* * For the full copyright and license information, please view the LICENSE * file that was distributed with this source code. */ namespace Symfony\Component\String; use Symfony\Component\String\Exception\ExceptionInterface; use Symfony\Component\String\Exception\InvalidArgumentException; use Symfony\Component\String\Exception\RuntimeException; /** * Represents a string of abstract Unicode characters. * * Unicode defines 3 types of "characters" (bytes, code points and grapheme clusters). * This class is the abstract type to use as a type-hint when the logic you want to * implement is Unicode-aware but doesn't care about code points vs grapheme clusters. * * @author Nicolas Grekas * * @throws ExceptionInterface */ abstract class AbstractUnicodeString extends \Symfony\Component\String\AbstractString { public const NFC = \Normalizer::NFC; public const NFD = \Normalizer::NFD; public const NFKC = \Normalizer::NFKC; public const NFKD = \Normalizer::NFKD; // all ASCII letters sorted by typical frequency of occurrence private const ASCII = " eiasntrolud][cmp'\ng|hv.fb,:=-q10C2*yx)(L9AS/P\"EjMIk3>5T>', '<', '>', '-', '-', '-', '-', '-', '-', '-', '-', '-', '||', '/', '[', ']', '*', ',', '.', '<', '>', '<<', '>>', '[', ']', '[', ']', '[', ']', ',', '.', '[', ']', '<<', '>>', '<', '>', ',', '[', ']', '((', '))', '.', ',', '*', '/', '-', '/', '\\', '|', '||', '<<', '>>', '((', '))']; private static $transliterators = []; private static $tableZero; private static $tableWide; /** * @return static */ public static function fromCodePoints(int ...$codes) : self { $string = ''; foreach ($codes as $code) { if (0x80 > ($code %= 0x200000)) { $string .= \chr($code); } elseif (0x800 > $code) { $string .= \chr(0xc0 | $code >> 6) . \chr(0x80 | $code & 0x3f); } elseif (0x10000 > $code) { $string .= \chr(0xe0 | $code >> 12) . \chr(0x80 | $code >> 6 & 0x3f) . \chr(0x80 | $code & 0x3f); } else { $string .= \chr(0xf0 | $code >> 18) . \chr(0x80 | $code >> 12 & 0x3f) . \chr(0x80 | $code >> 6 & 0x3f) . \chr(0x80 | $code & 0x3f); } } return new static($string); } /** * Generic UTF-8 to ASCII transliteration. * * Install the intl extension for best results. * * @param string[]|\Transliterator[]|\Closure[] $rules See "*-Latin" rules from Transliterator::listIDs() */ public function ascii(array $rules = []) : self { $str = clone $this; $s = $str->string; $str->string = ''; \array_unshift($rules, 'nfd'); $rules[] = 'latin-ascii'; if (\function_exists('transliterator_transliterate')) { $rules[] = 'any-latin/bgn'; } $rules[] = 'nfkd'; $rules[] = '[:nonspacing mark:] remove'; while (\strlen($s) - 1 > ($i = \strspn($s, self::ASCII))) { if (0 < --$i) { $str->string .= \substr($s, 0, $i); $s = \substr($s, $i); } if (!($rule = \array_shift($rules))) { $rules = []; // An empty rule interrupts the next ones } if ($rule instanceof \Transliterator) { $s = $rule->transliterate($s); } elseif ($rule instanceof \Closure) { $s = $rule($s); } elseif ($rule) { if ('nfd' === ($rule = \strtolower($rule))) { \normalizer_is_normalized($s, self::NFD) ?: ($s = \normalizer_normalize($s, self::NFD)); } elseif ('nfkd' === $rule) { \normalizer_is_normalized($s, self::NFKD) ?: ($s = \normalizer_normalize($s, self::NFKD)); } elseif ('[:nonspacing mark:] remove' === $rule) { $s = \preg_replace('/\\p{Mn}++/u', '', $s); } elseif ('latin-ascii' === $rule) { $s = \str_replace(self::TRANSLIT_FROM, self::TRANSLIT_TO, $s); } elseif ('de-ascii' === $rule) { $s = \preg_replace("/([AUO])̈(?=\\p{Ll})/u", '$1e', $s); $s = \str_replace(["ä", "ö", "ü", "Ä", "Ö", "Ü"], ['ae', 'oe', 'ue', 'AE', 'OE', 'UE'], $s); } elseif (\function_exists('transliterator_transliterate')) { if (null === ($transliterator = self::$transliterators[$rule] ?? (self::$transliterators[$rule] = \Transliterator::create($rule)))) { if ('any-latin/bgn' === $rule) { $rule = 'any-latin'; $transliterator = self::$transliterators[$rule] ?? (self::$transliterators[$rule] = \Transliterator::create($rule)); } if (null === $transliterator) { throw new InvalidArgumentException(\sprintf('Unknown transliteration rule "%s".', $rule)); } self::$transliterators['any-latin/bgn'] = $transliterator; } $s = $transliterator->transliterate($s); } } elseif (!\function_exists('iconv')) { $s = \preg_replace('/[^\\x00-\\x7F]/u', '?', $s); } else { $s = @\preg_replace_callback('/[^\\x00-\\x7F]/u', static function ($c) { $c = (string) \iconv('UTF-8', 'ASCII//TRANSLIT', $c[0]); if ('' === $c && '' === \iconv('UTF-8', 'ASCII//TRANSLIT', '²')) { throw new \LogicException(\sprintf('"%s" requires a translit-able iconv implementation, try installing "gnu-libiconv" if you\'re using Alpine Linux.', static::class)); } return 1 < \strlen($c) ? \ltrim($c, '\'`"^~') : ('' !== $c ? $c : '?'); }, $s); } } $str->string .= $s; return $str; } public function camel() : parent { $str = clone $this; $str->string = \str_replace(' ', '', \preg_replace_callback('/\\b.(?![A-Z]{2,})/u', static function ($m) use(&$i) { return 1 === ++$i ? 'İ' === $m[0] ? 'i̇' : \mb_strtolower($m[0], 'UTF-8') : \mb_convert_case($m[0], \MB_CASE_TITLE, 'UTF-8'); }, \preg_replace('/[^\\pL0-9]++/u', ' ', $this->string))); return $str; } /** * @return int[] */ public function codePointsAt(int $offset) : array { $str = $this->slice($offset, 1); if ('' === $str->string) { return []; } $codePoints = []; foreach (\preg_split('//u', $str->string, -1, \PREG_SPLIT_NO_EMPTY) as $c) { $codePoints[] = \mb_ord($c, 'UTF-8'); } return $codePoints; } public function folded(bool $compat = \true) : parent { $str = clone $this; if (!$compat || \PHP_VERSION_ID < 70300 || !\defined('Normalizer::NFKC_CF')) { $str->string = \normalizer_normalize($str->string, $compat ? \Normalizer::NFKC : \Normalizer::NFC); $str->string = \mb_strtolower(\str_replace(self::FOLD_FROM, self::FOLD_TO, $this->string), 'UTF-8'); } else { $str->string = \normalizer_normalize($str->string, \Normalizer::NFKC_CF); } return $str; } public function join(array $strings, ?string $lastGlue = null) : parent { $str = clone $this; $tail = null !== $lastGlue && 1 < \count($strings) ? $lastGlue . \array_pop($strings) : ''; $str->string = \implode($this->string, $strings) . $tail; if (!\preg_match('//u', $str->string)) { throw new InvalidArgumentException('Invalid UTF-8 string.'); } return $str; } public function lower() : parent { $str = clone $this; $str->string = \mb_strtolower(\str_replace('İ', 'i̇', $str->string), 'UTF-8'); return $str; } public function match(string $regexp, int $flags = 0, int $offset = 0) : array { $match = (\PREG_PATTERN_ORDER | \PREG_SET_ORDER) & $flags ? 'preg_match_all' : 'preg_match'; if ($this->ignoreCase) { $regexp .= 'i'; } \set_error_handler(static function ($t, $m) { throw new InvalidArgumentException($m); }); try { if (\false === $match($regexp . 'u', $this->string, $matches, $flags | \PREG_UNMATCHED_AS_NULL, $offset)) { $lastError = \preg_last_error(); foreach (\get_defined_constants(\true)['pcre'] as $k => $v) { if ($lastError === $v && '_ERROR' === \substr($k, -6)) { throw new RuntimeException('Matching failed with ' . $k . '.'); } } throw new RuntimeException('Matching failed with unknown error code.'); } } finally { \restore_error_handler(); } return $matches; } /** * @return static */ public function normalize(int $form = self::NFC) : self { if (!\in_array($form, [self::NFC, self::NFD, self::NFKC, self::NFKD])) { throw new InvalidArgumentException('Unsupported normalization form.'); } $str = clone $this; \normalizer_is_normalized($str->string, $form) ?: ($str->string = \normalizer_normalize($str->string, $form)); return $str; } public function padBoth(int $length, string $padStr = ' ') : parent { if ('' === $padStr || !\preg_match('//u', $padStr)) { throw new InvalidArgumentException('Invalid UTF-8 string.'); } $pad = clone $this; $pad->string = $padStr; return $this->pad($length, $pad, \STR_PAD_BOTH); } public function padEnd(int $length, string $padStr = ' ') : parent { if ('' === $padStr || !\preg_match('//u', $padStr)) { throw new InvalidArgumentException('Invalid UTF-8 string.'); } $pad = clone $this; $pad->string = $padStr; return $this->pad($length, $pad, \STR_PAD_RIGHT); } public function padStart(int $length, string $padStr = ' ') : parent { if ('' === $padStr || !\preg_match('//u', $padStr)) { throw new InvalidArgumentException('Invalid UTF-8 string.'); } $pad = clone $this; $pad->string = $padStr; return $this->pad($length, $pad, \STR_PAD_LEFT); } public function replaceMatches(string $fromRegexp, $to) : parent { if ($this->ignoreCase) { $fromRegexp .= 'i'; } if (\is_array($to) || $to instanceof \Closure) { if (!\is_callable($to)) { throw new \TypeError(\sprintf('Argument 2 passed to "%s::replaceMatches()" must be callable, array given.', static::class)); } $replace = 'preg_replace_callback'; $to = static function (array $m) use($to) : string { $to = $to($m); if ('' !== $to && (!\is_string($to) || !\preg_match('//u', $to))) { throw new InvalidArgumentException('Replace callback must return a valid UTF-8 string.'); } return $to; }; } elseif ('' !== $to && !\preg_match('//u', $to)) { throw new InvalidArgumentException('Invalid UTF-8 string.'); } else { $replace = 'preg_replace'; } \set_error_handler(static function ($t, $m) { throw new InvalidArgumentException($m); }); try { if (null === ($string = $replace($fromRegexp . 'u', $to, $this->string))) { $lastError = \preg_last_error(); foreach (\get_defined_constants(\true)['pcre'] as $k => $v) { if ($lastError === $v && '_ERROR' === \substr($k, -6)) { throw new RuntimeException('Matching failed with ' . $k . '.'); } } throw new RuntimeException('Matching failed with unknown error code.'); } } finally { \restore_error_handler(); } $str = clone $this; $str->string = $string; return $str; } public function reverse() : parent { $str = clone $this; $str->string = \implode('', \array_reverse(\preg_split('/(\\X)/u', $str->string, -1, \PREG_SPLIT_DELIM_CAPTURE | \PREG_SPLIT_NO_EMPTY))); return $str; } public function snake() : parent { $str = $this->camel(); $str->string = \mb_strtolower(\preg_replace(['/(\\p{Lu}+)(\\p{Lu}\\p{Ll})/u', '/([\\p{Ll}0-9])(\\p{Lu})/u'], 'ps_metrics_module_v4_0_5\\1_\\2', $str->string), 'UTF-8'); return $str; } public function title(bool $allWords = \false) : parent { $str = clone $this; $limit = $allWords ? -1 : 1; $str->string = \preg_replace_callback('/\\b./u', static function (array $m) : string { return \mb_convert_case($m[0], \MB_CASE_TITLE, 'UTF-8'); }, $str->string, $limit); return $str; } public function trim(string $chars = " \t\n\r\x00\v\f ") : parent { if (" \t\n\r\x00\v\f " !== $chars && !\preg_match('//u', $chars)) { throw new InvalidArgumentException('Invalid UTF-8 chars.'); } $chars = \preg_quote($chars); $str = clone $this; $str->string = \preg_replace("{^[{$chars}]++|[{$chars}]++\$}uD", '', $str->string); return $str; } public function trimEnd(string $chars = " \t\n\r\x00\v\f ") : parent { if (" \t\n\r\x00\v\f " !== $chars && !\preg_match('//u', $chars)) { throw new InvalidArgumentException('Invalid UTF-8 chars.'); } $chars = \preg_quote($chars); $str = clone $this; $str->string = \preg_replace("{[{$chars}]++\$}uD", '', $str->string); return $str; } public function trimPrefix($prefix) : parent { if (!$this->ignoreCase) { return parent::trimPrefix($prefix); } $str = clone $this; if ($prefix instanceof \Traversable) { $prefix = \iterator_to_array($prefix, \false); } elseif ($prefix instanceof parent) { $prefix = $prefix->string; } $prefix = \implode('|', \array_map('preg_quote', (array) $prefix)); $str->string = \preg_replace("{^(?:{$prefix})}iuD", '', $this->string); return $str; } public function trimStart(string $chars = " \t\n\r\x00\v\f ") : parent { if (" \t\n\r\x00\v\f " !== $chars && !\preg_match('//u', $chars)) { throw new InvalidArgumentException('Invalid UTF-8 chars.'); } $chars = \preg_quote($chars); $str = clone $this; $str->string = \preg_replace("{^[{$chars}]++}uD", '', $str->string); return $str; } public function trimSuffix($suffix) : parent { if (!$this->ignoreCase) { return parent::trimSuffix($suffix); } $str = clone $this; if ($suffix instanceof \Traversable) { $suffix = \iterator_to_array($suffix, \false); } elseif ($suffix instanceof parent) { $suffix = $suffix->string; } $suffix = \implode('|', \array_map('preg_quote', (array) $suffix)); $str->string = \preg_replace("{(?:{$suffix})\$}iuD", '', $this->string); return $str; } public function upper() : parent { $str = clone $this; $str->string = \mb_strtoupper($str->string, 'UTF-8'); if (\PHP_VERSION_ID < 70300) { $str->string = \str_replace(self::UPPER_FROM, self::UPPER_TO, $str->string); } return $str; } public function width(bool $ignoreAnsiDecoration = \true) : int { $width = 0; $s = \str_replace(["\x00", "\x05", "\x07"], '', $this->string); if (\false !== \strpos($s, "\r")) { $s = \str_replace(["\r\n", "\r"], "\n", $s); } if (!$ignoreAnsiDecoration) { $s = \preg_replace('/[\\p{Cc}\\x7F]++/u', '', $s); } foreach (\explode("\n", $s) as $s) { if ($ignoreAnsiDecoration) { $s = \preg_replace('/(?:\\x1B(?: \\[ [\\x30-\\x3F]*+ [\\x20-\\x2F]*+ [\\x40-\\x7E] | [P\\]X^_] .*? \\x1B\\\\ | [\\x41-\\x7E] )|[\\p{Cc}\\x7F]++)/xu', '', $s); } $lineWidth = $this->wcswidth($s); if ($lineWidth > $width) { $width = $lineWidth; } } return $width; } /** * @return static */ private function pad(int $len, self $pad, int $type) : parent { $sLen = $this->length(); if ($len <= $sLen) { return clone $this; } $padLen = $pad->length(); $freeLen = $len - $sLen; $len = $freeLen % $padLen; switch ($type) { case \STR_PAD_RIGHT: return $this->append(\str_repeat($pad->string, \intdiv($freeLen, $padLen)) . ($len ? $pad->slice(0, $len) : '')); case \STR_PAD_LEFT: return $this->prepend(\str_repeat($pad->string, \intdiv($freeLen, $padLen)) . ($len ? $pad->slice(0, $len) : '')); case \STR_PAD_BOTH: $freeLen /= 2; $rightLen = \ceil($freeLen); $len = $rightLen % $padLen; $str = $this->append(\str_repeat($pad->string, \intdiv($rightLen, $padLen)) . ($len ? $pad->slice(0, $len) : '')); $leftLen = \floor($freeLen); $len = $leftLen % $padLen; return $str->prepend(\str_repeat($pad->string, \intdiv($leftLen, $padLen)) . ($len ? $pad->slice(0, $len) : '')); default: throw new InvalidArgumentException('Invalid padding type.'); } } /** * Based on https://github.com/jquast/wcwidth, a Python implementation of https://www.cl.cam.ac.uk/~mgk25/ucs/wcwidth.c. */ private function wcswidth(string $string) : int { $width = 0; foreach (\preg_split('//u', $string, -1, \PREG_SPLIT_NO_EMPTY) as $c) { $codePoint = \mb_ord($c, 'UTF-8'); if (0 === $codePoint || 0x34f === $codePoint || 0x200b <= $codePoint && 0x200f >= $codePoint || 0x2028 === $codePoint || 0x2029 === $codePoint || 0x202a <= $codePoint && 0x202e >= $codePoint || 0x2060 <= $codePoint && 0x2063 >= $codePoint) { continue; } // Non printable characters if (32 > $codePoint || 0x7f <= $codePoint && 0xa0 > $codePoint) { return -1; } if (null === self::$tableZero) { self::$tableZero = (require __DIR__ . '/Resources/data/wcswidth_table_zero.php'); } if ($codePoint >= self::$tableZero[0][0] && $codePoint <= self::$tableZero[$ubound = \count(self::$tableZero) - 1][1]) { $lbound = 0; while ($ubound >= $lbound) { $mid = \floor(($lbound + $ubound) / 2); if ($codePoint > self::$tableZero[$mid][1]) { $lbound = $mid + 1; } elseif ($codePoint < self::$tableZero[$mid][0]) { $ubound = $mid - 1; } else { continue 2; } } } if (null === self::$tableWide) { self::$tableWide = (require __DIR__ . '/Resources/data/wcswidth_table_wide.php'); } if ($codePoint >= self::$tableWide[0][0] && $codePoint <= self::$tableWide[$ubound = \count(self::$tableWide) - 1][1]) { $lbound = 0; while ($ubound >= $lbound) { $mid = \floor(($lbound + $ubound) / 2); if ($codePoint > self::$tableWide[$mid][1]) { $lbound = $mid + 1; } elseif ($codePoint < self::$tableWide[$mid][0]) { $ubound = $mid - 1; } else { $width += 2; continue 2; } } } ++$width; } return $width; } }