Invalid replacement due to not using correct regular syntax in pattern

M66 2025-05-27

In PHP, the mb_eregi_replace function is used to perform case-insensitive multi-byte regular expression replacement operations. It is part of the mbstring extension designed specifically for multibyte encoded strings (such as UTF-8). Compared with traditional eregi_replace , mb_eregi_replace is more suitable for handling multi-byte characters such as Chinese, Japanese, and Korean.

However, many developers encounter the problem of invalid replacement when using mb_eregi_replace , and the root cause is that the correct regular syntax is not used in the regular expression pattern. This article will explain why unstandard patterns can cause replacement failures from the regular expression syntax and usage details of mb_eregi_replace .

Introduction to mb_eregi_replace function

The function definition is as follows:

 string mb_eregi_replace ( string $pattern , string $replacement , string $string [, string $option = "msr" ] )

$pattern : A regular expression for matching, case-insensitive.
$replacement : Replaced string.
$string : A pending string.
$option : The modifier of the regular expression, defaults to "msr".

It uses mbstring's multi-byte regular engine internally, so regular syntax and PCRE ( preg_* ) are slightly different.

Reasons for the replacement invalid due to regular pattern syntax problems

PCRE syntax is not supported

mb_eregi_replace uses the mbregex regular engine, not PCRE. Many developers are accustomed to using PCRE syntax (such as using /pattern/i to ignore case), while the $pattern of mb_eregi_replace cannot contain PCRE-style separators / , and does not support some PCRE syntax features.

Error example:

 mb_eregi_replace('/abc/', 'xyz', $str);

The correct usage should be:

 mb_eregi_replace('abc', 'xyz', $str);

There is no need to add / delimiter and i modifiers, because mb_eregi_replace itself is case-insensitive.

Incorrect escape characters

In mbregex, some special characters and escape methods are different from PCRE. For example, using \d to represent numeric classes may not be supported, and [0-9] must be used.

Error example:

 mb_eregi_replace('\d+', 'number', $str);

Correct example:

 mb_eregi_replace('[0-9]+', 'number', $str);

Multi-byte character matching misuse

Regular expressions should adapt to the characteristics of multi-byte characters, and if the pattern is written too simplified or the single-byte assumption is used, it will cause the match to fail.

Actual examples

Suppose we have a paragraph:

 $str = "Hello mb_eregi_replace example with M66.net URL.";

Error writing (using PCRE separator)

 echo mb_eregi_replace('/m66\.net/', 'example.com', $str);

This will not be replaced successfully because / is not recognized as delimiter.

Correct writing

 echo mb_eregi_replace('m66\.net', 'example.com', $str);

summary

mb_eregi_replace does not support PCRE-style separators and modifiers.
Regular expressions should follow the mbregex syntax specification.
Incorrect pattern causes the function to fail to match the target string, and natural replacement is invalid.
When matching multibyte characters, you should pay more attention to using the correct character set range and escape method.

Understanding the regular syntax specification of mb_eregi_replace can allow you to better use the function and avoid replacing invalid pitfalls.

 <?php
// Correct sample code：Put the domain name in the stringm66.netReplace withexample.com

$str = "Visit URL：http://m66.net/path?query=1";

// Notice：mb_eregi_replaceNo regular separator required，Default case insensitive
$result = mb_eregi_replace('m66\.net', 'example.com', $str);

echo $result;  // Output：Visit URL：http://example.com/path?query=1
?>

mb_eregi_replace