<?php

[0] Fix | Delete

[1] Fix | Delete

/**

[2] Fix | Delete

* IDNA URL encoder

[3] Fix | Delete

*

[4] Fix | Delete

* Note: Not fully compliant, as nameprep does nothing yet.

[5] Fix | Delete

*

[6] Fix | Delete

* @package Requests

[7] Fix | Delete

* @subpackage Utilities

[8] Fix | Delete

* @see https://tools.ietf.org/html/rfc3490 IDNA specification

[9] Fix | Delete

* @see https://tools.ietf.org/html/rfc3492 Punycode/Bootstrap specification

[10] Fix | Delete

*/

[11] Fix | Delete

class Requests_IDNAEncoder {

[12] Fix | Delete

/**

[13] Fix | Delete

* ACE prefix used for IDNA

[14] Fix | Delete

*

[15] Fix | Delete

* @see https://tools.ietf.org/html/rfc3490#section-5

[16] Fix | Delete

* @var string

[17] Fix | Delete

*/

[18] Fix | Delete

const ACE_PREFIX = 'xn--';

[19] Fix | Delete

[20] Fix | Delete

/**#@+

[21] Fix | Delete

* Bootstrap constant for Punycode

[22] Fix | Delete

*

[23] Fix | Delete

* @see https://tools.ietf.org/html/rfc3492#section-5

[24] Fix | Delete

* @var int

[25] Fix | Delete

*/

[26] Fix | Delete

const BOOTSTRAP_BASE = 36;

[27] Fix | Delete

const BOOTSTRAP_TMIN = 1;

[28] Fix | Delete

const BOOTSTRAP_TMAX = 26;

[29] Fix | Delete

const BOOTSTRAP_SKEW = 38;

[30] Fix | Delete

const BOOTSTRAP_DAMP = 700;

[31] Fix | Delete

const BOOTSTRAP_INITIAL_BIAS = 72;

[32] Fix | Delete

const BOOTSTRAP_INITIAL_N = 128;

[33] Fix | Delete

/**#@-*/

[34] Fix | Delete

[35] Fix | Delete

/**

[36] Fix | Delete

* Encode a hostname using Punycode

[37] Fix | Delete

*

[38] Fix | Delete

* @param string $string Hostname

[39] Fix | Delete

* @return string Punycode-encoded hostname

[40] Fix | Delete

*/

[41] Fix | Delete

public static function encode($string) {

[42] Fix | Delete

$parts = explode('.', $string);

[43] Fix | Delete

foreach ($parts as &$part) {

[44] Fix | Delete

$part = self::to_ascii($part);

[45] Fix | Delete

}

[46] Fix | Delete

return implode('.', $parts);

[47] Fix | Delete

}

[48] Fix | Delete

[49] Fix | Delete

/**

[50] Fix | Delete

* Convert a UTF-8 string to an ASCII string using Punycode

[51] Fix | Delete

*

[52] Fix | Delete

* @throws Requests_Exception Provided string longer than 64 ASCII characters (`idna.provided_too_long`)

[53] Fix | Delete

* @throws Requests_Exception Prepared string longer than 64 ASCII characters (`idna.prepared_too_long`)

[54] Fix | Delete

* @throws Requests_Exception Provided string already begins with xn-- (`idna.provided_is_prefixed`)

[55] Fix | Delete

* @throws Requests_Exception Encoded string longer than 64 ASCII characters (`idna.encoded_too_long`)

[56] Fix | Delete

*

[57] Fix | Delete

* @param string $string ASCII or UTF-8 string (max length 64 characters)

[58] Fix | Delete

* @return string ASCII string

[59] Fix | Delete

*/

[60] Fix | Delete

public static function to_ascii($string) {

[61] Fix | Delete

// Step 1: Check if the string is already ASCII

[62] Fix | Delete

if (self::is_ascii($string)) {

[63] Fix | Delete

// Skip to step 7

[64] Fix | Delete

if (strlen($string) < 64) {

[65] Fix | Delete

return $string;

[66] Fix | Delete

}

[67] Fix | Delete

[68] Fix | Delete

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

[69] Fix | Delete

}

[70] Fix | Delete

[71] Fix | Delete

// Step 2: nameprep

[72] Fix | Delete

$string = self::nameprep($string);

[73] Fix | Delete

[74] Fix | Delete

// Step 3: UseSTD3ASCIIRules is false, continue

[75] Fix | Delete

// Step 4: Check if it's ASCII now

[76] Fix | Delete

if (self::is_ascii($string)) {

[77] Fix | Delete

// Skip to step 7

[78] Fix | Delete

if (strlen($string) < 64) {

[79] Fix | Delete

return $string;

[80] Fix | Delete

}

[81] Fix | Delete

[82] Fix | Delete

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

[83] Fix | Delete

}

[84] Fix | Delete

[85] Fix | Delete

// Step 5: Check ACE prefix

[86] Fix | Delete

if (strpos($string, self::ACE_PREFIX) === 0) {

[87] Fix | Delete

throw new Requests_Exception('Provided string begins with ACE prefix', 'idna.provided_is_prefixed', $string);

[88] Fix | Delete

}

[89] Fix | Delete

[90] Fix | Delete

// Step 6: Encode with Punycode

[91] Fix | Delete

$string = self::punycode_encode($string);

[92] Fix | Delete

[93] Fix | Delete

// Step 7: Prepend ACE prefix

[94] Fix | Delete

$string = self::ACE_PREFIX . $string;

[95] Fix | Delete

[96] Fix | Delete

// Step 8: Check size

[97] Fix | Delete

if (strlen($string) < 64) {

[98] Fix | Delete

return $string;

[99] Fix | Delete

}

[100] Fix | Delete

[101] Fix | Delete

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

[102] Fix | Delete

}

[103] Fix | Delete

[104] Fix | Delete

/**

[105] Fix | Delete

* Check whether a given string contains only ASCII characters

[106] Fix | Delete

*

[107] Fix | Delete

* @internal (Testing found regex was the fastest implementation)

[108] Fix | Delete

*

[109] Fix | Delete

* @param string $string

[110] Fix | Delete

* @return bool Is the string ASCII-only?

[111] Fix | Delete

*/

[112] Fix | Delete

protected static function is_ascii($string) {

[113] Fix | Delete

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

[114] Fix | Delete

}

[115] Fix | Delete

[116] Fix | Delete

/**

[117] Fix | Delete

* Prepare a string for use as an IDNA name

[118] Fix | Delete

*

[119] Fix | Delete

* @todo Implement this based on RFC 3491 and the newer 5891

[120] Fix | Delete

* @param string $string

[121] Fix | Delete

* @return string Prepared string

[122] Fix | Delete

*/

[123] Fix | Delete

protected static function nameprep($string) {

[124] Fix | Delete

return $string;

[125] Fix | Delete

}

[126] Fix | Delete

[127] Fix | Delete

/**

[128] Fix | Delete

* Convert a UTF-8 string to a UCS-4 codepoint array

[129] Fix | Delete

*

[130] Fix | Delete

* Based on Requests_IRI::replace_invalid_with_pct_encoding()

[131] Fix | Delete

*

[132] Fix | Delete

* @throws Requests_Exception Invalid UTF-8 codepoint (`idna.invalidcodepoint`)

[133] Fix | Delete

* @param string $input

[134] Fix | Delete

* @return array Unicode code points

[135] Fix | Delete

*/

[136] Fix | Delete

protected static function utf8_to_codepoints($input) {

[137] Fix | Delete

$codepoints = array();

[138] Fix | Delete

[139] Fix | Delete

// Get number of bytes

[140] Fix | Delete

$strlen = strlen($input);

[141] Fix | Delete

[142] Fix | Delete

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

[143] Fix | Delete

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

[144] Fix | Delete

[145] Fix | Delete

// One byte sequence:

[146] Fix | Delete

if ((~$value & 0x80) === 0x80) {

[147] Fix | Delete

$character = $value;

[148] Fix | Delete

$length = 1;

[149] Fix | Delete

$remaining = 0;

[150] Fix | Delete

}

[151] Fix | Delete

// Two byte sequence:

[152] Fix | Delete

elseif (($value & 0xE0) === 0xC0) {

[153] Fix | Delete

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

[154] Fix | Delete

$length = 2;

[155] Fix | Delete

$remaining = 1;

[156] Fix | Delete

}

[157] Fix | Delete

// Three byte sequence:

[158] Fix | Delete

elseif (($value & 0xF0) === 0xE0) {

[159] Fix | Delete

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

[160] Fix | Delete

$length = 3;

[161] Fix | Delete

$remaining = 2;

[162] Fix | Delete

}

[163] Fix | Delete

// Four byte sequence:

[164] Fix | Delete

elseif (($value & 0xF8) === 0xF0) {

[165] Fix | Delete

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

[166] Fix | Delete

$length = 4;

[167] Fix | Delete

$remaining = 3;

[168] Fix | Delete

}

[169] Fix | Delete

// Invalid byte:

[170] Fix | Delete

else {

[171] Fix | Delete

throw new Requests_Exception('Invalid Unicode codepoint', 'idna.invalidcodepoint', $value);

[172] Fix | Delete

}

[173] Fix | Delete

[174] Fix | Delete

if ($remaining > 0) {

[175] Fix | Delete

if ($position + $length > $strlen) {

[176] Fix | Delete

throw new Requests_Exception('Invalid Unicode codepoint', 'idna.invalidcodepoint', $character);

[177] Fix | Delete

}

[178] Fix | Delete

for ($position++; $remaining > 0; $position++) {

[179] Fix | Delete

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

[180] Fix | Delete

[181] Fix | Delete

// If it is invalid, count the sequence as invalid and reprocess the current byte:

[182] Fix | Delete

if (($value & 0xC0) !== 0x80) {

[183] Fix | Delete

throw new Requests_Exception('Invalid Unicode codepoint', 'idna.invalidcodepoint', $character);

[184] Fix | Delete

}

[185] Fix | Delete

[186] Fix | Delete

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

[187] Fix | Delete

}

[188] Fix | Delete

$position--;

[189] Fix | Delete

}

[190] Fix | Delete

[191] Fix | Delete

if (

[192] Fix | Delete

// Non-shortest form sequences are invalid

[193] Fix | Delete

$length > 1 && $character <= 0x7F

[194] Fix | Delete

|| $length > 2 && $character <= 0x7FF

[195] Fix | Delete

|| $length > 3 && $character <= 0xFFFF

[196] Fix | Delete

// Outside of range of ucschar codepoints

[197] Fix | Delete

// Noncharacters

[198] Fix | Delete

|| ($character & 0xFFFE) === 0xFFFE

[199] Fix | Delete

|| $character >= 0xFDD0 && $character <= 0xFDEF

[200] Fix | Delete

|| (

[201] Fix | Delete

// Everything else not in ucschar

[202] Fix | Delete

$character > 0xD7FF && $character < 0xF900

[203] Fix | Delete

|| $character < 0x20

[204] Fix | Delete

|| $character > 0x7E && $character < 0xA0

[205] Fix | Delete

|| $character > 0xEFFFD

[206] Fix | Delete

)

[207] Fix | Delete

) {

[208] Fix | Delete

throw new Requests_Exception('Invalid Unicode codepoint', 'idna.invalidcodepoint', $character);

[209] Fix | Delete

}

[210] Fix | Delete

[211] Fix | Delete

$codepoints[] = $character;

[212] Fix | Delete

}

[213] Fix | Delete

[214] Fix | Delete

return $codepoints;

[215] Fix | Delete

}

[216] Fix | Delete

[217] Fix | Delete

/**

[218] Fix | Delete

* RFC3492-compliant encoder

[219] Fix | Delete

*

[220] Fix | Delete

* @internal Pseudo-code from Section 6.3 is commented with "#" next to relevant code

[221] Fix | Delete

* @throws Requests_Exception On character outside of the domain (never happens with Punycode) (`idna.character_outside_domain`)

[222] Fix | Delete

*

[223] Fix | Delete

* @param string $input UTF-8 encoded string to encode

[224] Fix | Delete

* @return string Punycode-encoded string

[225] Fix | Delete

*/

[226] Fix | Delete

public static function punycode_encode($input) {

[227] Fix | Delete

$output = '';

[228] Fix | Delete

# let n = initial_n

[229] Fix | Delete

$n = self::BOOTSTRAP_INITIAL_N;

[230] Fix | Delete

# let delta = 0

[231] Fix | Delete

$delta = 0;

[232] Fix | Delete

# let bias = initial_bias

[233] Fix | Delete

$bias = self::BOOTSTRAP_INITIAL_BIAS;

[234] Fix | Delete

# let h = b = the number of basic code points in the input

[235] Fix | Delete

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

[236] Fix | Delete

# copy them to the output in order

[237] Fix | Delete

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

[238] Fix | Delete

$extended = array();

[239] Fix | Delete

[240] Fix | Delete

foreach ($codepoints as $char) {

[241] Fix | Delete

if ($char < 128) {

[242] Fix | Delete

// Character is valid ASCII

[243] Fix | Delete

// TODO: this should also check if it's valid for a URL

[244] Fix | Delete

$output .= chr($char);

[245] Fix | Delete

$h++;

[246] Fix | Delete

}

[247] Fix | Delete

// Check if the character is non-ASCII, but below initial n

[248] Fix | Delete

// This never occurs for Punycode, so ignore in coverage

[249] Fix | Delete

// @codeCoverageIgnoreStart

[250] Fix | Delete

elseif ($char < $n) {

[251] Fix | Delete

throw new Requests_Exception('Invalid character', 'idna.character_outside_domain', $char);

[252] Fix | Delete

}

[253] Fix | Delete

// @codeCoverageIgnoreEnd

[254] Fix | Delete

else {

[255] Fix | Delete

$extended[$char] = true;

[256] Fix | Delete

}

[257] Fix | Delete

}

[258] Fix | Delete

$extended = array_keys($extended);

[259] Fix | Delete

sort($extended);

[260] Fix | Delete

$b = $h;

[261] Fix | Delete

# [copy them] followed by a delimiter if b > 0

[262] Fix | Delete

if (strlen($output) > 0) {

[263] Fix | Delete

$output .= '-';

[264] Fix | Delete

}

[265] Fix | Delete

# {if the input contains a non-basic code point < n then fail}

[266] Fix | Delete

# while h < length(input) do begin

[267] Fix | Delete

while ($h < count($codepoints)) {

[268] Fix | Delete

# let m = the minimum code point >= n in the input

[269] Fix | Delete

$m = array_shift($extended);

[270] Fix | Delete

//printf('next code point to insert is %s' . PHP_EOL, dechex($m));

[271] Fix | Delete

# let delta = delta + (m - n) * (h + 1), fail on overflow

[272] Fix | Delete

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

[273] Fix | Delete

# let n = m

[274] Fix | Delete

$n = $m;

[275] Fix | Delete

# for each code point c in the input (in order) do begin

[276] Fix | Delete

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

[277] Fix | Delete

$c = $codepoints[$num];

[278] Fix | Delete

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

[279] Fix | Delete

if ($c < $n) {

[280] Fix | Delete

$delta++;

[281] Fix | Delete

}

[282] Fix | Delete

# if c == n then begin

[283] Fix | Delete

elseif ($c === $n) {

[284] Fix | Delete

# let q = delta

[285] Fix | Delete

$q = $delta;

[286] Fix | Delete

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

[287] Fix | Delete

for ($k = self::BOOTSTRAP_BASE; ; $k += self::BOOTSTRAP_BASE) {

[288] Fix | Delete

# let t = tmin if k <= bias {+ tmin}, or

[289] Fix | Delete

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

[290] Fix | Delete

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

[291] Fix | Delete

$t = self::BOOTSTRAP_TMIN;

[292] Fix | Delete

}

[293] Fix | Delete

elseif ($k >= ($bias + self::BOOTSTRAP_TMAX)) {

[294] Fix | Delete

$t = self::BOOTSTRAP_TMAX;

[295] Fix | Delete

}

[296] Fix | Delete

else {

[297] Fix | Delete

$t = $k - $bias;

[298] Fix | Delete

}

[299] Fix | Delete

# if q < t then break

[300] Fix | Delete

if ($q < $t) {

[301] Fix | Delete

break;

[302] Fix | Delete

}

[303] Fix | Delete

# output the code point for digit t + ((q - t) mod (base - t))

[304] Fix | Delete

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

[305] Fix | Delete

$output .= self::digit_to_char($digit);

[306] Fix | Delete

# let q = (q - t) div (base - t)

[307] Fix | Delete

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

[308] Fix | Delete

# end

[309] Fix | Delete

}

[310] Fix | Delete

# output the code point for digit q

[311] Fix | Delete

$output .= self::digit_to_char($q);

[312] Fix | Delete

# let bias = adapt(delta, h + 1, test h equals b?)

[313] Fix | Delete

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

[314] Fix | Delete

# let delta = 0

[315] Fix | Delete

$delta = 0;

[316] Fix | Delete

# increment h

[317] Fix | Delete

$h++;

[318] Fix | Delete

# end

[319] Fix | Delete

}

[320] Fix | Delete

# end

[321] Fix | Delete

}

[322] Fix | Delete

# increment delta and n

[323] Fix | Delete

$delta++;

[324] Fix | Delete

$n++;

[325] Fix | Delete

# end

[326] Fix | Delete

}

[327] Fix | Delete

[328] Fix | Delete

return $output;

[329] Fix | Delete

}

[330] Fix | Delete

[331] Fix | Delete

/**

[332] Fix | Delete

* Convert a digit to its respective character

[333] Fix | Delete

*

[334] Fix | Delete

* @see https://tools.ietf.org/html/rfc3492#section-5

[335] Fix | Delete

* @throws Requests_Exception On invalid digit (`idna.invalid_digit`)

[336] Fix | Delete

*

[337] Fix | Delete

* @param int $digit Digit in the range 0-35

[338] Fix | Delete

* @return string Single character corresponding to digit

[339] Fix | Delete

*/

[340] Fix | Delete

protected static function digit_to_char($digit) {

[341] Fix | Delete

// @codeCoverageIgnoreStart

[342] Fix | Delete

// As far as I know, this never happens, but still good to be sure.

[343] Fix | Delete

if ($digit < 0 || $digit > 35) {

[344] Fix | Delete

throw new Requests_Exception(sprintf('Invalid digit %d', $digit), 'idna.invalid_digit', $digit);

[345] Fix | Delete

}

[346] Fix | Delete

// @codeCoverageIgnoreEnd

[347] Fix | Delete

$digits = 'abcdefghijklmnopqrstuvwxyz0123456789';

[348] Fix | Delete

return substr($digits, $digit, 1);

[349] Fix | Delete

}

[350] Fix | Delete

[351] Fix | Delete

/**

[352] Fix | Delete

* Adapt the bias

[353] Fix | Delete

*

[354] Fix | Delete

* @see https://tools.ietf.org/html/rfc3492#section-6.1

[355] Fix | Delete

* @param int $delta

[356] Fix | Delete

* @param int $numpoints

[357] Fix | Delete

* @param bool $firsttime

[358] Fix | Delete

* @return int New bias

[359] Fix | Delete

*/

[360] Fix | Delete

protected static function adapt($delta, $numpoints, $firsttime) {

[361] Fix | Delete

# function adapt(delta,numpoints,firsttime):

[362] Fix | Delete

# if firsttime then let delta = delta div damp

[363] Fix | Delete

if ($firsttime) {

[364] Fix | Delete

$delta = floor($delta / self::BOOTSTRAP_DAMP);

[365] Fix | Delete

}

[366] Fix | Delete

# else let delta = delta div 2

[367] Fix | Delete

else {

[368] Fix | Delete

$delta = floor($delta / 2);

[369] Fix | Delete

}

[370] Fix | Delete

# let delta = delta + (delta div numpoints)

[371] Fix | Delete

$delta += floor($delta / $numpoints);

[372] Fix | Delete

# let k = 0

[373] Fix | Delete

$k = 0;

[374] Fix | Delete

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

[375] Fix | Delete

$max = floor(((self::BOOTSTRAP_BASE - self::BOOTSTRAP_TMIN) * self::BOOTSTRAP_TMAX) / 2);

[376] Fix | Delete

while ($delta > $max) {

[377] Fix | Delete

# let delta = delta div (base - tmin)

[378] Fix | Delete

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

[379] Fix | Delete

# let k = k + base

[380] Fix | Delete

$k += self::BOOTSTRAP_BASE;

[381] Fix | Delete

# end

[382] Fix | Delete

}

[383] Fix | Delete

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

[384] Fix | Delete

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

[385] Fix | Delete

}

[386] Fix | Delete

}

[387] Fix | Delete