Function to intercept strings according to utf8 encoding rules (utf8 version of sub_str)

/*
* Function: The function is the same as substr, except that it will not cause garbled characters
* Parameters:
* Return:
*/
function utf8_substr( $str , $start , $length =null ){
// First intercept normally.
$res = substr( $str , $start , $length );
$strlen = strlen( $str );
/* Then determine whether the first and last 6 bytes are Complete (not incomplete) */
// If the parameter start is a positive number
if ( $start >= 0 ){
// intercept about 6 bytes forward
$next_start = $start + $length; // Initial Position
$next_len = $next_start + 6 <= $strlen ? 6 : $strlen - $next_start;
$next_segm = substr( $str , $next_start , $next_len );
// If the first byte is not complete The first byte of the character, and then intercept about 6 bytes
$prev_start = $start - 6 > 0 ? $start - 6 : 0;
$prev_segm = substr( $str , $prev_start , $start - $prev_start );
}
// start is a negative number
else{
// intercept about 6 bytes forward
$next_start = $strlen + $start + $length; // initial position
$next_len = $next_start + 6 < ;= $strlen ? 6 : $strlen - $next_start;
$next_segm = substr( $str , $next_start , $next_len );
// If the first byte is not the first byte of the complete character, intercept it later About 6 bytes.
$start = $strlen + $start;
$prev_start = $start - 6 > 0 ? $start - 6 : 0;
$prev_segm = substr( $str , $prev_start , $start - $ prev_start );
}
// Determine whether the first 6 bytes comply with utf8 rules
if ( preg_match( '@^([x80-xBF]{0,5})[xC0-xFD]?@' , $next_segm , $ bytes ) ){
if ( !empty( $bytes[1] ) ){
$bytes = $bytes[1];
$res .= $bytes;
}
}
// Determine whether the last 6 bytes match utf8 rules
$ord0 = ord( $res[0] );
if ( 128 <= $ord0 && 191 >= $ord0 ){
// Intercept from the back and add it in front of res.
if ( preg_match( '@[xC0-xFD][x80-xBF]{0,5}$@' , $prev_segm , $bytes ) ){
if ( !empty( $bytes[0] ) ){
$bytes = $ bytes[0];
$res = $bytes . $res;
}
}
}
return $res;
}
?>

Test ---

$str = 'dfjdjf test 13f test 65&2 datafｄdｊ（1 on mfe&...on';
var_dump( utf8_substr( $str , 22 , 12 ) ); echo '  ';
var_dump( utf8_substr( $str , 22 , -6 ) ); echo '  ';
var_dump( utf8_substr( $str , 9 , 12 ) ); echo '  ';
var_dump( utf8_substr( $str , 19 , 12 ) ); echo '  ';
var_dump( utf8_substr( $str , 28 , -6 ) ); echo ' ?>

Display results: (interception without garbled characters) string(12) "According to fdj" string(26) "According to fdj (1 is mfe&..." string(13) "13f try 65&2 number" string(12) "data fd" string(20) "dｊ（1justmfe&..."