| 1 |
<?php |
| 2 |
|
| 3 |
/** |
| 4 |
* This file is part of the Nette Framework (https://nette.org) |
| 5 |
* Copyright (c) 2004 David Grudl (https://davidgrudl.com) |
| 6 |
*/ |
| 7 |
declare (strict_types=1); |
| 8 |
namespace Packetery\Nette\Utils; |
| 9 |
|
| 10 |
use Packetery\Nette; |
| 11 |
use function is_array, is_object, strlen; |
| 12 |
/** |
| 13 |
* String tools library. |
| 14 |
*/ |
| 15 |
class Strings |
| 16 |
{ |
| 17 |
use \Packetery\Nette\StaticClass; |
| 18 |
public const TRIM_CHARACTERS = " \t\n\r\x00\v "; |
| 19 |
/** |
| 20 |
* Checks if the string is valid in UTF-8 encoding. |
| 21 |
*/ |
| 22 |
public static function checkEncoding(string $s) : bool |
| 23 |
{ |
| 24 |
return $s === self::fixEncoding($s); |
| 25 |
} |
| 26 |
/** |
| 27 |
* Removes all invalid UTF-8 characters from a string. |
| 28 |
*/ |
| 29 |
public static function fixEncoding(string $s) : string |
| 30 |
{ |
| 31 |
// removes xD800-xDFFF, x110000 and higher |
| 32 |
return \htmlspecialchars_decode(\htmlspecialchars($s, \ENT_NOQUOTES | \ENT_IGNORE, 'UTF-8'), \ENT_NOQUOTES); |
| 33 |
} |
| 34 |
/** |
| 35 |
* Returns a specific character in UTF-8 from code point (number in range 0x0000..D7FF or 0xE000..10FFFF). |
| 36 |
* @throws \Packetery\Nette\InvalidArgumentException if code point is not in valid range |
| 37 |
*/ |
| 38 |
public static function chr(int $code) : string |
| 39 |
{ |
| 40 |
if ($code < 0 || $code >= 0xd800 && $code <= 0xdfff || $code > 0x10ffff) { |
| 41 |
throw new \Packetery\Nette\InvalidArgumentException('Code point must be in range 0x0 to 0xD7FF or 0xE000 to 0x10FFFF.'); |
| 42 |
} elseif (!\extension_loaded('iconv')) { |
| 43 |
throw new \Packetery\Nette\NotSupportedException(__METHOD__ . '() requires ICONV extension that is not loaded.'); |
| 44 |
} |
| 45 |
return \iconv('UTF-32BE', 'UTF-8//IGNORE', \pack('N', $code)); |
| 46 |
} |
| 47 |
/** |
| 48 |
* Starts the $haystack string with the prefix $needle? |
| 49 |
*/ |
| 50 |
public static function startsWith(string $haystack, string $needle) : bool |
| 51 |
{ |
| 52 |
return \strncmp($haystack, $needle, strlen($needle)) === 0; |
| 53 |
} |
| 54 |
/** |
| 55 |
* Ends the $haystack string with the suffix $needle? |
| 56 |
*/ |
| 57 |
public static function endsWith(string $haystack, string $needle) : bool |
| 58 |
{ |
| 59 |
return $needle === '' || \substr($haystack, -strlen($needle)) === $needle; |
| 60 |
} |
| 61 |
/** |
| 62 |
* Does $haystack contain $needle? |
| 63 |
*/ |
| 64 |
public static function contains(string $haystack, string $needle) : bool |
| 65 |
{ |
| 66 |
return \strpos($haystack, $needle) !== \false; |
| 67 |
} |
| 68 |
/** |
| 69 |
* Returns a part of UTF-8 string specified by starting position and length. If start is negative, |
| 70 |
* the returned string will start at the start'th character from the end of string. |
| 71 |
*/ |
| 72 |
public static function substring(string $s, int $start, ?int $length = null) : string |
| 73 |
{ |
| 74 |
if (\function_exists('mb_substr')) { |
| 75 |
return \mb_substr($s, $start, $length, 'UTF-8'); |
| 76 |
// MB is much faster |
| 77 |
} elseif (!\extension_loaded('iconv')) { |
| 78 |
throw new \Packetery\Nette\NotSupportedException(__METHOD__ . '() requires extension ICONV or MBSTRING, neither is loaded.'); |
| 79 |
} elseif ($length === null) { |
| 80 |
$length = self::length($s); |
| 81 |
} elseif ($start < 0 && $length < 0) { |
| 82 |
$start += self::length($s); |
| 83 |
// unifies iconv_substr behavior with mb_substr |
| 84 |
} |
| 85 |
return \iconv_substr($s, $start, $length, 'UTF-8'); |
| 86 |
} |
| 87 |
/** |
| 88 |
* Removes control characters, normalizes line breaks to `\n`, removes leading and trailing blank lines, |
| 89 |
* trims end spaces on lines, normalizes UTF-8 to the normal form of NFC. |
| 90 |
*/ |
| 91 |
public static function normalize(string $s) : string |
| 92 |
{ |
| 93 |
// convert to compressed normal form (NFC) |
| 94 |
if (\class_exists('Normalizer', \false) && ($n = \Normalizer::normalize($s, \Normalizer::FORM_C)) !== \false) { |
| 95 |
$s = $n; |
| 96 |
} |
| 97 |
$s = self::normalizeNewLines($s); |
| 98 |
// remove control characters; leave \t + \n |
| 99 |
$s = self::pcre('preg_replace', ['#[\\x00-\\x08\\x0B-\\x1F\\x7F-\\x9F]+#u', '', $s]); |
| 100 |
// right trim |
| 101 |
$s = self::pcre('preg_replace', ['#[\\t ]+$#m', '', $s]); |
| 102 |
// leading and trailing blank lines |
| 103 |
$s = \trim($s, "\n"); |
| 104 |
return $s; |
| 105 |
} |
| 106 |
/** |
| 107 |
* Standardize line endings to unix-like. |
| 108 |
*/ |
| 109 |
public static function normalizeNewLines(string $s) : string |
| 110 |
{ |
| 111 |
return \str_replace(["\r\n", "\r"], "\n", $s); |
| 112 |
} |
| 113 |
/** |
| 114 |
* Converts UTF-8 string to ASCII, ie removes diacritics etc. |
| 115 |
*/ |
| 116 |
public static function toAscii(string $s) : string |
| 117 |
{ |
| 118 |
$iconv = \defined('ICONV_IMPL') ? \trim(\ICONV_IMPL, '"\'') : null; |
| 119 |
static $transliterator = null; |
| 120 |
if ($transliterator === null) { |
| 121 |
if (\class_exists('Transliterator', \false)) { |
| 122 |
$transliterator = \Transliterator::create('Any-Latin; Latin-ASCII'); |
| 123 |
} else { |
| 124 |
\trigger_error(__METHOD__ . "(): it is recommended to enable PHP extensions 'intl'.", \E_USER_NOTICE); |
| 125 |
$transliterator = \false; |
| 126 |
} |
| 127 |
} |
| 128 |
// remove control characters and check UTF-8 validity |
| 129 |
$s = self::pcre('preg_replace', ['#[^\\x09\\x0A\\x0D\\x20-\\x7E\\xA0-\\x{2FF}\\x{370}-\\x{10FFFF}]#u', '', $s]); |
| 130 |
// transliteration (by Transliterator and iconv) is not optimal, replace some characters directly |
| 131 |
$s = \strtr($s, ["„" => '"', "“" => '"', "”" => '"', "‚" => "'", "‘" => "'", "’" => "'", "°" => '^', "Я" => 'Ya', "я" => 'ya', "Ю" => 'Yu', "ю" => 'yu', "Ä" => 'Ae', "Ö" => 'Oe', "Ü" => 'Ue', "ẞ" => 'Ss', "ä" => 'ae', "ö" => 'oe', "ü" => 'ue', "ß" => 'ss']); |
| 132 |
// „ “ ” ‚ ‘ ’ ° Я я Ю ю Ä Ö Ü ẞ ä ö ü ß |
| 133 |
if ($iconv !== 'libiconv') { |
| 134 |
$s = \strtr($s, ["®" => '(R)', "©" => '(c)', "…" => '...', "«" => '<<', "»" => '>>', "£" => 'lb', "¥" => 'yen', "²" => '^2', "³" => '^3', "µ" => 'u', "¹" => '^1', "º" => 'o', "¿" => '?', "ˊ" => "'", "ˍ" => '_', "˝" => '"', "`" => '', "€" => 'EUR', "™" => 'TM', "℮" => 'e', "←" => '<-', "↑" => '^', "→" => '->', "↓" => 'V', "↔" => '<->']); |
| 135 |
// ® © … « » £ ¥ ² ³ µ ¹ º ¿ ˊ ˍ ˝ ` € ™ ℮ ← ↑ → ↓ ↔ |
| 136 |
} |
| 137 |
if ($transliterator) { |
| 138 |
$s = $transliterator->transliterate($s); |
| 139 |
// use iconv because The transliterator leaves some characters out of ASCII, eg → ʾ |
| 140 |
if ($iconv === 'glibc') { |
| 141 |
$s = \strtr($s, '?', "\x01"); |
| 142 |
// temporarily hide ? to distinguish them from the garbage that iconv creates |
| 143 |
$s = \iconv('UTF-8', 'ASCII//TRANSLIT//IGNORE', $s); |
| 144 |
$s = \str_replace(['?', "\x01"], ['', '?'], $s); |
| 145 |
// remove garbage and restore ? characters |
| 146 |
} elseif ($iconv === 'libiconv') { |
| 147 |
$s = \iconv('UTF-8', 'ASCII//TRANSLIT//IGNORE', $s); |
| 148 |
} else { |
| 149 |
// null or 'unknown' (#216) |
| 150 |
$s = self::pcre('preg_replace', ['#[^\\x00-\\x7F]++#', '', $s]); |
| 151 |
// remove non-ascii chars |
| 152 |
} |
| 153 |
} elseif ($iconv === 'glibc' || $iconv === 'libiconv') { |
| 154 |
// temporarily hide these characters to distinguish them from the garbage that iconv creates |
| 155 |
$s = \strtr($s, '`\'"^~?', "\x01\x02\x03\x04\x05\x06"); |
| 156 |
if ($iconv === 'glibc') { |
| 157 |
// glibc implementation is very limited. transliterate into Windows-1250 and then into ASCII, so most Eastern European characters are preserved |
| 158 |
$s = \iconv('UTF-8', 'WINDOWS-1250//TRANSLIT//IGNORE', $s); |
| 159 |
$s = \strtr($s, "\xa5\xa3\xbc\x8c\xa7\x8a\xaa\x8d\x8f\x8e\xaf\xb9\xb3\xbe\x9c\x9a\xba\x9d\x9f\x9e\xbf\xc0\xc1\xc2\xc3\xc4\xc5\xc6\xc7\xc8\xc9\xca\xcb\xcc\xcd\xce\xcf\xd0\xd1\xd2\xd3\xd4\xd5\xd6\xd7\xd8\xd9\xda\xdb\xdc\xdd\xde\xdf\xe0\xe1\xe2\xe3\xe4\xe5\xe6\xe7\xe8\xe9\xea\xeb\xec\xed\xee\xef\xf0\xf1\xf2\xf3\xf4\xf5\xf6\xf8\xf9\xfa\xfb\xfc\xfd\xfe\x96\xa0\x8b\x97\x9b\xa6\xad\xb7", 'ALLSSSSTZZZallssstzzzRAAAALCCCEEEEIIDDNNOOOOxRUUUUYTsraaaalccceeeeiiddnnooooruuuuyt- <->|-.'); |
| 160 |
$s = self::pcre('preg_replace', ['#[^\\x00-\\x7F]++#', '', $s]); |
| 161 |
} else { |
| 162 |
$s = \iconv('UTF-8', 'ASCII//TRANSLIT//IGNORE', $s); |
| 163 |
} |
| 164 |
// remove garbage that iconv creates during transliteration (eg Ý -> Y') |
| 165 |
$s = \str_replace(['`', "'", '"', '^', '~', '?'], '', $s); |
| 166 |
// restore temporarily hidden characters |
| 167 |
$s = \strtr($s, "\x01\x02\x03\x04\x05\x06", '`\'"^~?'); |
| 168 |
} else { |
| 169 |
$s = self::pcre('preg_replace', ['#[^\\x00-\\x7F]++#', '', $s]); |
| 170 |
// remove non-ascii chars |
| 171 |
} |
| 172 |
return $s; |
| 173 |
} |
| 174 |
/** |
| 175 |
* Modifies the UTF-8 string to the form used in the URL, ie removes diacritics and replaces all characters |
| 176 |
* except letters of the English alphabet and numbers with a hyphens. |
| 177 |
*/ |
| 178 |
public static function webalize(string $s, ?string $charlist = null, bool $lower = \true) : string |
| 179 |
{ |
| 180 |
$s = self::toAscii($s); |
| 181 |
if ($lower) { |
| 182 |
$s = \strtolower($s); |
| 183 |
} |
| 184 |
$s = self::pcre('preg_replace', ['#[^a-z0-9' . ($charlist !== null ? \preg_quote($charlist, '#') : '') . ']+#i', '-', $s]); |
| 185 |
$s = \trim($s, '-'); |
| 186 |
return $s; |
| 187 |
} |
| 188 |
/** |
| 189 |
* Truncates a UTF-8 string to given maximal length, while trying not to split whole words. Only if the string is truncated, |
| 190 |
* an ellipsis (or something else set with third argument) is appended to the string. |
| 191 |
*/ |
| 192 |
public static function truncate(string $s, int $maxLen, string $append = "…") : string |
| 193 |
{ |
| 194 |
if (self::length($s) > $maxLen) { |
| 195 |
$maxLen -= self::length($append); |
| 196 |
if ($maxLen < 1) { |
| 197 |
return $append; |
| 198 |
} elseif ($matches = self::match($s, '#^.{1,' . $maxLen . '}(?=[\\s\\x00-/:-@\\[-`{-~])#us')) { |
| 199 |
return $matches[0] . $append; |
| 200 |
} else { |
| 201 |
return self::substring($s, 0, $maxLen) . $append; |
| 202 |
} |
| 203 |
} |
| 204 |
return $s; |
| 205 |
} |
| 206 |
/** |
| 207 |
* Indents a multiline text from the left. Second argument sets how many indentation chars should be used, |
| 208 |
* while the indent itself is the third argument (*tab* by default). |
| 209 |
*/ |
| 210 |
public static function indent(string $s, int $level = 1, string $chars = "\t") : string |
| 211 |
{ |
| 212 |
if ($level > 0) { |
| 213 |
$s = self::replace($s, '#(?:^|[\\r\\n]+)(?=[^\\r\\n])#', '$0' . \str_repeat($chars, $level)); |
| 214 |
} |
| 215 |
return $s; |
| 216 |
} |
| 217 |
/** |
| 218 |
* Converts all characters of UTF-8 string to lower case. |
| 219 |
*/ |
| 220 |
public static function lower(string $s) : string |
| 221 |
{ |
| 222 |
return \mb_strtolower($s, 'UTF-8'); |
| 223 |
} |
| 224 |
/** |
| 225 |
* Converts the first character of a UTF-8 string to lower case and leaves the other characters unchanged. |
| 226 |
*/ |
| 227 |
public static function firstLower(string $s) : string |
| 228 |
{ |
| 229 |
return self::lower(self::substring($s, 0, 1)) . self::substring($s, 1); |
| 230 |
} |
| 231 |
/** |
| 232 |
* Converts all characters of a UTF-8 string to upper case. |
| 233 |
*/ |
| 234 |
public static function upper(string $s) : string |
| 235 |
{ |
| 236 |
return \mb_strtoupper($s, 'UTF-8'); |
| 237 |
} |
| 238 |
/** |
| 239 |
* Converts the first character of a UTF-8 string to upper case and leaves the other characters unchanged. |
| 240 |
*/ |
| 241 |
public static function firstUpper(string $s) : string |
| 242 |
{ |
| 243 |
return self::upper(self::substring($s, 0, 1)) . self::substring($s, 1); |
| 244 |
} |
| 245 |
/** |
| 246 |
* Converts the first character of every word of a UTF-8 string to upper case and the others to lower case. |
| 247 |
*/ |
| 248 |
public static function capitalize(string $s) : string |
| 249 |
{ |
| 250 |
return \mb_convert_case($s, \MB_CASE_TITLE, 'UTF-8'); |
| 251 |
} |
| 252 |
/** |
| 253 |
* Compares two UTF-8 strings or their parts, without taking character case into account. If length is null, whole strings are compared, |
| 254 |
* if it is negative, the corresponding number of characters from the end of the strings is compared, |
| 255 |
* otherwise the appropriate number of characters from the beginning is compared. |
| 256 |
*/ |
| 257 |
public static function compare(string $left, string $right, ?int $length = null) : bool |
| 258 |
{ |
| 259 |
if (\class_exists('Normalizer', \false)) { |
| 260 |
$left = \Normalizer::normalize($left, \Normalizer::FORM_D); |
| 261 |
// form NFD is faster |
| 262 |
$right = \Normalizer::normalize($right, \Normalizer::FORM_D); |
| 263 |
// form NFD is faster |
| 264 |
} |
| 265 |
if ($length < 0) { |
| 266 |
$left = self::substring($left, $length, -$length); |
| 267 |
$right = self::substring($right, $length, -$length); |
| 268 |
} elseif ($length !== null) { |
| 269 |
$left = self::substring($left, 0, $length); |
| 270 |
$right = self::substring($right, 0, $length); |
| 271 |
} |
| 272 |
return self::lower($left) === self::lower($right); |
| 273 |
} |
| 274 |
/** |
| 275 |
* Finds the common prefix of strings or returns empty string if the prefix was not found. |
| 276 |
* @param string[] $strings |
| 277 |
*/ |
| 278 |
public static function findPrefix(array $strings) : string |
| 279 |
{ |
| 280 |
$first = \array_shift($strings); |
| 281 |
for ($i = 0; $i < strlen($first); $i++) { |
| 282 |
foreach ($strings as $s) { |
| 283 |
if (!isset($s[$i]) || $first[$i] !== $s[$i]) { |
| 284 |
while ($i && $first[$i - 1] >= "\x80" && $first[$i] >= "\x80" && $first[$i] < "\xc0") { |
| 285 |
$i--; |
| 286 |
} |
| 287 |
return \substr($first, 0, $i); |
| 288 |
} |
| 289 |
} |
| 290 |
} |
| 291 |
return $first; |
| 292 |
} |
| 293 |
/** |
| 294 |
* Returns number of characters (not bytes) in UTF-8 string. |
| 295 |
* That is the number of Unicode code points which may differ from the number of graphemes. |
| 296 |
*/ |
| 297 |
public static function length(string $s) : int |
| 298 |
{ |
| 299 |
return \function_exists('mb_strlen') ? \mb_strlen($s, 'UTF-8') : strlen(\utf8_decode($s)); |
| 300 |
} |
| 301 |
/** |
| 302 |
* Removes all left and right side spaces (or the characters passed as second argument) from a UTF-8 encoded string. |
| 303 |
*/ |
| 304 |
public static function trim(string $s, string $charlist = self::TRIM_CHARACTERS) : string |
| 305 |
{ |
| 306 |
$charlist = \preg_quote($charlist, '#'); |
| 307 |
return self::replace($s, '#^[' . $charlist . ']+|[' . $charlist . ']+$#Du', ''); |
| 308 |
} |
| 309 |
/** |
| 310 |
* Pads a UTF-8 string to given length by prepending the $pad string to the beginning. |
| 311 |
*/ |
| 312 |
public static function padLeft(string $s, int $length, string $pad = ' ') : string |
| 313 |
{ |
| 314 |
$length = \max(0, $length - self::length($s)); |
| 315 |
$padLen = self::length($pad); |
| 316 |
return \str_repeat($pad, (int) ($length / $padLen)) . self::substring($pad, 0, $length % $padLen) . $s; |
| 317 |
} |
| 318 |
/** |
| 319 |
* Pads UTF-8 string to given length by appending the $pad string to the end. |
| 320 |
*/ |
| 321 |
public static function padRight(string $s, int $length, string $pad = ' ') : string |
| 322 |
{ |
| 323 |
$length = \max(0, $length - self::length($s)); |
| 324 |
$padLen = self::length($pad); |
| 325 |
return $s . \str_repeat($pad, (int) ($length / $padLen)) . self::substring($pad, 0, $length % $padLen); |
| 326 |
} |
| 327 |
/** |
| 328 |
* Reverses UTF-8 string. |
| 329 |
*/ |
| 330 |
public static function reverse(string $s) : string |
| 331 |
{ |
| 332 |
if (!\extension_loaded('iconv')) { |
| 333 |
throw new \Packetery\Nette\NotSupportedException(__METHOD__ . '() requires ICONV extension that is not loaded.'); |
| 334 |
} |
| 335 |
return \iconv('UTF-32LE', 'UTF-8', \strrev(\iconv('UTF-8', 'UTF-32BE', $s))); |
| 336 |
} |
| 337 |
/** |
| 338 |
* Returns part of $haystack before $nth occurence of $needle or returns null if the needle was not found. |
| 339 |
* Negative value means searching from the end. |
| 340 |
*/ |
| 341 |
public static function before(string $haystack, string $needle, int $nth = 1) : ?string |
| 342 |
{ |
| 343 |
$pos = self::pos($haystack, $needle, $nth); |
| 344 |
return $pos === null ? null : \substr($haystack, 0, $pos); |
| 345 |
} |
| 346 |
/** |
| 347 |
* Returns part of $haystack after $nth occurence of $needle or returns null if the needle was not found. |
| 348 |
* Negative value means searching from the end. |
| 349 |
*/ |
| 350 |
public static function after(string $haystack, string $needle, int $nth = 1) : ?string |
| 351 |
{ |
| 352 |
$pos = self::pos($haystack, $needle, $nth); |
| 353 |
return $pos === null ? null : \substr($haystack, $pos + strlen($needle)); |
| 354 |
} |
| 355 |
/** |
| 356 |
* Returns position in characters of $nth occurence of $needle in $haystack or null if the $needle was not found. |
| 357 |
* Negative value of `$nth` means searching from the end. |
| 358 |
*/ |
| 359 |
public static function indexOf(string $haystack, string $needle, int $nth = 1) : ?int |
| 360 |
{ |
| 361 |
$pos = self::pos($haystack, $needle, $nth); |
| 362 |
return $pos === null ? null : self::length(\substr($haystack, 0, $pos)); |
| 363 |
} |
| 364 |
/** |
| 365 |
* Returns position in characters of $nth occurence of $needle in $haystack or null if the needle was not found. |
| 366 |
*/ |
| 367 |
private static function pos(string $haystack, string $needle, int $nth = 1) : ?int |
| 368 |
{ |
| 369 |
if (!$nth) { |
| 370 |
return null; |
| 371 |
} elseif ($nth > 0) { |
| 372 |
if ($needle === '') { |
| 373 |
return 0; |
| 374 |
} |
| 375 |
$pos = 0; |
| 376 |
while (($pos = \strpos($haystack, $needle, $pos)) !== \false && --$nth) { |
| 377 |
$pos++; |
| 378 |
} |
| 379 |
} else { |
| 380 |
$len = strlen($haystack); |
| 381 |
if ($needle === '') { |
| 382 |
return $len; |
| 383 |
} elseif ($len === 0) { |
| 384 |
return null; |
| 385 |
} |
| 386 |
$pos = $len - 1; |
| 387 |
while (($pos = \strrpos($haystack, $needle, $pos - $len)) !== \false && ++$nth) { |
| 388 |
$pos--; |
| 389 |
} |
| 390 |
} |
| 391 |
return Helpers::falseToNull($pos); |
| 392 |
} |
| 393 |
/** |
| 394 |
* Splits a string into array by the regular expression. Parenthesized expression in the delimiter are captured. |
| 395 |
* Parameter $flags can be any combination of PREG_SPLIT_NO_EMPTY and PREG_OFFSET_CAPTURE flags. |
| 396 |
*/ |
| 397 |
public static function split(string $subject, string $pattern, int $flags = 0) : array |
| 398 |
{ |
| 399 |
return self::pcre('preg_split', [$pattern, $subject, -1, $flags | \PREG_SPLIT_DELIM_CAPTURE]); |
| 400 |
} |
| 401 |
/** |
| 402 |
* Checks if given string matches a regular expression pattern and returns an array with first found match and each subpattern. |
| 403 |
* Parameter $flags can be any combination of PREG_OFFSET_CAPTURE and PREG_UNMATCHED_AS_NULL flags. |
| 404 |
*/ |
| 405 |
public static function match(string $subject, string $pattern, int $flags = 0, int $offset = 0) : ?array |
| 406 |
{ |
| 407 |
if ($offset > strlen($subject)) { |
| 408 |
return null; |
| 409 |
} |
| 410 |
return self::pcre('preg_match', [$pattern, $subject, &$m, $flags, $offset]) ? $m : null; |
| 411 |
} |
| 412 |
/** |
| 413 |
* Finds all occurrences matching regular expression pattern and returns a two-dimensional array. Result is array of matches (ie uses by default PREG_SET_ORDER). |
| 414 |
* Parameter $flags can be any combination of PREG_OFFSET_CAPTURE, PREG_UNMATCHED_AS_NULL and PREG_PATTERN_ORDER flags. |
| 415 |
*/ |
| 416 |
public static function matchAll(string $subject, string $pattern, int $flags = 0, int $offset = 0) : array |
| 417 |
{ |
| 418 |
if ($offset > strlen($subject)) { |
| 419 |
return []; |
| 420 |
} |
| 421 |
self::pcre('preg_match_all', [$pattern, $subject, &$m, $flags & \PREG_PATTERN_ORDER ? $flags : $flags | \PREG_SET_ORDER, $offset]); |
| 422 |
return $m; |
| 423 |
} |
| 424 |
/** |
| 425 |
* Replaces all occurrences matching regular expression $pattern which can be string or array in the form `pattern => replacement`. |
| 426 |
* @param string|array $pattern |
| 427 |
* @param string|callable $replacement |
| 428 |
*/ |
| 429 |
public static function replace(string $subject, $pattern, $replacement = '', int $limit = -1) : string |
| 430 |
{ |
| 431 |
if (is_object($replacement) || is_array($replacement)) { |
| 432 |
if (!\is_callable($replacement, \false, $textual)) { |
| 433 |
throw new \Packetery\Nette\InvalidStateException("Callback '{$textual}' is not callable."); |
| 434 |
} |
| 435 |
return self::pcre('preg_replace_callback', [$pattern, $replacement, $subject, $limit]); |
| 436 |
} elseif (is_array($pattern) && \is_string(\key($pattern))) { |
| 437 |
$replacement = \array_values($pattern); |
| 438 |
$pattern = \array_keys($pattern); |
| 439 |
} |
| 440 |
return self::pcre('preg_replace', [$pattern, $replacement, $subject, $limit]); |
| 441 |
} |
| 442 |
/** @internal */ |
| 443 |
public static function pcre(string $func, array $args) |
| 444 |
{ |
| 445 |
$res = Callback::invokeSafe($func, $args, function (string $message) use($args) : void { |
| 446 |
// compile-time error, not detectable by preg_last_error |
| 447 |
throw new RegexpException($message . ' in pattern: ' . \implode(' or ', (array) $args[0])); |
| 448 |
}); |
| 449 |
if (($code = \preg_last_error()) && ($res === null || !\in_array($func, ['preg_filter', 'preg_replace_callback', 'preg_replace'], \true))) { |
| 450 |
throw new RegexpException((RegexpException::MESSAGES[$code] ?? 'Unknown error') . ' (pattern: ' . \implode(' or ', (array) $args[0]) . ')', $code); |
| 451 |
} |
| 452 |
return $res; |
| 453 |
} |
| 454 |
} |
| 455 |
|
