Decoder::__construct() - Code Metrics - Inspection of "Inlined integer overflow check" - s9e/Bencode - Measure and Improve Code Quality continuously with Scrutinizer

Passed

Push — master ( fde73f...e709b3 )

by Josh

created 2021-10-07 18:02 UTC

Decoder::__construct() A

↳ Parent: Decoder

Complexity

Conditions	1
Paths	1

Size

Total Lines	7
Code Lines	4

Duplication

Lines	0
Ratio	0 %

Code Coverage

Tests	5
CRAP Score	1

Importance

Changes	4
Bugs	0	Features	0

Metric	Value
cc	1
eloc	4
nc	1
nop	1
dl	0
loc	7
ccs	5
cts	5
cp	1
crap	1
rs	10
c	4
b	0
f	0

<?php declare(strict_types=1);

/**
* @package   s9e\Bencode
* @copyright Copyright (c) 2014-2021 The s9e authors
* @license   http://www.opensource.org/licenses/mit-license.php The MIT License
*/
namespace s9e\Bencode;

use ArrayObject;
use Throwable;
use TypeError;
use const PHP_INT_MAX, PHP_INT_MIN, false;
use function is_float, str_contains, strcmp, strlen, strspn, substr, substr_compare;
use s9e\Bencode\Exceptions\ComplianceError;
use s9e\Bencode\Exceptions\DecodingException;

class Decoder
{
	/**
	* @var string Bencoded string being decoded
	*/
	protected string $bencoded;

	/**
	* @var int Length of the bencoded string
	*/
	protected int $len;

	/**
	* @var int Safe rightmost boundary
	*/
	protected int $max;

	/**
	* @var int Position of the cursor while decoding
	*/
	protected int $offset = 0;

	public static function decode(string $bencoded): ArrayObject|array|int|string
	{
		$decoder = new static($bencoded);
		try
		{
			$value = $decoder->decodeAnything();
		}
		catch (TypeError $e)
		{
			throw self::convertTypeError($e, $decoder->offset);
		}

		$decoder->checkCursorPosition();

		return $value;
	}

	protected function __construct(string $bencoded)
	{
		$this->bencoded = $bencoded;
		$this->len      = strlen($bencoded);

		$this->computeSafeBoundary();
		$this->checkBoundary();
	}

	/**
	* Cast given string as an integer and check for clamping
	*/
	protected function castInteger(string $string, bool $negative): int
	{
		$clamp = PHP_INT_MAX;
		if ($negative)
		{
			$string = "-$string";
			$clamp  = PHP_INT_MIN;
		}

		$value = (int) $string;
		if ($value === $clamp && is_float(+$string))
		{
			throw new DecodingException('Integer overflow', $this->offset - 1 - strlen($string));
		}

		return $value;
	}

	protected function checkBoundary(): void
	{
		if ($this->max < 1)
		{
			throw match (substr($this->bencoded, 0, 1))
			{
				'-', 'e' => new DecodingException('Illegal character',     0),
				''       => new DecodingException('Premature end of data', 0),
				default  => new DecodingException('Premature end of data', $this->len - 1)
			};
		}
	}

	/**
	* Check the cursor's position after decoding is done
	*/
	protected function checkCursorPosition(): void
	{
		if ($this->offset !== $this->len)
		{
			if ($this->offset > $this->len)
			{
				throw new DecodingException('Premature end of data', $this->len - 1);
			}

			$this->complianceError('Superfluous content', $this->offset);
		}
	}

	protected function complianceError(string $message, int $offset): void
	{
		throw new ComplianceError($message, $offset);
	}

	/**
	* Adjust the rightmost boundary to the last safe character that can start a value
	*
	* Will rewind the boundary to skip the rightmost digits, optionally preceded by "i" or "i-"
	*/
	protected function computeSafeBoundary(): void
	{
		$boundary = $this->len - 1;
		do
		{
			$c = substr($this->bencoded, $boundary, 1);
		}
		while (str_contains('0123456789', $c) && --$boundary >= 0);

		$this->max = match ($c)
		{
			'-'     => $boundary - 2,
			'i'     => $boundary - 1,
			default => $boundary
		};
	}

	protected static function convertTypeError(TypeError $e, int $offset): Throwable
	{
		// A type error can occur in decodeString() if the string length exceeds an int
		$frame = $e->getTrace()[0];
		if ($frame['class'] === __CLASS__ && $frame['function'] === 'decodeString')
		{
			return new DecodingException('String length overflow', $offset - 1);
		}

		// Return any other error as-is
		return $e;
	}

	protected function decodeAnything(): ArrayObject|array|int|string
	{
		return match ($this->bencoded[$this->offset])
		{
			'i'     => $this->decodeInteger(),
			'd'     => $this->decodeDictionary(),
			'l'     => $this->decodeList(),
			default => $this->decodeString()
		};
	}

	protected function decodeDictionary(): ArrayObject
	{
		$values  = [];
		$lastKey = null;

		++$this->offset;
		while ($this->offset <= $this->max)
		{
			$c = $this->bencoded[$this->offset];
			if ($c === 'e')
			{
				++$this->offset;

				return new ArrayObject($values, ArrayObject::ARRAY_AS_PROPS);
			}

			// Quickly match the most common keys found in dictionaries
			$key = match ($c)
			{
				'4'     => $this->decodeFastString('4:path',   6, 'path'),
				'6'     => $this->decodeFastString('6:length', 8, 'length'),
				default => $this->decodeString()
			};
			if (isset($lastKey) && strcmp($key, $lastKey) <= 0)
			{
				$this->dictionaryComplianceError($key, $lastKey);
			}
			if ($this->offset > $this->max)
			{
				break;
			}
			$values[$key] = $this->decodeAnything();
			$lastKey      = $key;
		}

		throw new DecodingException('Premature end of data', $this->len - 1);
	}

	/**
	* @param string $match Bencoded string to match
	* @param int    $len   Length of the bencoded string
	* @param string $value String value to return if the string matches
	*/
	protected function decodeFastString(string $match, int $len, string $value): string
	{
		if (substr_compare($this->bencoded, $match, $this->offset, $len, false) === 0)
		{
			$this->offset += $len;

			return $value;
		}

		return $this->decodeString();
	}

	protected function decodeInteger(): int
	{
		$negative = ($this->bencoded[++$this->offset] === '-');
		if ($negative && $this->bencoded[++$this->offset] === '0')
		{
			$this->complianceError('Illegal character', $this->offset);
		}

		return $this->castInteger($this->readDigits('e'), $negative);
	}

	protected function decodeList(): array
	{
		++$this->offset;

		$list = [];
		while ($this->offset <= $this->max)
		{
			if ($this->bencoded[$this->offset] === 'e')
			{
				++$this->offset;

				return $list;
			}

			$list[] = $this->decodeAnything();
		}

		throw new DecodingException('Premature end of data', $this->len - 1);
	}

	protected function decodeString(): string
	{
		$len = (int) $this->readDigits(':');
		$string = substr($this->bencoded, $this->offset, $len);
		$this->offset += $len;

		return $string;
	}

	protected function dictionaryComplianceError(string $key, string $lastKey): void
	{
		// Compute the offset of the start of the string used as key
		$offset = $this->offset - strlen(strlen($key) . ':') - strlen($key);

		$msg = ($key === $lastKey) ? 'Duplicate' : 'Out of order';
		$this->complianceError($msg . " dictionary entry '" . $key . "'", $offset);
	}

	protected function readDigits(string $terminator): string
	{
		// Digits sorted by decreasing frequency as observed on a random sample of torrent files
		$spn = strspn($this->bencoded, '1463720859', $this->offset);
		if (!$spn)
		{
			throw new DecodingException('Illegal character', $this->offset);
		}
		if ($this->bencoded[$this->offset] === '0' && $spn !== 1)
		{
			$this->complianceError('Illegal character', 1 + $this->offset);
		}

		$string = substr($this->bencoded, $this->offset, $spn);

		$this->offset += $spn;
		if ($this->bencoded[$this->offset] !== $terminator)
		{
			throw new DecodingException('Illegal character', $this->offset);
		}
		++$this->offset;

		return $string;
	}
}

1		<?php declare(strict_types=1);
2
3		/**
4		* @package s9e\Bencode
5		* @copyright Copyright (c) 2014-2021 The s9e authors
6		* @license http://www.opensource.org/licenses/mit-license.php The MIT License
7		*/
8		namespace s9e\Bencode;
9
10		use ArrayObject;
11		use Throwable;
12		use TypeError;
13		use const PHP_INT_MAX, PHP_INT_MIN, false;
14		use function is_float, str_contains, strcmp, strlen, strspn, substr, substr_compare;
15		use s9e\Bencode\Exceptions\ComplianceError;
16		use s9e\Bencode\Exceptions\DecodingException;
17
18		class Decoder
19		{
20		/**
21		* @var string Bencoded string being decoded
22		*/
23		protected string $bencoded;
24
25		/**
26		* @var int Length of the bencoded string
27		*/
28		protected int $len;
29
30		/**
31		* @var int Safe rightmost boundary
32		*/
33		protected int $max;
34
35		/**
36		* @var int Position of the cursor while decoding
37		*/
38		protected int $offset = 0;
39
40	84	public static function decode(string $bencoded): ArrayObject\|array\|int\|string
41		{
42	84	$decoder = new static($bencoded);
43		try
44		{
45	72	$value = $decoder->decodeAnything();
46		}
47	30	catch (TypeError $e)
48		{
49	2	throw self::convertTypeError($e, $decoder->offset);
50		}
51
52	42	$decoder->checkCursorPosition();
53
54	34	return $value;
55		}
56
57	84	protected function __construct(string $bencoded)
58		{
59	84	$this->bencoded = $bencoded;
60	84	$this->len = strlen($bencoded);
61
62	84	$this->computeSafeBoundary();
63	84	$this->checkBoundary();
64	72	}
65
66		/**
67		* Cast given string as an integer and check for clamping
68		*/
69	25	protected function castInteger(string $string, bool $negative): int
70		{
71	25	$clamp = PHP_INT_MAX;
72	25	if ($negative)
73		{
74	4	$string = "-$string";
75	4	$clamp = PHP_INT_MIN;
76		}
77
78	25	$value = (int) $string;
79	25	if ($value === $clamp && is_float(+$string))
80		{
81	2	throw new DecodingException('Integer overflow', $this->offset - 1 - strlen($string));
82		}
83
84	23	return $value;
85		}
86
87	84	protected function checkBoundary(): void
88		{
89	84	if ($this->max < 1)
90		{
91	12	throw match (substr($this->bencoded, 0, 1))
92		{
93	2	'-', 'e' => new DecodingException('Illegal character', 0),
94	1	'' => new DecodingException('Premature end of data', 0),
95	9	default => new DecodingException('Premature end of data', $this->len - 1)
96		};
97		}
98	72	}
99
100		/**
101		* Check the cursor's position after decoding is done
102		*/
103	42	protected function checkCursorPosition(): void
104		{
105	42	if ($this->offset !== $this->len)
106		{
107	11	if ($this->offset > $this->len)
108		{
109	3	throw new DecodingException('Premature end of data', $this->len - 1);
110		}
111
112	8	$this->complianceError('Superfluous content', $this->offset);
113		}
114	34	}
115
116	13	protected function complianceError(string $message, int $offset): void
117		{
118	13	throw new ComplianceError($message, $offset);
119		}
120
121		/**
122		* Adjust the rightmost boundary to the last safe character that can start a value
123		*
124		* Will rewind the boundary to skip the rightmost digits, optionally preceded by "i" or "i-"
125		*/
126	84	protected function computeSafeBoundary(): void
127		{
128	84	$boundary = $this->len - 1;
129		do
130		{
131	84	$c = substr($this->bencoded, $boundary, 1);
132		}
133	84	while (str_contains('0123456789', $c) && --$boundary >= 0);
134
135	84	$this->max = match ($c)
136		{
137	8	'-' => $boundary - 2,
138	7	'i' => $boundary - 1,
139	69	default => $boundary
140		};
141	84	}
142
143	2	protected static function convertTypeError(TypeError $e, int $offset): Throwable
144		{
145		// A type error can occur in decodeString() if the string length exceeds an int
146	2	$frame = $e->getTrace()[0];
147	2	if ($frame['class'] === __CLASS__ && $frame['function'] === 'decodeString')
148		{
149	1	return new DecodingException('String length overflow', $offset - 1);
150		}
151
152		// Return any other error as-is
153	1	return $e;
154		}
155
156	72	protected function decodeAnything(): ArrayObject\|array\|int\|string
157		{
158	72	return match ($this->bencoded[$this->offset])
159		{
160	31	'i' => $this->decodeInteger(),
161	21	'd' => $this->decodeDictionary(),
162	16	'l' => $this->decodeList(),
163	59	default => $this->decodeString()
164		};
165		}
166
167	21	protected function decodeDictionary(): ArrayObject
168		{
169	21	$values = [];
170	21	$lastKey = null;
171
172	21	++$this->offset;
173	21	while ($this->offset <= $this->max)
174		{
175	21	$c = $this->bencoded[$this->offset];
176	21	if ($c === 'e')
177		{
178	12	++$this->offset;
179
180	12	return new ArrayObject($values, ArrayObject::ARRAY_AS_PROPS);
181		}
182
183		// Quickly match the most common keys found in dictionaries
184	20	$key = match ($c)
185		{
186	3	'4' => $this->decodeFastString('4:path', 6, 'path'),
187	1	'6' => $this->decodeFastString('6:length', 8, 'length'),
188	17	default => $this->decodeString()
189		};
190	19	if (isset($lastKey) && strcmp($key, $lastKey) <= 0)
191		{
192	8	$this->dictionaryComplianceError($key, $lastKey);
193		}
194	19	if ($this->offset > $this->max)
195		{
196	1	break;
197		}
198	18	$values[$key] = $this->decodeAnything();
199	15	$lastKey = $key;
200		}
201
202	1	throw new DecodingException('Premature end of data', $this->len - 1);
203		}
204
205		/**
206		* @param string $match Bencoded string to match
207		* @param int $len Length of the bencoded string
208		* @param string $value String value to return if the string matches
209		*/
210	3	protected function decodeFastString(string $match, int $len, string $value): string
211		{
212	3	if (substr_compare($this->bencoded, $match, $this->offset, $len, false) === 0)
213		{
214	1	$this->offset += $len;
215
216	1	return $value;
217		}
218
219	3	return $this->decodeString();
220		}
221
222	31	protected function decodeInteger(): int
223		{
224	31	$negative = ($this->bencoded[++$this->offset] === '-');
225	31	if ($negative && $this->bencoded[++$this->offset] === '0')
226		{
227	2	$this->complianceError('Illegal character', $this->offset);
228		}
229
230	30	return $this->castInteger($this->readDigits('e'), $negative);
231		}
232
233	16	protected function decodeList(): array
234		{
235	16	++$this->offset;
236
237	16	$list = [];
238	16	while ($this->offset <= $this->max)
239		{
240	15	if ($this->bencoded[$this->offset] === 'e')
241		{
242	9	++$this->offset;
243
244	9	return $list;
245		}
246
247	12	$list[] = $this->decodeAnything();
248		}
249
250	5	throw new DecodingException('Premature end of data', $this->len - 1);
251		}
252
253	46	protected function decodeString(): string
254		{
255	46	$len = (int) $this->readDigits(':');
256	38	$string = substr($this->bencoded, $this->offset, $len);
257	38	$this->offset += $len;
258
259	37	return $string;
260		}
261
262	4	protected function dictionaryComplianceError(string $key, string $lastKey): void
263		{
264		// Compute the offset of the start of the string used as key
265	4	$offset = $this->offset - strlen(strlen($key) . ':') - strlen($key);
266
267	4	$msg = ($key === $lastKey) ? 'Duplicate' : 'Out of order';
268	4	$this->complianceError($msg . " dictionary entry '" . $key . "'", $offset);
269		}
270
271	63	protected function readDigits(string $terminator): string
272		{
273		// Digits sorted by decreasing frequency as observed on a random sample of torrent files
274	63	$spn = strspn($this->bencoded, '1463720859', $this->offset);
275	63	if (!$spn)
276		{
277	7	throw new DecodingException('Illegal character', $this->offset);
278		}
279	58	if ($this->bencoded[$this->offset] === '0' && $spn !== 1)
280		{
281	6	$this->complianceError('Illegal character', 1 + $this->offset);
282		}
283
284	55	$string = substr($this->bencoded, $this->offset, $spn);
285
286	55	$this->offset += $spn;
287	55	if ($this->bencoded[$this->offset] !== $terminator)
288		{
289	5	throw new DecodingException('Illegal character', $this->offset);
290		}
291	50	++$this->offset;
292
293	50	return $string;
294		}
295		}

s9e / Bencode

Push — master ( fde73f...e709b3 )

Decoder::__construct() A

Complexity

Size

Duplication

Code Coverage

Importance

Duplication Side-by-Side

Filter issues like