$string = '你好123欢迎你abcd';
$cn = preg_replace('/[^\p{Han}]/u', '', $string); // 挑选出中文字符
echo $cn;
//结果：你好欢迎你

<?
$str = "测试中文";
echo $str;
echo "<hr>";
// if (preg_match("/^[".chr(0xa1)."-".chr(0xff)."]+$/", $str)) { //只能在GB2312情况下使用
// if (preg_match("/^[\x7f-\xff]+$/", $str)) { //兼容gb2312,utf-8 //判断字符串是否全是中文
if (preg_match("/[\x7f-\xff]/", $str)) { // 判断字符串中是否有中文
echo "正确输入";
} else {
echo "错误输入";
}
?>

1. GBK (GB2312/GB18030)
\x00-\xff GBK双字节编码范围
\x20-\x7f ASCII
\xa1-\xff 中文 gb2312
\x80-\xff 中文 gbk

2. UTF-8 (Unicode)
\u4e00-\u9fa5 (中文)
\x3130-\x318F (韩文
\xAC00-\xD7A3 (韩文)
\u0800-\u4e00 (日文)

<?php
// 确保字符串为utf-8
// 转换 GB2312 -> UTF-8
$str = mb_convert_encoding($str, 'UTF-8', 'GB2312');
echo preg_replace('~[^\p{Han}]~u', '', $str),"\n";

<?php
$str = "a 1b 2b<中文>。xxyy字符";
// 转换 GB2312 -> UTF-8
$str = mb_convert_encoding($str, 'UTF-8', 'GB2312');
preg_match_all('/[\x{4e00}-\x{9fff}]+/u', $str, $matches);
$str = join('', $matches[0]);
// 转换 UTF-8 -> GB2312
$str = mb_convert_encoding($str, 'GB2312', 'UTF-8');
echo $str; // 输出 中文字符
?>

这个是结合了下面3种方法
echo '只保留英文 : '.preg_replace("/[^a-zA-Z0-9_ \-+=.\{\}\(\)\/\\\]+/","", $a);

<?php
$str='AB12@#$%()_.-cd';
var_dump($str);
var_dump(preg_replace("/[^a-zA-Z0-9_.-]+/","", $str));
?>
输出:

string(15) "AB12@#$%()_.-cd"
string(9) "AB12_.-cd"

<?php
$text = 'In the 电影_后天 230809-people died.';

$preg = '/[^a-zA-Z0-9.\-_]/';
preg_replace($preg, "", $text);
// Inthe_230809-peopledied.
?>

$patt = '/[a-zA-Z0-9_\\-\\.]+/';
preg_match_all($patt, $content, $result);
echo implode('', $result);