Regex in Loop

Also Known As

TL;DR

Explanation

PHP compiles a regex pattern on each call to preg_match()/preg_replace() unless the JIT cache has it. Running the same pattern thousands of times per request wastes compilation overhead and can cause backtracking on complex patterns. The fix is to define the pattern as a constant or variable before the loop, then reference it inside. For simple membership tests on many strings, preg_grep() on the whole array is often faster than a per-element loop.

Common Misconception

Why It Matters

Common Mistakes

• Building regex patterns dynamically inside a loop — string concatenation plus compilation on every iteration.
• Using regex for simple string operations that strpos() or str_contains() handle faster.
• Not knowing that PHP caches compiled regexes in a PCRE cache — but the cache has a limited size and can be evicted.
• Applying complex regexes to unbounded user input without length limits — potential ReDoS.

Code Examples

// Pattern compiled on every iteration
foreach ($emails as $email) {
    if (preg_match('/^[\w.+-]+@[\w-]+\.[\w.]+$/', $email)) {
        $valid[] = $email;
    }
}

// PHP compiles and caches regex internally after first use in the same request,
// but pulling the constant pattern out makes intent clear and
// avoids accidental recompilation if the string is built dynamically
const EMAIL_PATTERN = '/^[\w.+-]+@[\w-]+\.[\w.]+$/';

$valid = array_filter($emails, fn($e) => preg_match(EMAIL_PATTERN, $e));

// For truly hot loops, pre-validate with filter_var (faster than regex)
$valid = array_filter($emails, fn($e) => filter_var($e, FILTER_VALIDATE_EMAIL));

Tags

🧠 Logic Engine

References

• https://www.php.net/manual/en/function.preg-match.php
• https://www.php.net/manual/en/function.preg-grep.php