strtobytes

strtobytes.php

<?php

/**
 * Parse a description of a number of bytes to an actual number of bytes.
 *
 * @author Craig Anderson <[email protected]>
 * @link https://gist.github.com/3759346
 */
function strtobytes($dataSize)
{
	$multipliers = array(
		"" => 1,
		"b" => 1,
		"byte" => 1,
		"k" => 1000,
		"kb" => 1000,
		"kilobyte" => 1000,
		"kib" => 1024,
		"kibibyte" => 1024,
		"m" => 1000 * 1000,
		"mb" => 1000 * 1000,
		"meg" => 1000 * 1000,
		"megabyte" => 1000 * 1000,
		"mib" => 1024 * 1024,
		"mibibyte" => 1024 * 1024,
		"g" => 1000 * 1000 * 1000,
		"gb" => 1000 * 1000 * 1000,
		"gig" => 1000 * 1000 * 1000,
		"gigabyte" => 1000 * 1000 * 1000,
		"gib" => 1024 * 1024 * 1024,
		"gibibyte" => 1024 * 1024 * 1024,
		"t" => 1000 * 1000 * 1000 * 1000,
		"tb" => 1000 * 1000 * 1000 * 1000,
		"terabyte" => 1000 * 1000 * 1000 * 1000,
		"tib" => 1024 * 1024 * 1024 * 1024,
		"tibibyte" => 1024 * 1024 * 1024 * 1024,
		"p" => 1000 * 1000 * 1000 * 1000 * 1000,
		"pb" => 1000 * 1000 * 1000 * 1000 * 1000,
		"petabyte" => 1000 * 1000 * 1000 * 1000 * 1000,
		"pib" => 1024 * 1024 * 1024 * 1024 * 1024,
		"pibibyte" => 1024 * 1024 * 1024 * 1024 * 1024,
	);
	
	$bytes = NAN;
	$dataSize = trim($dataSize);
	
	if(strlen($dataSize) < 1)
	{
		// ignore empty string
	}
	else if(preg_match("/^(\\-?[\\d\\.\\,]+)?(.*)?$/", $dataSize, $matches))
	{
		$num = $matches[1];
		$num = str_replace(",", "", $num); // remove thousands seperator (TODO: internationalise?)
		$unit = trim(strtolower($matches[2]));
		if(preg_match("/s$/", $unit) && $num == "")
		{
			// ignore plural units with no number
			// i.e. "kilobyte" makes sense (1000), but "kilobytes" doesn't.
		}
		else
		{
			if($num == "")
			{
				$num = 1;
			}
			$unit = rtrim($unit, "s"); // depluralise
			if(isset($multipliers[$unit]))
			{
				$bytes = round_out($num * $multipliers[$unit]);
			}
		}
	}
	return $bytes;
}

function round_out($num)
{
	$result = 0;
	if($num > 0)
	{
		$result = ceil($num);
	}
	else if($num < 0)
	{
		$result = floor($num);
	}
	return $result;
}

function num_decimals($num)
{
	$numDecimals = strlen(strstr($num, ".")) - 1;
	if($numDecimals < 0)
	{
		$numDecimals = 0;
	}
	return $numDecimals;
}

function test_strtobytes()
{
	function assert_strtobytes($input, $expected)
	{
		$fail = true;
		printf("strtobytes(\"%s\") = \"%s\"?\n", $input, $expected);
		$actual = strtobytes($input);
		if($actual == $expected || (is_nan($actual) && is_nan($expected)))
		{
			$fail = false;
		}
		else
		{
			printf("\nFAIL!\nstrtobytes(\"%s\") != \"%s\"\nstrtobytes(\"%s\") == \"%s\"\n", $input, $expected, $input, $actual);
			die();
		}
		return $fail;
	}
	
	$dataMeasurements = array(
		"b" => 1,
		"byte" => 1,
		"k" => 1000,
		"kb" => 1000,
		"kilobyte" => 1000,
		"kib" => 1024,
		"kibibyte" => 1024,
		"m" => 1000 * 1000,
		"mb" => 1000 * 1000,
		"meg" => 1000 * 1000,
		"megabyte" => 1000 * 1000,
		"mib" => 1024 * 1024,
		"mibibyte" => 1024 * 1024,
		"g" => 1000 * 1000 * 1000,
		"gb" => 1000 * 1000 * 1000,
		"gig" => 1000 * 1000 * 1000,
		"gigabyte" => 1000 * 1000 * 1000,
		"gib" => 1024 * 1024 * 1024,
		"gibibyte" => 1024 * 1024 * 1024,
		"t" => 1000 * 1000 * 1000 * 1000,
		"tb" => 1000 * 1000 * 1000 * 1000,
		"terabyte" => 1000 * 1000 * 1000 * 1000,
		"tib" => 1024 * 1024 * 1024 * 1024,
		"tibibyte" => 1024 * 1024 * 1024 * 1024,
		"p" => 1000 * 1000 * 1000 * 1000 * 1000,
		"pb" => 1000 * 1000 * 1000 * 1000 * 1000,
		"petabyte" => 1000 * 1000 * 1000 * 1000 * 1000,
		"pib" => 1024 * 1024 * 1024 * 1024 * 1024,
		"pibibyte" => 1024 * 1024 * 1024 * 1024 * 1024,
	);
	
	$nums = array(1, 0, 5, 12345, -1, -5, -12345, 0.1, 1.44, -0.1, ".1");
	
	foreach($nums as $num)
	{
		assert_strtobytes($num, round_out($num));
	}
	
	foreach($dataMeasurements as $measurement => $numBytes)
	{
		assert_strtobytes(sprintf("%s", $measurement), $numBytes);
		assert_strtobytes(sprintf("%s", strtoupper($measurement)), $numBytes);
		assert_strtobytes(sprintf("%ss", $measurement), NAN);
		assert_strtobytes(sprintf("%sS", strtoupper($measurement)), NAN);
		foreach($nums as $num)
		{
			assert_strtobytes(sprintf("%s%s", $num, $measurement), round_out($numBytes * $num));
			assert_strtobytes(sprintf("%s%s", $num, strtoupper($measurement)), round_out($numBytes * $num));
			assert_strtobytes(sprintf("%s %s", $num, $measurement), round_out($numBytes * $num));
			assert_strtobytes(sprintf("%s %s", $num, strtoupper($measurement)), round_out($numBytes * $num));
			assert_strtobytes(sprintf("%s%ss", $num, $measurement), round_out($numBytes * $num));
			assert_strtobytes(sprintf("%s%sS", $num, strtoupper($measurement)), round_out($numBytes * $num));
			assert_strtobytes(sprintf("%s %ss", $num, $measurement), round_out($numBytes * $num));
			assert_strtobytes(sprintf("%s %sS", $num, strtoupper($measurement)), round_out($numBytes * $num));
			assert_strtobytes(sprintf(" %s %s ", $num, $measurement), round_out($numBytes * $num));
			assert_strtobytes(sprintf("%s %s", number_format($num, num_decimals($num)), $measurement), round_out($numBytes * $num));
		}
	}
	
	assert_strtobytes("", NAN);
	assert_strtobytes(" ", NAN);
	assert_strtobytes("1 thing", NAN);
	assert_strtobytes("1 x", NAN);
}