Recently, I learned from thinkphp’s ‘RUNTIME_ALLINONE’ experience: not afraid of cache, afraid of call chaos, simply all commonly used files are merged into one file, isn’t it beautiful…

function strip_whitespace($content) {
    $stripStr = '';
    // Analysis of the php The source code
    $tokens =   token_get_all ($content);
    $last_space = false;
    for ($i = 0, $j = count ($tokens); $i < $j; $i++){
        if (is_string ($tokens[$i])){
            $last_space = false;
            $stripStr .= $tokens[$i];
        }
        else{
            switch ($tokens[$i][0]){
                // Filter a variety of PHP annotation
                case T_COMMENT:
                case T_DOC_COMMENT:
                    break;
                // Filtering the blank space
                case T_WHITESPACE:
                    if (!$last_space){
                        $stripStr .= ' ';
                        $last_space = true;
                    }
                    break;
                default:
                    $last_space = false;
                    $stripStr .= $tokens[$i][1];
            }
        }
    }
    return $stripStr;
}

This custom function effectively solves the problem that the php_strip_whitespace system built-in annotation space function cannot be understood correctly < < < Question of EOT (heredoc) Method of use

$str = strip_whitespace('<?php'.$str);

The first 1 has to concatenate this, and I don’t get it, because if you don’t concatenate it, you’re going to get the wrong result php_strip_whitespace string php_strip_whitespace (string$filename ) If it is just a single file and there is no heredoc, the fast php_strip_whitespace function is recommended