Файл: vendor/symfony/polyfill-intl-normalizer/Normalizer.php
Строк: 197
<?php
/*
* This file is part of the Symfony package.
*
* (c) Fabien Potencier <fabien@symfony.com>
*
* For the full copyright and license information, please view the LICENSE
* file that was distributed with this source code.
*/
namespace SymfonyPolyfillIntlNormalizer;
/**
* Normalizer is a PHP fallback implementation of the Normalizer class provided by the intl extension.
*
* It has been validated with Unicode 6.3 Normalization Conformance Test.
* See http://www.unicode.org/reports/tr15/ for detailed info about Unicode normalizations.
*
* @author Nicolas Grekas <p@tchwork.com>
*
* @internal
*/
class Normalizer
{
public const FORM_D = Normalizer::FORM_D;
public const FORM_KD = Normalizer::FORM_KD;
public const FORM_C = Normalizer::FORM_C;
public const FORM_KC = Normalizer::FORM_KC;
public const NFD = Normalizer::NFD;
public const NFKD = Normalizer::NFKD;
public const NFC = Normalizer::NFC;
public const NFKC = Normalizer::NFKC;
private static $C;
private static $D;
private static $KD;
private static $cC;
private static $rawD;
private static $rawKD;
private static $ulenMask = ["xC0" => 2, "xD0" => 2, "xE0" => 3, "xF0" => 4];
private static $ASCII = "x20x65x69x61x73x6Ex74x72x6Fx6Cx75x64x5Dx5Bx63x6Dx70x27x0Ax67x7Cx68x76x2Ex66x62x2Cx3Ax3Dx2Dx71x31x30x43x32x2Ax79x78x29x28x4Cx39x41x53x2Fx50x22x45x6Ax4Dx49x6Bx33x3Ex35x54x3Cx44x34x7Dx42x7Bx38x46x77x52x36x37x55x47x4Ex3Bx4Ax7Ax56x23x48x4Fx57x5Fx26x21x4Bx3Fx58x51x25x59x5Cx09x5Ax2Bx7Ex5Ex24x40x60x7Fx00x01x02x03x04x05x06x07x08x0Bx0Cx0Dx0Ex0Fx10x11x12x13x14x15x16x17x18x19x1Ax1Bx1Cx1Dx1Ex1F";
public static function isNormalized(string $s, int $form = self::FORM_C)
{
if (!in_array($form, [self::NFD, self::NFKD, self::NFC, self::NFKC])) {
return false;
}
if (!isset($s[strspn($s, self::$ASCII)])) {
return true;
}
if (self::NFC == $form && preg_match('//u', $s) && !preg_match('/[^x00-x{2FF}]/u', $s)) {
return true;
}
return self::normalize($s, $form) === $s;
}
public static function getRawDecomposition(string $s, int $form = self::FORM_C)
{
if ('' === $s || !preg_match('//u', $s)) {
return null;
}
$ulen = $s[0] < "x80" ? 1 : (self::$ulenMask[$s[0] & "xF0"] ?? 0);
if (!$ulen || strlen($s) !== $ulen) {
return null;
}
if (self::NFC !== $form && self::NFD !== $form && self::NFKC !== $form && self::NFKD !== $form) {
return '';
}
if ($s >= "xEAxB0x80" && $s <= "xEDx9ExA3") {
$u = unpack('C*', $s);
$j = (($u[1] - 224) << 12) + (($u[2] - 128) << 6) + $u[3] - 0xAC80;
if ($t = $j % 28) {
$j -= $t;
$lv = 0xAC00 + $j;
$r = chr(0xE0 | $lv >> 12).chr(0x80 | $lv >> 6 & 0x3F).chr(0x80 | $lv & 0x3F);
$r .= $t < 25
? ("xE1x86".chr(0xA7 + $t))
: ("xE1x87".chr(0x67 + $t));
return $r;
}
return "xE1x84".chr(0x80 + (int) ($j / 588))
."xE1x85".chr(0xA1 + (int) (($j % 588) / 28));
}
if (null === self::$rawD) {
self::$rawD = self::getData('rawCanonicalDecomposition');
}
if (isset(self::$rawD[$s])) {
return self::$rawD[$s];
}
if (self::NFKC === $form || self::NFKD === $form) {
if (null === self::$rawKD) {
self::$rawKD = self::getData('rawCompatibilityDecomposition');
}
return self::$rawKD[$s] ?? null;
}
return null;
}
public static function normalize(string $s, int $form = self::FORM_C)
{
if (!preg_match('//u', $s)) {
return false;
}
switch ($form) {
case self::NFC: $C = true; $K = false; break;
case self::NFD: $C = false; $K = false; break;
case self::NFKC: $C = true; $K = true; break;
case self::NFKD: $C = false; $K = true; break;
default:
if (defined('Normalizer::NONE') && Normalizer::NONE == $form) {
return $s;
}
if (80000 > PHP_VERSION_ID) {
return false;
}
throw new ValueError('normalizer_normalize(): Argument #2 ($form) must be a a valid normalization form');
}
if ('' === $s) {
return '';
}
if ($K && null === self::$KD) {
self::$KD = self::getData('compatibilityDecomposition');
}
if (null === self::$D) {
self::$D = self::getData('canonicalDecomposition');
self::$cC = self::getData('combiningClass');
}
if (null !== $mbEncoding = (2 /* MB_OVERLOAD_STRING */ & (int) ini_get('mbstring.func_overload')) ? mb_internal_encoding() : null) {
mb_internal_encoding('8bit');
}
$r = self::decompose($s, $K);
if ($C) {
if (null === self::$C) {
self::$C = self::getData('canonicalComposition');
}
$r = self::recompose($r);
}
if (null !== $mbEncoding) {
mb_internal_encoding($mbEncoding);
}
return $r;
}
private static function recompose($s)
{
$ASCII = self::$ASCII;
$compMap = self::$C;
$combClass = self::$cC;
$ulenMask = self::$ulenMask;
$result = $tail = '';
$i = $s[0] < "x80" ? 1 : $ulenMask[$s[0] & "xF0"];
$len = strlen($s);
$lastUchr = substr($s, 0, $i);
$lastUcls = isset($combClass[$lastUchr]) ? 256 : 0;
while ($i < $len) {
if ($s[$i] < "x80") {
// ASCII chars
if ($tail) {
$lastUchr .= $tail;
$tail = '';
}
if ($j = strspn($s, $ASCII, $i + 1)) {
$lastUchr .= substr($s, $i, $j);
$i += $j;
}
$result .= $lastUchr;
$lastUchr = $s[$i];
$lastUcls = 0;
++$i;
continue;
}
$ulen = $ulenMask[$s[$i] & "xF0"];
$uchr = substr($s, $i, $ulen);
if ($lastUchr < "xE1x84x80" || "xE1x84x92" < $lastUchr
|| $uchr < "xE1x85xA1" || "xE1x85xB5" < $uchr
|| $lastUcls) {
// Table lookup and combining chars composition
$ucls = $combClass[$uchr] ?? 0;
if (isset($compMap[$lastUchr.$uchr]) && (!$lastUcls || $lastUcls < $ucls)) {
$lastUchr = $compMap[$lastUchr.$uchr];
} elseif ($lastUcls = $ucls) {
$tail .= $uchr;
} else {
if ($tail) {
$lastUchr .= $tail;
$tail = '';
}
$result .= $lastUchr;
$lastUchr = $uchr;
}
} else {
// Hangul chars
$L = ord($lastUchr[2]) - 0x80;
$V = ord($uchr[2]) - 0xA1;
$T = 0;
$uchr = substr($s, $i + $ulen, 3);
if ("xE1x86xA7" <= $uchr && $uchr <= "xE1x87x82") {
$T = ord($uchr[2]) - 0xA7;
0 > $T && $T += 0x40;
$ulen += 3;
}
$L = 0xAC00 + ($L * 21 + $V) * 28 + $T;
$lastUchr = chr(0xE0 | $L >> 12).chr(0x80 | $L >> 6 & 0x3F).chr(0x80 | $L & 0x3F);
}
$i += $ulen;
}
return $result.$lastUchr.$tail;
}
private static function decompose($s, $c)
{
$result = '';
$ASCII = self::$ASCII;
$decompMap = self::$D;
$combClass = self::$cC;
$ulenMask = self::$ulenMask;
if ($c) {
$compatMap = self::$KD;
}
$c = [];
$i = 0;
$len = strlen($s);
while ($i < $len) {
if ($s[$i] < "x80") {
// ASCII chars
if ($c) {
ksort($c);
$result .= implode('', $c);
$c = [];
}
$j = 1 + strspn($s, $ASCII, $i + 1);
$result .= substr($s, $i, $j);
$i += $j;
continue;
}
$ulen = $ulenMask[$s[$i] & "xF0"];
$uchr = substr($s, $i, $ulen);
$i += $ulen;
if ($uchr < "xEAxB0x80" || "xEDx9ExA3" < $uchr) {
// Table lookup
if ($uchr !== $j = $compatMap[$uchr] ?? ($decompMap[$uchr] ?? $uchr)) {
$uchr = $j;
$j = strlen($uchr);
$ulen = $uchr[0] < "x80" ? 1 : $ulenMask[$uchr[0] & "xF0"];
if ($ulen != $j) {
// Put trailing chars in $s
$j -= $ulen;
$i -= $j;
if (0 > $i) {
$s = str_repeat(' ', -$i).$s;
$len -= $i;
$i = 0;
}
while ($j--) {
$s[$i + $j] = $uchr[$ulen + $j];
}
$uchr = substr($uchr, 0, $ulen);
}
}
if (isset($combClass[$uchr])) {
// Combining chars, for sorting
if (!isset($c[$combClass[$uchr]])) {
$c[$combClass[$uchr]] = '';
}
$c[$combClass[$uchr]] .= $uchr;
continue;
}
} else {
// Hangul chars
$uchr = unpack('C*', $uchr);
$j = (($uchr[1] - 224) << 12) + (($uchr[2] - 128) << 6) + $uchr[3] - 0xAC80;
$uchr = "xE1x84".chr(0x80 + (int) ($j / 588))
."xE1x85".chr(0xA1 + (int) (($j % 588) / 28));
if ($j %= 28) {
$uchr .= $j < 25
? ("xE1x86".chr(0xA7 + $j))
: ("xE1x87".chr(0x67 + $j));
}
}
if ($c) {
ksort($c);
$result .= implode('', $c);
$c = [];
}
$result .= $uchr;
}
if ($c) {
ksort($c);
$result .= implode('', $c);
}
return $result;
}
private static function getData($file)
{
if (file_exists($file = __DIR__.'/Resources/unidata/'.$file.'.php')) {
return require $file;
}
return false;
}
}