[Solution] PHP ereg() Deprecated — Replace with preg_match() Migration

The ereg() function was removed from PHP because PCRE (Perl Compatible Regular Expressions) provides better performance, richer features, and broader industry support. Migrating to preg_match() is straightforward once you understand the key differences in syntax. The PCRE engine used by preg_match() is faster and more secure than the old POSIX regex functions.

What You’ll See

If your code still calls ereg(), eregi(), ereg_replace(), or eregi_replace() on PHP 7.0+, you will see:

Deprecated: Function ereg() is deprecated in /path/to/script.php on line X

On PHP 7.0+, the error becomes fatal:

Fatal error: Uncaught Error: Call to undefined function ereg()

Why Deprecated

PHP deprecated all POSIX regex functions (ereg, eregi, ereg_replace, eregi_replace, split) in PHP 5.3 and removed them entirely in PHP 7.0. The PCRE extension (preg_* functions) replaced them because:

• Performance: PCRE is significantly faster for complex patterns.
• Features: PCRE supports non-greedy matching, lookaheads, named groups, and other advanced syntax.
• Industry standard: PCRE is based on Perl’s regex engine and is used across many languages and platforms.
• Security: The POSIX regex engine in PHP had known vulnerabilities with certain crafted patterns.

Old Code (Deprecated)

// Basic match with ereg
$email = "user@example.com";
if (ereg("^[a-zA-Z0-9._]+@[a-zA-Z0-9._]+$", $email)) {
    echo "Valid email";
}

// Case-insensitive match with eregi
if (eregi("^hello", "Hello World")) {
    echo "Starts with hello";
}

// Replace with ereg_replace
$date = "2024-01-15";
$formatted = ereg_replace("([0-9]{4})-([0-9]{2})-([0-9]{2})", "\\3/\\2/\\1", $date);
echo $formatted; // 15/01/2024

// Sub-pattern capture with ereg
if (ereg("([a-z]+)@([a-z]+)", $email, $matches)) {
    echo "User: " . $matches[1] . ", Domain: " . $matches[2];
}

New Code (Replacement)

// Basic match with preg_match — note the delimiters and anchoring
$email = "user@example.com";
if (preg_match("/^[a-zA-Z0-9._]+@[a-zA-Z0-9._]+$/", $email)) {
    echo "Valid email";
}

// Case-insensitive match with preg_match — use the 'i' modifier
if (preg_match("/^hello/i", "Hello World")) {
    echo "Starts with hello";
}

// Replace with preg_match_replace
$date = "2024-01-15";
$formatted = preg_replace("/(\d{4})-(\d{2})-(\d{2})/", "$3/$2/$1", $date);
echo $formatted; // 15/01/2024

// Sub-pattern capture with preg_match
if (preg_match("/([a-z]+)@([a-z]+)/", $email, $matches)) {
    echo "User: " . $matches[1] . ", Domain: " . $matches[2];
}

// Named capture groups (PCRE advantage)
if (preg_match("/(?P<user>[a-z]+)@(?P<domain>[a-z]+)/", $email, $matches)) {
    echo "User: " . $matches['user'] . ", Domain: " . $matches['domain'];
}

Key Syntax Differences

The most common mistake is forgetting to wrap the pattern in delimiters. Use /, ~, #, or any non-alphanumeric character that does not appear in your pattern.

Migration Steps

1. Find all POSIX regex calls in your codebase:

grep -rn "ereg\|eregi\|ereg_replace\|eregi_replace" --include="*.php" /path/to/project/

2. Wrap every pattern in delimiters (/pattern/). Escape any / characters inside the pattern.

3. Replace eregi() calls by adding the i modifier after the closing delimiter instead.

4. Replace back-references in ereg_replace(). Change \\1 to $1 in the replacement string.

5. Test thoroughly. PCRE and POSIX regex have subtle differences in character class behavior and anchoring.

6. Search for leftover split() calls since split() was also deprecated alongside ereg():

grep -rn "\bsplit\s*(" --include="*.php" /path/to/project/

For large codebases, consider using a scripted sed replacement or a tool like Rector to automate the conversion.