Requests_IDNAEncoder

Complexity

Total Complexity

55

Size/Duplication

Total Lines	376
Duplicated Lines	8.24 %

Coupling/Cohesion

Components	1
Dependencies	1

Importance

Changes

0

8 Methods

Rating	Name	Duplication	Size	Complexity
A	encode()	0	7	2
C	to_ascii()	16	43	7
A	is_ascii()	0	3	1
A	nameprep()	0	3	1
D	utf8_to_codepoints()	15	80	25
D	punycode_encode()	0	104	13
A	digit_to_char()	0	10	3
B	adapt()	0	26	3

<?php

/**
 * IDNA URL encoder
 *
 * Note: Not fully compliant, as nameprep does nothing yet.
 *
 * @package Requests
 * @subpackage Utilities
 * @see https://tools.ietf.org/html/rfc3490 IDNA specification
 * @see https://tools.ietf.org/html/rfc3492 Punycode/Bootstrap specification
 */
class Requests_IDNAEncoder {

PSR1 recommends that each class must be in a namespace of at least one level to avoid collisions.

This class is not in CamelCase format.

	/**
	 * ACE prefix used for IDNA
	 *
	 * @see https://tools.ietf.org/html/rfc3490#section-5
	 * @var string
	 */
	const ACE_PREFIX = 'xn--';

	/**#@+
	 * Bootstrap constant for Punycode
	 *
	 * @see https://tools.ietf.org/html/rfc3492#section-5
	 * @var int
	 */
	const BOOTSTRAP_BASE         = 36;
	const BOOTSTRAP_TMIN         = 1;
	const BOOTSTRAP_TMAX         = 26;
	const BOOTSTRAP_SKEW         = 38;
	const BOOTSTRAP_DAMP         = 700;
	const BOOTSTRAP_INITIAL_BIAS = 72;
	const BOOTSTRAP_INITIAL_N    = 128;
	/**#@-*/

	/**
	 * Encode a hostname using Punycode
	 *
	 * @param string $string Hostname
	 * @return string Punycode-encoded hostname
	 */
	public static function encode($string) {
		$parts = explode('.', $string);
		foreach ($parts as &$part) {
			$part = self::to_ascii($part);
		}
		return implode('.', $parts);
	}

	/**
	 * Convert a UTF-8 string to an ASCII string using Punycode
	 *
	 * @throws Requests_Exception Provided string longer than 64 ASCII characters (`idna.provided_too_long`)
	 * @throws Requests_Exception Prepared string longer than 64 ASCII characters (`idna.prepared_too_long`)
	 * @throws Requests_Exception Provided string already begins with xn-- (`idna.provided_is_prefixed`)
	 * @throws Requests_Exception Encoded string longer than 64 ASCII characters (`idna.encoded_too_long`)
	 *
	 * @param string $string ASCII or UTF-8 string (max length 64 characters)
	 * @return string ASCII string
	 */
	public static function to_ascii($string) {

This method is not in camel caps format.

		// Step 1: Check if the string is already ASCII
		if (self::is_ascii($string)) {

This code seems to be duplicated across your project.

			// Skip to step 7
			if (strlen($string) < 64) {
				return $string;
			}

			throw new Requests_Exception('Provided string is too long', 'idna.provided_too_long', $string);
		}

		// Step 2: nameprep
		$string = self::nameprep($string);

Consider using a different name than the parameter $string. This often makes code more readable.

		// Step 3: UseSTD3ASCIIRules is false, continue

36% of this comment could be valid code. Did you maybe forget this after debugging?

		// Step 4: Check if it's ASCII now
		if (self::is_ascii($string)) {

This code seems to be duplicated across your project.

			// Skip to step 7
			if (strlen($string) < 64) {
				return $string;
			}

			throw new Requests_Exception('Prepared string is too long', 'idna.prepared_too_long', $string);
		}

		// Step 5: Check ACE prefix
		if (strpos($string, self::ACE_PREFIX) === 0) {
			throw new Requests_Exception('Provided string begins with ACE prefix', 'idna.provided_is_prefixed', $string);
		}

		// Step 6: Encode with Punycode
		$string = self::punycode_encode($string);

Consider using a different name than the parameter $string. This often makes code more readable.

		// Step 7: Prepend ACE prefix
		$string = self::ACE_PREFIX . $string;

Consider using a different name than the parameter $string. This often makes code more readable.

		// Step 8: Check size
		if (strlen($string) < 64) {
			return $string;
		}

		throw new Requests_Exception('Encoded string is too long', 'idna.encoded_too_long', $string);
	}

	/**
	 * Check whether a given string contains only ASCII characters
	 *
	 * @internal (Testing found regex was the fastest implementation)
	 *
	 * @param string $string
	 * @return bool Is the string ASCII-only?
	 */
	protected static function is_ascii($string) {

This method is not in camel caps format.

		return (preg_match('/(?:[^\x00-\x7F])/', $string) !== 1);
	}

	/**
	 * Prepare a string for use as an IDNA name
	 *
	 * @todo Implement this based on RFC 3491 and the newer 5891

Comment refers to a TODO task

	 * @param string $string
	 * @return string Prepared string
	 */
	protected static function nameprep($string) {
		return $string;
	}

	/**
	 * Convert a UTF-8 string to a UCS-4 codepoint array
	 *
	 * Based on Requests_IRI::replace_invalid_with_pct_encoding()
	 *
	 * @throws Requests_Exception Invalid UTF-8 codepoint (`idna.invalidcodepoint`)
	 * @param string $input
	 * @return array Unicode code points
	 */
	protected static function utf8_to_codepoints($input) {

This method is not in camel caps format.

		$codepoints = array();

		// Get number of bytes
		$strlen = strlen($input);

		for ($position = 0; $position < $strlen; $position++) {

Loop incrementor ($position) jumbling with inner loop

			$value = ord($input[$position]);

			// One byte sequence:
			if ((~$value & 0x80) === 0x80) {
				$character = $value;
				$length = 1;

Equals sign not aligned with surrounding assignments; expected 4 spaces but found 1 space

				$remaining = 0;
			}
			// Two byte sequence:
			elseif (($value & 0xE0) === 0xC0) {

This code seems to be duplicated across your project.

				$character = ($value & 0x1F) << 6;
				$length = 2;

Equals sign not aligned with surrounding assignments; expected 4 spaces but found 1 space

				$remaining = 1;
			}
			// Three byte sequence:
			elseif (($value & 0xF0) === 0xE0) {

This code seems to be duplicated across your project.

				$character = ($value & 0x0F) << 12;
				$length = 3;

Equals sign not aligned with surrounding assignments; expected 4 spaces but found 1 space

				$remaining = 2;
			}
			// Four byte sequence:
			elseif (($value & 0xF8) === 0xF0) {

This code seems to be duplicated across your project.

				$character = ($value & 0x07) << 18;
				$length = 4;

Equals sign not aligned with surrounding assignments; expected 4 spaces but found 1 space

				$remaining = 3;
			}
			// Invalid byte:
			else {
				throw new Requests_Exception('Invalid Unicode codepoint', 'idna.invalidcodepoint', $value);
			}

			if ($remaining > 0) {
				if ($position + $length > $strlen) {
					throw new Requests_Exception('Invalid Unicode codepoint', 'idna.invalidcodepoint', $character);
				}
				for ($position++; $remaining > 0; $position++) {
					$value = ord($input[$position]);

					// If it is invalid, count the sequence as invalid and reprocess the current byte:
					if (($value & 0xC0) !== 0x80) {
						throw new Requests_Exception('Invalid Unicode codepoint', 'idna.invalidcodepoint', $character);
					}

					$character |= ($value & 0x3F) << (--$remaining * 6);
				}
				$position--;
			}

			if (
				// Non-shortest form sequences are invalid
				   $length > 1 && $character <= 0x7F
				|| $length > 2 && $character <= 0x7FF
				|| $length > 3 && $character <= 0xFFFF
				// Outside of range of ucschar codepoints
				// Noncharacters
				|| ($character & 0xFFFE) === 0xFFFE
				|| $character >= 0xFDD0 && $character <= 0xFDEF
				|| (
					// Everything else not in ucschar
					   $character > 0xD7FF && $character < 0xF900
					|| $character < 0x20
					|| $character > 0x7E && $character < 0xA0
					|| $character > 0xEFFFD
				)
			) {
				throw new Requests_Exception('Invalid Unicode codepoint', 'idna.invalidcodepoint', $character);
			}

			$codepoints[] = $character;
		}

		return $codepoints;
	}

	/**
	 * RFC3492-compliant encoder
	 *
	 * @internal Pseudo-code from Section 6.3 is commented with "#" next to relevant code
	 * @throws Requests_Exception On character outside of the domain (never happens with Punycode) (`idna.character_outside_domain`)

This line exceeds maximum limit of 120 characters; contains 129 characters

	 *
	 * @param string $input UTF-8 encoded string to encode
	 * @return string Punycode-encoded string
	 */
	public static function punycode_encode($input) {

This method is not in camel caps format.

		$output = '';
#		let n = initial_n

Perl-style comments are not allowed. Use "// Comment." or "/* comment */" instead.

		$n = self::BOOTSTRAP_INITIAL_N;

Avoid variables with short names like $n. Configured minimum length is 3.

#		let delta = 0

Perl-style comments are not allowed. Use "// Comment." or "/* comment */" instead.

		$delta = 0;
#		let bias = initial_bias

Perl-style comments are not allowed. Use "// Comment." or "/* comment */" instead.

		$bias = self::BOOTSTRAP_INITIAL_BIAS;
#		let h = b = the number of basic code points in the input

Perl-style comments are not allowed. Use "// Comment." or "/* comment */" instead.

		$h = $b = 0; // see loop

$b is not used, you could remove the assignment.

Avoid variables with short names like $h. Configured minimum length is 3.

Avoid variables with short names like $b. Configured minimum length is 3.

#		copy them to the output in order

Perl-style comments are not allowed. Use "// Comment." or "/* comment */" instead.

		$codepoints = self::utf8_to_codepoints($input);
		$extended = array();

Equals sign not aligned with surrounding assignments; expected 3 spaces but found 1 space

		foreach ($codepoints as $char) {
			if ($char < 128) {
				// Character is valid ASCII
				// TODO: this should also check if it's valid for a URL

Comments for TODO tasks are often forgotten in the code; it might be better to use a dedicated issue tracker.

				$output .= chr($char);
				$h++;
			}
			// Check if the character is non-ASCII, but below initial n
			// This never occurs for Punycode, so ignore in coverage
			// @codeCoverageIgnoreStart
			elseif ($char < $n) {
				throw new Requests_Exception('Invalid character', 'idna.character_outside_domain', $char);
			}
			// @codeCoverageIgnoreEnd
			else {
				$extended[$char] = true;
			}
		}
		$extended = array_keys($extended);
		sort($extended);
		$b = $h;
#		[copy them] followed by a delimiter if b > 0

Perl-style comments are not allowed. Use "// Comment." or "/* comment */" instead.

		if (strlen($output) > 0) {
			$output .= '-';
		}
#		{if the input contains a non-basic code point < n then fail}

Perl-style comments are not allowed. Use "// Comment." or "/* comment */" instead.

#		while h < length(input) do begin

Perl-style comments are not allowed. Use "// Comment." or "/* comment */" instead.

		while ($h < count($codepoints)) {
#			let m = the minimum code point >= n in the input

Perl-style comments are not allowed. Use "// Comment." or "/* comment */" instead.

			$m = array_shift($extended);

Avoid variables with short names like $m. Configured minimum length is 3.

			//printf('next code point to insert is %s' . PHP_EOL, dechex($m));
#			let delta = delta + (m - n) * (h + 1), fail on overflow

Perl-style comments are not allowed. Use "// Comment." or "/* comment */" instead.

			$delta += ($m - $n) * ($h + 1);
#			let n = m

Perl-style comments are not allowed. Use "// Comment." or "/* comment */" instead.

			$n = $m;
#			for each code point c in the input (in order) do begin

Perl-style comments are not allowed. Use "// Comment." or "/* comment */" instead.

			for ($num = 0; $num < count($codepoints); $num++) {

It seems like you are calling the size function count() as part of the test condition. You might want to compute the size beforehand, and not on each iteration.

				$c = $codepoints[$num];

Avoid variables with short names like $c. Configured minimum length is 3.

#				if c < n then increment delta, fail on overflow

Perl-style comments are not allowed. Use "// Comment." or "/* comment */" instead.

				if ($c < $n) {
					$delta++;
				}
#				if c == n then begin

Perl-style comments are not allowed. Use "// Comment." or "/* comment */" instead.

				elseif ($c === $n) {
#					let q = delta

Perl-style comments are not allowed. Use "// Comment." or "/* comment */" instead.

					$q = $delta;

Avoid variables with short names like $q. Configured minimum length is 3.

#					for k = base to infinity in steps of base do begin

Perl-style comments are not allowed. Use "// Comment." or "/* comment */" instead.

					for ($k = self::BOOTSTRAP_BASE; ; $k += self::BOOTSTRAP_BASE) {
#						let t = tmin if k <= bias {+ tmin}, or

Perl-style comments are not allowed. Use "// Comment." or "/* comment */" instead.

#								tmax if k >= bias + tmax, or k - bias otherwise

Perl-style comments are not allowed. Use "// Comment." or "/* comment */" instead.

						if ($k <= ($bias + self::BOOTSTRAP_TMIN)) {
							$t = self::BOOTSTRAP_TMIN;

Avoid variables with short names like $t. Configured minimum length is 3.

						}
						elseif ($k >= ($bias + self::BOOTSTRAP_TMAX)) {
							$t = self::BOOTSTRAP_TMAX;
						}
						else {
							$t = $k - $bias;
						}
#						if q < t then break

Perl-style comments are not allowed. Use "// Comment." or "/* comment */" instead.

						if ($q < $t) {
							break;
						}
#						output the code point for digit t + ((q - t) mod (base - t))

Perl-style comments are not allowed. Use "// Comment." or "/* comment */" instead.

						$digit = $t + (($q - $t) % (self::BOOTSTRAP_BASE - $t));

Equals sign not aligned with surrounding assignments; expected 3 spaces but found 1 space

						$output .= self::digit_to_char($digit);
#						let q = (q - t) div (base - t)

Perl-style comments are not allowed. Use "// Comment." or "/* comment */" instead.

						$q = floor(($q - $t) / (self::BOOTSTRAP_BASE - $t));
#					end

Perl-style comments are not allowed. Use "// Comment." or "/* comment */" instead.

					}
#					output the code point for digit q

Perl-style comments are not allowed. Use "// Comment." or "/* comment */" instead.

					$output .= self::digit_to_char($q);
#					let bias = adapt(delta, h + 1, test h equals b?)

Perl-style comments are not allowed. Use "// Comment." or "/* comment */" instead.

					$bias = self::adapt($delta, $h + 1, $h === $b);
#					let delta = 0

Perl-style comments are not allowed. Use "// Comment." or "/* comment */" instead.

					$delta = 0;
#					increment h

Perl-style comments are not allowed. Use "// Comment." or "/* comment */" instead.

					$h++;
#				end

Perl-style comments are not allowed. Use "// Comment." or "/* comment */" instead.

				}
#			end

Perl-style comments are not allowed. Use "// Comment." or "/* comment */" instead.

			}
#			increment delta and n

Perl-style comments are not allowed. Use "// Comment." or "/* comment */" instead.

			$delta++;
			$n++;
#		end

Perl-style comments are not allowed. Use "// Comment." or "/* comment */" instead.

		}

		return $output;
	}

	/**
	 * Convert a digit to its respective character
	 *
	 * @see https://tools.ietf.org/html/rfc3492#section-5
	 * @throws Requests_Exception On invalid digit (`idna.invalid_digit`)
	 *
	 * @param int $digit Digit in the range 0-35
	 * @return string Single character corresponding to digit
	 */
	protected static function digit_to_char($digit) {

This method is not in camel caps format.

		// @codeCoverageIgnoreStart
		// As far as I know, this never happens, but still good to be sure.
		if ($digit < 0 || $digit > 35) {
			throw new Requests_Exception(sprintf('Invalid digit %d', $digit), 'idna.invalid_digit', $digit);
		}
		// @codeCoverageIgnoreEnd
		$digits = 'abcdefghijklmnopqrstuvwxyz0123456789';
		return substr($digits, $digit, 1);
	}

	/**
	 * Adapt the bias
	 *
	 * @see https://tools.ietf.org/html/rfc3492#section-6.1
	 * @param int $delta
	 * @param int $numpoints
	 * @param bool $firsttime
	 * @return int New bias

Should the return type not be double?

	 */
	protected static function adapt($delta, $numpoints, $firsttime) {
#	function adapt(delta,numpoints,firsttime):

Perl-style comments are not allowed. Use "// Comment." or "/* comment */" instead.

#		if firsttime then let delta = delta div damp

Perl-style comments are not allowed. Use "// Comment." or "/* comment */" instead.

		if ($firsttime) {
			$delta = floor($delta / self::BOOTSTRAP_DAMP);

Consider using a different name than the parameter $delta. This often makes code more readable.

		}
#		else let delta = delta div 2

Perl-style comments are not allowed. Use "// Comment." or "/* comment */" instead.

		else {
			$delta = floor($delta / 2);

Consider using a different name than the parameter $delta. This often makes code more readable.

		}
#		let delta = delta + (delta div numpoints)

Perl-style comments are not allowed. Use "// Comment." or "/* comment */" instead.

		$delta += floor($delta / $numpoints);
#		let k = 0

Perl-style comments are not allowed. Use "// Comment." or "/* comment */" instead.

		$k = 0;

Avoid variables with short names like $k. Configured minimum length is 3.

#		while delta > ((base - tmin) * tmax) div 2 do begin

Perl-style comments are not allowed. Use "// Comment." or "/* comment */" instead.

		$max = floor(((self::BOOTSTRAP_BASE - self::BOOTSTRAP_TMIN) * self::BOOTSTRAP_TMAX) / 2);
		while ($delta > $max) {
#			let delta = delta div (base - tmin)

Perl-style comments are not allowed. Use "// Comment." or "/* comment */" instead.

			$delta = floor($delta / (self::BOOTSTRAP_BASE - self::BOOTSTRAP_TMIN));

Consider using a different name than the parameter $delta. This often makes code more readable.

#			let k = k + base

Perl-style comments are not allowed. Use "// Comment." or "/* comment */" instead.

			$k += self::BOOTSTRAP_BASE;
#		end

Perl-style comments are not allowed. Use "// Comment." or "/* comment */" instead.

		}
#		return k + (((base - tmin + 1) * delta) div (delta + skew))

Perl-style comments are not allowed. Use "// Comment." or "/* comment */" instead.

		return $k + floor(((self::BOOTSTRAP_BASE - self::BOOTSTRAP_TMIN + 1) * $delta) / ($delta + self::BOOTSTRAP_SKEW));
	}
}