519 lines
21 KiB
PHP
519 lines
21 KiB
PHP
<?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 Symfony\Polyfill\Intl\Grapheme;
|
||
|
||
\define('SYMFONY_GRAPHEME_CLUSTER_RX', ((float) \PCRE_VERSION >= 10.44) ? '\X' : Grapheme::GRAPHEME_CLUSTER_RX);
|
||
|
||
/**
|
||
* Partial intl implementation in pure PHP.
|
||
*
|
||
* Implemented:
|
||
* - grapheme_extract - Extract a sequence of grapheme clusters from a text buffer, which must be encoded in UTF-8
|
||
* - grapheme_stripos - Find position (in grapheme units) of first occurrence of a case-insensitive string
|
||
* - grapheme_stristr - Returns part of haystack string from the first occurrence of case-insensitive needle to the end of haystack
|
||
* - grapheme_strlen - Get string length in grapheme units
|
||
* - grapheme_strpos - Find position (in grapheme units) of first occurrence of a string
|
||
* - grapheme_strripos - Find position (in grapheme units) of last occurrence of a case-insensitive string
|
||
* - grapheme_strrpos - Find position (in grapheme units) of last occurrence of a string
|
||
* - grapheme_strstr - Returns part of haystack string from the first occurrence of needle to the end of haystack
|
||
* - grapheme_substr - Return part of a string
|
||
* - grapheme_str_split - Splits a string into an array of individual or chunks of graphemes
|
||
* - grapheme_levenshtein - Calculate the grapheme-unit Levenshtein distance between two strings
|
||
* - grapheme_strrev - Reverse a string by grapheme clusters
|
||
*
|
||
* @author Nicolas Grekas <p@tchwork.com>
|
||
*
|
||
* @internal
|
||
*/
|
||
final class Grapheme
|
||
{
|
||
// (CRLF|RI RI?|(([ZWNJ-ZWJ]|T+|L*(LV?V+|LV|LVT)T*|L+|[^Control])[Extend]*)(ZWJ([ZWNJ-ZWJ]|T+|L*(LV?V+|LV|LVT)T*|L+|[^Control])[Extend]*)*|[Control])
|
||
// This regular expression is a work around for http://bugs.exim.org/1279
|
||
public const GRAPHEME_CLUSTER_RX = '(?:\r\n|[\x{1F1E6}-\x{1F1FF}][\x{1F1E6}-\x{1F1FF}]?|(?:[ -~\x{200C}\x{200D}]|[ᆨ-ᇹ]+|[ᄀ-ᅟ]*(?:[가개갸걔거게겨계고과괘괴교구궈궤귀규그긔기까깨꺄꺠꺼께껴꼐꼬꽈꽤꾀꾜꾸꿔꿰뀌뀨끄끠끼나내냐냬너네녀녜노놔놰뇌뇨누눠눼뉘뉴느늬니다대댜댸더데뎌뎨도돠돼되됴두둬뒈뒤듀드듸디따때땨떄떠떼뗘뗴또똬뙈뙤뚀뚜뚸뛔뛰뜌뜨띄띠라래랴럐러레려례로롸뢔뢰료루뤄뤠뤼류르릐리마매먀먜머메며몌모뫄뫠뫼묘무뭐뭬뮈뮤므믜미바배뱌뱨버베벼볘보봐봬뵈뵤부붜붸뷔뷰브븨비빠빼뺘뺴뻐뻬뼈뼤뽀뽜뽸뾔뾰뿌뿨쀄쀠쀼쁘쁴삐사새샤섀서세셔셰소솨쇄쇠쇼수숴쉐쉬슈스싀시싸쌔쌰썌써쎄쎠쎼쏘쏴쐐쐬쑈쑤쒀쒜쒸쓔쓰씌씨아애야얘어에여예오와왜외요우워웨위유으의이자재쟈쟤저제져졔조좌좨죄죠주줘줴쥐쥬즈즤지짜째쨔쨰쩌쩨쪄쪠쪼쫘쫴쬐쬬쭈쭤쮀쮜쮸쯔쯰찌차채챠챼처체쳐쳬초촤쵀최쵸추춰췌취츄츠츼치카캐캬컈커케켜켸코콰쾌쾨쿄쿠쿼퀘퀴큐크킈키타태탸턔터테텨톄토톼퇘퇴툐투퉈퉤튀튜트틔티파패퍄퍠퍼페펴폐포퐈퐤푀표푸풔풰퓌퓨프픠피하해햐햬허헤혀혜호화홰회효후훠훼휘휴흐희히]?[ᅠ-ᆢ]+|[가-힣])[ᆨ-ᇹ]*|[ᄀ-ᅟ]+|[^\p{Cc}\p{Cf}\p{Zl}\p{Zp}])[\p{Mn}\p{Mc}\p{Me}\x{09BE}\x{09D7}\x{0B3E}\x{0B57}\x{0BBE}\x{0BD7}\x{0CC2}\x{0CD5}\x{0CD6}\x{0D3E}\x{0D57}\x{0DCF}\x{0DDF}\x{200C}\x{1D165}\x{1D16E}-\x{1D172}\x{1F3FB}-\x{1F3FF}\x{FE0E}-\x{FE0F}\x{E0020}-\x{E007F}]*(?:\x{200D}(?:[\x{1F1E6}-\x{1F1FF}]|[ -~\x{200C}\x{200D}]|[ᆨ-ᇹ]+|[ᄀ-ᅟ]*(?:[가-힣]?[ᅠ-ᆢ]+|[가-힣])[ᆨ-ᇹ]*|[ᄀ-ᅟ]+|[^\p{Cc}\p{Cf}\p{Zl}\p{Zp}])[\p{Mn}\p{Mc}\p{Me}\x{09BE}\x{09D7}\x{0B3E}\x{0B57}\x{0BBE}\x{0BD7}\x{0CC2}\x{0CD5}\x{0CD6}\x{0D3E}\x{0D57}\x{0DCF}\x{0DDF}\x{200C}\x{1D165}\x{1D16E}-\x{1D172}\x{1F3FB}-\x{1F3FF}\x{FE0E}-\x{FE0F}\x{E0020}-\x{E007F}]*)*|[\p{Cc}\p{Cf}\p{Zl}\p{Zp}])';
|
||
|
||
// Table 3-7 of the Unicode standard for the well-formed byte sequences (group 1),
|
||
// then the maximal subparts an ill-formed sequence is made of (group 2).
|
||
private const UTF8_CHAR_OR_SUBPART_RX = '/([\x00-\x7F]|[\xC2-\xDF][\x80-\xBF]|\xE0[\xA0-\xBF][\x80-\xBF]|[\xE1-\xEC][\x80-\xBF]{2}|\xED[\x80-\x9F][\x80-\xBF]|[\xEE-\xEF][\x80-\xBF]{2}|\xF0[\x90-\xBF][\x80-\xBF]{2}|[\xF1-\xF3][\x80-\xBF]{3}|\xF4[\x80-\x8F][\x80-\xBF]{2})|(\xF0[\x90-\xBF][\x80-\xBF]?|[\xF1-\xF3][\x80-\xBF]{1,2}|\xF4[\x80-\x8F][\x80-\xBF]?|\xE0[\xA0-\xBF]|[\xE1-\xEC][\x80-\xBF]|\xED[\x80-\x9F]|[\xEE-\xEF][\x80-\xBF]|[\x80-\xFF])/s';
|
||
|
||
private const CASE_FOLD = [
|
||
['µ', 'ſ', "\xCD\x85", 'ς', "\xCF\x90", "\xCF\x91", "\xCF\x95", "\xCF\x96", "\xCF\xB0", "\xCF\xB1", "\xCF\xB5", "\xE1\xBA\x9B", "\xE1\xBE\xBE"],
|
||
['μ', 's', 'ι', 'σ', 'β', 'θ', 'φ', 'π', 'κ', 'ρ', 'ε', "\xE1\xB9\xA1", 'ι'],
|
||
];
|
||
|
||
// The characters that extend a grapheme cluster and that ICU 74 ignores when it compares strings
|
||
private const IGNORABLE_RX = '/^[\x{034F}\x{0488}-\x{0489}\x{0591}-\x{05AF}\x{05BD}\x{05C4}-\x{05C5}\x{0610}-\x{061A}\x{06D6}-\x{06DC}\x{06DF}-\x{06E4}\x{06E7}-\x{06E8}\x{06EA}-\x{06ED}\x{0740}\x{0743}-\x{0744}\x{0747}-\x{074A}\x{0898}-\x{089D}\x{08CA}-\x{08E1}\x{08EA}-\x{08EF}\x{08F3}\x{0951}-\x{0952}\x{0F18}-\x{0F19}\x{0F35}\x{0F37}\x{0F3E}-\x{0F3F}\x{0F86}-\x{0F87}\x{0FC6}\x{17B4}-\x{17B5}\x{17D3}\x{180B}-\x{180D}\x{180F}\x{1A7F}\x{1B6B}-\x{1B73}\x{1CD0}-\x{1CD2}\x{1CD4}-\x{1CE8}\x{1CF4}\x{1CF7}-\x{1CF9}\x{200C}-\x{200D}\x{2D7F}\x{A670}-\x{A672}\x{A8E0}-\x{A8F1}\x{FE00}-\x{FE0F}\x{FE21}\x{FE23}-\x{FE26}\x{FE28}\x{FE2A}-\x{FE2D}\x{FE2F}\x{102E0}\x{10EFD}-\x{10EFF}\x{11366}-\x{1136C}\x{11370}-\x{11374}\x{13440}\x{13447}-\x{13455}\x{16FE4}\x{1CF00}-\x{1CF2D}\x{1CF30}-\x{1CF46}\x{1D165}-\x{1D169}\x{1D16D}-\x{1D172}\x{1D17B}-\x{1D182}\x{1D185}-\x{1D18B}\x{1D1AA}-\x{1D1AD}\x{1D242}-\x{1D244}\x{1DA00}-\x{1DA36}\x{1DA3B}-\x{1DA6C}\x{1DA75}\x{1DA84}\x{1DA9B}-\x{1DA9F}\x{1DAA1}-\x{1DAAF}\x{1E8D0}-\x{1E8D6}\x{E0020}-\x{E007F}\x{E0100}-\x{E01EF}]*+\z/u';
|
||
|
||
// indexed by the $mode argument of grapheme_position()
|
||
private const POSITION_FUNCTIONS = ['grapheme_strpos', 'grapheme_stripos', 'grapheme_strrpos', 'grapheme_strripos'];
|
||
|
||
public static function grapheme_extract($s, $size, $type = \GRAPHEME_EXTR_COUNT, $start = 0, &$next = 0)
|
||
{
|
||
if (0 > $start) {
|
||
$start = \strlen($s) + $start;
|
||
}
|
||
|
||
if (!\is_scalar($s)) {
|
||
$hasError = false;
|
||
set_error_handler(static function () use (&$hasError) { $hasError = true; });
|
||
$next = substr($s, $start);
|
||
restore_error_handler();
|
||
if ($hasError) {
|
||
substr($s, $start);
|
||
$s = '';
|
||
} else {
|
||
$s = $next;
|
||
}
|
||
} else {
|
||
$s = substr($s, $start);
|
||
}
|
||
$size = (int) $size;
|
||
$type = (int) $type;
|
||
$start = (int) $start;
|
||
|
||
if (\GRAPHEME_EXTR_COUNT !== $type && \GRAPHEME_EXTR_MAXBYTES !== $type && \GRAPHEME_EXTR_MAXCHARS !== $type) {
|
||
if (80000 > \PHP_VERSION_ID) {
|
||
return false;
|
||
}
|
||
|
||
throw new \ValueError('grapheme_extract(): Argument #3 ($type) must be one of GRAPHEME_EXTR_COUNT, GRAPHEME_EXTR_MAXBYTES, or GRAPHEME_EXTR_MAXCHARS');
|
||
}
|
||
|
||
if (!isset($s[0]) || 0 > $size || 0 > $start) {
|
||
return false;
|
||
}
|
||
if (0 === $size) {
|
||
return '';
|
||
}
|
||
|
||
$next = $start;
|
||
|
||
$s = preg_split('/('.SYMFONY_GRAPHEME_CLUSTER_RX.')/u', "\r\n".$s, $size + 1, \PREG_SPLIT_NO_EMPTY | \PREG_SPLIT_DELIM_CAPTURE);
|
||
|
||
if (!isset($s[1])) {
|
||
return false;
|
||
}
|
||
|
||
$i = 1;
|
||
$ret = '';
|
||
|
||
do {
|
||
if (\GRAPHEME_EXTR_COUNT === $type) {
|
||
--$size;
|
||
} elseif (\GRAPHEME_EXTR_MAXBYTES === $type) {
|
||
$size -= \strlen($s[$i]);
|
||
} else {
|
||
$size -= iconv_strlen($s[$i], 'UTF-8//IGNORE');
|
||
}
|
||
|
||
if ($size >= 0) {
|
||
$ret .= $s[$i];
|
||
}
|
||
} while (isset($s[++$i]) && $size > 0);
|
||
|
||
$next += \strlen($ret);
|
||
|
||
return $ret;
|
||
}
|
||
|
||
public static function grapheme_strlen($s)
|
||
{
|
||
preg_replace('/'.SYMFONY_GRAPHEME_CLUSTER_RX.'/u', '', $s, -1, $len);
|
||
|
||
return 0 === $len && '' !== $s ? null : $len;
|
||
}
|
||
|
||
public static function grapheme_substr($s, $start, $len = null)
|
||
{
|
||
if (null === $len) {
|
||
$len = 2147483647;
|
||
}
|
||
|
||
if (!self::isUtf8($s)) {
|
||
return false;
|
||
}
|
||
|
||
preg_match_all('/'.SYMFONY_GRAPHEME_CLUSTER_RX.'/u', $s, $s);
|
||
|
||
$slen = \count($s[0]);
|
||
$start = (int) $start;
|
||
|
||
if (0 > $start) {
|
||
$start += $slen;
|
||
}
|
||
if (0 > $start) {
|
||
if (\PHP_VERSION_ID < 80000) {
|
||
return false;
|
||
}
|
||
|
||
$start = 0;
|
||
}
|
||
if ($start >= $slen) {
|
||
return \PHP_VERSION_ID >= 80000 ? '' : false;
|
||
}
|
||
|
||
$rem = $slen - $start;
|
||
|
||
if (0 > $len) {
|
||
$len += $rem;
|
||
}
|
||
if (0 === $len) {
|
||
return '';
|
||
}
|
||
if (0 > $len) {
|
||
return \PHP_VERSION_ID >= 80000 ? '' : false;
|
||
}
|
||
if ($len > $rem) {
|
||
$len = $rem;
|
||
}
|
||
|
||
return implode('', \array_slice($s[0], $start, $len));
|
||
}
|
||
|
||
public static function grapheme_strpos($s, $needle, $offset = 0)
|
||
{
|
||
return self::grapheme_position($s, $needle, $offset, 0);
|
||
}
|
||
|
||
public static function grapheme_stripos($s, $needle, $offset = 0)
|
||
{
|
||
return self::grapheme_position($s, $needle, $offset, 1);
|
||
}
|
||
|
||
public static function grapheme_strrpos($s, $needle, $offset = 0)
|
||
{
|
||
return self::grapheme_position($s, $needle, $offset, 2);
|
||
}
|
||
|
||
public static function grapheme_strripos($s, $needle, $offset = 0)
|
||
{
|
||
return self::grapheme_position($s, $needle, $offset, 3);
|
||
}
|
||
|
||
public static function grapheme_stristr($s, $needle, $beforeNeedle = false)
|
||
{
|
||
if (!self::isUtf8($s) || !self::isUtf8($needle)) {
|
||
return false;
|
||
}
|
||
|
||
return mb_stristr($s, $needle, $beforeNeedle, 'UTF-8');
|
||
}
|
||
|
||
public static function grapheme_strstr($s, $needle, $beforeNeedle = false)
|
||
{
|
||
if (!self::isUtf8($s) || !self::isUtf8($needle)) {
|
||
return false;
|
||
}
|
||
|
||
return mb_strstr($s, $needle, $beforeNeedle, 'UTF-8');
|
||
}
|
||
|
||
public static function grapheme_str_split($s, $len = 1)
|
||
{
|
||
if (0 >= $len || 1073741823 < $len) {
|
||
throw new \ValueError('grapheme_str_split(): Argument #2 ($length) must be greater than 0 and less than or equal to 1073741823');
|
||
}
|
||
|
||
if ('' === $s) {
|
||
return [];
|
||
}
|
||
|
||
if (preg_match_all('/('.SYMFONY_GRAPHEME_CLUSTER_RX.')/u', $s, $matches)) {
|
||
$graphemes = $matches[0];
|
||
} else {
|
||
$graphemes = self::splitIllFormed($s);
|
||
}
|
||
|
||
if (!$graphemes) {
|
||
return false;
|
||
}
|
||
|
||
if (1 === $len) {
|
||
return $graphemes;
|
||
}
|
||
|
||
$chunks = array_chunk($graphemes, $len);
|
||
|
||
foreach ($chunks as &$chunk) {
|
||
$chunk = implode('', $chunk);
|
||
}
|
||
|
||
return $chunks;
|
||
}
|
||
|
||
public static function grapheme_levenshtein($s1, $s2, $insertion_cost = 1, $replacement_cost = 1, $deletion_cost = 1, $locale = '')
|
||
{
|
||
if ($insertion_cost <= 0 || $insertion_cost > 1073741823) {
|
||
throw new \ValueError('grapheme_levenshtein(): Argument #3 ($insertion_cost) must be greater than 0 and less than or equal to 1073741823');
|
||
}
|
||
if ($replacement_cost <= 0 || $replacement_cost > 1073741823) {
|
||
throw new \ValueError('grapheme_levenshtein(): Argument #4 ($replacement_cost) must be greater than 0 and less than or equal to 1073741823');
|
||
}
|
||
if ($deletion_cost <= 0 || $deletion_cost > 1073741823) {
|
||
throw new \ValueError('grapheme_levenshtein(): Argument #5 ($deletion_cost) must be greater than 0 and less than or equal to 1073741823');
|
||
}
|
||
|
||
if (!preg_match('//u', $s1) || !preg_match('//u', $s2)) {
|
||
return false;
|
||
}
|
||
|
||
preg_match_all('/'.SYMFONY_GRAPHEME_CLUSTER_RX.'/u', $s1, $s1);
|
||
preg_match_all('/'.SYMFONY_GRAPHEME_CLUSTER_RX.'/u', $s2, $s2);
|
||
|
||
$s1 = $s1[0];
|
||
$s2 = $s2[0];
|
||
|
||
$l1 = \count($s1);
|
||
$l2 = \count($s2);
|
||
|
||
// Keep the rows as short as possible. Reversing the transformation
|
||
// swaps the meaning of insertion and deletion.
|
||
if ($l1 < $l2) {
|
||
[$s1, $s2] = [$s2, $s1];
|
||
[$l1, $l2] = [$l2, $l1];
|
||
[$insertion_cost, $deletion_cost] = [$deletion_cost, $insertion_cost];
|
||
}
|
||
|
||
if (0 === $l2) {
|
||
return $l1 * $deletion_cost;
|
||
}
|
||
|
||
if (\extension_loaded('intl')) {
|
||
// Graphemes are equal when the collator says so, as with ucol_strcoll() in intl
|
||
try {
|
||
$collator = new \Collator('' === $locale ? 'root' : $locale);
|
||
} catch (\IntlException $e) {
|
||
return false;
|
||
}
|
||
$s1 = array_map([$collator, 'getSortKey'], $s1);
|
||
$s2 = array_map([$collator, 'getSortKey'], $s2);
|
||
} elseif (class_exists(\Normalizer::class)) {
|
||
// Without a collator, at least treat canonically equivalent graphemes as equal
|
||
foreach ($s1 as $i => $g) {
|
||
$s1[$i] = \Normalizer::normalize($g, \Normalizer::FORM_D);
|
||
}
|
||
foreach ($s2 as $i => $g) {
|
||
$s2[$i] = \Normalizer::normalize($g, \Normalizer::FORM_D);
|
||
}
|
||
}
|
||
|
||
$previousRow = $currentRow = array_fill(0, $l2 + 1, 0);
|
||
for ($j = 1; $j <= $l2; ++$j) {
|
||
$previousRow[$j] = $previousRow[$j - 1] + $insertion_cost;
|
||
}
|
||
|
||
for ($i = 1; $i <= $l1; ++$i) {
|
||
$currentRow[0] = $previousRow[0] + $deletion_cost;
|
||
|
||
for ($j = 1; $j <= $l2; ++$j) {
|
||
$cost = ($s1[$i - 1] === $s2[$j - 1]) ? 0 : $replacement_cost;
|
||
$currentRow[$j] = min($previousRow[$j] + $deletion_cost, $currentRow[$j - 1] + $insertion_cost, $previousRow[$j - 1] + $cost);
|
||
}
|
||
|
||
[$previousRow, $currentRow] = [$currentRow, $previousRow];
|
||
}
|
||
|
||
return $previousRow[$l2];
|
||
}
|
||
|
||
private static function grapheme_position($s, $needle, $offset, $mode)
|
||
{
|
||
$needle = (string) $needle;
|
||
if (80000 > \PHP_VERSION_ID && !preg_match('/./us', $needle)) {
|
||
return false;
|
||
}
|
||
$s = (string) $s;
|
||
// let the empty string through: it accepts no offset but 0, which is checked below
|
||
if ('' !== $s && !preg_match('/./us', $s)) {
|
||
return false;
|
||
}
|
||
if ($offset && ($offset > ($len = self::grapheme_strlen($s)) || $offset < -$len)) {
|
||
if (80000 > \PHP_VERSION_ID) {
|
||
return false;
|
||
}
|
||
|
||
throw new \ValueError(self::POSITION_FUNCTIONS[$mode].'(): Argument #3 ($offset) must be contained in argument #1 ($haystack)');
|
||
}
|
||
if ('' === $s) {
|
||
return false;
|
||
}
|
||
if ($offset > 0) {
|
||
$s = self::grapheme_substr($s, $offset);
|
||
} elseif ($offset < 0) {
|
||
if (2 > $mode) {
|
||
$offset += $len;
|
||
$s = self::grapheme_substr($s, $offset);
|
||
if (0 > $offset) {
|
||
$offset = 0;
|
||
}
|
||
} elseif (0 > $offset += self::grapheme_strlen($needle)) {
|
||
$s = self::grapheme_substr($s, 0, $offset);
|
||
$offset = 0;
|
||
} else {
|
||
$offset = 0;
|
||
}
|
||
}
|
||
|
||
// As UTF-8 is self-synchronizing, and we have ensured the strings are valid UTF-8,
|
||
// we can use normal binary string functions here. For case-insensitive searches,
|
||
// case fold the strings first.
|
||
$caseInsensitive = $mode & 1;
|
||
$reverse = $mode & 2;
|
||
if ($caseInsensitive) {
|
||
// Use the same case folding mode as mbstring does for mb_stripos().
|
||
// Stick to SIMPLE case folding to avoid changing the length of the string, which
|
||
// might result in offsets being shifted.
|
||
$mode = \defined('MB_CASE_FOLD_SIMPLE') ? \MB_CASE_FOLD_SIMPLE : \MB_CASE_LOWER;
|
||
$s = mb_convert_case($s, $mode, 'UTF-8');
|
||
$needle = mb_convert_case($needle, $mode, 'UTF-8');
|
||
|
||
if (!\defined('MB_CASE_FOLD_SIMPLE')) {
|
||
$s = str_replace(self::CASE_FOLD[0], self::CASE_FOLD[1], $s);
|
||
$needle = str_replace(self::CASE_FOLD[0], self::CASE_FOLD[1], $needle);
|
||
}
|
||
}
|
||
$needleLen = \strlen($needle);
|
||
$needlePos = $reverse ? strrpos($s, $needle) : strpos($s, $needle);
|
||
|
||
while (false !== $needlePos) {
|
||
if (false !== $i = self::graphemeIndex($s, $needlePos, $needlePos + $needleLen)) {
|
||
return $i + $offset;
|
||
}
|
||
|
||
if (!$reverse) {
|
||
$needlePos = strpos($s, $needle, $needlePos + 1);
|
||
} elseif ($needlePos) {
|
||
$needlePos = strrpos(substr($s, 0, $needlePos + $needleLen - 1), $needle);
|
||
} else {
|
||
$needlePos = false;
|
||
}
|
||
}
|
||
|
||
return false;
|
||
}
|
||
|
||
/**
|
||
* Returns the position in grapheme units of the bytes from $start to $end,
|
||
* or false when they split a grapheme cluster on characters that are not ignorable.
|
||
*/
|
||
private static function graphemeIndex($s, $start, $end)
|
||
{
|
||
if (self::isAsciiBoundary($s, $start) && self::isAsciiBoundary($s, $end)) {
|
||
return self::grapheme_strlen(substr($s, 0, $start));
|
||
}
|
||
|
||
$i = $pos = 0;
|
||
|
||
if ($start) {
|
||
// count the clusters that start before $start, then find where the last one starts
|
||
$prefix = substr($s, 0, $start);
|
||
preg_replace('/'.SYMFONY_GRAPHEME_CLUSTER_RX.'/u', '', $prefix, -1, $i);
|
||
|
||
if (!$i) {
|
||
return false;
|
||
}
|
||
|
||
$pos = 1 < $i ? $start - \strlen(preg_replace('/'.SYMFONY_GRAPHEME_CLUSTER_RX.'/u', '', $prefix, $i - 1)) : 0;
|
||
}
|
||
|
||
do {
|
||
if (!preg_match('/\G'.SYMFONY_GRAPHEME_CLUSTER_RX.'/u', $s, $cluster, 0, $pos)) {
|
||
return false;
|
||
}
|
||
|
||
$next = $pos + \strlen($cluster[0]);
|
||
|
||
if ($pos < $start && $start < $next && !preg_match(self::IGNORABLE_RX, substr($s, $start, min($next, $end) - $start))) {
|
||
return false;
|
||
}
|
||
|
||
if ($pos < $end && $end < $next && !preg_match(self::IGNORABLE_RX, substr($s, $end, $next - $end))) {
|
||
return false;
|
||
}
|
||
} while (($pos = $next) < $end);
|
||
|
||
return $i;
|
||
}
|
||
|
||
private static function isAsciiBoundary($s, $i)
|
||
{
|
||
return 0 === $i || !isset($s[$i]) || ($s[$i - 1] < "\x80" && $s[$i] < "\x80" && ("\r" !== $s[$i - 1] || "\n" !== $s[$i]));
|
||
}
|
||
|
||
public static function grapheme_strrev(string $string)
|
||
{
|
||
if (\extension_loaded('intl')) {
|
||
$iterator = \IntlBreakIterator::createCharacterInstance();
|
||
$iterator->setText($string);
|
||
$reversed = '';
|
||
$end = $iterator->last();
|
||
|
||
while (\IntlBreakIterator::DONE !== $start = $iterator->previous()) {
|
||
$reversed .= substr($string, $start, $end - $start);
|
||
$end = $start;
|
||
}
|
||
|
||
return $reversed;
|
||
}
|
||
|
||
return implode('', array_reverse(self::grapheme_str_split($string)));
|
||
}
|
||
|
||
private static function isUtf8(?string $s)
|
||
{
|
||
$s = $s ?? '';
|
||
|
||
return '' === $s || preg_match('/./us', $s);
|
||
}
|
||
|
||
/**
|
||
* Splits a string that is not valid UTF-8 into grapheme clusters.
|
||
*
|
||
* The intl extension keeps the ill-formed bytes and breaks around them the
|
||
* way it would around the U+FFFD each maximal subpart stands for. Standing
|
||
* one in for real makes the cluster regexp applicable, then each cluster is
|
||
* mapped back to the bytes it was made of.
|
||
*/
|
||
private static function splitIllFormed($s)
|
||
{
|
||
preg_match_all(self::UTF8_CHAR_OR_SUBPART_RX, $s, $matches, \PREG_SET_ORDER);
|
||
|
||
$segments = [];
|
||
$segmentAt = [];
|
||
$sanitized = '';
|
||
|
||
foreach ($matches as $m) {
|
||
$segmentAt[\strlen($sanitized)] = \count($segments);
|
||
|
||
if (isset($m[2]) && '' !== $m[2]) {
|
||
$segments[] = $m[2];
|
||
$sanitized .= "\xEF\xBF\xBD";
|
||
} else {
|
||
$segments[] = $m[1];
|
||
$sanitized .= $m[1];
|
||
}
|
||
}
|
||
|
||
$segmentAt[\strlen($sanitized)] = \count($segments);
|
||
|
||
preg_match_all('/'.SYMFONY_GRAPHEME_CLUSTER_RX.'/u', $sanitized, $clusters, \PREG_OFFSET_CAPTURE);
|
||
|
||
$graphemes = [];
|
||
|
||
foreach ($clusters[0] as $cluster) {
|
||
$start = $segmentAt[$cluster[1]];
|
||
$graphemes[] = implode('', \array_slice($segments, $start, $segmentAt[$cluster[1] + \strlen($cluster[0])] - $start));
|
||
}
|
||
|
||
return $graphemes;
|
||
}
|
||
}
|