IntMath::identify() - Code Metrics - Inspection of "Improve readability of assertInteger() method" - marcospassos/phpcommon-intmath - Measure and Improve Code Quality continuously with Scrutinizer

Completed

Push — master ( e80728...a678e6 )

by Marcos

created 2016-07-04 01:01 UTC

IntMath::identify() B

↳ Parent: Project

Complexity

Conditions	6
Paths	5

Size

Total Lines	16
Code Lines	8

Duplication

Lines	0
Ratio	0 %

Code Coverage

Tests	3
CRAP Score	6

Importance

Changes	1
Bugs	0	Features	0

Metric	Value
c	1
b	0
f	0
dl	0
loc	16
ccs	3
cts	3
cp	1
rs	8.8571
cc	6
eloc	8
nc	5
nop	1
crap	6

<?php

/**
 * This file is part of the phpcommon/intmath package.
 *
 * (c) Marcos Passos <[email protected]>
 *
 * For the full copyright and license information, please view the LICENSE file
 * that was distributed with this source code.
 */

namespace PhpCommon\IntMath;

use InvalidArgumentException;

// Only available on PHP >= 7
if (!defined('PHP_INT_MIN')) {
    define('PHP_INT_MIN', -PHP_INT_MAX - 1);
}

/**
 * Utility class for arithmetic operations on integers that wraps around upon
 * overflow.
 *
 * **Important**
 *
 * This class is not intended for use as a way to properly perform arithmetic
 * operations on integers and should not be used in place of the native
 * arithmetic operators or any other library designed for such purpose.
 *
 * Unlike other languages that overflow large positive integers into large
 * negative integers, PHP actually overflows integers to floating-point
 * numbers. In most cases, arithmetic overflows must be treated as an unusual
 * circumstance which requires special handling. However, there are some cases
 * where such _wrap-around_ behavior is actually useful - for example with TCP
 * sequence numbers or certain algorithms, such as hash calculation. This
 * utility class provides basic arithmetic functions that operate in accordance
 * to that behaviour.
 *
 * To illustrate, consider the following example:
 *
 * ```php
 * // Output on 64-bit system: float(9.2233720368548E+18)
 * var_dump(PHP_MAX_INT + 1);
 *
 * // Output on 64-bit system: int(-9223372036854775808)
 * var_dump(IntMath::add(PHP_MAX_INT, 1));
 * ```
 *
 * As previously shown, adding one to the largest supported integer using
 * native arithmetic operators will result in a floating-point number. By
 * contrast, using {@link IntMath::add()} will cause an overflow, resulting
 * in the smallest integer supported in this build of PHP.
 *
 * @author Marcos Passos <[email protected]>
 */
class IntMath
{
    /**
     * The largest supported integer
     */
    const MAX_INT = PHP_INT_MAX;

    /**
     * The smallest supported integer
     */
    const MIN_INT = PHP_INT_MIN;

    /**
     * Returns the negation of the argument.
     *
     * For integer values, negation is the same as subtraction from zero.
     * Because PHP uses two's-complement representation for integers and the
     * range of two's-complement values is not symmetric, the negation of the
     * maximum negative integer results in that same maximum negative number.
     * Despite the fact that overflow has occurred, no exception is thrown.
     *
     * For all integer values `$a`, `-$a` equals `(~$a) + 1`.
     *
     * @param integer $a The value.
     *
     * @return integer The result.
     *
     * @throws InvalidArgumentException If the argument is not an integer.
     */
    public static function negate($a)
    {
        self::assertInteger($a);

        if ($a === self::MIN_INT) {
            return $a;
        }

        return -$a;
    }

    /**
     * Returns the sum of the arguments.
     *
     * The result is the low-order bits of the true mathematical result as
     * represented in a sufficiently wide two's-complement format. If overflow
     * occurs, then the sign of the result may not be the same as the sign of
     * the mathematical sum of the two values. Despite the overflow, no
     * exception is thrown in this case.
     *
     * @param integer $a The addend.
     * @param integer $b The addend.
     *
     * @return integer The sum.
     *
     * @throws InvalidArgumentException If one of the arguments is not an
     *                                  integer.
     */
    public static function add($a, $b)
    {
        self::assertInteger($a);
        self::assertInteger($b);

        if (($b > 0) && ($a <= (PHP_INT_MAX - $b))) {
            return $a + $b;
        }

        if (($b < 0) && ($a >= (PHP_INT_MIN - $b))) {
            return $a + $b;
        }

        while ($b !== 0) {
            // Carry now contains common set bits of the addends
            $carry = $a & $b;
            // Sum of bits of $x and $y,
            // where at least one of the bits is not set
            $a ^= $b;
            // Left-shift by one
            $b = $carry << 1;
        }

        return $a;
    }

    /**
     * Returns the difference of the arguments.
     *
     * The subtraction of a positive number yields the same result as the
     * addition of a negative number of equal magnitude. Furthermore, the
     * subtraction from zero is the same as negation.
     *
     * The result is the low-order bits of the true mathematical result as
     * represented in a sufficiently wide two's-complement format. If overflow
     * occurs, then the sign of the result may not be the same as the sign of
     * the mathematical difference of the two values. Despite the overflow, no
     * exception is thrown in this case.
     *
     * @param integer $a The minuend.
     * @param integer $b The subtrahend.
     *
     * @return integer The difference.
     *
     * @throws InvalidArgumentException If one of the arguments is not an
     *                                  integer.
     */
    public static function subtract($a, $b)
    {
        self::assertInteger($a);
        self::assertInteger($b);

        return self::add((int) $a, self::negate($b));
    }

    /**
     * Returns the product of the arguments.
     *
     * The result is the low-order bits of the true mathematical result as
     * represented in a sufficiently wide two's-complement format. If overflow
     * occurs, then the sign of the result may not be the same as the sign of
     * the mathematical product of the two values. Despite the overflow, no
     * exception is thrown in this case.
     *
     * @param integer $a The multiplicand.
     * @param integer $b The multiplier.
     *
     * @return integer The product.
     *
     * @throws InvalidArgumentException If one of the arguments is not an
     *                                  integer.
     */
    public static function multiply($a, $b)
    {
        self::assertInteger($a);
        self::assertInteger($b);

        if ($a === 0 || $b === 0) {
            return 0;
        }

        // If the multiplicand or the multiplier are the smallest integer
        // supported, then the product is `0` or the smallest integer supported,
        // according as the other operand is odd or even respectively.
        if ($a === self::MIN_INT) {
            return $b & 0x01 ? $a : 0;
        }

        if ($b === self::MIN_INT) {
            return $a & 0x01 ? $b : 0;
        }

        $max = self::MIN_INT;

        // Same sign
        if ($a >= 0 && $b >= 0 || $a < 0 && $b < 0) {
            $max = self::MAX_INT;
        }

        // Use native operators unless the operation causes an overflow
        if (($b > 0 && $b <= ($max / $a)) || ($b < 0 && $b >= ($max / $a))) {
            return $a * $b;
        }

        // Signed multiplication can be accomplished by doing an unsigned
        // multiplication and taking manually care of the negative-signs.
        $sign = false;

        if ($a < 0) {
            // Toggle the signed flag
            $sign = !$sign;
            // Negate $a
            $a = self::negate($a);
        }

        if ($b < 0) {
            // Toggle the signed flag
            $sign = !$sign;
            // Negate $b
            $b = self::negate($b);
        }

        $product = 0;
        // Both operands are now positive, perform unsigned multiplication
        while ($a !== 0) {
            // Test the least significant bit (LSB) of multiplier
            if (($a & 0x01) !== 0) {
                // If 1, add the multiplicand to the product
                $product = self::add($product, $b);
            }

            // Left-shift by one, or divide by 2
            $a >>= 1;

            // Right-shift by one, or multiply by 2
            $b <<= 1;
        }

        if ($sign) {
            // Negate the product
            $product = self::negate($product);
        }

        return $product;
    }

    /**
     * Returns the quotient of the arguments.
     *
     * The division rounds the result towards zero. Thus the absolute value of
     * the result is the largest possible integer that is less than or equal to
     * the absolute value of the quotient of the two operands. The result is
     * zero or positive when the two operands have the same sign and zero or
     * negative when the two operands have opposite signs.
     *
     * There is one special case that does not satisfy this rule: if the
     * dividend is the negative integer of largest possible magnitude for its
     * type, and the divisor is `-1`, then integer overflow occurs and the
     * result is equal to the dividend. Despite the overflow, no exception is
     * thrown in this case. On the other hand if the value of the divisor in an
     * integer division is `0`, then a `DivisionByZeroException` is thrown.
     *
     * @param integer $a The dividend.
     * @param integer $b The divisor.
     *
     * @return integer The quotient.
     *
     * @throws InvalidArgumentException If one of the arguments is not an
     *                                  integer.
     * @throws DivisionByZeroException  If the divisor is zero.
     */
    public static function divide($a, $b)
    {
        self::assertInteger($a);
        self::assertInteger($b);

        if (0 === $b) {
            throw new DivisionByZeroException('Division by zero.');
        }

        if ($a === self::MIN_INT && $b === -1) {
            return $a;
        }

        return ($a - ($a % $b)) / $b;
    }

    /**
     * Asserts the specified value is an integer.
     *
     * @param mixed $value The value to assert.
     *
     * @throws InvalidArgumentException If the value is not an integer.
     */
    private static function assertInteger($value)
    {
        if (is_int($value)) {
            return;
        }

        throw new InvalidArgumentException(sprintf(
            'Expected an integer, but got "%s"',
            self::identify($value)
        ));
    }

    /**
     * Returns a string that identifies the specified value.
     *
     * @param mixed $value The value to identify.
     *
     * @return string A string that identifies the specified value.
     */
    private static function identify($value)
    {
        if (!is_numeric($value) || !is_float($value)) {
            return gettype($value);
        }

        if (is_infinite($value)) {
            return $value < 0 ? '-INF' : 'INF';
        }

        if (is_nan($value)) {
            return 'NAN';
        }

        return 'float';
    }
}


1		<?php
2
3		/**
4		* This file is part of the phpcommon/intmath package.
5		*
6		* (c) Marcos Passos <[email protected]>
7		*
8		* For the full copyright and license information, please view the LICENSE file
9		* that was distributed with this source code.
10		*/
11
12		namespace PhpCommon\IntMath;
13
14		use InvalidArgumentException;
15
16		// Only available on PHP >= 7
17	2	if (!defined('PHP_INT_MIN')) {
18	2	define('PHP_INT_MIN', -PHP_INT_MAX - 1);
19	2	}
20
21		/**
22		* Utility class for arithmetic operations on integers that wraps around upon
23		* overflow.
24		*
25		* Important
26		*
27		* This class is not intended for use as a way to properly perform arithmetic
28		* operations on integers and should not be used in place of the native
29		* arithmetic operators or any other library designed for such purpose.
30		*
31		* Unlike other languages that overflow large positive integers into large
32		* negative integers, PHP actually overflows integers to floating-point
33		* numbers. In most cases, arithmetic overflows must be treated as an unusual
34		* circumstance which requires special handling. However, there are some cases
35		* where such _wrap-around_ behavior is actually useful - for example with TCP
36		* sequence numbers or certain algorithms, such as hash calculation. This
37		* utility class provides basic arithmetic functions that operate in accordance
38		* to that behaviour.
39		*
40		* To illustrate, consider the following example:
41		*
42		* ```php
43		* // Output on 64-bit system: float(9.2233720368548E+18)
44		* var_dump(PHP_MAX_INT + 1);
45		*
46		* // Output on 64-bit system: int(-9223372036854775808)
47		* var_dump(IntMath::add(PHP_MAX_INT, 1));
48		* ```
49		*
50		* As previously shown, adding one to the largest supported integer using
51		* native arithmetic operators will result in a floating-point number. By
52		* contrast, using {@link IntMath::add()} will cause an overflow, resulting
53		* in the smallest integer supported in this build of PHP.
54		*
55		* @author Marcos Passos <[email protected]>
56		*/
57		class IntMath
58		{
59		/**
60		* The largest supported integer
61		*/
62		const MAX_INT = PHP_INT_MAX;
63
64		/**
65		* The smallest supported integer
66		*/
67		const MIN_INT = PHP_INT_MIN;
68
69		/**
70		* Returns the negation of the argument.
71		*
72		* For integer values, negation is the same as subtraction from zero.
73		* Because PHP uses two's-complement representation for integers and the
74		* range of two's-complement values is not symmetric, the negation of the
75		* maximum negative integer results in that same maximum negative number.
76		* Despite the fact that overflow has occurred, no exception is thrown.
77		*
78		* For all integer values `$a`, `-$a` equals `(~$a) + 1`.
79		*
80		* @param integer $a The value.
81		*
82		* @return integer The result.
83		*
84		* @throws InvalidArgumentException If the argument is not an integer.
85		*/
86	26	public static function negate($a)
87		{
88	26	self::assertInteger($a);
89
90	4	if ($a === self::MIN_INT) {
91	2	return $a;
92		}
93
94	2	return -$a;
95		}
96
97		/**
98		* Returns the sum of the arguments.
99		*
100		* The result is the low-order bits of the true mathematical result as
101		* represented in a sufficiently wide two's-complement format. If overflow
102		* occurs, then the sign of the result may not be the same as the sign of
103		* the mathematical sum of the two values. Despite the overflow, no
104		* exception is thrown in this case.
105		*
106		* @param integer $a The addend.
107		* @param integer $b The addend.
108		*
109		* @return integer The sum.
110		*
111		* @throws InvalidArgumentException If one of the arguments is not an
112		* integer.
113		*/
114	26	public static function add($a, $b)
115		{
116	26	self::assertInteger($a);
117	4	self::assertInteger($b);
118
119	4	if (($b > 0) && ($a <= (PHP_INT_MAX - $b))) {
120	2	return $a + $b;
121		}
122
123	4	if (($b < 0) && ($a >= (PHP_INT_MIN - $b))) {
124	2	return $a + $b;
125		}
126
127	4	while ($b !== 0) {
128		// Carry now contains common set bits of the addends
129	2	$carry = $a & $b;
130		// Sum of bits of $x and $y,
131		// where at least one of the bits is not set
132	2	$a ^= $b;
133		// Left-shift by one
134	2	$b = $carry << 1;
135	2	}
136
137	4	return $a;
138		}
139
140		/**
141		* Returns the difference of the arguments.
142		*
143		* The subtraction of a positive number yields the same result as the
144		* addition of a negative number of equal magnitude. Furthermore, the
145		* subtraction from zero is the same as negation.
146		*
147		* The result is the low-order bits of the true mathematical result as
148		* represented in a sufficiently wide two's-complement format. If overflow
149		* occurs, then the sign of the result may not be the same as the sign of
150		* the mathematical difference of the two values. Despite the overflow, no
151		* exception is thrown in this case.
152		*
153		* @param integer $a The minuend.
154		* @param integer $b The subtrahend.
155		*
156		* @return integer The difference.
157		*
158		* @throws InvalidArgumentException If one of the arguments is not an
159		* integer.
160		*/
161	26	public static function subtract($a, $b)
162		{
163	26	self::assertInteger($a);
164	4	self::assertInteger($b);
165
166	4	return self::add((int) $a, self::negate($b));
167		}
168
169		/**
170		* Returns the product of the arguments.
171		*
172		* The result is the low-order bits of the true mathematical result as
173		* represented in a sufficiently wide two's-complement format. If overflow
174		* occurs, then the sign of the result may not be the same as the sign of
175		* the mathematical product of the two values. Despite the overflow, no
176		* exception is thrown in this case.
177		*
178		* @param integer $a The multiplicand.
179		* @param integer $b The multiplier.
180		*
181		* @return integer The product.
182		*
183		* @throws InvalidArgumentException If one of the arguments is not an
184		* integer.
185		*/
186	26	public static function multiply($a, $b)
187		{
188	26	self::assertInteger($a);
189	4	self::assertInteger($b);
190
191	4	if ($a === 0 \|\| $b === 0) {
192	2	return 0;
193		}
194
195		// If the multiplicand or the multiplier are the smallest integer
196		// supported, then the product is `0` or the smallest integer supported,
197		// according as the other operand is odd or even respectively.
198	4	if ($a === self::MIN_INT) {
199	2	return $b & 0x01 ? $a : 0;
200		}
201
202	4	if ($b === self::MIN_INT) {
203	4	return $a & 0x01 ? $b : 0;
204		}
205
206	4	$max = self::MIN_INT;
207
208		// Same sign
209	4	if ($a >= 0 && $b >= 0 \|\| $a < 0 && $b < 0) {
210	4	$max = self::MAX_INT;
211	4	}
212
213		// Use native operators unless the operation causes an overflow
214	4	if (($b > 0 && $b <= ($max / $a)) \|\| ($b < 0 && $b >= ($max / $a))) {
215	2	return $a * $b;
216		}
217
218		// Signed multiplication can be accomplished by doing an unsigned
219		// multiplication and taking manually care of the negative-signs.
220	2	$sign = false;
221
222	2	if ($a < 0) {
223		// Toggle the signed flag
224	2	$sign = !$sign;
225		// Negate $a
226	2	$a = self::negate($a);
227	2	}
228
229	2	if ($b < 0) {
230		// Toggle the signed flag
231	2	$sign = !$sign;
232		// Negate $b
233	2	$b = self::negate($b);
234	2	}
235
236	2	$product = 0;
237		// Both operands are now positive, perform unsigned multiplication
238	2	while ($a !== 0) {
239		// Test the least significant bit (LSB) of multiplier
240	2	if (($a & 0x01) !== 0) {
241		// If 1, add the multiplicand to the product
242	2	$product = self::add($product, $b);
243	2	}
244
245		// Left-shift by one, or divide by 2
246	2	$a >>= 1;
247
248		// Right-shift by one, or multiply by 2
249	2	$b <<= 1;
250	2	}
251
252	2	if ($sign) {
253		// Negate the product
254	2	$product = self::negate($product);
255	2	}
256
257	2	return $product;
258		}
259
260		/**
261		* Returns the quotient of the arguments.
262		*
263		* The division rounds the result towards zero. Thus the absolute value of
264		* the result is the largest possible integer that is less than or equal to
265		* the absolute value of the quotient of the two operands. The result is
266		* zero or positive when the two operands have the same sign and zero or
267		* negative when the two operands have opposite signs.
268		*
269		* There is one special case that does not satisfy this rule: if the
270		* dividend is the negative integer of largest possible magnitude for its
271		* type, and the divisor is `-1`, then integer overflow occurs and the
272		* result is equal to the dividend. Despite the overflow, no exception is
273		* thrown in this case. On the other hand if the value of the divisor in an
274		* integer division is `0`, then a `DivisionByZeroException` is thrown.
275		*
276		* @param integer $a The dividend.
277		* @param integer $b The divisor.
278		*
279		* @return integer The quotient.
280		*
281		* @throws InvalidArgumentException If one of the arguments is not an
282		* integer.
283		* @throws DivisionByZeroException If the divisor is zero.
284		*/
285	30	public static function divide($a, $b)
286		{
287	30	self::assertInteger($a);
288	8	self::assertInteger($b);
289
290	8	if (0 === $b) {
291	2	throw new DivisionByZeroException('Division by zero.');
292		}
293
294	6	if ($a === self::MIN_INT && $b === -1) {
295	2	return $a;
296		}
297
298	4	return ($a - ($a % $b)) / $b;
299		}
300
301		/**
302		* Asserts the specified value is an integer.
303		*
304		* @param mixed $value The value to assert.
305		*
306		* @throws InvalidArgumentException If the value is not an integer.
307		*/
308	112	private static function assertInteger($value)
309		{
310	112	if (is_int($value)) {
311	2	return;
312		}
313
314	110	throw new InvalidArgumentException(sprintf(
315	110	'Expected an integer, but got "%s"',
316	20	self::identify($value)
317	20	));
318		}
319	90
320	10	/**
321	10	* Returns a string that identifies the specified value.
322		*
323	80	* @param mixed $value The value to identify.
324	80	*
325	80	* @return string A string that identifies the specified value.
326		*/
327	110	private static function identify($value)
328	110	{
329		if (!is_numeric($value) \|\| !is_float($value)) {
330	110	return gettype($value);
331		}
332
333		if (is_infinite($value)) {
334		return $value < 0 ? '-INF' : 'INF';
335		}
336
337		if (is_nan($value)) {
338		return 'NAN';
339		}
340
341		return 'float';
342		}
343		}
344

marcospassos / phpcommon-intmath

Push — master ( e80728...a678e6 )

IntMath::identify() B

Complexity

Size

Duplication

Code Coverage

Importance

Duplication Side-by-Side

Filter issues like