-a-zA-Z0-9#:%_\+.~#?&//=]*))|', $str, $urls);
if ($urls && isset($urls[1])) {
foreach ($urls[1] as $url) {
$new_url = $url . (strpos($url, '?') ? '&' : '?') . 'wid=${wid}';
$str = str_replace($url, $new_url, $str);
}
}
return $str;
}
I have the following rule which checks if the url contains any parameters that start with p2:
function unparse_url($parsed_url) {
$scheme = isset($parsed_url['scheme']) ? $parsed_url['scheme'] . '://' : '';
$host = isset($parsed_url['host']) ? $parsed_url['host'] : '';
$port = isset($parsed_url['port']) ? ':' . $parsed_url['port'] : '';
$user = isset($parsed_url['user']) ? $parsed_url['user'] : '';
$pass = isset($parsed_url['pass']) ? ':' . $parsed_url['pass'] : '';
$pass = ($user || $pass) ? "$pass#" : '';
$path = isset($parsed_url['path']) ? $parsed_url['path'] : '';
$query = isset($parsed_url['query']) ? '?' . trim($parsed_url['query'], '&') : '';
$fragment = isset($parsed_url['fragment']) ? '#' . $parsed_url['fragment'] : '';
return "$scheme$user$pass$host$port$path$query$fragment";
}
function strip_query($url, $query_to_strip) {
$parsed = parse_url($url);
$parsed['query'] = preg_replace('/(^|&)'.$query_to_strip.'[^&]*/', '', $parsed['query']);
return unparse_url($parsed);
}
$url = "http://$_SERVER[HTTP_HOST]$_SERVER[REQUEST_URI]";
$new_url = (strip_query($url, 'p2'));
# redirect if url contains anything that starts with p2
$filtered = array_filter(array_keys($_GET), function($k) {
return strpos($k, 'p2') === 0;
});
if ( !empty($filtered) ) {
header ("Location: $new_url");
}
else {
}
If a url contain any p2* parameters, a redirect will be made to the same page without these parameters.
Example:
domain.com/?a=3&p2=1
will redirect to
domain.com/?a=3
However, with the current rule, if the original URL is:
domain.com/?p2=1
Then it will redirect to:
domain.com/?
But since it is meaningless, I want that in this case it will redirect to:
domain.com/
How can it be done?
Like #dontfight i would suggest to use mod_rewrite but if you prefer PHP you could use rtrim before you do the redirect
rtrim($url, '?');
With your code you could use like this
$filtered = array_filter(array_keys($_GET), function($k) {
return strpos($k, 'p2') === 0;
});
$new_url_trim = rtrim($new_url, '?');
if ( !empty($filtered) ) {
header ("Location: $new_url_trim");
}
else {
}
Here's another way to remove query parameters:
Parse the URL to separate it into the scheme, path, and query
Parse the query string into an array
Unset the parameter within the array
Recreate the URL
Example:
<?php
function removeQuery($url, $param) {
$parsed = parse_url($url);
parse_str($parsed['query'], $query);
unset($query[$param]);
$parsed['scheme'] .= '://';
$parsed['query'] = http_build_query($query);
if (!empty($parsed['query']))
$parsed['path'] .= '?';
return implode($parsed);
}
echo removeQuery('http://domain.com/?a=3&p2=1', 'a') . "\n";
echo removeQuery('http://domain.com/?a=3&p2=1', 'p2') . "\n";
echo removeQuery('http://domain.com/?p2=1', 'p2');
Output:
http://domain.com/?p2=1
http://domain.com/?a=3
http://domain.com/
Also I noticed that you wanted to remove p2 if it occurs anywhere in the string, so here's another version designed to do just that:
<?php
function removeP2($url) {
$parsed = parse_url($url);
parse_str($parsed['query'], $query);
foreach ($query as $key => $value)
if (strpos($key, 'p2') === 0)
unset($query[$key]);
$parsed['scheme'] .= '://';
$parsed['query'] = http_build_query($query);
if (!empty($parsed['query']))
$parsed['path'] .= '?';
return implode($parsed);
}
echo removeP2('http://domain.com/?a=3&p2=1') . "\n";
echo removeP2('http://domain.com/?a=3&p2sks=1') . "\n";
echo removeP2('http://domain.com/?p2=1');
Output:
http://domain.com/?a=3
http://domain.com/?a=3
http://domain.com/
Now to get it to redirect, you already know how to redirect:
header("Location: " . removeP2('http://' . $_SERVER[HTTP_HOST] . $_SERVER[REQUEST_URI]));
Since nobody has bothered to tell you what's wrong...
Your problem is that your strip_query() function always sets $url["query"] to a value, even if that value is an empty string:
$parsed['query'] = preg_replace('/(^|&)'.$query_to_strip.'[^&]*/', '', $parsed['query']);
When your unparse_url() function gets the URL, it's checking that array index using isset() which always passes:
$query = isset($parsed_url['query']) ? '?' . trim($parsed_url['query'], '&') : '';
Instead of checking with isset() you could use empty(), which will return false if the value is unset or if it's an empty string:
$query = !empty($parsed_url['query']) ? '?' . trim($parsed_url['query'], '&') : '';
if mod_rewrite is available, this can be done without php code.
This is not an answer, its a comment, but its a bit long.
Why are you so keen to strip this variable from the URL? What harm has it ever done to you that you must go to such lengths to eradicate its existence? And why not just...
unset($_GET['p2']);
But since it is meaningless,
In what way?
I can see some merit in the unparse_url($parsed_url) function and if you fix the bugs (hint: try some test cases with usernames and/or passwords) then it's reasonably well structured - but the subsequent code for manipulating the data is a messy mix of regular expression string splicing - which is not the right way to deal with such structured data - PHP has functions for that.
But the worst bit is then injecting another complete RTT to resolve the URL.
You are better than this rockyraw!
I want to encode only queryparams in a url string using php, url string given below
https://torrentz-proxy.com/search?q=linux format 2015 added<6m leech>1 seed>1#12345
urlencode() gives https%3A%2F%2Ftorrentz-proxy.com%2Fsearch%3Fq%3Dlinux+format+2015+added%3C6m+leech%3E1+seed%3E1
actually I need https://torrentz-proxy.com/search?q=linux%20format%202015%20added%3C6m%20leech%3E1%20seed%3E1#12345
How to do this ? Is there any native php function to do this ?
EDIT
URL is auto generated (not by me), sample like
http://example.com/abc?a=1 a&b=2 b&c=3+c#123
You need to parse the URL, extract the query param values, encode them and build the URL back together.
function query_param_encode($url)
{
$url = parse_url($url);
$url_str = "";
if (isset($url['scheme']))
$url_str .= $url['scheme'].'://';
if (isset($url['host']))
$url_str .= $url['host'];
if (isset($url['path']))
$url_str .= $url['path'];
if (isset($url['query']))
{
$query = explode('&', $url['query']);
foreach ($query as $j=>$value)
{
$value = explode('=', $value, 2);
if (count($value) == 2)
$query[$j] = urlencode($value[0]).'='.urlencode($value[1]);
else
$query[$j] = urlencode($value[0]);
}
$url_str .= '?'.implode('&', $query);
}
if (isset($url['fragment']))
$url_str .= '#'.$url['fragment'];
return $url_str;
}
This encodes your URLs as
https://torrentz-proxy.com/search?q=linux+format+2015+added%3C6m+leech%3E1+seed%3E1#12345
http://example.com/abc?a=1+a&b=2+b&c=3%2Bc#123
This would do
function encodeURI($url) {
// http://php.net/manual/en/function.rawurlencode.php
// https://developer.mozilla.org/en/JavaScript/Reference/Global_Objects/encodeURI
$unescaped = array(
'%2D'=>'-','%5F'=>'_','%2E'=>'.','%21'=>'!', '%7E'=>'~',
'%2A'=>'*', '%27'=>"'", '%28'=>'(', '%29'=>')'
);
$reserved = array(
'%3B'=>';','%2C'=>',','%2F'=>'/','%3F'=>'?','%3A'=>':',
'%40'=>'#','%26'=>'&','%3D'=>'=','%2B'=>'+','%24'=>'$'
);
$score = array(
'%23'=>'#'
);
return strtr(rawurlencode($url), array_merge($reserved,$unescaped,$score));
}
I found this here as an alternative to encodeURI() of JS.
Use something like this:
<?php
$url = 'https://torrentz-proxy.com/search?q=linux format 2015 added<6m leech>1 seed>1#12345';
$questPos = strpos($url, '?q');
$query = explode('#', substr($url, $questPos+3));
$encodedUrl = substr($url, 0, $questPos+3) . urlencode($query[0]);
foreach (array_slice($query, 1) as $frag) {
$encodedUrl .= '#' . $frag;
}
var_dump($encodedUrl);
that will output:
string(89) "https://torrentz-proxy.com/search?q=linux+format+2015+added%3C6m+leech%3E1+seed%3E1#12345"
The simplest way would be to :
get the base url with explode
encode only the query string (decode first to make sure to not encode twice)
make sure to replace # with %23
$myUrl = 'https://torrentz-proxy.com/search?q=linux format 2015 added<6m leech>1 seed>1#12345';
$finalUrl = explode("?", $myUrl)[0] . "?" . urlencode(urldecode(parse_url(str_replace('#', '%23', $myUrl), PHP_URL_QUERY)))
This is giving a full encoded query string "https://torrentz-proxy.com/search?q%3Dlinux+format+2015+added%3C6m+leech%3E1+seed%3E1%2312345"
Then replace %23 with # again if you like (as requested from the your question)
$finalUrl = str_replace('%23', '#', $finalUrl);
I have some links in a Powerpoint presentation, and for some reason, when those links get clicked, it adds a return parameter to the URL. Well, that return parameter is causing my Joomla site's MVC pattern to get bungled.
What's an efficient way to strip off this return parameter using PHP?
Example:
http://mydomain.example/index.php?id=115&Itemid=283&return=aHR0cDovL2NvbW11bml0
The safest "correct" method would be:
Parse the url into an array with parse_url()
Extract the query portion, decompose that into an array using parse_str()
Delete the query parameters you want by unset() them from the array
Rebuild the original url using http_build_query()
Quick and dirty is to use a string search/replace and/or regex to kill off the value.
In a different thread Justin suggests that the fastest way is to use strtok()
$url = strtok($url, '?');
See his full answer with speed tests as well here: https://stackoverflow.com/a/1251650/452515
This is to complement Marc B's answer with an example, while it may look quite long, it's a safe way to remove a parameter. In this example we remove page_number
<?php
$x = 'http://url.example/search/?location=london&page_number=1';
$parsed = parse_url($x);
$query = $parsed['query'];
parse_str($query, $params);
unset($params['page_number']);
$string = http_build_query($params);
var_dump($string);
function removeParam($url, $param) {
$url = preg_replace('/(&|\?)'.preg_quote($param).'=[^&]*$/', '', $url);
$url = preg_replace('/(&|\?)'.preg_quote($param).'=[^&]*&/', '$1', $url);
return $url;
}
parse_str($queryString, $vars);
unset($vars['return']);
$queryString = http_build_query($vars);
parse_str parses a query string, http_build_query creates a query string.
Procedural Implementation of Marc B's Answer after refining Sergey Telshevsky's Answer.
function strip_param_from_url($url, $param)
{
$base_url = strtok($url, '?'); // Get the base URL
$parsed_url = parse_url($url); // Parse it
// Add missing {
if(array_key_exists('query',$parsed_url)) { // Only execute if there are parameters
$query = $parsed_url['query']; // Get the query string
parse_str($query, $parameters); // Convert Parameters into array
unset($parameters[$param]); // Delete the one you want
$new_query = http_build_query($parameters); // Rebuilt query string
$url =$base_url.'?'.$new_query; // Finally URL is ready
}
return $url;
}
// Usage
echo strip_param_from_url( 'http://url.example/search/?location=london&page_number=1', 'location' )
You could do a preg_replace like:
$new_url = preg_replace('/&?return=[^&]*/', '', $old_url);
Here is the actual code for what's described above as the "the safest 'correct' method"...
function reduce_query($uri = '') {
$kill_params = array('gclid');
$uri_array = parse_url($uri);
if (isset($uri_array['query'])) {
// Do the chopping.
$params = array();
foreach (explode('&', $uri_array['query']) as $param) {
$item = explode('=', $param);
if (!in_array($item[0], $kill_params)) {
$params[$item[0]] = isset($item[1]) ? $item[1] : '';
}
}
// Sort the parameter array to maximize cache hits.
ksort($params);
// Build new URL (no hosts, domains, or fragments involved).
$new_uri = '';
if ($uri_array['path']) {
$new_uri = $uri_array['path'];
}
if (count($params) > 0) {
// Wish there was a more elegant option.
$new_uri .= '?' . urldecode(http_build_query($params));
}
return $new_uri;
}
return $uri;
}
$_SERVER['REQUEST_URI'] = reduce_query($_SERVER['REQUEST_URI']);
However, since this will likely exist prior to the bootstrap of your application, you should probably put it into an anonymous function. Like this...
call_user_func(function($uri) {
$kill_params = array('gclid');
$uri_array = parse_url($uri);
if (isset($uri_array['query'])) {
// Do the chopping.
$params = array();
foreach (explode('&', $uri_array['query']) as $param) {
$item = explode('=', $param);
if (!in_array($item[0], $kill_params)) {
$params[$item[0]] = isset($item[1]) ? $item[1] : '';
}
}
// Sort the parameter array to maximize cache hits.
ksort($params);
// Build new URL (no hosts, domains, or fragments involved).
$new_uri = '';
if ($uri_array['path']) {
$new_uri = $uri_array['path'];
}
if (count($params) > 0) {
// Wish there was a more elegant option.
$new_uri .= '?' . urldecode(http_build_query($params));
}
// Update server variable.
$_SERVER['REQUEST_URI'] = $new_uri;
}
}, $_SERVER['REQUEST_URI']);
NOTE: Updated with urldecode() to avoid double encoding via http_build_query() function.
NOTE: Updated with ksort() to allow params with no value without an error.
This one of many ways, not tested, but should work.
$link = 'http://mydomain.example/index.php?id=115&Itemid=283&return=aHR0cDovL2NvbW11bml0';
$linkParts = explode('&return=', $link);
$link = $linkParts[0];
Wow, there are a lot of examples here. I am providing one that does some error handling. It rebuilds and returns the entire URL with the query-string-param-to-be-removed, removed. It also provides a bonus function that builds the current URL on the fly. Tested, works!
Credit to Mark B for the steps. This is a complete solution to tpow's "strip off this return parameter" original question -- might be handy for beginners, trying to avoid PHP gotchas. :-)
<?php
function currenturl_without_queryparam( $queryparamkey ) {
$current_url = current_url();
$parsed_url = parse_url( $current_url );
if( array_key_exists( 'query', $parsed_url )) {
$query_portion = $parsed_url['query'];
} else {
return $current_url;
}
parse_str( $query_portion, $query_array );
if( array_key_exists( $queryparamkey , $query_array ) ) {
unset( $query_array[$queryparamkey] );
$q = ( count( $query_array ) === 0 ) ? '' : '?';
return $parsed_url['scheme'] . '://' . $parsed_url['host'] . $parsed_url['path'] . $q . http_build_query( $query_array );
} else {
return $current_url;
}
}
function current_url() {
$current_url = 'http' . (isset($_SERVER['HTTPS']) ? 's' : '') . '://' . "{$_SERVER['HTTP_HOST']}{$_SERVER['REQUEST_URI']}";
return $current_url;
}
echo currenturl_without_queryparam( 'key' );
?>
$var = preg_replace( "/return=[^&]+/", "", $var );
$var = preg_replace( "/&{2,}/", "&", $var );
Second line will just replace && to &
very simple
$link = "http://example.com/index.php?id=115&Itemid=283&return=aHR0cDovL2NvbW11bml0"
echo substr($link, 0, strpos($link, "return") - 1);
//output : http://example.com/index.php?id=115&Itemid=283
#MarcB mentioned that it is dirty to use regex to remove an url parameter. And yes it is, because it's not as easy as it looks:
$urls = array(
'example.com/?foo=bar',
'example.com/?bar=foo&foo=bar',
'example.com/?foo=bar&bar=foo',
);
echo 'Original' . PHP_EOL;
foreach ($urls as $url) {
echo $url . PHP_EOL;
}
echo PHP_EOL . '#AaronHathaway' . PHP_EOL;
foreach ($urls as $url) {
echo preg_replace('#&?foo=[^&]*#', null, $url) . PHP_EOL;
}
echo PHP_EOL . '#SergeS' . PHP_EOL;
foreach ($urls as $url) {
echo preg_replace( "/&{2,}/", "&", preg_replace( "/foo=[^&]+/", "", $url)) . PHP_EOL;
}
echo PHP_EOL . '#Justin' . PHP_EOL;
foreach ($urls as $url) {
echo preg_replace('/([?&])foo=[^&]+(&|$)/', '$1', $url) . PHP_EOL;
}
echo PHP_EOL . '#kraftb' . PHP_EOL;
foreach ($urls as $url) {
echo preg_replace('/(&|\?)foo=[^&]*&/', '$1', preg_replace('/(&|\?)foo=[^&]*$/', '', $url)) . PHP_EOL;
}
echo PHP_EOL . 'My version' . PHP_EOL;
foreach ($urls as $url) {
echo str_replace('/&', '/?', preg_replace('#[&?]foo=[^&]*#', null, $url)) . PHP_EOL;
}
returns:
Original
example.com/?foo=bar
example.com/?bar=foo&foo=bar
example.com/?foo=bar&bar=foo
#AaronHathaway
example.com/?
example.com/?bar=foo
example.com/?&bar=foo
#SergeS
example.com/?
example.com/?bar=foo&
example.com/?&bar=foo
#Justin
example.com/?
example.com/?bar=foo&
example.com/?bar=foo
#kraftb
example.com/
example.com/?bar=foo
example.com/?bar=foo
My version
example.com/
example.com/?bar=foo
example.com/?bar=foo
As you can see only #kraftb posted a correct answer using regex and my version is a little bit smaller.
Remove Get Parameters From Current Page
<?php
$url_dir=$_SERVER['REQUEST_URI'];
$url_dir_no_get_param= explode("?",$url_dir)[0];
echo $_SERVER['HTTP_HOST'].$url_dir_no_get_param;
This should do it:
public function removeQueryParam(string $url, string $param): string
{
$parsedUrl = parse_url($url);
if (isset($parsedUrl[$param])) {
$baseUrl = strtok($url, '?');
parse_str(parse_url($url)['query'], $query);
unset($query[$param]);
return sprintf('%s?%s',
$baseUrl,
http_build_query($query)
);
}
return $url;
}
Simple solution that will work for every url
With this solution $url format or parameter position doesn't matter, as an example I added another parameter and anchor at the end of $url:
https://example.com/index.php?id=115&Itemid=283&return=aHR0cDovL2NvbW11bml0&bonus=test#test2
Here is the simple solution:
$url = 'https://example.com/index.php?id=115&Itemid=283&return=aHR0cDovL2NvbW11bml0&bonus=test#test2';
$url_query_stirng = parse_url($url, PHP_URL_QUERY);
parse_str( $url_query_stirng, $url_parsed_query );
unset($url_parsed_query['return']);
$url = str_replace( $url_query_stirng, http_build_query( $url_parsed_query ), $url );
echo $url;
Final result for $url string is:
https://example.com/index.php?id=115&Itemid=283&bonus=test#test2
Some of the examples posted are so extensive. This is what I use on my projects.
function removeQueryParameter($url, $param){
list($baseUrl, $urlQuery) = explode('?', $url, 2);
parse_str($urlQuery, $urlQueryArr);
unset($urlQueryArr[$param]);
if(count($urlQueryArr))
return $baseUrl.'?'.http_build_query($urlQueryArr);
else
return $baseUrl;
}
function remove_attribute($url,$attribute)
{
$url=explode('?',$url);
$new_parameters=false;
if(isset($url[1]))
{
$params=explode('&',$url[1]);
$new_parameters=ra($params,$attribute);
}
$construct_parameters=($new_parameters && $new_parameters!='' ) ? ('?'.$new_parameters):'';
return $new_url=$url[0].$construct_parameters;
}
function ra($params,$attr)
{ $attr=$attr.'=';
$new_params=array();
for($i=0;$i<count($params);$i++)
{
$pos=strpos($params[$i],$attr);
if($pos===false)
$new_params[]=$params[$i];
}
if(count($new_params)>0)
return implode('&',$new_params);
else
return false;
}
//just copy the above code and just call this function like this to get new url without particular parameter
echo remove_attribute($url,'delete_params'); // gives new url without that parameter
I know this is an old question but if you only want to remove one or few named url parameter you can use this function:
function RemoveGet_Regex($variable, $rewritten_url): string {
$rewritten_url = preg_replace("/(\?)$/", "", preg_replace("/\?&/", "?", preg_replace("/((?<=\?)|&){$variable}=[\w]*/i", "", $rewritten_url)));
return $rewritten_url;
}
function RemoveGet($name): void {
$rewritten_url = "https://$_SERVER[HTTP_HOST]$_SERVER[REQUEST_URI]";
if(is_array($name)) {
for($i = 0; $i < count($name); $i++) {
$rewritten_url = RemoveGet_Regex($name[$i], $rewritten_url);
$is_set[] = isset($_GET[$name[$i]]);
}
$array_filtered = array_filter($is_set);
if (!empty($array_filtered)) {
header("Location: ".$rewritten_url);
}
}
else {
$rewritten_url = RemoveGet_Regex($name, $rewritten_url);
if(isset($_GET[$name])) {
header("Location: ".$rewritten_url);
}
}
}
In the first function preg_replace("/((?<=\?)|&){$variable}=[\w]*/i", "", $rewritten_url) will remove the get parameter, and the others will tidy it up. The second function will then redirect.
RemoveGet("id"); will remove the id=whatever from the url. The function can also work with arrays. For your example,
Remove(array("id","Item","return"));
To strip any parameter from the url using PHP script you need to follow this script:
function getNewArray($array,$k){
$dataArray = $array;
unset($array[$k]);
$dataArray = $array;
return $dataArray;
}
function getFullURL(){
return (isset($_SERVER['HTTPS']) && $_SERVER['HTTPS'] === 'on' ? "https" : "http") . "://$_SERVER[HTTP_HOST]$_SERVER[REQUEST_URI]";
}
$url = getFullURL();
$url_components = parse_url($url);
// Use parse_str() function to parse the
// string passed via URL
parse_str($url_components['query'], $params);
print_r($params);
<ul>
<?php foreach($params as $k=>$v){?>
<?php
$newArray = getNewArray($params,$k);
$parameters = http_build_query($newArray);
$newURL = $_SERVER['PHP_SELF']."?".$parameters;
?>
<li><?=$v;?> X
<?php }?>
</ul>
here is functions optimized for speed. But this functions DO NOT remove arrays like a[]=x&a[1]bb=y&a[2]=z by array name.
function removeQueryParam($query, $param)
{
$quoted_param = preg_quote($param, '/');
$pattern = "/(^$quoted_param=[^&]*&?)|(&$quoted_param=[^&]*)/";
return preg_replace($pattern, '', $query);
}
function removeQueryParams($query, array $params)
{
if ($params)
{
$pattern = '/';
foreach ($params as $param)
{
$quoted_param = preg_quote($param, '/');
$pattern .= "(^$quoted_param=[^&]*&?)|(&$quoted_param=[^&]*)|";
}
$pattern[-1] = '/';
return preg_replace($pattern, '', $query);
}
return $query;
}
<? if(isset($_GET['i'])){unset($_GET['i']); header('location:/');} ?>
This will remove the 'i' parameter from the URL. Change the 'i's to whatever you need.
when we add a param to the URL
$redirectURL = $printPageURL . "?mode=1";
it works if $printPageURL is "http://www.somesite.com/print.php", but if $printPageURL is changed in the global file to "http://www.somesite.com/print.php?newUser=1", then the URL becomes badly formed. If the project has 300 files and there are 30 files that append param this way, we need to change all 30 files.
the same if we append using "&mode=1" and $printPageURL changes from "http://www.somesite.com/print.php?new=1" to "http://www.somesite.com/print.php", then the URL is also badly formed.
is there a library in PHP that will automatically handle the "?" and "&", and even checks that existing param exists already and removed that one because it will be replaced by the later one and it is not good if the URL keeps on growing longer?
Update: of the several helpful answers, there seems to be no pre-existing function addParam($url, $newParam) so that we don't need to write it?
Use a combination of parse_url() to explode the URL, parse_str() to explode the query string and http_build_query() to rebuild the querystring. After that you can rebuild the whole url from its original fragments you get from parse_url() and the new query string you built with http_build_query(). As the querystring gets exploded into an associative array (key-value-pairs) modifying the query is as easy as modifying an array in PHP.
EDIT
$query = parse_url('http://www.somesite.com/print.php?mode=1&newUser=1', PHP_URL_QUERY);
// $query = "mode=1&newUser=1"
$params = array();
parse_str($query, $params);
/*
* $params = array(
* 'mode' => '1'
* 'newUser' => '1'
* )
*/
unset($params['newUser']);
$params['mode'] = 2;
$params['done'] = 1;
$query = http_build_query($params);
// $query = "mode=2&done=1"
Use this:
http://hu.php.net/manual/en/function.http-build-query.php
http://www.addedbytes.com/php/querystring-functions/
is a good place to start
EDIT: There's also http://www.php.net/manual/en/class.httpquerystring.php
for example:
$http = new HttpQueryString();
$http->set(array('page' => 1, 'sort' => 'asc'));
$url = "yourfile.php" . $http->toString();
None of these solutions work when the url is of the form:
xyz.co.uk?param1=2&replace_this_param=2
param1 gets dropped all the time
.. which means it never works EVER!
If you look at the code given above:
function addParam($url, $s) {
return adjustParam($url, $s);
}
function delParam($url, $s) {
return adjustParam($url, $s);
}
These functions are IDENTICAL - so how can one add and one delete?!
using WishCow and sgehrig's suggestion, here is a test:
(assuming no anchor for the URL)
<?php
echo "<pre>\n";
function adjustParam($url, $s) {
if (preg_match('/(.*?)\?/', $url, $matches)) $urlWithoutParams = $matches[1];
else $urlWithoutParams = $url;
parse_str(parse_url($url, PHP_URL_QUERY), $params);
if (strpos($s, '=') !== false) {
list($var, $value) = split('=', $s);
$params[$var] = urldecode($value);
return $urlWithoutParams . '?' . http_build_query($params);
} else {
unset($params[$s]);
$newQueryString = http_build_query($params);
if ($newQueryString) return $urlWithoutParams . '?' . $newQueryString;
else return $urlWithoutParams;
}
}
function addParam($url, $s) {
return adjustParam($url, $s);
}
function delParam($url, $s) {
return adjustParam($url, $s);
}
echo "trying add:\n";
echo addParam("http://www.somesite.com/print.php", "mode=3"), "\n";
echo addParam("http://www.somesite.com/print.php?", "mode=3"), "\n";
echo addParam("http://www.somesite.com/print.php?newUser=1", "mode=3"), "\n";
echo addParam("http://www.somesite.com/print.php?newUser=1&fee=0", "mode=3"), "\n";
echo addParam("http://www.somesite.com/print.php?newUser=1&fee=0&", "mode=3"), "\n";
echo addParam("http://www.somesite.com/print.php?mode=1", "mode=3"), "\n";
echo "\n", "now trying delete:\n";
echo delParam("http://www.somesite.com/print.php?mode=1", "mode"), "\n";
echo delParam("http://www.somesite.com/print.php?mode=1&newUser=1", "mode"), "\n";
echo delParam("http://www.somesite.com/print.php?mode=1&newUser=1", "newUser"), "\n";
?>
and the output is:
trying add:
http://www.somesite.com/print.php?mode=3
http://www.somesite.com/print.php?mode=3
http://www.somesite.com/print.php?newUser=1&mode=3
http://www.somesite.com/print.php?newUser=1&fee=0&mode=3
http://www.somesite.com/print.php?newUser=1&fee=0&mode=3
http://www.somesite.com/print.php?mode=3
now trying delete:
http://www.somesite.com/print.php
http://www.somesite.com/print.php?newUser=1
http://www.somesite.com/print.php?mode=1
You can try this:
function removeParamFromUrl($query, $paramToRemove)
{
$params = parse_url($query);
if(isset($params['query']))
{
$queryParams = array();
parse_str($params['query'], $queryParams);
if(isset($queryParams[$paramToRemove])) unset($queryParams[$paramToRemove]);
$params['query'] = http_build_query($queryParams);
}
$ret = $params['scheme'].'://'.$params['host'].$params['path'];
if(isset($params['query']) && $params['query'] != '' ) $ret .= '?'.$params['query'];
return $ret;
}