当字符串包含扩展 ASCII 代码时，为什么 `mb_eregi_replace` return NULL？

Question

使用PHP v7.1.5，为什么PHP功能mb_eregi_replace return NULL if the string contains extended ASCII code A0 ("Non-breaking space" according to www.ascii-code.com)？

$ php -a
Interactive shell

php > $t = mb_eregi_replace('d', '', "do\xA0not");
php > echo $t;
php > if( $t === null ) { echo "is null"; } else { echo "replace worked"; }
is null


php > $t = mb_eregi_replace('d', '', "do not");
php > echo $t;
o not
php > if( $t === null ) { echo "is null"; } else { echo "replace worked"; }
replace worked

Answer 1

在将它传递给 mb_eregi_replace 之前阅读 this article (specifically the "What to Do to Convert Extended ASCII to UTF-8" section) shows that passing the string to utf8_encode 解决了问题。

$ php -a
Interactive shell

php > $t = mb_eregi_replace('d', '', utf8_encode("do\xA0not"));
php > echo $t;
o not

当字符串包含扩展 ASCII 代码时，为什么 `mb_eregi_replace` return NULL？

Why does `mb_eregi_replace` return NULL when the string contains extended ASCII codes?

php

ascii

multibyte