Wordpress minify html only in the main page

Question

I was using this code to minify HTML output in wordpress. It works perfect on the main page and on the post page, but in the admin section it causes a lot of problems.

function minify_html(){
    ob_start('html_compress');
}

function html_compress($buffer){

    $search = array(
        '/\n/',         // replace end of line by a space
        '/\>[^\S ]+/s',     // strip whitespaces after tags, except space
        '/[^\S ]+\</s',     // strip whitespaces before tags, except space
        '/(\s)+/s',     // shorten multiple whitespace sequences,
        '~<!--//(.*?)-->~s' //html comments
    );

    $replace = array(
        ' ',
        '>',
        '<',
        '\\1',
        ''
    );

    $buffer = preg_replace($search, $replace, $buffer);

    return $buffer;
}

add_action('wp_loaded','minify_html');

Using 'the_post' instead of 'wp_loaded' minifies only the posts, but I'd like to be able to minify at 100% the main page, and the post page, but nothing in the admin section. How can I combine the actions in order to manage it?

Thank you!

Sycee

That is an excellent piece of code to add to your theme functions instead of having to manage yet another plugin. Thanks for that! – Seamark

Answer 1

Nice Code, exclude admin :

if (!(is_admin() )) {
function minify_html(){
    ob_start('html_compress');
}

function html_compress($buffer){

    $search = array(
        '/\n/',         // replace end of line by a space
        '/\>[^\S ]+/s',     // strip whitespaces after tags, except space
        '/[^\S ]+\</s',     // strip whitespaces before tags, except space
        '/(\s)+/s',     // shorten multiple whitespace sequences,
        '~<!--//(.*?)-->~s' //html comments
    );

    $replace = array(
        ' ',
        '>',
        '<',
        '\\1',
        ''
    );

    $buffer = preg_replace($search, $replace, $buffer);

    return $buffer;
}

add_action('wp_loaded','minify_html'); }

It works well WP admin !

Swett

And if you have users who login to your site and those pages are having problems adding if(!(is_admin() && !is_user_logged_in())){ your stuff }; works pretty good too. – Seamark

Update Alternative code : gist.github.com/loorlab/f47cc2cdd3552b798925d764c5ebe812 – Swett

Answer 2

There is a troubleshoot case where replacing empty space with preg_replace causes invalid characters with UTF-8. (eg: à encoding would fail and result in �). This is due to a mismatch between the current locale and the defined locale. We can use the /u flag as a way to properly filter UTF-8 characters.

In some instances, removing new lines and multiple spaces from inline <script> tags will cause the script to break. Isolating the <script> tag's content will prevent this from happening which is done by splitting the buffer into an array delimited by opening and closing script tags through preg_split().

<?php

/**
 * Turn on output buffering, prettify and minimy the HTML output.
 * 
 * @see https://www.php.net/manual/en/function.ob-start.php
 * 
 * @since 1.0.0
 */
add_action( 'wp_loaded', function () {

    ob_start( function ( $buffer ) {

        /**
         * preg_replace() UTF-8 troubleshooting.
         * 
         * Replacing empty space with preg_replace causes invalid characters with UTF-8.
         * As preg_replace() depends on the current defined locale, characters not supported will be returned as � invalid.
         * The /u flag is used to make regex unicode aware.
         * 
         * @see https://mcmap.net/q/95241/-replacing-empty-space-with-preg_replace-causes-invalid-characters-with-utf-8
         */
        $replace = array(
            '/\n/smu'                   => '',      //Remove new lines.
            '/(\s)+/smu'                => '\\1',   //Replace multiple spaces with a single one.
            '/\s*<\s*/smu'              => '<',     //Remove spaces before and after opening tags.
            '/\s*>\s*/smu'              => '>',     //Remove spaces before and after closing tags.
            '/\s*\/>\s*/smu'            => '>',     //Remove spaces before and after closing tags for self closing tags.
            '/<!--(.|\s)*?-->/'         => '',      //Remove HTML comments.
            '/\s*=\s*\'([^\']*)\'/smu'  => '="$1"', //Replace single quotes with double quotes within attributes.
        );
        
        /**
         * preg_replace() inline <script> troubleshooting.
         * 
         * In some instances, removing new lines and multiple spaces from inline <script> tags will cause the script to break.
         * Isolating the <script> tag's content will prevent this from happening which is done by splitting the buffer into an array delimited by opening and closing script tags through preg_split().
         * 
         * @see https://mcmap.net/q/94935/-removing-new-lines-except-in-lt-pre-gt
         */
        $buffer = preg_split( '/(<\/?script[^>]*>)/', $buffer, null, PREG_SPLIT_DELIM_CAPTURE );

        foreach ( $buffer as $key => $value ) {

            //If the $key is a <script> opening tag, $key + 1 is the script tag's content, $key + 2 is the script closing tag.
            if ( false !== stripos( $value, '<script' ) ) {

                $k = $key;

            };

            //$key + 1 is the script tag's content, which we want to ignore.
            if ( $k + 1 === $key ) {

                unset( $k );

                continue;

            };

            $buffer[ $key ] = preg_replace( array_keys( $replace ), array_values( $replace ), $value );

        };

        return implode( '', $buffer );

    } );

} );

/**
 * Get current buffer contents and delete current output buffer.
 * 
 * @see https://www.php.net/manual/en/function.ob-get-clean.php
 * 
 * @since 1.0.0
 */
add_action( 'shutdown', function () {

    ob_get_clean();

} );

Onfroi