URL Implode

gsnedders · 12-21-2005, 06:32 PM

PHP Code:


			
function url_implode($parsed)

{

    $output = '';

    if (isset($parsed['scheme']))

        $output .= "$parsed[scheme]://";

    if (isset($parsed['user']) && isset($parsed['pass']))

        $output .= "$parsed[user]:$parsed[pass]@";

    elseif (isset($parsed['user']))

        $output .= "$parsed[user]@";

    if (isset($parsed['host']))

        $output .= rawurlencode(rawurldecode($parsed['host']));

    if (isset($parsed['path']))

        $output .= "$parsed[path]";

    if (isset($parsed['query']))

        $output .= '?' . urlencode(urldecode($parsed['query']));

    if (isset($parsed['fragment']))

        $output .= "#$parsed[fragment]";

    return $output;

}

This function will take an parse_url array, and return a URL.

You may be wondering what the point of it is, as (normally) if you have a parse_url array, you have the URL. I was creating a PHP script to grab news from a site (using preg_match_all()) and produce an Atom feed. An Atom feed is XML, therefore must be well-formed, so all the parts of a URL must be properly escaped.

Element · 12-21-2005, 09:30 PM

Quote:

Originally Posted by Error 404

PHP Code:


			
function url_implode($parsed)

{

    $output = '';

    if (isset($parsed['scheme']))

        $output .= "$parsed[scheme]://";

    if (isset($parsed['user']) && isset($parsed['pass']))

        $output .= "$parsed[user]:$parsed[pass]@";

    elseif (isset($parsed['user']))

        $output .= "$parsed[user]@";

    if (isset($parsed['host']))

        $output .= rawurlencode(rawurldecode($parsed['host']));

    if (isset($parsed['path']))

        $output .= "$parsed[path]";

    if (isset($parsed['query']))

        $output .= '?' . urlencode(urldecode($parsed['query']));

    if (isset($parsed['fragment']))

        $output .= "#$parsed[fragment]";

    return $output;

}

This function will take an parse_url array, and return a URL.

You may be wondering what the point of it is, as (normally) if you have a parse_url array, you have the URL. I was creating a PHP script to grab news from a site (using preg_match_all()) and produce an Atom feed. An Atom feed is XML, therefore must be well-formed, so all the parts of a URL must be properly escaped.

Indeed, this is very useful for that. Its also good for script like that one topsite (Gold something? I don't remember the name.) that stored urls all broken up by parse_url() (For some pointless reason.) And they manually put the URLs back together for redirection, but this would reduce their code alot by using a simple, useful function. Thanks for this.

gsnedders · 12-26-2005, 04:10 PM

New version:

PHP Code:


			
function cleanup_url($url)

{

    $parsed = parse_url($url);

    $output = '';

    if (isset($parsed['scheme']))

        $output .= "$parsed[scheme]://";

    else

        $output .= 'http://';

    if (isset($parsed['user']) && isset($parsed['pass']))

        $output .= "$parsed[user]:$parsed[pass]@";

    elseif (isset($parsed['user']))

        $output .= "$parsed[user]@";

    if (isset($parsed['host']))

        $output .= rawurlencode(rawurldecode($parsed['host']));

    if (isset($parsed['path']))

        $output .= "$parsed[path]";

    else

        $output .= '/';

    if (isset($parsed['query']))

        $output .= '?' . urlencode(urldecode($parsed['query']));

    if (isset($parsed['fragment']))

        $output .= "#$parsed[fragment]";

    if (isset($parsed['scheme']))

        return $output;

    else

        return cleanup_url($output);

}

What's new:

If no scheme is specified (eg. http, https, ftp) "http" is assumed
If no path is specified (eg. /newreply.php) "/" is assumed