Hardcoding values like "\n" or 2147483647 in your PHP code is a subtle way to introduce platform-specific bugs. PHP’s predefined constants exist to save you from these headaches.

Why This Matters

Consider this seemingly innocent code:

file_put_contents('output.txt', "Line 1\nLine 2\nLine 3");

On Linux, this works perfectly. On Windows, it creates a file that confuses text editors because Windows expects \r\n line endings, not just \n. Your logs look fine on your development machine but break when deployed.

Or imagine checking if a number exceeds the maximum integer:

if ($value > 2147483647) {
    // Handle large number
}

This breaks on 64-bit systems where integers can be much larger. Your “safe” boundary check becomes meaningless.

The Essential Constants

Platform-Specific Constants

PHP_EOL — The correct line ending for the current platform

• \n on Unix/Linux/macOS
• \r\n on Windows
• Available since PHP 5.0.2

// ✅ Portable
$log = "Error occurred" . PHP_EOL . "Stack trace:" . PHP_EOL;

// ❌ Platform-specific
$log = "Error occurred\n" . "Stack trace:\n";

Integer Boundaries

PHP_INT_MAX — Largest possible integer

• 2,147,483,647 on 32-bit systems
• 9,223,372,036,854,775,807 on 64-bit systems
• Available since PHP 5.0.5

PHP_INT_MIN — Smallest possible integer (usually -PHP_INT_MAX - 1)

• Available since PHP 7.0.0

PHP_INT_SIZE — Integer size in bytes (4 or 8)

• Available since PHP 5.0.5

// Check if a value will overflow
if ($value > PHP_INT_MAX) {
    // Use arbitrary precision math instead
    $result = bcmul($value, $multiplier);
}

Floating Point Precision

PHP_FLOAT_EPSILON — The smallest difference between 1.0 and the next representable float

• Useful for comparing floats with tolerance
• Available since PHP 7.2.0

PHP_FLOAT_MIN / PHP_FLOAT_MAX — Smallest and largest representable floats

• Available since PHP 7.2.0

// ✅ Correct float comparison
if (abs($a - $b) < PHP_FLOAT_EPSILON) {
    // Numbers are equal within precision limits
}

// ❌ Unreliable
if ($a == $b) {
    // Can fail due to floating point rounding
}

When to Use These

Use PHP_EOL when:

• Writing to text files that humans will read
• Generating CSV files
• Creating log files
• Building email bodies

Use integer constants when:

• Validating numeric input ranges
• Deciding between int and string storage
• Implementing overflow-safe arithmetic
• Supporting both 32-bit and 64-bit environments

Use float constants when:

• Comparing decimal numbers
• Implementing scientific calculations
• Detecting precision loss

References

• Predefined Constants
• When do I use the PHP constant “PHP_EOL”