php 中文和编码判断代码_php技巧

编码范围1. GBK (GB2312/GB18030)
\x00-\xff GBK双字节编码范围
\x20-\x7f ASCII
\xa1-\xff 中文
\x80-\xff 中文
2. UTF-8 (Unicode)
\u4e00-\u9fa5 (中文)
\x3130-\x318F (韩文
\xAC00-\xD7A3 (韩文)
\u0800-\u4e00 (日文)
ps: 韩文是大于[\u9fa5]的字符
正则例子:
preg_replace(”/([\x80-\xff])/”,””,$str);
preg_replace(”/([u4e00-u9fa5])/”,””,$str);
二、代码例子

 <BR>//判断内容里有没有中文-GBK (PHP) <BR>function check_is_chinese($s){ <BR>return preg_match('/[\x80-\xff]./', $s); <BR>} <BR>//获取字符串长度-GBK (PHP) <BR>function gb_strlen($str){ <BR>$count = 0; <BR>for($i=0; $i<strlen($str); $i++){ <BR>$s = substr($str, $i, 1); <BR>if (preg_match("/[\x80-\xff]/", $s)) ++$i; <BR>++$count; <BR>} <BR>return $count; <BR>} <BR>//截取字符串字串-GBK (PHP) <BR>function gb_substr($str, $len){ <BR>$count = 0; <BR>for($i=0; $i<<a style="color:transparent">来@源gao*daima.com搞@代#码网</a><strong>搞gaodaima代码</strong>strlen($str); $i++){ <BR>if($count == $len) break; <BR>if(preg_match("/[\x80-\xff]/", substr($str, $i, 1))) ++$i; <BR>++$count; <BR>} <BR>return substr($str, 0, $i); <BR>} <BR>//统计字符串长度-UTF8 (PHP) <BR>function utf8_strlen($str) { <BR>$count = 0; <BR>for($i = 0; $i < strlen($str); $i++){ <BR>$value = ord($str[$i]); <BR>if($value > 127) { <BR>$count++; <BR>if($value >= 192 && $value <= 223) $i++; <BR>elseif($value >= 224 && $value <= 239) $i = $i + 2; <BR>elseif($value >= 240 && $value <= 247) $i = $i + 3; <BR>else die('Not a UTF-8 compatible string'); <BR>} <BR>$count++; <BR>} <BR>return $count; <BR>} <BR>//截取字符串-UTF8(PHP) <BR>function utf8_substr($str,$position,$length){ <BR>$start_position = strlen($str); <BR>$start_byte = 0; <BR>$end_position = strlen($str); <BR>$count = 0; <BR>for($i = 0; $i < strlen($str); $i++){ <BR>if($count >= $position && $start_position > $i){ <BR>$start_position = $i; <BR>$start_byte = $count; <BR>} <BR>if(($count-$start_byte)>=$length) { <BR>$end_position = $i; <BR>break; <BR>} <BR>$value = ord($str[$i]); <BR>if($value > 127){ <BR>$count++; <BR>if($value >= 192 && $value <= 223) $i++; <BR>elseif($value >= 224 && $value <= 239) $i = $i + 2; <BR>elseif($value >= 240 && $value <= 247) $i = $i + 3; <BR>else die('Not a UTF-8 compatible string'); <BR>} <BR>$count++; <BR>} <BR>return(substr($str,$start_position,$end_position-$start_position)); <BR>} <BR>//判断是否是有韩文-UTF-8 (JavaScript) <BR>function checkKoreaChar(str) { <BR>for(i=0; i<str.length; i++) { <BR>if(((str.charCodeAt(i) > 0x3130 && str.charCodeAt(i) = 0xAC00 && str.charCodeAt(i) <= 0xD7A3))) { <BR>return true; <BR>} <BR>} <BR>return false; <BR>} <BR>//判断是否有中文字符-GBK (JavaScript) <BR>function check_chinese_char(s){ <BR>return (s.length != s.replace(/[^\x00-\xff]/g,"**").length); <BR>} <BR>

喜欢 (0)赏

[搞代码]

分享 (0)