Mercurial > hg > ywww
comparison goodreads/UTFUtil.php @ 6:077b0a0a3e6d
remaining originals according to dependency walk
author | Robert Boland <robert@markup.co.uk> |
---|---|
date | Thu, 16 Feb 2017 22:29:02 +0000 |
parents | |
children |
comparison
equal
deleted
inserted
replaced
5:55445b456ad0 | 6:077b0a0a3e6d |
---|---|
1 <?PHP | |
2 function convert_utf8($str){ | |
3 if(!seems_utf8($str)) | |
4 return mb_convert_encoding($str, 'UTF-8'); | |
5 return $str; | |
6 } | |
7 | |
8 function seems_utf8($Str) { | |
9 for ($i=0; $i<strlen($Str); $i++) { | |
10 if (ord($Str[$i]) < 0x80) continue; # 0bbbbbbb | |
11 elseif ((ord($Str[$i]) & 0xE0) == 0xC0) $n=1; # 110bbbbb | |
12 elseif ((ord($Str[$i]) & 0xF0) == 0xE0) $n=2; # 1110bbbb | |
13 elseif ((ord($Str[$i]) & 0xF8) == 0xF0) $n=3; # 11110bbb | |
14 elseif ((ord($Str[$i]) & 0xFC) == 0xF8) $n=4; # 111110bb | |
15 elseif ((ord($Str[$i]) & 0xFE) == 0xFC) $n=5; # 1111110b | |
16 else return false; # Does not match any model | |
17 for ($j=0; $j<$n; $j++) { # n bytes matching 10bbbbbb follow ? | |
18 if ((++$i == strlen($Str)) || ((ord($Str[$i]) & 0xC0) != 0x80)) | |
19 return false; | |
20 } | |
21 } | |
22 return true; | |
23 } | |
24 ?> |