(PHP 5, PHP 7, PHP 8)
iconv_strlen — 返回字符串的字符数统计
$str
, string $charset = ini_get("iconv.internal_encoding")
) : int
和 strlen() 不同的是，iconv_strlen() 统计了给定的字节序列 str 中出现字符数的统计，基于指定的字符集，其产生的结果不一定和字符字节数相等。
返回 str 字符数的统计，是整型。
Notice there is a disconnect:
>If charset`parameter is omitted, str is assumed to be encoded in iconv.internal_encoding.
But clicking on the iconv.internal_encoding link (https://www.php.net/manual/en/iconv.configuration.php), the docs indicate that iconv.internal_encoding is deprecated since 5.6.
If iconv_strlen is passed a UTF-8 string containing badly formed sequences, it will return FALSE. This is in contrast to mb_strlen of the behaviour of utf8_decode, which strip out any bad sequences;
<?php
# UTF-8 string containing bad sequence: \xe9
$str = "I?t?rn?ti?n\xe9?liz?ti?n";
print "mb_strlen: ".mb_strlen($str,'UTF-8')."\n";
print "strlen/utf8_decode: ".strlen(utf8_decode($str))."\n";
print "iconv_strlen: ".iconv_strlen($str,'UTF-8')."\n";
?>
Displays;
mb_strlen: 20
strlen/utf8_decode: 20
iconv_strlen:
(PHP 5.0.5)
As such it is being "stricter" than mb_strlen and it may mean you need to check for invalid sequences first. A quick way to check is to exploit the behaviour of the PCRE extension (see notes on pattern modifiers);
<?php
if (preg_match('/^.{1}/us',$str,$ar) != 1) {
die("string contains invalid UTF-8");
}
?>
A slower but stricter check (regex) can be found at: http://www.w3.org/International/questions/qa-forms-utf-8
Similiar applies to iconv_substr, iconv_strpos and iconv_strrpos